Resource Niche and Trade-offs 167The seedlin gsurvival data of Rose (2000;
Figure 10.2a) across a large gradient of gap size
show that rank reversals are observed only above
gap size greater than 1000 m^2 , below which
species exhibit a perfect rank concordance; the
four shade-tolerant species survive better than the
three light-demanding species. Two of the four
shade-tolerant species maintain almost 100%
survival across the entire gap-size gradient, while
the other two exhibit lower survival at larger
gaps. Three light-demanding species, in contrast,
respond strongly and positively to increasing gap
size. Yet, survival probabilities of three light-
demandin gspecies do not surpass that of two
shade-tolerant species even in the largest gaps.
Similarly, Kobe (1999) found that seedlin gsur-
vival of four species responded positively to higher
light availability, except survival of one of the two
shade-tolerant species decreased above 20% full
sun (i.e., light level typically found in treefall gaps).
The fact that survival rank reversals are observed
at light levels higher than typical for treefall gaps
speaks against the relevance of rank reversals
for light partitioning between shade-tolerant and
gap-dependent species.
Seedlin gsurvivorship of nine species studied
on Barro Colorado Island (BCI) exhibits only
five out of the maximum possible 36 rank
reversals between shaded understory and gaps
(Figure 10.3a, Augspurger 1984a). This fre-
quency is lower than expected by a statistical null
model (rejected atP=0.009 with a Monte Carlo
analysis with 1000 shuffles, assumin g50% as a
chance of rank reversal between any randomly
selected pair of species; see Box 10.1). Hence,
species that survive well relative to other species
in shade also survive well in gaps (Figure 10.3b;
Pearson’s correlation coefficientr =0.76,P<
0.02, and Kendall’sτ=0.70,P<0.009). A pos-
itive correlation of survival rate was also found