Etf. Witkowski et Rd. Garner, Spatial distribution of soil seed banks of three African savanna woody species at two contrasting sites, PLANT ECOL, 149(1), 2000, pp. 91-106
In southern African savannas, bush encroachment is a major problem for rang
e managers. However, little is understood of the actual regeneration proces
ses leading to it, and in particular the role of soil seed banks. The horiz
ontal (between microsites) and vertical (with depth in litter and soil) dis
tribution of soil seed banks of the microphyllous woody species, Acacia tor
tilis, A. nilotica and Dichrostachys cinerea (all legumes of the Mimosoidea
e), were quantified in an area with low intensity grazing (reserve), and a
bordering cattle farm with high intensity grazing (farm). Species differed
in seed bank densities between microsites and sites. Seed densities for all
species were highest below parent tree canopies and decreased with distanc
e beyond the canopy, and with soil depth. D. cinerea had the smallest seed
bank associated with parent trees, particularly on the farm (8 vs. 1643 see
ds/tree on the reserve), A. tortilis had the largest (6357, 31910), with A.
nilotica intermediate (1789, 1906). The proportion of current (recently fa
llen) versus old (greater than or equal to 1 year old) seeds differed betwe
en species and sites. These species form at least short-term persistent see
d banks with the old seeds largely representing the persistent seed bank. S
eed densities in the open (inter-canopy) and those dispersed under either o
f the other two (non-parental) study species were much lower than those ass
ociated with parent trees. The latter were mostly found under the acacias (
single-stemmed) rather than D. cinerea (multistemmed). Total seed store per
parent plant increased with plant size (best fits were mostly power curves
of canopy area). A large proportion of intact seeds were viable, namely 81
-84% for A. tortilis, 68-77% for A. nilotica and 63-78% for D. cinerea, wit
h no differences between sites. Viability tended to increase with depth of
burial, except for A. nilotica seeds at the 3-5 cm depths on the farm. At t
he landscape scale there were 1.5 million and 140 000 A. tortilis seeds/ha
on the reserve and farm respectively, with corresponding values of 2000 and
31 000 for D. cinerea, and 23 000 and 86 000 for A. nilotica.