Agriculture Reference
In-Depth Information
recruitment success appear to vary geographically, with marked effects in the
western and southern Cape but no apparent effects in the eastern Cape under a
non-MTC (Heelemann
et al.
2008
). Postfire recruitment may also be poorer after
very intense fires that appear to incinerate seeds, but little data are available.
Some protea species also show intrinsic variation in recruitment success due to
density-dependent effects. Whereas the “law of constant yield” states that most
plant populations produce constant seed output across a wide range of densities
(Harper
1977
), some dense protea stands show a decline in stand-level seed
output. Protea flower heads are large structures borne on terminal shoots in many
species. Dense crowding may lead to etiolation of shoots, reduced numbers of
thick shoots, and therefore reduced developmental potential to produce the
flower heads (Bond
et al.
1995
). For candidate species, fires in mature (12-20
year old) stands, and in the most favorable season for seedlings, can produce
very dense populations that, after the next fire, have very poor seedling recruit-
ment. Repeated fires in the most favorable season for seedlings would lead to
oscillating populations, which is a rare example of chaotic dynamics in plants
(Bond
et al.
1995
).
Unlike northern hemisphere MTC shrubland species, serotinous protea popu-
lations often fluctuate greatly from one fire to the next, whether from strong
density dependence or from unfavorable fires. Local extinction is not uncommon
after a particularly unfavorable fire. Thus, fynbos proteas not only produce
transient overstories in space (being killed by fire) but also in time where an entire
stand may be eliminated by a single fire. They are rare examples of plant meta-
populations with quite frequent local extinction of subpopulations in some species
but rapid recolonization. Whereas most fynbos species have very limited dispersal
(1-100 m), serotinous proteas disperse hundreds to thousands of meters in the
few days after a fire. They do so primarily by seeds rolling over the soil surface,
with the distance traveled dependent on seed attributes, wind velocity and
terrain roughness (Bond
1988
; Schurr
et al.
2005
). The combination of high
vulnerability to local extinction and remarkable dispersal ability makes for highly
labile populations of serotinous proteoids, the dominant component of many
fynbos communities.
Serotiny contrasts strongly with myrmecochory, a dispersal syndrome charac-
terized by specialized seeds that attract ant dispersers. Ants move seeds only a few
meters but seed life span in the soil is long. Some myrmecochorous proteas have
disappeared only to reappear decades later after fire from long-lived seedbanks.
The combination of seed dispersal by ants and long-lived dormant seeds is thus an
alternative strategy of surviving highly variable fynbos fire regimes (Christian &
Stanton
2004
; see also
Chapter 9
).
Mimetes stokoei
is a spectacular example (Sling-
sby & Johns
2009
). A single specimen of this very rare 3-m-tall protea was first
discovered in 1922. A second population of just five plants was found in 1925. The
populations regenerated after several fires until the last flowering plant died in
1950 (it set no seed). Following disturbance of the ground for a planned planting
of proteas, in 1965 a single seedling appeared but died before flowering. No plants
