The Elephant Dilemma: A Long-term Perspective on the Management of African Savannas - Biodiversity Conservation and Environmental Change - page 26

Further simulations suggest that savannas in areas of sufficient rainfall may undergo hys-
teretic transitions to forest conditions with increasing CO 2 ; above a tree density threshold,
ground cover is insufficient to carry fire, and thus shade tolerant, fire-sensitive forest trees can
establish (Warman and Moles 2009, Hirota et al. 2011, Mayer and Khalyani 2011, Higgins and
Scheiter 2012, Parr et al. 2012). Once this occurs, such a transition is likely to be irreversible,
because dense shade prevents fire from returning to the ecosystem, and a permanent forest
condition is likely to establish (Higgins and Scheiter 2012). This would be excellent for carbon
storage and provision of fuel wood, but poor for grazing resources and for wildlife adapted to
open, grassy conditions (Parr 2014).
Palaeoecological data from savannas can help in tracking changes in tree density over
time, understanding the role of fire, herbivory, and land management in driving transitions
between woodland grassland and forest vegetation (Ekblom and Gillson 2010c, Ekblom et al.
2012), thereby providing clues as to how the potential impact of rising CO 2 can be mitigated
through the management of fire and herbivores.
Palaeoecology and thresholds of potential concern
The paucity of precolonial baseline data in Africa, and the scarcity of palaeoecological sites in
savannas, means that managers of even the oldest national parks have no records of elephant
habitats prior to their decimation at the hands of European hunters and ivory traders. This
means that choosing baselines and targets based on the twentieth century landscapes, when
many National Parks and game reserves were founded, can be extremely misleading. The
arguments used to justify elephant culling rest on atypical snapshots in time and aim to
restore an equilibrium that never existed. Therefore, it is essential that long-term data are
embedded into the management of savannas, and that the adaptive management process
includes a temporal dimension that provides knowledge of the history, local context, and
range of natural variability.
Palaeoecological and other long-term information can contribute to ecological under-
standing of variability and resilience, helping to identify ecological thresholds and develop-
ing realistic management thresholds that incorporate variability over time through concepts
such as the historical range of variability, limits of acceptable change, and TPCs (Figure 2.6)
(Cyr et al. 2009, Long 2009, Battarbee and Bennion 2011, Bennion et al. 2011, McLoughlin et al.
2011, Gell et al. 2012, Wolfe et al. 2012). These concepts accept heterogeneity and flux as inher-
ent aspects of ecological systems and aim to work within a range of ecosystem variability,
rather than aiming towards a specific ecosystem state.
For TPCs to be ecological realistic, long-term knowledge is needed about the variability
and resilience of vegetation, and the thresholds at which biodiversity may be threatened.
Identifying these thresholds is the first step towards defining the desired range of variability
for tree density in savannas, which may become converted to grasslands or forests due to the
interacting effects of climate, fire, herbivory and land-use. Then, a safety margin must be
incorporated, allowing managers time to intervene, for example by manipulating fire or