impacts of contrasting land-use history on composition ... · impacts of contrasting land-use...

Impacts of contrasting land-use history on composition, soils,and development of mixed-oak, coastal plain forests on Shelter

Island, New York

Marc D. Abrams1,2 and Vanessa L. W. HayesSchool of Forest Resources, 307 Forest Resources Building, Penn State University,

University Park, PA 16802

ABRAMS, MARC D. AND VANESSA L. W. HAYES (School of Forest Resources, Forest Resources Building,Penn State University, University Park, PA 16802). Impacts of contrasting land-use history on composition,soils, and development of mixed-oak, coastal plain forests on Shelter Island, New York. J. Torrey Bot. Soc.135: 37–52. 2008.—The Mashomack Preserve on Shelter Island, New York contains one of the premier oak-dominated coastal plain forests in the northeastern U.S. It represents a unique opportunity to study woodyspecies in juxtaposed forests that differ primarily in past land-use (logging versus agriculture). We researchedhow contrasting land-use history affected: 1) tree species composition, size, and age structure; 2) soilchemistry and morphology; 3) the dominance of Smilax, a native invasive shrub; and 4) the historicalecology and successional pathways. Randomly located plots were sampled for vegetation and soils in theunplowed interior forest versus the maritime forest that was cleared and/or plowed for agriculture andgrazing in the 18th and 19th centuries and then abandoned after 1870. The upper soil profile was examined forsoil nutrients and pH, the presence/absence of a plow layer (Ap) and for soil charcoal. Tree cores (N 5 130)were taken across both forest types to include the full range of species and diameter size classes to assesstemporal and spatial variation in recruitment. Maritime forests are closer to the ocean shore and aredominated by Quercus velutina, Sassafras albidum, Carya glabra, and Q. alba, whereas the interior forests aredominated by Q. velutina, Acer rubrum, C. glabra, and Q. prinus. The interior forest plots appear to havecontinually supported forests, have soils that are somewhat doughtier and less fertile, and lack a plowedhorizon, but have frequent soil charcoal as compared with the maritime forest. Formerly plowed and grazedmaritime soils are more fertile and Q. alba has greater importance, while Q. velutina has lower importance.Following agricultural abandonment, the maritime forest understory became dominated by thickets ofSmilax and presently has lower tree diversity than interior forests. Tree regeneration is sparse in both foresttypes due, at least in part, to intense browsing by white-tailed deer (Odocoileus virginianus). The interiorforest exhibits a predictable successional pathway from mixed-oak to A. rubrum and Fagus grandifolia, whichis not apparent in the maritime black oak forests. We believe that differences in overstory and understorycomposition, species diversity, and successional pathways between the two forest types can be mainlyattributed to contrasting land-use history. Smilax is a super-dominant species that has profoundly influencedmany ecological processes on the preserve and if left unchecked will continue to do so long into the future.

Key words: biogeography, disturbance, fire, invasive, plowed layer, Smilax, white-tailed deer.

The natural distribution or biogeography of

plant species around the world has primarily

been studied in relation to variation in climate,

soils, and topography (Eyre 1963, Whittaker

1975, Ellenberg 1995). For example, the

Holdridge (1967) triangle separates world-

wide vegetation biomes by precipitation,

temperature, and humidity province. Eyre

(1963) describes the relationship of vegetation

and soils for various regions of the world,

including temperate, boreal, and tropical

biomes. Using gradient analysis, Whittaker

(1975) was able to separate vegetation types by

topography, elevation, and soils in montane

regions of the eastern and western U.S. More

recently, however, researchers have identified

land-use history as an important determinant

of vegetation pattern in many world-wide

ecosystems (Foster 1992, Orwig and Abrams

1994, Verheyen et al. 1999, Dahlstrom et al.

2006). In the northeastern U.S., forest that

developed on plowed soils following agricul-

tural abandonment support different trees

species and abundances when compared with

nearby unplowed soils (Glitzenstein et al.

1990, Motzkin et al. 1996). Land-use history

has also been reported to impact soil chemistry

and nutrient cycling processes (Glatzel 1991,

Connell et al. 1994, Verheyen et al. 1999). In

1 Author for correspondence. E-mail: [email protected]

2 The Nature Conservancy Mashomack Preserveand Eppley Foundation provided funding for thisstudy. We want to thank Barnaby Friedman forproducing several of the GIS maps used in thisstudy, Mike Scheibel for his help with experimentaldesign and field data collection, and along withMichael Laspia and Marilyn Jordan, for their helpwith data interpretation.

Received for publication October 27, 2007, and inrevised form January 2, 2008.

Journal of the Torrey Botanical Society 135(1), 2008, pp. 37–52

37

western Belgium, cultivated soils had higher

pH, Ca, and P than forests and grasslands

(Verheyen et al. 1999). However, impact of

land-use history on soil chemistry seems to be

more profound in European studies than it is

in the U.S. (Motzkin et al. 1996, Koerner et al.

1997, Flinn et al. 2005). This may be due to the

comparatively short history of post-European

settlement (150–400 years) in the U.S. relative

to Europe (5000+ years; Foster 1992). None-

theless, there is a relative scarcity of detailed

studies on the impacts of land-use history on

plant biogeography and soils, particularly in

the U.S.

Oak forests dominate much of the eastern

deciduous forest biome and have done so for

much of the Holocene epoch (Braun 1950,

Barnes 1991, Webb 1988). The mechanism

underlying the long-term importance of oak in

eastern forests has been the subject of consid-

erable amount of research (Abrams 2003,

McEwan et al. 2007). It has been hypothesized

that oak forests have developed and been

maintained by periodic, low intensity, under-

story burning (Lorimer 1985, Abrams 1992).

Periodic fire, coupled with logging activities,

has allowed the relatively fire resistant and

light demanding species, such as oak, hickory,

and pine to survive (Abrams 1996, 2003).

Moreover, later successional, fire sensitive tree

species readily invade and dominate most

upland oak forests in which fire has been

suppressed during the 20th century (Lorimer

1984, Abrams 1992, 1998). The magnitude of

anthropogenic disturbances in North Ameri-

can forests changed dramatically following

European settlement. These included extensive

logging and land clearing, catastrophic fire

followed by fire suppression, agriculture often

followed by agricultural abandonment, the

introduction of exotic insects and diseases, and

spread of exotic and native invasive plant

species (Foster 1992, Abrams 2003). All of

these have led to unprecedented and rapid

changes in forest composition and structure.

In many respects, the landscape has undergone

a nearly complete transformation in the years

following European settlement (Whitney 1994,

Foster et al. 1998). Thus, land-use history

apparently represents a highly pervasive, yet

understudied, force affecting forests of the

eastern U.S. (Russell 1997).

The Nature Conservancy’s Mashomack

Preserve on Shelter Island, New York contains

one of the premier oak-dominated coastal

plain forests in the northeastern U.S., but little

is known about its forest composition, ecolog-

ical history, or the impacts of land-use history

on vegetation, plant succession, and soils.

Indeed, very little detailed information of this

type exists for northeastern coastal plain oak

forests (Ehrenfeld 1982, Clark 1986, Motzkin

et al. 2002). Two forested areas exist on

Mashomack that primarily differ in their

land-use history (logging versus agriculture).

After agricultural abandonment starting in

1870, one forest area formed with a dense

understory of greenbriar (Smilax spp.), a

native invasive shrub. A second forest area

on Mashomack was repeatedly cut and

exposed to period fire, but was never plowed.

The contrasting land-use in edaphically

similar and juxtaposed forests creates a unique

opportunity to study the impacts of anthro-

pogenic disturbances on the plant geography

and succession. The objectives of this research

are to characterize the impacts of land-use

history on plant geography and soils and

includes: 1) tree species composition, size, and

age structure; 2) soil type, texture, and

chemistry, including the presence or absence

of an Ap (plowed) surface horizon and soil

charcoal from past fires; 3) the impact of

Smilax cover on forest composition; and 4) the

historical development and successional trends

in the study area forests.

Materials and Methods. STUDY AREA DE-

SCRIPTION. The 810 ha Mashomack Preserve is

located on the coastal plain of Shelter Island,

New York (41.106uN, 72.233uW) off the

eastern end of Long Island (Fig. 1). Two

major oak-dominated areas on Mashomack

are the interior forest (270 ha) and the

maritime forest (235 ha; Fig. 1). The interior

forest forms the central core of Mashomack

and was not used extensively, if at all, for

agriculture. This is evident from an 1855

survey showing the forested interior of Ma-

shomack surrounded by cleared fields used for

agriculture (Fig. 1). Nonetheless, the interior

forest was logged and burned several times

since European settlement starting in the early

1600s (Weaver 1990). In contrast, the maritime

forest developed on those cleared plots as a

result of agricultural abandonment after 1870.

The extent of the maritime forest is easily

recognizable today because it is covered in

dense thickets of greenbriar (Smilax spp.) not

present in the interior forest. The naming of

38 JOURNAL OF THE TORREY BOTANICAL SOCIETY [VOL. 135

FIG. 1. Distribution of the major forests types on the Mashomack Preserve, showing the location ofShelter Island on the east end of Long Island, New York, the present boundary between the interior andmaritime forest types, the location of the 53 permanent plots within the interior forest (designated by I) andmaritime forest (M), and the 1855 forested (shaded interior) and cleared plots (white) with boundaries basedon a geodetic survey of Mashomack (lower left panel; Shalowitz 1964).

2008] ABRAMS AND HAYES: IMPACTS OF LAND-USE HISTORY 39

the two forests types on Mashomack as

interior versus maritime is done only for the

convenience of this study; in the broader sense

both types are, in fact, maritime or coastal

plain forests.

The average annual high and low tempera-

ture on Shelter Island is 21uC and 7uC,

respectively. Average annual precipitation

totals 116.6 cm, which is evenly dispersed

throughout the year (monthly precipitation

averages range from 8.0 to 11.7 cm). Influen-

tial weather disturbances include high winds

and hurricanes, the most severe of which

occurred in 1893, 1938, and 1985 (Gornitz et

al., 2002). The average elevation on the island

is 15 m above sea level. All of the soils found

on Shelter Island are typical of terminal

moraines, including Montauk, Riverhead,

and Plymouth sandy loams, as well as Carver

and Plymouth sands (Warner et al. 1975). In

general these soils series are well-drained to

excessively well-drained, have low natural

fertility, and moderately coarse to coarse

texture. The Montauk sandy loam soil is the

most prevalent throughout the Mashomack

Preserve, while the Carver and Plymouth

sands are generally found on higher elevation

sites.

All field work was completed between

March and December 2006. Fifty-three ran-

domly located sampling points in the interior

forest (30 plots) and maritime forest (23 plots)

were used for vegetation and soil analysis

(Fig. 1). Fewer plots were used in the maritime

forest because it covered less area on the

preserve. A 400 m2 circular plot was used at

each sampling point to inventory all tree

species $ 2.5 cm dbh (diameter at breast

height) and saplings (tree species . 1.5 m in

height and dbh , 2.5 cm). Species, diameter,

and crown class were recorded for all trees.

Classification of tree crowns into four catego-

ries (dominant, codominant, intermediate, and

overtopped) was based on canopy position.

One hundred and thirteen tree cores were

taken at 0.5–1.0 m height across both forest

types to include the full range of species and

diameter size classes. This information was

used to determine age structure, temporal, and

spatial variation in tree recruitment, and forest

succession. All cores were brought back to

Penn State University, sanded with increas-

ingly fine sand paper, and analyzed under a

microscope for total age to pith (first year

growth) by counting tree rings. In cores where

the pith was missed (i.e., pith located outside

the exact center of the tree), simple geometry

and the early growth rate of the tree were used

to estimate total age. Using concentric plots at

the center of each sampling point, tall

seedlings $ 0.5 m and , 2.0 m in height were

counted by species in 50 m2 plots at each

sample point. Shrub cover was estimated into

cover classes (0–5%, 5–25%, 25–50%, 50–75%,

75–95%, and 95–100%) by species in the 50 m2

plots. For each forest type, tree data were used

to calculate a relative importance value from

the average of the relative frequency, relative

density, and relative dominance (basal area;

Cottam and Curtis, 1956). Species richness

(the average number of species per plot) and

the Shannon-Weaver Index of Diversity (2S

pi log pi, where pi 5 tree species basal area)

were calculated for the two forest types

(Whittaker 1975).

A soil sample was taken from the top of the

B horizon (at approximately 10 cm in depth in

the zone of illuviation) at each of 53 sample

points used for vegetation sampling. These

samples were air dried and then brought to the

soil analysis laboratory at Penn State Univer-

sity for chemical analysis of pH, hydrogen,

phosphorus, potassium, magnesium, calcium,

and cation exchange capacity (CEC). Cation

exchange capacity was determined by the

summation methods, pH was determined in

water, and P, K, Mg, and Ca were determined

using the Mehlich 3 technique (Wolf and

Beegle, 1995). In addition, at 27 of the sample

plots distributed throughout both forest types,

the upper soil profile was examined for the

presence/absence of a plow layer (Ap) and for

coarse soil charcoal fragments $ 2 mm. Soils

and vegetation data were tested for normality

and then analyzed using ANOVA, t-tests, and

Pearson product-moment correlation. Plant

nomenclature follows Gleason and Cronquist

1991.

Results. FOREST COMPOSITION AND STRUC-

TURE. The interior forest of the Mashomack

Preserve is dominated by black oak (Quercus

velutina Lam.), red maple (Acer rubrum L.),

pignut hickory (Carya glabra (Mill.) Sweet),

and chestnut oak (Quercus prinus L.; Table 1).

A total of 12 tree species were recorded in the

30 sample plots. The total density and basal

area is 547 trees per ha and 27 m2 per ha,

respectively. Black oak has a relative impor-

tance value (RIV) of 20%, occurring in 22 of


the 30 sample plots, and has the highest

dominance (or basal area) of any tree species.

Black oak’s density is second only to red

maple. Red maple has a RIV of nearly 20%, it

occurred in 23 sample plots and has a density

of 150 trees per ha. Pignut hickory has a

relatively high importance value, but a low

dominance is a result of its being primarily a

small diameter tree. In contrast, the impor-

tance of chestnut oak and white oak (Q. alba

L.), is due to having mostly large diameter

trees. Dogwood (Cornus florida L.) is the fifth

rank dominant due to having a large number

of small diameter trees. Beech (Fagus grand-

ifolia Ehrh.) is the seventh most important

species. The five remaining tree species (north-

ern red oak, Quercus rubra L.; sassafras;

Sassafras albidum (Nutt.) Nees; sweet birch;

Betula lenta L.; shagbark hickory, Carya ovata

(Mill.) K. Koch; and black cherry, Prunus

serotina Ehrh.) occur relatively infrequently.

The interior forest at Mashomack has attri-

butes of both the oak-beech and oak-hickory-

dogwood types previously described for Long

Island (Greller 1977, Greller et al. 1982).

A total of 12 tree species were also recorded

in the maritime forest that is dominated by

black oak, sassafras, pignut hickory, white

oak, and black cherry (Table 2). The total

density and basal area of all surveyed trees is

346 trees per ha and 21.5 m2 per ha, respec-

tively. These values are much lower than those

recorded in the coastal interior forest. Black

oak has a RIV of 32.7%, occurring in 20 of the

23 sample plots, and had the highest domi-

nance of any tree species. Sassafras has the

Table 1. Total and relative frequency (number of sample plots), density (total number of trees), anddominance (basal area) and relative importance value for all tree species recorded in the interior forest at theMashomack Preserve. Relative importance value is the sum of the relative frequency, density, anddominance divided by three for each tree species.

SpeciesFrequency(30 plots)

Density(stems/ha)

Dominance(m2 ha21)

Relativefrequency

Relativedensity

Relativedominance RIV

Q. velutina 22 94.17 7.40 15.49 17.04 27.45 20.00A. rubrum 23 150.83 4.43 16.20 27.30 16.42 19.97C. glabra 22 87.50 2.83 15.49 15.84 10.48 13.94Q. prinus 19 52.50 4.95 13.38 9.50 18.36 13.75C. florida 21 88.33 0.58 14.79 15.99 2.16 10.98Q. alba 17 21.67 3.58 11.97 3.92 13.27 9.72F. grandifolia 8 33.33 2.02 5.63 6.03 7.48 6.38Q. rubra 4 0.83 0.69 2.82 1.21 2.57 2.20S. albidum 2 9.17 0.15 1.41 1.66 0.55 1.20B. lenta 1 5.83 0.26 0.70 1.06 0.97 0.91C. ovalis 2 1.67 0.08 1.41 0.30 0.28 0.66P. serotina 1 0.83 0.003 0.70 0.15 0.01 0.20Total 142 546.67 26.96 100% 100% 100% 100%

Table 2. Total and relative frequency (number of sample plots), density (total number of trees), anddominance (basal area) and relative importance value for all tree species recorded in the 23 sample plots inthe maritime forest at the Mashomack Preserve. Relative importance value is the sum of the relativefrequency, density, and dominance divided by three for each tree species.

Frequency(23 plots)

Density(stems/ha)

Dominance(m2 ha21)

Relativefrequency

Relativedensity

Relativedominance RIV

Q. velutina 20.00 81.52 10.97 23.53 23.58 51.05 32.72S. albidum 15.00 85.87 2.67 17.65 24.84 12.44 18.31C. glabra 9.00 45.65 2.30 10.59 13.21 10.68 11.49Q. alba 7.00 36.96 3.23 8.24 10.69 15.03 11.32P. serotina 13.00 41.30 0.51 15.29 11.95 2.37 9.87C. florida 8.00 18.48 0.18 9.41 5.35 0.85 5.20A. rubrum 4.00 16.30 0.39 4.71 4.72 1.83 3.75Q. prinus 3.00 7.61 0.69 3.53 2.20 3.21 2.98Q. stellata 2.00 4.35 0.24 2.35 1.26 1.13 1.58A. laevis 2.00 5.43 0.06 2.35 1.57 0.26 1.40F. grandifolia 1.00 1.09 0.23 1.18 0.31 1.09 0.86C. ovata 1.00 1.09 0.02 1.18 0.31 0.07 0.52Total 85.00 345.65 21.50 100% 100% 100% 100%


second highest RIV from a high density of

small to intermediate size trees. Pignut hickory

has the third highest RIV of 11.5%, and is

present mainly as a relatively small diameter

tree. White oak ranks fourth over-all in RIV

although its dominance is the second highest

of any species as a result of having mostly

large diameter trees. Black cherry is the fifth

most important species, but has a very low

basal area of 0.5 m2 ha21. The seven remaining

tree species (dogwood, red maple, chestnut

oak, post oak (Quercus stellata Wangenh.),

serviceberry (Amelanchier laevis (Michx. f.

(Fern.)), beech and shagbark hickory) have

relative importance values ranging from 0.5 to

5.2%. The maritime black oak forest has

increased black oak and sassafras and de-

creased red maple, dogwood, chestnut oak,

and beech when compared with the interior

forest. It has elements of other maritime

forests described for Long Island, including

the presence of Smilax (Greller 1977).

The diversity of overstory trees in the

interior forest is significantly higher (P 5

0.006), with a species richness of 4.77 6 0.20

(versus 3.74 6 0.29 in the maritime forest) and

a Shannon-Weaver Index of 0.915 (versus

0.290 in the maritime forest).

In the interior forest, the three largest

diameter classes (50–80 cm) contain relatively

few trees and are dominated by beech, black

oak, white oak, and chestnut oak (Fig. 2). The

three intermediate diameter classes (20 to 49.9

cm) are dominated by black oak, chestnut

oak, hickory, and red maple. The two smallest

diameter classes (2.5 to 19.9 cm) are primarily

red maple, dogwood, and hickory; although,

black oak has 32 trees in the 10–19.9 cm class

and beech has 18 trees (most appeared to be

root suckers) in the 2.5 to 9.9 cm class.

Chestnut oak and white oak have very

low representation in the smallest diameter

classes.

In the maritime forest, one very large white

oak is represented in the 90–99.9 cm dbh class

(Fig. 3). The 40–69.9 cm classes are dominat-

ed by black oak, with lesser amounts of white

oak, sassafras, and pignut hickory. Black oak

is also well represented in the 30–39.9 cm dbh

class, but shares dominance with sassafras as

well as pignut hickory, chestnut oak, and

white oak. The 10–29.9 cm dbh classes are

dominated by sassafras, pignut hickory, white

oak, and red maple, with lesser amounts of

black oak. The 2.5–9.9 cm dbh class is dom-

inated by sassafras and black cherry, followed

FIG. 2. Diameter class distribution for all surveyed trees in the 30 samples plots in the interior forest onthe Mashomack Preserve.


by pignut hickory, dogwood, and very few

black oak and white oak.

The dominant canopy class in the interior

forest is mainly comprised of black oak,

chestnut oak, hickory, white oak, and red

maple, followed by beech (Fig. 4). The co-

dominant class is primarily red maple and

black oak, with moderate amounts of chestnut

oak and pignut hickory. Small numbers of

white oak and beech also occur in the co-

dominant class. Between the dominant and co-

dominant classes there is a large increase in red

maple, and moderate declines in black oak and

white oak. The intermediate canopy class is

dominated by red maple, with moderate

amounts of pignut hickory, black oak, dog-

wood, and beech. The over-topped canopy

class is primarily dogwood, followed by pignut

hickory, red maple, and beech. White oak and

chestnut oak are almost nonexistent in the

overtopped and intermediate canopy classes.

In the maritime forest, the dominant canopy

class has a large number of black oak, a

moderate amount of white oak, and small

numbers of chestnut oak, sassafras, and pignut

hickory (Fig. 5). The co-dominant class is

dominated by sassafras, black oak, and white

oak, with moderate amounts of pignut hick-

ory, chestnut oak, and red maple. Sassafras

also dominates the intermediate canopy class,

followed by red maple, pignut hickory, white

oak, black oak, dogwood, and black cherry.

Sassafras and black cherry dominate the over-

topped class, followed by dogwood, pignut

hickory, and serviceberry. A steady decline in

the number of black oak and white oak is

evident in the progressively lower canopy

classes that are dominated by sassafras, black

cherry, dogwood, and red maple.

The oldest trees cored in the interior forest

are a 161 year old beech, a 144 year old white

oak and a 134 year old black oak (Fig. 6). The

age structure of the remaining trees reveals

several interesting developmental patterns.

Chestnut oak dominated recruitment (n 5

14) between 1875 and 1915; in fact all the

chestnut oak sampled in the forest recruited

during this time period. Other species recruit-

ing then were white oak, black oak and beech.

Between 1915 and 1940, tree recruitment was

dominated by red maple and beech. After

1940, red maple and black oak dominated

recruitment, followed by hickory and beech.

Beech age ranged from 26 and 161 years, and

it recruited more-or-less continuously within

that timeframe.

FIG. 3. Diameter class distribution for all surveyed trees in the 23 samples plots in the maritime forest onthe Mashomack Preserve.


FIG. 4. Canopy class distribution for all surveyed trees in the 30 samples plots in interior forest on theMashomack Preserve.

FIG. 5. Canopy class distribution for all surveyed trees in the 23 samples plots in the maritime forests onthe Mashomack Preserve.


In the maritime forest, black oak dominated

recruitment (n 5 22) between 1880 and 1945,

followed by chestnut oak, pignut hickory, and

white oak (Fig. 7). Between 1945 and 1975,

only one black oak tree in our sample

recruited in the forest, but an additional four

black oak recruited from 1975 to 1987.

Chestnut oak and white oak recruited between

1880–1945 and 1915–1924, respectively. Pig-

nut hickory did not obtain the maximum ages

seen in black oak and chestnut oak, but

exhibited fairly continuous recruitment from

1895 until 1980. During 1945–1987, tree

recruitment included sassafras, pignut hickory,

black cherry, and red maple, in addition to the

black oak mentioned above. Thus, recent

recruitment includes more sassafras and black

cherry, and less black oak, chestnut oak, and

white oak when compared with the early

history of the forest.

Tree regeneration numbers are very low in

both forest types. In the interior forest plots,

only eleven beech, eight dogwood, and one red

maple saplings per ha and no tall seedlings

were recorded. In the maritime forest, only 49

saplings per ha were recorded, mainly sassa-

fras and black cherry with no oaks and very

few hickories. Seventeen tall seedlings per ha

were recorded, mostly black cherry and no

oaks. In the interior forest, shrub cover

averages 41%, with huckleberry (Gaylussacia

spp.) at 26% cover and greenbriar (Smilax

rotundifolia), and blueberry (Vaccinium spp.)

each at 7% cover (Fig. 8). In the maritime

forest, shrub cover averages 80%, dominated

by greenbriar at 67%; blueberry is the second

with 8% cover, followed by bayberry (Myrica

spp.) at 2% cover. A dense thicket of green-

briar approximately a meter or more in height

exists throughout most of the maritime forest.

Because the maximum cover of a single species

within a plot is 95–100%, our sampling

protocol does not fully capture the over-

whelming density of overlapping stems for

greenbriar throughout the maritime forest.

IMPACTS OF SOIL CHEMISTRY, TYPE AND

TEXTURE. Soil pH averages 4.6 and 4.4 in the

interior forest and maritime forest, respective-

ly (Table 3). Hydrogen content (a related and

inverse measure of pH) is higher in the

maritime versus interior forest, with the

FIG. 6. Age-diameter distribution for all cored trees in the interior forest on the Mashomack Preserve.


exception of the Plymouth sands in the latter.

Potassium ranged from 30.0 to 38.7 ppm

across all soil types and is not significantly

different between forest types. Magnesium,

Ca, Na, and CEC is typically higher in the

maritime versus interior plots, but similar

across the various soil types within each forest

type.

Overall, both forest types are edaphically

similar. However, the interior forest contain a

greater proportion of the more xeric Plymouth

loamy sands and Carver sands than do the

maritime forests (Table 3), while the maritime

forests contain more of the mesic Montauk

loamy sands. Both forest types contain a

similar acreage of Montauk sandy loam and

Riverhead sandy loam. The maritime forests

have a lower elevation (3–13 m above sea

level) compared to the interior forests at 13–

20 m asl. Overall, plot-by-plot variation in soil

chemistry and soil type/texture had no statis-

tically significant impact on tree distribution

within or between the two forest types.

In 16 sample plots in the interior forest,

eight contained soil charcoal but all of the

tested plots were unplowed (lack an Ap

horizon; Fig. 9). In the maritime forest, only

three of eleven plots had charcoal fragments,

but all eleven plots had an Ap horizon. This

indicates that the maritime forest was primar-

ily used for agriculture or grazing and

experienced less burning. The interior forest

existed prior to 1855 (Fig. 1), was not used

extensively for agriculture, was burned peri-

odically, at least in part, and probably

remained forested since the time of European

settlement, albeit with several cutting cycles.

Within the maritime forest, soils containing a

plowed (Ap) horizon have significantly (P ,

0.05) higher concentrations of P, K, Ca, and

CEC, Mg and Na (Table 4). The importance

of white oak is 10% on plowed plots and 0%

on unplowed plots. In contrast, black oak has

significantly higher importance on unplowed

(40% importance) versus plowed (31%) plots.

Discussion. Agricultural clearing on Shelter

Island began shortly after European settle-

ment in 1640 and eventually lead to tenant

farming and livestock rearing (Weaver 1990).

The English military continued to clear the

Mashomack forest during the Revolutionary

FIG. 7. Age-diameter distribution for all cored trees in the maritime forest on the Mashomack Preserve.


War (1776–1786). During this time, at least

3500 cords of wood (mostly white oak and

chestnut oak) were extracted for the produc-

tion of barrels. Large quantities of lumber

were also cut from 1784–1827, and by the

middle of the 19th century 2000 ha of Shelter

Island were cleared by various owners and

tenant farmers for agriculture alone. Heavy

commercial logging occurred in the 1870s–

1890s and produced slash that likely fueled a

catastrophic fire of 1902 (Weaver 1990).

These land-use history events greatly influ-

enced the present-day forests on the Masho-

mack Preserve. The ecology of oak forests in

the northeast region is thought to be closely

tied to periodic fire, past logging, and land

clearing (Lorimer 1985, Abrams 1992). In-

deed, oak species dominated tree recruitment

from approximately 1875 to 1915 in the

interior forest following widespread logging

and fire. All the chestnut oak and most of the

white oak trees in the study area date to this

FIG. 8. Dense greenbriar (Smilax) cover in amaritime black oak forest understory at Masho-mack (top panel). More open and patchy understorydominated by blueberry (Vaccinium) in the coastalinterior forest. In both examples, note the lack oftree seedlings.

Tab

le3.

Mea

nan

dst

an

dard

erro

rvalu

esfo

rso

ilp

H,

cati

on

exch

an

ge

cap

aci

ty(C

EC

),an

dca

tio

ns

for

the

majo

rso

ilty

pes

inth

ein

teri

or

an

dm

ari

tim

eb

lack

oak

fore

sts

on

the

Mash

om

ack

Pre

serv

e.N

um

ber

sw

ith

ina

row

shari

ng

the

sam

ele

tter

are

no

tsi

gn

ific

an

tly

dif

fere

nt

at

P,

0.0

5.

Inte

rio

rfo

rest

Mari

tim

efo

rest

Mo

nta

uk

fin

esa

nd

ylo

am

Riv

erh

ead

san

dy

loa

mM

on

tau

klo

am

ysa

nd

Ply

mo

uth

san

ds

Carv

era

nd

Ply

mo

uth

san

ds

Mo

nta

uk

fin

esa

nd

ylo

am

Mo

nta

uk

loa

my

san

dR

iver

hea

dP

lym

ou

thC

arv

ersa

nd

s

pH

4.5

96

0.0

5a

4.8

60.1

1a

4.5

26

0.0

9a

4.3

86

0.0

6a

4.6

26

0.0

9a

4.3

66

0.1

1a

4.4

56

0.0

8a

4.4

06

.016a

P(p

pm

)7.4

46

0.8

7a

7.8

60.6

3a

8.2

61.0

2ab

10.3

36

1.4

6b

6.2

61.0

7a

10.8

16

0.9

5b

10.5

06

1.0

6b

16.6

76

2.0

3c

K(p

pm

)30.0

06

3.9

1a

34.2

63.8

0a

31.2

63.4

0a

38.0

06

6.4

8a

33.4

63.6

8a

37.5

06

3.9

4a

38.7

06

5.6

3a

33.0

06

7.9

9a

Mg

(pp

m)

14.0

06

1.6

3a

17.8

64.7

6a

12.6

60.8

1a

17.1

76

2.3

6a

12.6

61.4

4a

32.1

96

3.0

8b

41.4

06

3.6

3b

43.0

06

4.6

3b

Ca

(pp

m)

18.0

16

3.6

1a

39.9

26

8.0

6b

18.4

26

5.2

4a

17.9

76

2.3

6a

16.8

26

3.8

7a

52.4

96

6.6

4b

58.2

16

5.2

5b

61.1

36

4.3

4b

Na

(pp

m)

10.1

86

0.4

6a

10.2

66

0.7

2a

10.0

26

1.0

9a

10.6

86

0.2

9a

8.9

60.9

0a

16.9

66

1.1

5b

27.8

86

4.0

3c

12.3

36

2.1

9a

H(m

eq/1

00g)

6.5

06

0.3

0a

6.5

46

0.3

1a

6.4

26

0.1

2a

8.5

06

0.7

0ab

6.7

86

0.5

8a

10.3

56

0.9

7b

9.4

26

0.7

5ab

11.3

06

1.4

0b

CE

C(m

eq/1

00g)

6.7

60.3

0a

6.9

66

0.4

0a

6.7

60.1

3a

8.8

36

0.7

4ab

7.0

86

0.5

8a

10.6

96

0.7

9b

10.1

66

0.8

7b

10.5

76

0.9

1b

Are

a(h

a)

85

25

35

64

70

87

63

74


time period. Many black oak trees recruited

between 1885 and 1890, but after a long hiatus

this species started recruiting again between

1950 and 1985. The two-phase recruitment

pattern of black oak at Mashomack is

intriguing, and suggests that black oak can

establish following large scale disturbances

(clearcutting and fire) as well as under canopy

gaps created in the mature forest (Orwig and

Abrams 1995, Abrams 1996). Indeed, high

wind is a predominant ecological factor at

Mashomack, frequently blowing down trees

and creating canopy openings (pers. obs.).

The ecology of most hickory species is

similar to that of oaks in many respects; they

are long-lived (over 300 years), have inter-

mediate shade tolerance, are early to mid-

successional, and are thought to require

periodic understory fire for long-term recruit-

ment (Burns and Honkala 1990, Orwig and

Abrams 1994). Most of the pignut hickory in

both forest types recruited during the middle

and latter stages of stand development. Recent

hickory establishment may be due to their

higher than normal shade tolerance as young

trees and/or their ability to recruit in canopy

gaps. Alternatively, the absence of older

hickory trees in the forest may be related to

their short lifespan in the mature forest canopy

at Mashomack.

Most beech recruitment in the interior forest

occurred between 1885 and 1940. Several

beech trees recruited along with the many

oaks during the earliest stage of forest

development, which indicates the ability of

beech to compete with the earlier successional

oaks. Beech is considered a late successional,

shade tolerant species and can replace oak in

the absence of fire and other disturbances

(Quarterman and Keever 1962, Monette and

Ware 1983, Abrams and Downs 1990). How-

ever, the ability of beech to replace oak may

be limited to moister sites (Seischab 1985).

Despite the drought nature of the sandy soils

at Mashomack, the low elevation and high

water table appears to be conducive to beech

success. Only three of the beech trees we aged

are less than 60 years old, and we believe that

this is due to intense white-tailed deer brows-

ing (Horsley et al. 2003). Indeed, beech

saplings dominate a deer exclosure established

in 1982, suggesting that more young beech

FIG. 9. Soil profile with Ap (plowed) horizonpresent with a homogeneous mixing of soil horizonsin the top 12–15 cm (top panel) and a soil profilewithout Ap present with a distinct upper A horizon(dark brown), middle E horizon (grey), and lower Bhorizon (light brown) (bottom panel).

Table 4. Mean and standard error values for soil pH, cation exchange capacity (CEC), and cations forplowed (Ap present) and unplowed soils within the Montauk soils group in the maritime forests on theMashomack Preserve. Numbers within a row sharing the same letter are not significantly different at P, 0.05.

Plowed Unplowed

pH 4.79 6 0.17a 4.56 6 0.07aP (ppm) 14.50 6 2.59a 5.57 6 0.72bK (ppm) 48.38 6 6.06a 25.50 6 3.65bMg (ppm) 45.00 6 9.19a 12.83 6 1.91bCa (ppm) 131.53 6 16.69a 15.70 6 3.80bCEC (meq/100g) 9.26 6 0.34a 7.30 6 0.25bNa (ppm) 16.27 6 0.87a 10.53 6 0.64b


trees would be present if deer browsing was

reduced (Abrams and Hayes, unpublished

data).

Red maple in the interior forest dates from

1905 to 1980, with peak recruitment from 1925

to 1960. None of the existing red maple date

back to before 1900, despite the fact that this

species can live up to 200–300 years. A large

increase in red maple trees has occurred

throughout much of the eastern forest after

1900, which has been primarily attributed to

the suppression of forest fire and widespread

logging coupled with being avoided by early

loggers (Abrams 1998, 2003). It now domi-

nates many upland oak forests and is replacing

oak based on its higher shade tolerance,

among other physiological attributes (Abrams

1998). Thus, the post-1900 increase in red

maple at Mashomack is highly consistent with

this pattern seen throughout the eastern oak

forests. Red maple foliage is not a preferred

browse species for white-tailed deer, although

they use twigs as a winter food source

(Abrams 1998, Horsley et al. 2003). In

contrast, oak species are highly preferred by

deer in all seasons and continual summer

browsing most likely results in oak seedling

and sapling mortality and limits canopy tree

recruitment (Canham et al. 1994, Abrams

1998).

Despite the differences in land-use history,

the age structure of the maritime forest closely

matches that of the interior forest, including

the maximum ages of tree in plowed versus

unplowed soils. Thus, it appears that the vast

majority of Mashomack oak forests were cut

(in the interior) or developed following agri-

cultural abandonment (maritime forest) in the

1870s time period. In the maritime forest,

black oak, pignut hickory, and chestnut oak

dominated tree recruitment from approxi-

mately 1875 to 1940. Red maple and beech

occur only at low levels, and may be limited by

dense greenbriar cover and oceanic influences

of salt spray, high winds, and sand blasting in

maritime forests. Thus, there is no obvious

developmental pathway leading from oak-

hickory to a later successional beech-maple

type in the maritime forest as seen in the

interior forest. Indeed, black oak and pignut

hickory have found opportunities to recruit

in the maritime forest in recent decades de-

spite pressure from deer browsing and high

shrub cover and may continue to do so in the

future.

The number of young sassafras, black

cherry, and serviceberry in the maritime black

oak forest is remarkable in the face of deer

browsing and dense greenbriar cover. Sassa-

fras and black cherry are even less shade

tolerant and shorter lived than black oak and

pignut hickory and are more likely to die off as

the forest matures and the canopy becomes

more closed (Burns and Honkala 1990). They

are moderately preferred browse species for

white-tailed deer, whereas the shade tolerant

serviceberry is highly preferred (Elias 1980,

Johnson et al. 1995). It is possible that the high

density of greenbriar limits the movement of

deer in the maritime forest to a system of trails

they created (pers. obs.) and therefore pro-

vides protection from browsing of tree regen-

eration away from the trails.

The most prevalent soil types on Masho-

mack are the Montauk sandy loam and

Montauk loamy sand (Warner et al. 1975).

These soils tend to be acidic, droughty and

nutrient poor. Indeed, our soil chemical

analysis of the upper B horizon from the

forest plots supports this description. Relative

to most other soils in the eastern U.S., soil

cations, and pH at Mashomack are low

(Pritchett 1979, David and Lawrence 1996,

Sims and Gartley 1996). However, when

compared to other northern coastal plain

soils, the Mashomack values are quite typical.

For example, pH values ranged from 4.3 to 5.3

for upland oak and pine forests on the coastal

plain of Cape Cod, Massachusetts (Motzkin et

al. 2002). In that study, Ca ranged from 29 to

80 ppm, Mg ranged from 10 to 28 ppm, K

ranged from 8 to 22 ppm, and CEC ranged

from 1 to 13 meq/100 g, similar to values we

report for Mashomack. Comparable values

are also reported for the Pine Barrens soils in

New Jersey and for coastal plain soils in

Delaware (Foreman 1979, Sims and Gartley

1996). Despite low soil fertility at Mashomack,

the trees of most species are growing quite

well, most in the range of 7–10 mm yr21 in

diameter increment (Abrams and Hayes,

unpublished data).

The maritime soil types appear to be

somewhat superior and better suited for

agriculture compared with the coastal interior

forest soils, which contain more sands. Mar-

itime soils containing a plowed horizon are

more fertile and retain more nutrients (higher

CEC) than unplowed soils. This suggests the

possibility that agricultural activities along


with livestock grazing further improved the

nutrient content of the maritime soils. Plowed

soils at Mashomack support more of the mesic

white oak and less of the drought-tolerant

black oak than unplowed soils. In central

Massachusetts, plowed soils had similar chem-

istry to unplowed soils, but they supported

forest with higher pitch pine (Pinus rigida

Mill.), white oak, Rubus spp., and birch

(Motzkin et al. 1996). Forests on abandoned

agricultural lands in eastern New York are

dominated by white oak, black oak, and

pignut hickory, and also had a similar nutrient

content to unplowed soils (Glitzenstein et al.

1990). Agricultural soil chemistry did not

differ significantly from primary and second-

ary forests soils in central New York (Flinn et

al. 2005). Thus, the plowed soils at Masho-

mack seem to be somewhat unique among

U.S. studies in that they are more fertile than

unplowed soils. These results are similar to

several European studies in which previous

agricultural lands had higher pH, P, and/or Ca

(Koerner et al. 1997; Verheyen et al. 1999).

Most of the post-agricultural maritime

forest understory is covered in a dense thicket

of greenbriar. High shrub cover is a likely

important contributing factor to the lack of

forest regeneration there (O’Brien et al. 2006).

This in turn probably explains the lower tree

density and species diversity in the maritime

forest, mostly from a lack of beech, red maple,

and dogwood, as compared with the interior

forest. It is interesting to note that the

combination of high deer density and high

cover of a preferred browse species (green-

briar) is unusual for the eastern U.S (Healy

1997). Intensive white-tailed deer browsing

typically leads to either a very sparse forest

understory (like the interior forest) or the

domination by shrub species typically avoided

by deer (e.g., Kalmia and Rhododendron;

Moser et al. 1996). High cover of greenbriar

at Mashomack suggests several different but

not mutually exclusive scenarios involving

deer, including the early history of greenbriar

establishment, limited deer access in dense

thickets and predator satiation.

Smilax represents a super-dominant species

that has the potential to influence a vast array

of ecological processes. High levels of green-

briar also exist in other areas of Long Island,

such as in black oak-vine forests (Greller

1977). Oak and pine coastal plain forests on

Cape Cod Massachusetts include a vegetation

type called oak-briar (Motzkin et al. 2002).

The frequency of greenbriar there is highest in

open sites (63%), but is also high in plowed

(41%), disturbed (23%), and woodlot (36%)

communities. Thus, it appears that the factors

relating to high cover of greenbriar in other

anthropogenic-disturbed maritime environ-

ments of Long Island and Cape Cod are

similar to that on Mashomack. We believe

that most of the differences in forest compo-

sition between the two forest types at Masho-

mack can be attributed to contrasting land-use

history. Indeed, our analysis of soil type and

chemistry on tree species distribution within

and between the two forest types yielded no

significant differences except on plowed versus

unplowed soils within the maritime forest. It is

important to note, however, that differences in

microclimate between the maritime and inte-

rior forest not measured here may explain

some of the compositional variation.

Varying degrees of deer browsing now and

in the future may curtail the recruitment of

most tree species in both forest types, but

particularly in the interior forest where access

is not limited by the dense greenbriar.

Alternatively, greenbriar imposes its own

competitive pressure on tree regeneration in

the maritime forest. We believe that many of

the differences in the overstory and understory

composition, successional pathways, and soil

morphology (Ap horizon) and chemistry

between the two forest types can be attributed

to their contrasting land-use history. Green-

briar thickets formed after agricultural aban-

donment appear to have the ability to

profoundly influence many past, present and

future ecological processes in the maritime

forest. The results of this study contribute to

the rather sparse ecological literature describ-

ing land-use history impacts on plant geogra-

phy, forest dynamics and soils, and in this case

the creation of a super-dominant understory

species.

Literature Cited

ABRAMS, M. D. 1992. Fire and the development ofoak forests. BioScience 42: 346–353.

ABRAMS, M. D. 1996. Distribution, historicaldevelopment and ecophysiological attributes ofoak species in the eastern United States. Ann.Sci. For. 53: 487–512.

ABRAMS, M. D. 1998. The red maple paradox.BioScience 48: 355–364.

ABRAMS, M. D. 2003. Where has all the white oakgone? BioScience 53: 927–939.


ABRAMS, M. D. AND J. A. DOWNS. 1990. Successionalreplacement of old-growth white oak by mixed-mesophytic hardwood species in southwesternPennsylvania. Can. J. For. Res. 20: 1864–1870.

BARNES, B. V. 1991. Deciduous forests of NorthAmerica, p. 219–344. In E. Rohrig and B. Ulrich[eds.], Ecosystems of the World: 7 TemperateDeciduous Forests Elsevier, Amsterdam, theNetherlands.

BRAUN, E. L. 1950. Deciduous forests of easternNorth America. MacMillan, New York, NY.596 p.

BURNS, R. M. AND B. H. HONKALA. 1990. Silvics ofNorth America. Vol. 1. Conifers and Vol. 2:Hardwoods. U.S.D.A. Agricultural Handbook654, Washington, D.C.

CANHAM, C. D., J. B. MCANINCH, AND D. M. WOOD.1994. Effects of the frequency, timing andintensity of simulated browsing on growth andmortality of tree seedlings. Can. J. For. Res. 24:817–825.

CLARK, J. S. 1986. Coastal forest tree populations ina changing environment, southeastern LongIsland, New York. Ecol. Monogr. 56: 259–277.

CONNELL, M. J., R. J. RAISON, AND P. K. KHANNA.1995. Nitrogen mineralization in relation to sitehistory and soil properties for a range ofAustralian forest soils. Biol. Fert. Soils 20:213–220.

COTTAM, G. AND J. T. CURTIS. 1956. The use ofdistance in phytosociological sampling. Ecology37: 451–460.

DAHLSTROM, A., S. A. O. COUSINS, AND O. ERIKSSON.2006. The history (1620–2003) of land-use,people and livestock, and the relationship topresent plant species diversity in a rural land-scape in Sweden. Environ. Hist. 12: 191–212.

DAVID, M. B. AND G. B. LAWRENCE. 1996. Soil andsoil solution chemistry under red spruce standsacross the northeastern United States. Soil Sci.161: 314–328.

EHRENFELD, J. G. 1982. The history of the vegetationand the land of Morristown National HistoricalPark, New Jersey, since 1700. Bull. N. J. Acad.Sci. 27: 1–19.

ELIAS, T. S. 1980. Trees of North America. VanNostrand Reinhold Press, New York, NY. 948 p.

ELLENBERG, H. 1995. Vegetation Mitteleuropas Mitden Alpen, 5 Auflage. Ulmer, Stuttgart, Ger-many. 945 p.

EYRE, S. R. 1963. Vegetation and soils, a worldpicture. Edward Arnold, London, and AldinePubl. Co., Chicago, IL. 324 p.

FLINN, K. M., M. VELLEND, AND P. L. MARKS. 2005.Environmental causes and consequences of forestclearance and agricultural abandonment in cen-tral New York, USA. J. Biogeog. 32: 439–452.

FOREMAN, R. T. T. 1979. Pine barrens: ecosystemand landscape. Academic Press, New York, NY.585 p.

FOSTER, D. R. 1992. Land-use history (1730–1990)and vegetation dynamics in central New Eng-land, USA. J. Ecol. 80: 753–772.

FOSTER, D. R., G. MOTZKIN, AND B. SLATER. 1998.Land-use history as long-term broad scaledisturbance: regional forest dynamics in centralNew England. Ecosystems 1: 96–119.

GLATZEL, G. 1991. The impact of historic land useand modern forestry on nutrient relations ofcentral European forest ecosystems. Fert. Res.27: 1–8.

GLEASON, H. A. AND A. CRONQUIST. 1991. Manual ofvascular plants of northeastern United States andadjacent Canada, 2nd Edition. New York Botan-ical Garden, Bronx, NY.

GLITZENSTEIN, J. S., C. D. CANHAM, M. J. MCDON-

NELL, AND D. R. STRENG. 1990. Effects ofenvironment and land-use history on uplandforests of the Cary Arboretum, Hudson Valley,New York. Bull. Torrey Bot. Club 117: 106–122.

GORNITZ, V., S. COUCH, AND E. K. HARTIG. 2002.Impacts of sea level rise in the New York Citymetropolitan area. Global Planet. Changes 32:61–88.

GRELLER, A. M. 1977. A classification of matureforests on Long Island, New York. Bull. TorreyBot. Club 104: 376–382.

GRELLER, A. M., J. M. MANSKY, AND R. E.CALHOON. 1982. An oak, hickory-dogwood foreston central Long Island, New York. Bull. TorreyBot. Club 109: 219–225.

HEALY, W. M. 1997. Influence of deer on thestructure and composition of oak forests incentral Massachusetts, p. 249–266. In W. J.McShea, H. B. Underwood, and J. H. Rappole[eds.], The science of overabundance: deerecology and population management. Smithso-nian Institution Press, Washington, D.C.

HORSLEY, S. B., S. L. STOUT, AND D. S. DECALESTA.2003. White-tailed deer impact on the vegetationdynamics of a northern hardwood forest. Ecol.Appl. 13: 98–118.

HOLDRIDGE, L. R. 1967. Life zone ecology. TropicalScience Center, San Jose, Costa Rica. 206 p.

JOHNSON, A. S., P. E. HALE, W. M. FORD, J. M.WENTWORTH, J. R. FRENCH, O. F. ANDERSON,AND G. B. PULLEN. 1995. White-tailed deerforaging in relation to successional stage, over-story type and management of Southern Appa-lachian Forests. Am. Mid. Nat. 133: 18–35.

KOERNER, W., J. L. DUPOUEY, E. DAMBRINE, AND M.BENOIT. 1997. Influence of past land use on thevegetation and soils of present day forest in theVosges Mountains, France. J. Ecol. 85: 351–358.

LORIMER, C. G. 1984. Development of the red mapleunderstory in northeastern oak forests. For. Sci.30: 3–22.

LORIMER, C. G. 1985. The role of fire in theperpetuation of oak forests. In J. E. Johnson[ed.], Challenges in oak management and utili-zation. Cooperative Extension Service, Universi-ty of Wisconsin, Madison, WI.

MCEWAN, R. W., R. LONG, T. F. HUTCHINSON, R. P.LONG, R. D. FORD, AND B. C. MCCARTHY. 2007.Temporal and spatial patterns of fire occurrenceduring the establishment of mixed-oak forests ineastern North America. J. Veg. Sci. 18: 655–664.

MONNETTE, R. AND S. WARE. 1983. Early forestsuccession in the Virginia Coastal Plain. Bull.Torrey Bot. Club 110: 80–86.

MOSER, W. K., M. J. DUCEY, AND P. M. S. ASHTON.1996. Effects of fire intensity on competitivedynamics between red and black oaks and


mountain laurel. North. J. Appl. For. 13:119–123.

MOTZKIN, G., D. R. FOSTER, A. ALLEN, J. HARROD,AND R. D. BOONE. 1996. Controlling site toevaluate history: vegetation patterns of a NewEngland sand plain. Ecol. Monogr. 66: 345–365.

MOTZKIN, G., R. EBERHARDT, B. HALL, D. R.FOSTER, J. HARROD, AND D. MACDONALD. 2002.Vegetation variation across Cape Cod, Massa-chusetts: environmental and historical determi-nants. J. Biogeog. 29: 1439–1454.

O’BRIEN, M. J., K. L. O’HARA, N. ERBILGIN, AND D.L. WOOD. 2007. Overstory and shrub effects onnatural regeneration processes in native Pinusradiata stands. For. Ecol. Manage. 240: 178–185.

ORWIG, D. A. AND M. D. ABRAMS. 1994. Land-usehistory (1720–1992), composition, and dynamicsof oak-pine forests within the Piedmont andCoastal Plain of northern Virginia. Can. J. For.Res. 24: 2141–2149.

PRITCHETT, W. L. 1979. Properties and managementof forest soils. John Wiley and Sons, New York,NY. 461 p.

QUATERMAN, E. AND C. KEEVER. 1962. Southernmixed hardwood forest: climax in the southeast-ern coastal plain, U.S.A. Ecol. Monogr. 32:167–185.

RUSSELL, E. W. B. 1997. People and the landthrough time: linking ecology and history. YaleUniversity Press, New Haven, CT. 306 p.

SHALOWITZ, A. L. 1964. Shore and sea boundaries:with special reference to the interpretation anduse of coast and geodetic survey data. Publica-tion Number 10-1. U.S. Dept. of Commerce,Coast and Geodetic Survey National Oceanicand Atmospheric Administration, WashingtonD.C.

SEISCHAB, F. 1985. An analysis of the forests of theBristol Hills of New York. Am. Midl. Nat. 114:77–83.

SIMS, J. T. AND K. L. GARTLEY. 1996. Nutrientmanagement handbook for Delaware. Coop.Bull. No. 59. University of Delaware Agricultur-al Experiment Station, Newark, DE. 77 p.

VERHEYEN, K., B. BOSSUYT, M. HERMY, AND G.TACK. 1999. The land use history (1278–1990) ofa mixed hardwood forest in western Belgium andits relationship with chemical soil characteristics.J. Biogeog. 26: 1115–1128.

WARNER, J. W., JR., W. E. HANNA, R. J. LANDRY, J.P. WULFROST, J. A. NEELEY, R. L. HOLMES, AND

C. E. RICE. 1975. Soil survey of Suffolk County,New York. U.S.D.A. Soil Conservation Service,New York. 101 p. plus maps.

WEAVER, M. P. 1990. Where they go by water. TheNature Conservancy, U.S.A. 128 p.

WEBB, T., III. 1998. Glacial and Holocene vegeta-tion history: eastern North America, p. 385–414.In B. Huntley and T. Webb, III [eds.], VegetationHistory. Kluwer Academic Publications, Am-sterdam, the Netherlands.

WHITNEY, G. G. 1994. From coastal wilderness tofruited plain. Cambridge University Press, Cam-bridge, U.K. 451 p.

WHITTAKER, R. H. 1975. Communities and ecosys-tems (2nd edition). Macmillan, New York, NY.385 p.

WOLF, A. M. AND D. B. BEEGLE. 1995. Recom-mended soil tests for macronutrients: phospho-rus, potassium, calcium, and magnesium,p. 25–34. In J. Thomas Sims and A. Wolf[eds.], Recommended soil testing procedures forthe Northeastern United States. Northeast Re-gional Bulletin #493. Agricultural ExperimentStation, University of Delaware, Newark, DE.


impacts of contrasting land-use history on composition ... · impacts of contrasting land-use...

Documents