Tropical Forest Community Ecology

104 Jess K. Zimmermanetal.

0.000001

0.00001

0.0001

0.001

0.01

0.1

1

Rank abundance

Proportional abundance

Luquillo

Sinharaja BCI

Korup Pasoh

Figure 7.1 Dominance–diversity curves for large FDPs from Luquillo, Puerto Rico (16 ha; J. Thompsonet al.
unpublished data), Sinharaja, Sri Lanka (25 ha; Gunatillekeet al. 2004), Barro Colorado Island, Panama (BCI; 50 ha;
Conditet al. 1996), Korup, Cameroon (50 ha; Thomaset al. 2003), and Pasoh, Malaysia (50 ha;
Manokaranet al. 1992).

have the same structure and show that all the
forests have a few common species and rela-
tively large numbers of rare species. The extent
of each curve depends, in part, on the forest diver-
sity and the size of the plot (Figure 7.1). These
patterns suggested to Hubbell (2001) that simi-
lar factors underlie the structuring of the forest
communities. In the following section we discuss
three potential explanations for the patterns of
species diversity and community structure of dif-
ferent tropical forests that have been tested using
large FDPs.

LARGE FDPSAND ECOLOGICAL

THEORY

Neutral theory – a null model for
communities

In developing neutral theory, Hubbell (2001,
Chapter 9, this volume) united ideas on island
biogeography (MacArthur and Wilson 1967)
and relative species abundances, incorporating
assumptions regarding speciation, immigration,
and extinction at local (community) and global
(metacommunity) scales. Borrowing ideas origi-
nally and independently discovered by population
geneticists (Hubbell 2001, Chave 2004), neutral

theory can generate expected distributions of rel-

ative species abundances in local communities

very similar to the pattern of curves observed for

natural forest communities (Figure 7.1). Another

important aspect of neutral theory is the dis-

tinction between the local community and the

metacommunity (or regional species pool) and

the degree to which dispersal limitation allows

the local extinction of species (i.e., by prevent-

ing a species’ presence in a local community to

be “rescued” by dispersal from outside the local

community). One resulting contribution of neu-

tral theory is that it makes clear that the logseries

distribution (Fisheret al. 1943), which in contrast

to neutral theory predicts a greater number of rare

species relative to common species in the com-

munity, actually applies to the metacommunity,

not the measured, local community. Further, neu-

tral theory results in the replacement of Preston’s

(1948, 1962) symmetrical lognormal distribution

with a type of multinomial distribution, thereby

explaining the asymmetrical shape of the dis-

tribution of relative species abundances in local

communities (Figure 7.2). Forest dynamics plots

are relatively large and sample a large portion of

the “local” community, and as a result are able to

reveal the true shape of the distribution of rela-

tive species abundances.The FDP results appeared

to confirm one of the key predictions of neutral