Tropical Forest Community Ecology

328 Gregory H. Adler

1.75

1.50

1.25

Island 4

Island 8

Island 14

Island 52

0.85 0.95 1.05 1.15

Log density of fruiting trees

1.25 1.35

Log density of spiny rats

1.00

0.75

0.50

Figure 19.1 Relationship between mean annual spiny rat density and mean annual density of fruiting trees and
lianas (from Adler and Beatty 1997).

reproduction–density space, that is, the statistical
space resulting when a measure of reproduc-
tive output is plotted versus population density
(Adler and Beatty 1997). Thus, plots of repro-
ductive output, expressed at the individual level
as births per adult female or at the population
level as length of the breeding season, versus den-
sity demonstrate clockwise trajectories through
reproduction–density space. The clockwise tra-
jectories conform to theoretical predictions and
strongly implicate intrinsic regulation of pop-
ulation density by density-dependent changes
in reproductive output (Schaffer and Tamarin
1973). Thus, reproductive effort was adjusted
with changes in density, apparently as individuals
attempted to maximize their fitness in a seasonal
environment with attendant changes in resource
abundance.
Reproductive output therefore is influenced
both directly and indirectly by climatic seasonal-
ity and resource abundance. The role of resource
abundance may be summarized in a simple flow
diagram (Figure 19.2). According to this sce-
nario, climatic seasonality influences the timing
of plant reproduction and therefore community-
wide fruiting phenology and seasonal fruit and
seed abundance. In turn, resource abundance
determines to a great extent annual and longer-
term changes in the demography ofProechimys
semispinosus. Thus, resource abundance influ-
ences density, which drives reproductive effort

(albeit with a time lag because reproductive

responses to changing density and resource abun-

dance are not instantaneous), and reproductive

effort feeds back directly onto density by either

increased effort and more recruitment when den-

sitiesareloworreducedeffortandlessrecruitment

whendensitiesarehigh.Resourceabundancealso

maydirectlyinfluencereproductiveeffortirrespec-

tive of density when per capita resource abun-

dance is insufficient to support the physiological

demands of reproduction.

Experimental tests of resource

limitation

To test the role of food resources in limiting

populations of Proechimys semispinosus, I con-

ducted two food-provisioning experiments. The

first experiment tested the hypothesis that insular

populations that are largely released from top-

down limitation are limited by food resources even

during the season of resource abundance (Adler

1998). The second experiment tested the hypoth-

esis that such populations are limited during the

season of resource scarcity (Adler unpublished

data). By using small islands as experimental

replicates, we could control for the “commuter

effect,” whereby individuals living near the exper-

imental area temporarily increase density in that

area (Adler 1998). This effect is common in