Aj. Rebertus et Br. Burns, THE IMPORTANCE OF GAP PROCESSES IN THE DEVELOPMENT AND MAINTENANCE OFOAK SAVANNAS AND DRY FORESTS, Journal of Ecology, 85(5), 1997, pp. 635-645
1 We examined the dynamics of a Quercus-dominated dry forest-savanna c
omplex in the Ozark Highlands of the central United States, focusing o
n regeneration patterns within gaps. The study area includes one of ve
ry few temperate Quercus ecosystems not degraded by fire suppression,
logging or excessive grazing. 2 We compared patterns of tree seedling
establishment, tree height growth and postfire survival in gaps betwee
n savanna, found only on south and west aspects in our study area, and
dry forest, found only on north and east aspects, to determine which
factors contributed to maintaining structural differences. 3 Canopy ga
ps constituted 42% of savanna area compared with 17% for dry forest. M
ean canopy gap size in savanna was 316 m(2) compared with 185 m(2) for
dry forest, but gap sizes were highly variable and the difference was
not significant. Remnants of canopy trees were found in 37 of 38 gaps
, indicating that gaps in both savanna and dry forest can usually supp
ort trees. 4 Savanna and dry forest gaps were generally favourable for
seedlings and growth of small trees. Seedlings of most woody species
were widespread in the understoreys of both savanna and dry forest but
were more abundant in gaps than under closed canopies, Species had di
fferent height growth patterns depending on site, with growth rates or
dered as follows: Quercus velutina in savanna > Q. velutina in forest
> Q. stellata in forest > Q. stellata in savanna. Site differences wer
e significant for e. stellata but not for Q. velutina. 5 Tree survival
in two surface fires was significantly lower in gaps in savanna compa
red with gaps in dry forest. Survival within individual gaps was inver
sely related to grass cover. In savanna, survival of small trees (2.5-
7 cm d.b.h.) ranged from > 80% for gaps with < 5% grass cover to < 50%
survival in gaps with > 90% grass cover. 6 Canopy gaps in the Ozarks
are created primarily by windfall and drought-related dieback, but fir
e influences the timing and rate of gap infilling. Understanding the i
nteractions among these disturbances may help clarify the dynamics of
many temperate woodlands once dominated by surface fires.