Agriculture Reference
In-Depth Information
(Le Maitre & Midgley
1992
; Bond & Keeley
2005
). Fynbos is rich in species
with fire-stimulated recruitment but this is extremely rare or absent in savannas.
Many woody fynbos plants are obligate seeders (see
Chapter 3
), that is they
lack resprouting capacity after fire and seedling recruitment is restricted to a
single postfire pulse. In contrast, all South African savanna trees and shrubs
have some capacity to resprout after fire and none restrict seedling recruitment
to a single postfire burst. Thus, though both fire-prone biomes experience predict-
able fire, the different fire regimes in the two biomes have selected for entirely
different life history traits.
Fynbos-like shrublands dominate MTC landscapes and also occur in isolated
patches throughout the mesic grassy biomes of summer rainfall areas (O'Connor &
Bredenkamp
1997
). A climate-only hypothesis might propose that the extent of
drought-prone habitat favoring fynbos shrubs is greater in MTC landscapes and
the patchy occurrence in summer rainfall regimes is due to a more limited distri-
bution of appropriate habitat. However, landscapes in both of these climatic
regimes are fire-prone, but with very different regimes that could be a factor in
determining shrubland dominance in these two regions. These communities have
likely assembled due to interactions between climate, fire and perhaps geology (see
Fig. 1.4
). In this light we might hypothesize that the high fire frequency main-
tained by C
4
grasses is sufficient to exclude sclerophyllous fynbos shrubs over
much of the summer rainfall landscape. This model involves a close integration
between climate and fire, as very high fire frequency is the proximal factor
excluding fynbos shrubs in C
4
grasslands, but this is ultimately a response to the
monsoon climate selecting for rapid fuel production capability of C
4
grasses and
high fire frequency dependent on frequent lightning storms. Geology also plays a
role in that fuel continuity and fire spread are enhanced by the broad open plains
in a manner similar to the North American prairies where woodlands are
restricted to escarpments by frequent fires (Wells
1965
).
Fire exclusion experiments in the summer rain grassland biome support this
model with repeated instances of invasion of woody sclerophyllous MTV where
fire has been excluded for a decade or more (Titshall
et al.
2000
; Bond
et al.
2003
).
Indeed, shrubby Ericaceae, Rosaceae (
Cliffortia
) and Asteraceae (
Stoebe
)with
fynbos affinities have been perceived as woody invasives by livestock farmers,
and fire regimes have been developed to eliminate them and promote grasslands
(Trollope
1973
). In the absence of fire, both fynbos and the C
4
grasslands in more
humid regions have the climate potential to support forest (Bond
et al.
2003
). It
seems probable that both fire-prone biomes exist because of fire, and differ in the
traits of dominant plants in part because of different fire regimes.
If fire regime determines the boundaries of fynbos and grassy systems, why do
they have such different fire regimes? Why are fires so much less frequent in
fynbos than in C
4
grasslands? Fynbos was much more extensive in South Africa
in the last glacial period extending into areas that are now pure C
4
grasslands or
savannas (Chase and Meadows
2007
). How might climate change potentially
change the fire regime and therefore the distribution of MTC ecosystems?
