Dense tree stage of mostly birch, aspen, but also some
fens on the Tanana Flats and related them to perma-
white spruce after 1530 years.
Mature hardwood stage, with quaking aspen and pa-
Alaska Range. Permafrost underneath degrading birch
per birch after 50150 years.
forests, found adjacent to collapse scar fens, has been
Spruce stage after 100200 years.
found to be extremely ice-rich, in contrast to perma-
frost under black spruce forests, which tends to be ice-
The successional sequence on black spruce sites is
poor (Walters et al. 1998). While permafrost degrada-
similar in structure but varies some in species compo-
tion on the Tanana Flats has been found to be wide-
sition, and includes:
spread (50% of frozen or previously frozen areas are in
some stage of permafrost degradation) and rapid
Newly burned stage, with resprouting ericaceous
(Racine et al. 1998), little permafrost degradation has
shrubs during 01 year.
occurred on Fort Greely, presumably because of the
Mossherb stage, when fast growing mosses, herbs,
cooler climate associated with the higher elevations and
and tree seedling become established after 15 years.
the prevalence of thaw-stable, gravelly soils.
Tall shrub-sapling stage, occurring 530 years after
Humans. Human disturbances include cut-and-fill
Dense tree stage of mostly birch, aspen, and black
associated with the construction of roads and pads, land
spruce after 3055 years.
clearing, excavation for impoundments, trail develop-
Mixed hardwood-spruce stage, with black spruce,
paper birch, and quaking aspen after 5590 years.
Of these disturbances, only roads, pads, clearings, and
Spruce stage with black spruce and Sphagnum mosses
excavations were large or distinct enough to be mapped.
after 90200+ years.
Of the entire study area, 0.6% was Human Modified.
Although little is known about the response of subarc-
Fire frequencies as high as every 3055 years have
tic ecosystems to disturbance because most research in
been reported for some forest types in interior Alaska
Alaska has focused on tundra ecosystems (Van Cleve
(Yarie 1981). Between 1940 and 1970, nearly 1% of
1977, Walker et al. 1987, Slaughter et al. 1989), we
interior Alaska forestland burned annually (Barney
provide brief descriptions of types of human distur-
1971), whereas since 1970, 0.6% of forested land has
bances and references to pertinent literature below.
burned annually (Viereck and Schandelmeier 1980).
The effects of roads on forest ecosystems have been
Based on fires recorded on Fort Greely since 1950, 1.2%
assessed briefly by Brown and Berg (1980), but major
of the area has burned annually.
studies on ecological effects are lacking. In addition to
the direct effects, the indirect impact of dust also has
Thermokarst. While a relatively large portion of
significant ecological effects (Walker and Everett 1987).
the landscape has permafrost, surface forms indicate
Trails resulting from training exercises and recre-
that only a relatively small proportion of the area (no-
ational activities are common, but little is known about
tably thaw ponds) has been affected by permafrost deg-
the ecological changes and recovery potential for bo-
radation. A map of active layer depths, an indicator of
real ecosystems (Sparrow et al. 1978, Racine and
permafrost presence, reveals that permafrost distribu-
Ahlstrand 1991). In addition, generalization of the eco-
tion is highly patchy, particularly in morainal areas with
logical effects and recovery potential is made more dif-
abrupt changes in slope and aspect (Fig. 13). In the study
ficult by the complex interactions of ecosystem char-
area, the ecotypes that generally have developed in re-
acteristics, seasonality of impacts, number of passes,
sponse to thermokarst include Lacustrine Fen Meadow
type of vehicle or foot traffic, and soil and permafrost
(1.6% of plots, although most patches were associated
with kettle lakes), Lowland Fen Meadow (0.8% of plots,
A wide range of contaminants has been found on
although some plots were associated with swales), and
Fort Wainwright (Kennedy et al. 1997) and we assume
thaw ponds (0.1% of area). Overall, we estimate that
that many of these also are present to some extent on
less than 1% of the study area has undergone some de-
gree of permafrost degradation.
cated in the main cantonment area and include pesti-
Successional relationships related to permafrost deg-
cides, dioxin/furans, heavy metals, petroleum products,
radation are poorly understood. Drury (1956) first de-
and other organic compounds. Most of this contamina-
scribed thermokarst processes in the upper Kuskokwim
tion normally is associated with leakage at buildings,
River region and the changes in vegetation associated
tank farms, landfills, fire-training pits, drum burial sites,
with them, but little attention has been paid to this dis-
and coal storage. Little is known, however, about the
turbance regime. Racine and Walters (1994) described