To understand the relationship between seasonal
patterns of solitaire abundance and fruit removal
rates, I examined the correlation between solitaire
capture rate and daily fruit removal rate from May
1982 to April 1983. Daily fruit removal rate (propor-
tion per day, arcsin-transformed) is correlated with
the size of the fruit crop (log-transformed; Murray
1987). I computed partial correlations of removal rate
with the monthly solitaire capture rate, controlling for
crop size. In P. rivinoides, daily fruit removal rate was
correlated with solitaire capture rate (r = 0.635, 8 df,
p < 0.05), which suggests the potential for lower re-
productive success of individuals ripening their fruits
when solitaires were absent. The partial correlations
for W. meiantha and W. coccoloboides indicated the
same relationship but were not statistically signifi-
cant. Thus, disperser seasonality might act as a selec-
tive agent on fruiting phenology but not a strong one;
removal rates were only slightly depressed during the
months when solitaires were entirely absent from the
Monteverde area.
Because seeds of the three plant species germinate
well only in very young gaps (<8 months), and be-
cause gap formation itself is highly seasonal (Murray
1988), there should be a selective advantage for seed
germination schedules that coincide with periods of
peak gap formation (the "windy-misty" season from
October through January; Murray 1988, K. G. Murray,
unpubl. data). All three species display "enforced"
seed dormancy (sensu Harper 1977); seeds remain
dormant on or in the soil until a gap forms overhead
if they are initially deposited beneath an intact
canopy (Murray 1988). This characteristic effectively
decouples germination from dispersal, so the exact
timing of fruit ripening might be relatively unimpor-
tant in these plants.
To estimate the consequences of fruiting phenol-
ogy for the likelihood of colonizing recent gaps, I used
a computer simulation model that combined empiri-
cal data on rates of gap formation, germination re-
quirements, seed dormancy, and bird-generated seed
shadows (Murray 1986a, b, 1988) to yield an estimate
of "lifetime reproductive potential" (number of off-
spring potentially produced during an individual's
lifetime). By shifting the observed phenologies (Fig.
8.21) so that ripening peaked at different times of year,
I inferred the consequences of ripening phenology
with all other variables held constant. For all runs, I
used the estimated seed shadow produced by Black-
faced Solitaires (Fig. 8 in Murray 1988), a dormancy
capability of 24 months (seeds of all three species can
remain viable in the soil for at least 27 months with
no loss of viability; Fig. 3 in Murray 1988), and the
empirically measured germination response to gaps
of different ages. These estimates of reproductive
potential were similar for plants fruiting at different
times of year, primarily because of the decoupling
of seed germination from dispersal.
A confounding factor is seed predation and parasit-
ism. The physiological ability to wait for a gap to oc-
cur overhead is irrelevant if seeds are eaten soon after
dispersal. Although all three species are taken by preda-
tors, removal rates are not so high that the temporal
decoupling of dispersal and germination is negated
(K. G. Murray, unpubl. data). Rather, it is likely that
fruiting seasons that coincide with peak gap formation
would carry a greater advantage than implied by the
results of the simulations described above.
The fact that none of these factors can explain the
seasonal patterns of fruiting in P. rivinoides, W. mei-
antha, and W. coccoloboidesbegs other explanations.
Strongly seasonal patterns of fruit ripening mightFigure 8.22. Black-faced solitaire
capture rates and breeding activity
from July 1981 to June 1983. Bars
at bottom of figure indicate the
presence of breeding adults (i.e.,
with brood patches) and young of
the year on study plots (lower bar).285 Plant-Animal Interactions