Mitigation and Adaptation Strategies for Global Change (2006) 11: 829–846
C©
Springer 2006
DOI: 10.1007/s11027-005-9015-4
HOW CLIMATE AND VEGETATION INFLUENCE THE FIRE REGIME
OF THE ALASKAN BOREAL BIOME: THE HOLOCENE
PERSPECTIVE
FENG SHENG HU
1,∗
, LINDA B. BRUBAKER
2
, DANIEL G. GAVIN
1,4
,
PHILIP E. HIGUERA
2
, JASON A. LYNCH
1,4
, T. SCOTT RUPP
3
and WILLY TINNER
1,4
1
Departments of Plant Biology and Geology, and Program in Ecology and Evolutionary Biology,
University of Illinois, Urbana, IL 61801
2
College of Forest Resources, University of Washington, Seattle, WA 98915
3
Department of Forest Sciences, University of Alaska, Fairbanks, AK 99775
4
Present addresses: D.G. Gavin, Department of Botany and Agricultural Biochemistry, University
of Vermont, Burlington VT 05405; J.A. Lynch, Department of Biology, North Central College,
Naperville, IL 60540; W. Tinner, Institut fu¨r Pflanzenwissenschaften, Universita¨t Bern,
Altenbergrain 21, CH 3013 Bern
(
∗
Author for correspondence: Tel: 217-244-2982; Fax: 217-244-7246, E-mail: fshu@life.uiuc.edu)
(Accepted in final form 24 June 2004)
Abstract. We synthesize recent results from lake-sediment studies of Holocene fire-climate-
vegetation interactions in Alaskan boreal ecosystems. At the millennial time scale, the most robust
feature of these records is an increase in fire occurrence with the establishment of boreal forests dom-
inated by Picea mariana: estimated mean fire-return intervals decreased from ≥300 yrs to as low as
∼80 yrs. This fire-vegetation relationship occurred at all sites in interior Alaska with charcoal
†
-based
fire reconstructions, regardless of the specific time of P. mariana arrival during the Holocene. The
establishment of P. mariana forests was associated with a regional climatic trend toward cooler/wetter
conditions. Because such climatic change should not directly enhance fire occurrence, the increase
in fire frequency most likely reflects the influence of highly flammable P. mariana forests, which
are more conducive to fire ignition and spread than the preceding vegetation types (tundra, and
woodlands/forests dominated by Populus or Picea glauca). Increased lightning associated with al-
tered atmospheric circulation may have also played a role in certain areas where fire frequency
increased around 4000 calibrated
‡
years before present (BP) without an apparent increase in the
abundance of P. mariana. When viewed together, the paleo-fire records reveal that fire histories dif-
fered among sites in the same modern fire regime and that the fire regime and plant community
similar to those of today became established at different times. Thus the spatial array of regional fire
regimes was non-static through the Holocene. However, the patterns and causes of the spatial variation
remain largely unknown. Advancing our understanding of climate-fire-vegetation interactions in the
Alaskan boreal biome will require a network of charcoal records across various ecoregions, quantita-
tive paleoclimate reconstructions, and improved knowledge of how sedimentary charcoal records fire
events.
Keywords: Alaska, boreal forests, charcoal records, climate change, fire regime, Holocene
†
In this paper, charcoal refers to macroscopic (≥180 µm) as opposed to microscopic (<180 µm)
particles unless indicated otherwise.
‡
Radiocarbon ages were converted to calibrated years before AD 1950 using the atmospheric
calibration data set (Stuiver et al. 1998).

830 F.S. HU ET AL.
1. Introduction
Historic observations, computer simulations, and paleoecological analyses pro-
vided compelling evidence for strong climate-fire relationships in the boreal-
forest biome of North America (Larsen and MacDonald 1998a; Hess et al. 2001;
Kasischke et al. 2002; Stocks et al. 2002; Lynch et al. 2004b; Duffy et al. 2005).
Most of the area burned in Alaska over the past 50 years resulted from fire activity in
a limited number of years with particularly warm/dry growing seasons (Kasischke
et al. 2002), and large-scale climatic forcings (e.g., ENSO and PDO) explained a
significant portion of the variance in fire occurrence in that region (Hess et al. 2001;
Duffy et al. 2005). Modeling suggests that future warming will lead to increases in
the frequency, severity, and extent of boreal-forest fires in western North America
(Flannigan and Van Wagner 1991; Starfield and Chapin 1996; Weber and Flannigan
1997). The sensitivity of the boreal-fire regime to climatic change is also evident
in the paleorecord, which has revealed major fire-frequency shifts in response to
centennial- and millennial-scale climatic variation (e.g., Larsen and MacDonald
1998a; Carcaillet et al. 2001; Lynch et al. 2003; Lynch et al. 2004a).
However, climate-fire relationships are complex, and predicting responses of
boreal fire regimes to future climate change is not straightforward. This complex-
ity stems from the multiple biological and physical controls of fire occurrence
whose relative importance may vary across a wide range of spatial and temporal
scales (Figure 1). For example, weather conditions may override the importance
of fuels as a regulator of fire occurrence at sub-annual time scales, but the direct
impacts of climate on the fire regime may be dwarfed by the effects of vegetation
composition at longer time scales, which can exert a key control through changing
the abundance, structure, and combustibility of fuels (Rupp et al. 2002). Paleoeco-
logical research can help elucidate the patterns and controls of fire-regime change
by providing records of fire occurrence at stand to regional scales and decadal
to millennial scales. The long-term perspective of paleoecological analysis is par-
ticularly useful for ecosystems characterized by long fire return intervals, such
as boreal forests. Furthermore, paleoecological studies offer insights into climate-
fire-vegetation interactions unavailable from observational and modeling studies.
Existing observational data do not capture the whole spectrum of climatic vari-
ability in the past and anticipated for the future, and simulation models are often
constrained by knowledge from short-term observations.
In Alaska, several decades of paleoecological studies have resulted in a large
pollen database for vegetational and climatic reconstructions (Anderson et al. 2003).
Paleoecologists working in that region have generally acknowledged the role of fire
disturbances in modulating Holocene vegetational response to climatic change (Hu
et al. 1993; Anderson and Brubaker 1994; Hu et al. 1996). However, the study of fire
history is in its infancy in this region. Quantitative fire-frequency reconstructions
based on charcoal records did not begin until recently (Lynch et al. 2003; Lynch
et al. 2004b). This paper reviews our current state of knowledge and discusses