Amazonian forests are the largest and most diverse in the tropics. and much
of the mystery surrounding their ecology can be traced to attempts to unde
rstand them through tiny local inventories. In this paper we bring together
a large number of such inventories scattered across immense areas of weste
rn Amazonia in order to address simple questions about the distribution and
abundance of tropical tree species in lowland terra firme forests there. T
he goal is to describe patterns of commonness and rarity at local (1 ha), l
andscape (similar to 10(4) km(2)), and regional (> 10(6) km(2)) scales, and
to fuse the results into a more complete picture of how tropical tree comm
unities are structured. We present estimates of landscape-scale densities f
or similar to 1400 taxa. based on data from tree plots scattered over large
tracts of terra firme forest in eastern Ecuador and southeastern Peru. A d
atabase of morphological. ecological, and other traits of > 1000 of these s
pecies compiled from the taxonomic literature is then used to explore how s
pecies that are common in the inventories differ from species that are rare
.
Although most species show landscape-scale densities of < 1 individual/ha,
most trees in both forests belong to a small set of ubiquitous common speci
es. These common species combine high frequency with high local abundance,
forming predictable oligarchies that dominate several thousand square kilom
eters of forest at each site.
The common species comprising these oligarchies are a nonrandom subset of t
he two floras. At both sites a disproportionate number of common species ar
e concentrated in the families Arecaceae, Moraceae, Myristicaceae, and Viol
aceae, and large-statured tree species are more likely to be common than sm
all ones. Nearly a third of the 150 most common tree species in the Ecuador
ean forest are also found among the 150 most common tree species in the Per
uvian forest. For the 254 tree species shared by the two data sets, abundan
ce in Ecuador is positively and significantly correlated with abundance sim
ilar to 1400 km away in Peru.
These findings challenge popular depictions of Amazonian vegetation as a sm
all-scale mosaic of unpredictable composition and structure. Instead, they
provide additional evidence that tropical tree communities are not qualitat
ively different from their temperate counterparts., where a few common spec
ies concentrated in a few higher taxa can dominate immense areas of forest.
We hypothesize that most Amazonian forests are dominated at large scales b
y oligarchies similar in nature to the ones observed in Ecuador and Peru, a
nd we argue that the patterns are more indicative of regulation of relative
abundances by ecological factors than of nonequilibrium chance-based dynam
ics. The paper concludes with a discussion of the practical applications of
predictable oligarchies over large areas of unexplored forest.