A Burning Question: Can Long-term Data Inform Fire Management in the Twenty-first Century? - Biodiversity Conservation and Environmental Change - page 78

forest stands of jack pine that can be encouraged through frequent, managed burns (Long
2009, McKenzie et al. 2004). Lynx need mature forest for their dens with adjacent early suc-
cession habitat for hunting (McKenzie et al 2004). The effects of post-fire management must
also be considered; post-fire logging, once considered beneficial, is now discouraged because
it disrupts forest regeneration by removing seed sources. It also has negative effects on spe-
cies adapted to conditions that occur immediately after fire, and may amplify fire risk by
increasing surface fuel loads (Donato et al. 2006,, Noss et al. 2006). Protecting sensitive areas
from extreme fires, controlling invasive species, conserving genetic variability, and maintain-
ing a mosaic of postfire ages, fire patch sizes, and microclimates and landscape connectivity,
will all help to build resistance, resilience, and adaptive capacity (Millar et al. 2007).
Integrate knowledge of future scenarios using modelling and hindcasting
Future changes in rainfall and temperature may alter fire regimes, but the magnitude and
ecological implications of these affects are highly uncertain. Fire-management decisions can
exacerbate or ameliorate the effect of changing climate, and must be tailored to changing
environmental and socioecological conditions. Looking at how fire changed in past warm cli-
mates such as the MWP and MHA might help in predicting how ecosystems and their associ-
ated fire regimes might change in the future, thereby informing management choices.
Specifically, such events can be used to test the outputs of models, thereby helping to validate
predictions of future fire regimes and potentially informing fire-management plans under
different future climate and management scenarios (Keane et al. 2002, Pratt et al. 2006, Power
et al. 2010, Kehrwald et al. 2013).
Model simulations suggest that fires will become more frequent and widespread due to
anthropogenically driven climate warming in the twenty-first century, although some trop-
ical areas may experience decreases in fire, unless thresholds in rainfall and forest clearance
are reached (Cowling et al. 2004, Scholze et al. 2006, IPCC 2007, Marlon et al. 2009, Bowman
and Murphy 2011, Moritz et al. 2012, Huntingford et al. 2013). Different types of models can
simulate the empirical relationships between fire and climate (correlative models), and/or
the ecological processes by which fire, vegetation, and climate interact (mechanistic models)
(Moritz et al. 2012). Mechanistic models allow managers to test out the effects of various fire
frequencies and configurations. Dynamic vegetation models (DGVMs) can now incorporate
the effects of fire, simulating vegetation-fire feedbacks, and the influence of rising CO 2 at
both physiological and ecosystem levels (Gavin et al. 2007, Moritz et al. 2012, Pfeiffer and
Kaplan 2012). The effects of different fire frequencies, patterns and postfire treatments can
also be evaluated using modelling software (North and Keeton 2008).
Future climate projections are uncertain, and regional climate systems, changes in rainfall,
fire conditions, local topography and hydrology, and feedbacks with vegetation cover and
land-use change, all influence fire regimes at regional-local scales (Whitlock et al. 2010, Mor-
itz et al. 2012). Furthermore, weather conditions and management interventions such as fuel
reduction, tree thinning, and controlled burning interact to influence the frequency and
intensity of fires. There is therefore considerable uncertainty about future fire regimes, and