276
dates, which in conjunction with aerial photographs
and forest-type maps were the basis for production
of stand origin and fire year maps for the period
1595 to 1972.
3. Kitzberger et al. (1997) derived fire chronologies from fire-scarred trees in northern Patagonia,
Argentina, for the period 1820-1974 and detected
a strong influence of annual climatic variation on
fire frequency and extent, although human activities were also found to have affected fire occurrence.
4. Amo et al. (1993) developed a method for
characterizing disturbance patterns in lodgepole
pine forests of the northern Rocky Mountains by
sampling plots in a grid system within a landscape,
including fire-scar sampling and determination of
age classes of trees. Synthesis of disturbance histories for grid points enabled estimation of the spatial extent and severity of past fires in the landscape.
5. Comparisons of reconstructed stand ages following stand-replacing fires were used to discern
differences in fire frequency and rate of secondary
succession on uplands versus valley bottoms in a
subalpine watershed in the Medicine Bow Mountains, Wyoming, where fire-scarred trees are rare
(Romme and Knight, 1981).
6. Veblen et al. (1991) incorporated the analysis of growth patterns in tree rings to assess the relative effects of dated fire and windthrow events on
subalpine old-growth stand dynamics in the Front
Range of Colorado.
7. The long-term preservation of woody stems
in peatland enabled Arsenault and Payette (1997)
to document a fire-induced shift from conifer to
tundra vegetation in Quebec in the 16th century.
Tree-ring chronologies were constructed for remains of spruce trees living prior to and following
a fire in 1567-1568. A well-drained site responded
immediately to fire-induced deforestation and consequent modification of snow accumulation, but a
delayed response to landscape changes was observed for a protected site.
8. McCord (1996) studied stream-bank trees
scarred by flood events to reconstruct the flood history of Frijoles Canyon, New Mexico, which was
examined in conjunction with fire history data to
detect possible correlation of floods with largescale fires.
9. Baker (1988) determined stand ages using
tree-ring analysis in riparian woodland stands on a
reach of the Animas River in Colorado. Stand structure was found to be strongly influenced by the timing and character of large-scale flooding disturbances.
Methods for Determining Historical Range of Variability
19.3 Pollen Content Analysis
Pollen and other plant materials such as macrofossils and charcoal may accumulate over time and be
preserved in lake sediments and wetlands (Delcourt
and Delcourt, 1988). When extracted from sediment cores, pollen data can be used to reconstruct
the vegetation of an area over periods of up to thousands of years (Whitlock, 1992). The methodology
involves collection and extrusion of sediment cores
that are subsampled for pollen (as well as macrofossils and charcoal; see Sections 19.4 and 19.5) at
measured stratigraphic intervals along the core
(Foster and Zebryk, 1993). Berglund (1986) provides a comprehensive summary of paleoecologic
techniques. A book edited by Bryant and Holloway
(1985) contains summaries of pollen analyses
across North America.
The number and spacing of samples taken from
the sediment core determines the temporal resolution of the pollen analysis (Grimm, 1988; Whitlock, 1992). Paleoecological studies that examine
changes over periods of thousands of years usually
have a relatively coarse temporal resolution of 300to 1000-year intervals (Whitlock, 1992). The finest
temporal resolution is provided by sites with annually laminated (varved) sediments, but varved
lakes are relatively uncommon (Millspaugh and
Whitlock, 1995). Methods for determining chronological frameworks for sediment cores include radiocarbon and lead 210 dating of bulk sediment or
organic fossils (Whitlock, 1992). In addition,
varves, if present, can be counted, and volcanic ash
layers can be associated with eruptions of known
age (Clark, 1990; Whitlock, 1992). The pollen
source area and hence the spatial extent sampled
by pollen analysis is a function of the size of the
lake or wetland; for example, small- to mediumsized lakes (1 to 50 ha) collect pollen from an area
of 100 to 1000 km 2 (Jacobson and Bradshaw, 1981;
Whitlock, 1992). The smaller the lake or wetland,
the stronger the local pollen signal will be compared to regional pollen (Foster and Zebryk, 1993).
Tracer pollen is added to samples in known
quantities to calculate pollen concentrations and
pollen accumulation rates (Whitlock, 1993). Pollen
grains are identified by comparison with published
atlases and reference collections (Whitlock, 1993).
Identification of pollen to species is often not possible, but greater taxonomic resolution may be
achieved by combining pollen analysis with identification of associated plant macrofossils (see Section 19.4; Whitlock, 1993; Fall, 1997). Plant
species differ in the amount of pollen produced and
pollen susceptibility to destruction (Fall, 1997).
dates, which in conjunction with aerial photographs
and forest-type maps were the basis for production
of stand origin and fire year maps for the period
1595 to 1972.
3. Kitzberger et al. (1997) derived fire chronologies from fire-scarred trees in northern Patagonia,
Argentina, for the period 1820-1974 and detected
a strong influence of annual climatic variation on
fire frequency and extent, although human activities were also found to have affected fire occurrence.
4. Amo et al. (1993) developed a method for
characterizing disturbance patterns in lodgepole
pine forests of the northern Rocky Mountains by
sampling plots in a grid system within a landscape,
including fire-scar sampling and determination of
age classes of trees. Synthesis of disturbance histories for grid points enabled estimation of the spatial extent and severity of past fires in the landscape.
5. Comparisons of reconstructed stand ages following stand-replacing fires were used to discern
differences in fire frequency and rate of secondary
succession on uplands versus valley bottoms in a
subalpine watershed in the Medicine Bow Mountains, Wyoming, where fire-scarred trees are rare
(Romme and Knight, 1981).
6. Veblen et al. (1991) incorporated the analysis of growth patterns in tree rings to assess the relative effects of dated fire and windthrow events on
subalpine old-growth stand dynamics in the Front
Range of Colorado.
7. The long-term preservation of woody stems
in peatland enabled Arsenault and Payette (1997)
to document a fire-induced shift from conifer to
tundra vegetation in Quebec in the 16th century.
Tree-ring chronologies were constructed for remains of spruce trees living prior to and following
a fire in 1567-1568. A well-drained site responded
immediately to fire-induced deforestation and consequent modification of snow accumulation, but a
delayed response to landscape changes was observed for a protected site.
8. McCord (1996) studied stream-bank trees
scarred by flood events to reconstruct the flood history of Frijoles Canyon, New Mexico, which was
examined in conjunction with fire history data to
detect possible correlation of floods with largescale fires.
9. Baker (1988) determined stand ages using
tree-ring analysis in riparian woodland stands on a
reach of the Animas River in Colorado. Stand structure was found to be strongly influenced by the timing and character of large-scale flooding disturbances.
Methods for Determining Historical Range of Variability
19.3 Pollen Content Analysis
Pollen and other plant materials such as macrofossils and charcoal may accumulate over time and be
preserved in lake sediments and wetlands (Delcourt
and Delcourt, 1988). When extracted from sediment cores, pollen data can be used to reconstruct
the vegetation of an area over periods of up to thousands of years (Whitlock, 1992). The methodology
involves collection and extrusion of sediment cores
that are subsampled for pollen (as well as macrofossils and charcoal; see Sections 19.4 and 19.5) at
measured stratigraphic intervals along the core
(Foster and Zebryk, 1993). Berglund (1986) provides a comprehensive summary of paleoecologic
techniques. A book edited by Bryant and Holloway
(1985) contains summaries of pollen analyses
across North America.
The number and spacing of samples taken from
the sediment core determines the temporal resolution of the pollen analysis (Grimm, 1988; Whitlock, 1992). Paleoecological studies that examine
changes over periods of thousands of years usually
have a relatively coarse temporal resolution of 300to 1000-year intervals (Whitlock, 1992). The finest
temporal resolution is provided by sites with annually laminated (varved) sediments, but varved
lakes are relatively uncommon (Millspaugh and
Whitlock, 1995). Methods for determining chronological frameworks for sediment cores include radiocarbon and lead 210 dating of bulk sediment or
organic fossils (Whitlock, 1992). In addition,
varves, if present, can be counted, and volcanic ash
layers can be associated with eruptions of known
age (Clark, 1990; Whitlock, 1992). The pollen
source area and hence the spatial extent sampled
by pollen analysis is a function of the size of the
lake or wetland; for example, small- to mediumsized lakes (1 to 50 ha) collect pollen from an area
of 100 to 1000 km 2 (Jacobson and Bradshaw, 1981;
Whitlock, 1992). The smaller the lake or wetland,
the stronger the local pollen signal will be compared to regional pollen (Foster and Zebryk, 1993).
Tracer pollen is added to samples in known
quantities to calculate pollen concentrations and
pollen accumulation rates (Whitlock, 1993). Pollen
grains are identified by comparison with published
atlases and reference collections (Whitlock, 1993).
Identification of pollen to species is often not possible, but greater taxonomic resolution may be
achieved by combining pollen analysis with identification of associated plant macrofossils (see Section 19.4; Whitlock, 1993; Fall, 1997). Plant
species differ in the amount of pollen produced and
pollen susceptibility to destruction (Fall, 1997).