Tropical Forest Community Ecology

Response of Seedlings and AMF Spores to Vertebrates 301

Table 17.2 Logistic regression analysis of seedling survival versus conspecific density and total seedling
densitySat the scale of 1 m×4 m subplots within exclosure and open experimental plots.

Model of

seedling

survival

versus

Treatment Model fit Parameter estimates

d.f. No. of

seedlings

χ^2 Probability Intercept Slope 1 SE of

slope

Conspecific Exclosure 1 3895 5.77 0.016 0.007 −0.002 0.001

density Open 1 2861 0.73 0.373 −0.313 −0.001 0.001

Total density Exclosure 1 3895 0.07 0.782 −0.045 0.0001 0.0001

Open 1 2861 3.25 0.073 −0.514 0.002 0.001

Note :Analysis was based on seedlings present in 1996 plus those recruiting between 1996 and 2000, with ultimate fate (survived
versus died) based on the 2001 seedling census.

Although strong density-dependent mortality
could slow the loss of uncommon species, we
found little evidence for these effects. For example,
logistic regression analysis of seedling survival
versus conspecific density and total density at the
scale of 1 m×4 m plots showed relatively small
but significant effects of conspecific density on
survival in exclosure plots but not on open plots
(Table 17.2). Instead, the higher seedling mortal-
ity on open plots apparently led to a more rapid
rate of species extinction.
Taken together, the effects of terrestrial verte-
brates on the dynamics of seedling colonization
and extinction can be summarized as no signifi-
cant increase in the rate of species accumulation
due to an increased species pool, combined with
increased seedling mortality that was not strongly
density dependent. The result was a lower net
rate of species accumulation on plots open to
terrestrial vertebrates (Figure 17.4).
In contrast, the significantly higher abundance
and richness of AMF spores on open plots sug-
gests that vertebrates acted to enhance the species
pool available on these plots and thereby increase
species accumulation. AMF spores have been
shown to pass through the guts of many of the
species on our study plot in viable form (all of
the rodents, cassowaries, bandicoots, and musky
rat kangaroo; Reddellet al. 1997). Consistent
with this hypothesis, we found that AMF species
that produced spores in macroscopic sporocarps
(and were therefore more likely to be attractive
to mycophagous vertebrates) were more likely

to occur on plots open to terrestrial vertebrates

(Gehringet al. 2002). For example, spores of

Glomus rubiformewere found in all open plots

but in only half of the vertebrate exclosures,

suggesting that these spores were more widely

distributed in the presence of vertebrates. Lacking

in the case of spores, however, is any indica-

tion that terrestrial vertebrates act as predators

that would reduce spore abundance and thereby

reduce colonization of plant roots. While verte-

brates might indirectly negatively impact AMF

through either herbivory, changes in soil nutri-

ent status, or altered host plant abundance and/or

species composition, we found no evidence of this

(Gehringetal. 2002). As a result, in contrast to the

effects of terrestrial vertebrates on seedling extinc-

tion, we hypothesize that terrestrial vertebrates

had little impact on local extinction of species of

AMF. As a result, we hypothesize that terrestrial

vertebrates enhanced the species pool available to

open plots through spore dispersal and thereby

increased the rate of species accumulation while

having little impact on spore mortality and species

extinction (Figure 17.5).

IMPLICATIONS FOR THE ROLE OF

TERRESTRIAL VERTEBRATES IN

TROPICAL SYSTEMS

The results outlined above suggest that commu-

nity response to terrestrial vertebrates will be

complex, depending to a large extent on the