Jw. Hughes et Wb. Cass, PATTERN AND PROCESS OF A FLOODPLAIN FOREST, VERMONT, USA - PREDICTED RESPONSES OF VEGETATION TO PERTURBATION, Journal of Applied Ecology, 34(3), 1997, pp. 594-612
1. We evaluated the potential diversity of vegetation in a lowland flo
odplain forest as a way of predicting the range of possible communitie
s that might develop when a natural system is subjected to flood contr
ol or other common perturbations. The potential floras (newly fallen s
eeds, seeds stored in soil, seeds transported by water) were quantifie
d and compared with the standing flora at different distances from the
stream. 2. The species composition of newly fallen seeds was similar
to that of the standing flora but concordance of the standing and pote
ntial floras was otherwise low. The composition of seeds in flood wate
r was most different from the standing flora: only 14 of the 40 flood
water species were present in tile standing flora. Overall, only 25 of
tithe 73 species that germinated from the combined potential floras w
ere present in the standing flora. 3. Concordance of species compositi
on among the potential floras was also low. For example, the density o
f seeds germinating from flood water collections was related to distan
ce from stream, with forb density highest near the stream, but no simi
lar trend was noted for seedlings emerging from soil collections. Equa
l numbers of annual and non-natives species germinated from the soil a
nd flood water but the species composition was different. 4. Seeds sto
red in the soil that were subjected to hydric germinating conditions g
enerated a somewhat different complement of species and densities than
replicate samples subjected to mesic germinating conditions. A total
of 48 species germinated from the soil collections (hydric & mesic com
bined), but only 29 species were common to bath. 5. The composition an
d distribution of the standing vegetation at different distances from
the stream appeared to be defined mainly by tile flooding regime, but
selective disturbance agents such as beaver activity and Dutch elm dis
ease probably also affected the spatial distribution of species within
the forest. Collectively, these factors generated a highly diverse mo
saic of possible floras. Flood control can be expected to homogenize t
his mosaic and reduce the forest's potential diversity, thereby compro
mising the value nf floodplain forests as storage banks nf biotic dive
rsity.