Tropical Forest Community Ecology

Patterns of Herbivory and Defense in Tropical Forests 75

Lonchocarpus

Leaf area eaten (%)

0

2

4

6

8

10

12

14

16

(^18) Dry forest
Rain forest
Cordia
Erythroxylum
Trichilia
SapindaceaeApocynaceaeEuphorbiaceae
Piper
Figure 5.4 Levels of herbivory in pairs of taxa comparin gTDF (black bars) and TRF ( gray bars) species within the
same genus or within the same family. Herbivory data are derived from the measurement of a bulk sample of
50 randomly collected leaves from each of three plants of each of the species.
of them the difference is negligible, and the com-
parison between the two species ofLonchocarpusis
slightly higher in the rain forest species. An over-
all paired comparison indicated that differences
are statistically significant (signed-rank test=15,
P=0.04). A more controlled study and one with
a larger number of comparisons, such as that
performed by Fineet al. (2004), should be useful
to assess the relevance of phylogenetic inertia in
studies of water availability/phenology related to
herbivory.
Due to their scarcity, the available data are
far from satisfactory for teasin gapart the rel-
ative importance of phylogeny in the patterns
we uncovered. Nevertheless, it is of interest that
the limited information points in the direction of
water availability/phenology having an important
role for our predicted patterns, independent of, or
in addition to, phylogenetic inertia.

CONCLUSIONS

Some studies of ecogeographic variation in plant
attributes have looked for patterns of variation

and then assessed the possible mechanisms

responsible for such patterns. Our approach was

to develop a line of reasonin gas to why rainfall

seasonality and phenology of TRF and TDF plants

brin gabout fundamental differences in the avail-

ability of foliage for folivores and how these dif-

ferences may in turn lead to predictable patterns

of herbivory and defense. We then confronted the

predictions with the available data and found a

strong suggestionthathigherconstancyof foliage,

implyin g greater risk and impact of herbivory

in TRF plants in ecological time, may lead to a

greaterevolutionaryhistoryof herbivory,favoring

greater selection for increased defense and lower

herbivory. The predicted patterns were evident

when we controlled for interspecific heterogene-

ity in herbivory and defense within both TDF

and TRF. In addition, the expected patterns were

mirrored usin gmore controlled intra-site (dry

forest) comparisons lookin gat plants of contrast-

in gphenolo gies. Moreover, patterns seem to hold

independently of, or in addition to, phylogenetic

influence.

As is well known, local and ecogeographic pat-

terns of herbivory are influenced not only by