A Burning Question: Can Long-term Data Inform Fire Management in the Twenty-first Century? - Biodiversity Conservation and Environmental Change - page 62

Such tipping points have important consequences for biodiversity and ecosystem ser-
vices, as they occur rapidly and leave little time for adaptation. However, ecological studies
seldom cover long enough timescales to understand and predict threshold effects, and com-
plex interactions between fire-vegetation-climate-CO 2 (Bond and Midgley 2000, Scheiter
and Higgins 2009, Higgins and Scheiter 2012). The palaeoecological record can be used to
disentangle the role of vegetation-fire feedbacks from climatic drivers and identify the tip-
ping points between forest and savanna ecosystems. For example, Rull et al. (2013) studied
savanna forest dynamics in the Gran Sabana region of Venezuela. They found that upland
savannas and the savanna-forest mosaic were not in equilibrium with climate, but were
largely conditioned by fire. They highlighted species that might be most sensitive to change
and worthy of individual conservation attention. Shifts from forest to savanna appear irre-
versible, because of feedbacks between vegetation and fire (see Figure 4.4) (Rull 2009, Rull
et al. 2013). Similarly, in the Noel Kempff Mercado National Park (NKMNP), Bolivia, fossil
pollen and charcoal records show the expansion of rainforest into savanna areas over the
past 3,000 years, associated with decreased fire (Mayle et al. 2007). Relict savanna tree taxa
in older rainforests provide further evidence of forest encroachment in Amazonia, a process
that is of conservation concern because there are more endangered species in the savanna
vegetation (Mayle et al. 2007).
Predicting the future of savanna-forest dynamics is uncertain, because rising CO 2 and
increasing temperature have opposing effects. CO 2 favours tree recruitment and forest expan-
sion, but warmer temperatures may reduce available moisture and enhance fire, thereby
causing savanna expansion. Modelling experiments suggest that current conditions are near
a tipping point, where some forests will expand while in others the balance tips in favour of
savanna expansion and increased fire, causing tropical forests to change from carbon sinks to
carbon sources (Cowling et al. 2004, Rull et al. 2013). For example, Amazonia may tip from
carbon sink to carbon source by 2100, though the rainforests of Asia and Africa might be more
resilient to climate change, and could even potentially expand at the expense of savannas
and grasslands (Hirota et al. 2011, Staver et al. 2011, Higgins and Scheiter 2012, Huntingford
et al. 2013).
Fire-vegetation feedbacks have also had major impacts outside of the tropical savanna
biome. For example, the hardwood forest known as the 'Big Woods, in Minnesota, USA, per-
sist today because of the same mechanism of shading and fuel reduction, described above. In
the MWP, open oak savannas were associated with frequent fires, but drought conditions in
the LIA reduced fuel load and connectivity, thereby reducing fire frequency and allowing the
expansion of forest (Umbanhowar 2004, Shuman et al. 2009). Initially, oak abundance
increased, followed by other forest species like elm ( Ulmus ), lime ( Tilia ), maple ( Acer ) and
hornbeam ( Ostrya ). Once established, the canopy shaded out understory grasses and
reduced fire frequency, thus becoming a self-sustaining stable state (Umbanhowar 2004,
Jackson and Hobbs 2009, Shuman et al. 2009). In boreal forests, Ohlson et al. (2011) assem-
bled a spatially comprehensive data set of 75 macroscopic charcoal records from northern
Europe. They found that the late-Holocene invasion of Norway spruce ( Picea abies ), a new
forest dominant in northern Europe, significantly reduced wildfire activity, allowing canopy