Tropical Forest Community Ecology

Response of Seedlings and AMF Spores to Vertebrates 303

− 10

− 8

− 6

− 4

− 2

0

2

4

6

8

10

Years since exclosure

Change in species number

Figure 17.5 Hypothesized relationship for accumulation of mycorrhizal spore species through time on open plots

(open squares) and exclosure plots (closed squares). Loss of spore species due to extinction caused by terrestrial

vertebrates is assumed to be zero for both open and exclosure plots (closed diamonds). Vertebrates increase the species

pool of spores arriving on open plots, resulting in the net rate of species accumulation being higher on those plots.

The tree community at our site had relatively
few species with seeds dispersed primarily by ter-
restrial vertebrates, and therefore exclusion of
those vertebrates had little overall effect on species
accumulation. Tree communities with a larger
proportion of species with seeds dispersed by ter-
restrial vertebrates will have a larger difference in
the species pool available to colonize a site, and
therefore loss of terrestrial vertebrates would be
predicted to have a greater effect. For example,
seeds of many tree species on our plots are dis-
persed by arboreal birds and bats, and we predict
that excluding input from these animals would
more strongly decrease the species pool available
to exclosure plots and thereby cause a greater
difference in species accumulation between treat-
ments.
Second, even in those communities with a large
number of species dependent on vertebrate disper-
sal, if seedling mortality due to vertebrates is high
enough, and not overall more strongly density
dependent than in their absence, the net effect of
vertebrates may still be to reduce local species rich-
ness due to increased extinction rates. Increased
local species extinction may be more likely when
terrestrial vertebrates act as random mortality

agents, destroying seeds and seedlings by uproot-

ing or trampling. On our plots, this mortality

often impacted wind- and arboreal bird-dispersed

species that produced small seedlings with shallow

root systems or those with large cotyledons that

made the seedling more susceptible to toppling

or breakage. Combined with vertebrate predation

on seeds and seedlings, overall mortality due to

vertebrates was thereby spread across almost all

seedling species on our site, resulting in higher

rates of species extinction. In communities where

more tree species have seeds or seedlings that are

directly preyed upon by vertebrates, the potential

for vertebrates to act in density-dependent ways

would be greater and the overall impact of verte-

brates on extinction rates may not be as great as it

was on our site. Even so, this suggests that several

seedling species in a community can show strong

density-dependent mortality in the presence of

vertebrates, but if many more species in the com-

munity experience higher, non-density-dependent

mortality, the net effect of vertebrates could still

be to reduce species richness.

Third, our results suggest that terrestrial ver-

tebrates increase local dispersal limitation by

reducing the probability that species arriving at