Naturally regenerated seedlings of longleaf pine are typically observe
d to cluster in the center of tree fall canopy gaps and be encircled b
y a wide zone from which they are generally excluded. Twelve represent
ative canopy gaps distributed across 600 ha of a naturally regenerated
uneven-aged longleaf pine forest in the sandhills of north central Fl
orida were examined to determine which aboveground and belowground fac
tors are responsible for development of this seedling exclusionary zon
e. Within 12 m of adult trees growing along the gap edge, significantl
y fewer longleaf pine seedlings were present, The canopy of overstory
trees, however, extended only 4-5 m into the gap. The relatively open
structure of the longleaf pine canopy (57% cover) allowed photosynthet
ically active radiation (PAR) to be evenly distributed upon the forest
floor across each canopy gap. Thus, light avail-ability was not relat
ed to pine seedling clustering near gap centers, Significantly greater
forest litter mass beneath adult trees (5 Mg ha(-1)) could result in
fires more intense than would be supported by the litter mass near gap
centers (2.5 Mg ha(-1)). However, litter mass was significantly eleva
ted only within 4 m of the gap edge. The fine root biomass within 12 m
of the gap edge (3-4.5 Mg ha(-1)) was two to six times that measured
near the gap center and most closely coincided with the width of the s
eedling exclusionary zone along the margin of each gap. Thus, while th
e canopy of adjacent adult trees may indirectly influence seedling mor
tality through deposition of needle litter and greater fire intensity
within 4 m of the gap edge, the root systems of these adults also appe
ar to directly compete with seedlings within 12-16 m of the gap edge f
or limited site resources, To effectively regenerate and sustain longl
eaf pine wiregrass ecosystems, caution should be used in prescribing s
ingle-tree selection harvest methods so as not to create gap openings
so small (< 30 m diameter) that intraspecific competition from adult t
oes totally excludes seedlings. Group selection methods may prove more
effective in creating canopy gaps of sizes suitable (greater than or
equal to 40 m diameter) for ensuring that intraspecific competition fr
om adults is maintained at a level which allows for survival and growt
h of seedlings. Shelterwood and deferment methods may also prove usefu
l where more open canopy conditions are desired. However, the pine see
dlings' need for light cannot be used as an appropriate rationale for
application of clearcutting methods. (C) 1998 Elsevier Science B.V.