280
iments following fire. Close agreement of a 450year dendrochronologic fire history with charcoal
peaks was used as justification to extend the charcoal chronology to 750 years. The results indicated
that the fire regime was characterized by punctuated episodes of severe fires.
2. The same 750-year period was analyzed using varved lake sediments as the source of annual
charcoal records for small lakes in northwestern
Minnesota (Clark, 1990). Charcoal peaks were
compared to dated fire-scar records. Fire was found
to be most frequent in the relatively warm and dry
15th and 16th centuries, and longer fire intervals
were observed after A.D. 1600 during the Little Ice
Age.
3. In Indiana, charcoal analysis over the past
4000 years showed an increase in abundance of
charcoal as vegetation shifted from pine forest to
oak savanna and prairie, as well as extremely high
amounts of charcoal during the period 1850 to 1900
associated with intense fires that accompanied European settlement (Cole and Taylor, 1995).
4. Larsen and MacDonald (1998) combined
pollen and charcoal records from varved lake sediments to determine boreal forest fire and vegetation history over an 840-year period in Wood Buffalo National Park, Alberta. A general sequence of
herb, shrub, deciduous tree, and conifer pollen was
observed following fires, but peak values for taxa
differed among fires, suggesting that a site may not
consistently undergo the same vegetation sequence
following fires. This may be due to climate change,
fires with varying characteristics, or rearrangement
of patches of forest dominants around the lake
(Larsen and MacDonald, 1998).
19.6 Land Survey Records
Land survey records can provide information about
the biotic and environmental conditions that existed
at the time the survey was conducted. Such information may include vegetation composition, occurrence of disturbance, and land-use patterns. Survey records generally represent points in time and
a set of sample points in space whose spatial extent is determined by the extent of the survey. In
North America, several sources of information describe the vegetation of an area prior to European
settlement, including colonial surveys, warrant
maps, and public land surveys. Key components of
each of these sources are witness trees, or trees that
mark boundary comers and points along survey
lines (Loeb, 1987). Surveys of unsettled lands in
the United States often included the identification
Methods for Determining Historical Range of Variability
of witness trees in areas where trees were present.
Old survey records are available for most of the
continental United States, and when surveys were
conducted prior to settlement, witness tree records
provide a source of information about presettlement
tree species distributions and associations. However, care must be taken in the use of witness tree
records due to possible surveyor bias in the selection of the species and sizes of witness trees (discussed further in Section 19.6.3). In addition, unreliable survey records have resulted in some cases
from careless work or outright fraud (Bourdo,
1956).
19.6.1 Colonial Surveys
Land surveys conducted during the colonial era of
the United States, as well as most surveys conducted east of Ohio, did not follow standardized
procedures and described properties that were irregular in size and shape, unlike the rectangular
public land surveys described later (Galatowitsch,
1990). However, such surveys can be a source of
information about pre settlement vegetation using
witness tree records from surveys of unsettled
lands. Loeb (1987) compiled witness tree records
for unappropriated lands prior to 1790 for eastern
New Jersey and southern New York and compared
the resulting coarse-scale presettlement species distributions with a current forest-type map and with
pollen records.
19.6.2 Warrant Maps
County surveyors produced warrant maps at the
time of first settlement of an area to describe a tract
of land. Witness trees were used along with posts
and stone monuments to mark property comers
prior to issuing a warrant conveying the property
to an owner (Abrams and Ruffner, 1995). Witness
tree records obtained from warrant maps were used
to determine presettlement tree species distributions over physiographic units and landforms in
Pennsylvania and West Virginia and to assess
differences in overall percent species composition compared to present-day forest composition
(Abrams and Ruffner, 1995; Abrams and McCay,
1996).
19.6.3 Public Land Surveys
The U.S. General Land Office conducted rectangular surveys of public land to delineate townships
(36 square miles in size) and sections (1 square
iments following fire. Close agreement of a 450year dendrochronologic fire history with charcoal
peaks was used as justification to extend the charcoal chronology to 750 years. The results indicated
that the fire regime was characterized by punctuated episodes of severe fires.
2. The same 750-year period was analyzed using varved lake sediments as the source of annual
charcoal records for small lakes in northwestern
Minnesota (Clark, 1990). Charcoal peaks were
compared to dated fire-scar records. Fire was found
to be most frequent in the relatively warm and dry
15th and 16th centuries, and longer fire intervals
were observed after A.D. 1600 during the Little Ice
Age.
3. In Indiana, charcoal analysis over the past
4000 years showed an increase in abundance of
charcoal as vegetation shifted from pine forest to
oak savanna and prairie, as well as extremely high
amounts of charcoal during the period 1850 to 1900
associated with intense fires that accompanied European settlement (Cole and Taylor, 1995).
4. Larsen and MacDonald (1998) combined
pollen and charcoal records from varved lake sediments to determine boreal forest fire and vegetation history over an 840-year period in Wood Buffalo National Park, Alberta. A general sequence of
herb, shrub, deciduous tree, and conifer pollen was
observed following fires, but peak values for taxa
differed among fires, suggesting that a site may not
consistently undergo the same vegetation sequence
following fires. This may be due to climate change,
fires with varying characteristics, or rearrangement
of patches of forest dominants around the lake
(Larsen and MacDonald, 1998).
19.6 Land Survey Records
Land survey records can provide information about
the biotic and environmental conditions that existed
at the time the survey was conducted. Such information may include vegetation composition, occurrence of disturbance, and land-use patterns. Survey records generally represent points in time and
a set of sample points in space whose spatial extent is determined by the extent of the survey. In
North America, several sources of information describe the vegetation of an area prior to European
settlement, including colonial surveys, warrant
maps, and public land surveys. Key components of
each of these sources are witness trees, or trees that
mark boundary comers and points along survey
lines (Loeb, 1987). Surveys of unsettled lands in
the United States often included the identification
Methods for Determining Historical Range of Variability
of witness trees in areas where trees were present.
Old survey records are available for most of the
continental United States, and when surveys were
conducted prior to settlement, witness tree records
provide a source of information about presettlement
tree species distributions and associations. However, care must be taken in the use of witness tree
records due to possible surveyor bias in the selection of the species and sizes of witness trees (discussed further in Section 19.6.3). In addition, unreliable survey records have resulted in some cases
from careless work or outright fraud (Bourdo,
1956).
19.6.1 Colonial Surveys
Land surveys conducted during the colonial era of
the United States, as well as most surveys conducted east of Ohio, did not follow standardized
procedures and described properties that were irregular in size and shape, unlike the rectangular
public land surveys described later (Galatowitsch,
1990). However, such surveys can be a source of
information about pre settlement vegetation using
witness tree records from surveys of unsettled
lands. Loeb (1987) compiled witness tree records
for unappropriated lands prior to 1790 for eastern
New Jersey and southern New York and compared
the resulting coarse-scale presettlement species distributions with a current forest-type map and with
pollen records.
19.6.2 Warrant Maps
County surveyors produced warrant maps at the
time of first settlement of an area to describe a tract
of land. Witness trees were used along with posts
and stone monuments to mark property comers
prior to issuing a warrant conveying the property
to an owner (Abrams and Ruffner, 1995). Witness
tree records obtained from warrant maps were used
to determine presettlement tree species distributions over physiographic units and landforms in
Pennsylvania and West Virginia and to assess
differences in overall percent species composition compared to present-day forest composition
(Abrams and Ruffner, 1995; Abrams and McCay,
1996).
19.6.3 Public Land Surveys
The U.S. General Land Office conducted rectangular surveys of public land to delineate townships
(36 square miles in size) and sections (1 square