A Burning Question: Can Long-term Data Inform Fire Management in the Twenty-first Century? - Biodiversity Conservation and Environmental Change - page 64

Case studies from New Zealand, Madagascar, and the tropics illustrate how difficult it is to
interpret mosaic landscapes without long-term data. A forest-savanna mosaic can be ancient,
or a product of human management, and it is not always clear whether humans have cleared
or encouraged forest growth. Some fire-prone savannas are natural, some anthropogenic,
and when fire is excluded from savannas, forest vegetation can re-establish in areas of high
enough rainfall (Fairhead and Leach 1996, King et al. 1997, Bond and Keeley 2005). A long-
term perspective can dramatically change how we perceive landscapes and affect the conser-
vation value that we place on them, affecting their future management and protecting
animals and plants that are adapted to open, grass-dominated habitats (Bond and Silander
2008, Parr et al. 2014).
New Zealand provides a classic case of forest clearance and anthropogenic burning, but
this model has been over-applied and all forest-grassland mosaics do not share the same
landscape history. In New Zealand, palaeoecological records show a strong correlation
between human arrival and fire history, and much of New Zealand's flora has been heavily
transformed by anthropogenic activities and the introduction of alien species (Figure 4.5a)
(McWethy et al. 2010b). When Polynesian settlers (M¯ori), arrived in New Zealand, they
introduced fires to forest areas that had previously only burned one or twice every thousand
years (Bowman and Haberle 2010). Palaeo lake records from throughout the South Island
show peaks in charcoal beginning in the thirteenth century (Figure 4.5a), causing the
replacement of 40% of forest cover with grassland and fern shrubland (Bowman and Haberle
2010, McWethy et al. 2010b, Perry et al. 2012). European settlement in the mid-nineteenth
century brought a second wave of burning, associated with erosion, causing the spread in
non-native plant species. As the ecosystem had not evolved with fire, there were no adapta-
tions that allowed trees to regenerate rapidly—for example by fire-adapted seeds, or the
ability to resprout or coppice from burnt stems. Thus, the grasslands and shrublands both
caused, and were maintained by, fire regimes that began in the thirteenth century, creating
a grassland-forest mosaic that has no past analogue (Bowman and Haberle 2010, McWethy
et al. 2010b).
Some of New Zealand's smaller islands escaped the worst effects of forest clearing and pro-
vide reference conditions for replanting on the mainland (Wilmshurst et al. 2013). They are
important havens for habitat for relict and reintroduced endemic species. Even on small
islands, however, long-term data are needed to understand the legacies of past anthropo-
genic change. A recent study from Twahiti Rahi, the largest of the Poor Knight's Islands, which
lie of the east coast of New Zealand's North Island, shows that present-day forest compos-
ition does not reflect pre-anthropogenic conditions (Wilmshurst et al. 2013). Twahiti Rahi is
currently home to one of New Zealand's largest pohutukawa ( Metrosideros excelsa— the 'New
Zealand Christmas Tree') forests, long considered to be the natural climax vegetation. How-
ever, the fossil pollen record shows that this forest composition is less than 100 years old, and
represents extensive, secondary forest recovery since abandonment of the island in 1820, fol-
lowing an intertribal massacre of M¯ori inhabitants. he island forests encountered by the
first Polynesian settlers were dominated by palm trees, ferns, and podocarps (coniferous
trees found predominantly in the Southern Hemisphere). Active management interventions