suiting in highly clumped seed distributions away
from the parent plant. During the breeding season,
male bellbirds spend 80% of the day calling from one
or a few display perches, making short foraging trips
to nearby fruiting trees (Snow 1977). Most of the
seeds they carry are deposited at very high density
directly below the display perch, although far from
the parent plant. To determine the consequences of
such behavior for seed mortality, Wheelwright
(1988a) scattered seeds of Ocotea tonduzii (ca. 20/
m^2 ) at 5-day intervals beneath three bellbird perches
and at randomly chosen sites about 20m from each
of the three perches. Removal (and, presumably,
mortality) of seeds beneath perches was 100% in
every trial, presumably because seed-eating mam-
mals were visiting the areas beneath the perches
regularly in anticipation of the seeds dropped by
displaying male bellbirds. Seeds only 20 m away
went undiscovered by seed predators until the fourth
trial, after which 100% were removed by animals
who had habituated to the artificially created seed
cache. Areas beneath the display perches of male
Long-tailed Manakins also collect large numbers of
seeds during the breeding season (Wheelwright
1988a, N. Wheelwright and D. McDonald, unpubl.
data). The spatial distribution of seeds dispersed by
nonbreeding individuals of these bird species has not
been measured, but they are probably spread more
widely, and thus have a greater chance of escape
from predators.
Only two studies have directly measured the pre-
dation rates suffered by large seeds at the lower den-
sities typically produced by most dispersers at
Monteverde. To understand the benefits of dispersal
via both reduced seed density and distance from seed
source, seeds of Nectandra davidsoniana (Laura-
ceae) were placed at different densities (clumped vs.
dispersed) and at distances from 0 to more than 30
m from conspecific fruiting trees in forested tracts
within the Monteverde community. Contrary to ex-
pectation, no significant density or distance effects
were observed because virtually all of the 520 seeds
were removed within 24 hr (chiefly by the Spiny
Pocket Mouse), and none survived more than 4 days
(Wheelwright 1988a).
A study of rodent predation documented that ef-
fects can be severe even on dispersed, low-density
seeds in some species (see Wenny, "What Happens,"
pp. 286—287). Dispersal benefits most large-seeded
species by removing them from zones of predictably
high predation beneath fruiting trees. However, the
effects of consumption by frugivores can vary widely
with disperser species, habitat, and season. Research
is needed to understand the fitness consequences of
dispersal for large-seeded plants.
Colonization of patchily distributed habitats by pioneer
plants. Howe and Smallwood's (1982) "colonization
hypothesis" provided the framework for studies on
three species of pioneer plants in the MCFP (Murray
1986a, b, 1988). Phytolacca rivinoides (Phytolac-
caceae), Witheringia meiantha, and W. coccoloboides
(Solanaceae; Fig. 8.6) establish only in the high light
environment of recently formed canopy gaps. Because
such gaps occur at Monteverde at a rate of only 1.5%
of land area per year, colonization sites for pioneers
are rare and spatially unpredictable. To understand
the consequences of dispersal by different species of
birds for plant fitness, Murray (1988) compared the
seed shadows produced by birds with the spatial
and temporal distributions of suitable germination
sites. The probability of germination varies as a func-
tion of gap size and age in the three plant species.
Phytolacca rivinoides, for example, requires larger,
younger gaps than does W. meiantha to stimulate
germination.
The three most important dispersers of the plants
(Black-faced Solitaires, Black-and-yellow Silky Fly-Figure 8.6. Fruits of Witheringia coccoloboides. Photo-
graph by Nathaniel Wheelwright.264 Plant-Animal Interactions