up to an hour, and some toucanets stayed in the same
tree for up to 4 hr, but such cases were rare. Visit
lengths tended to be longer in large-fruited tree spe-
cies, perhaps because larger fruits required longer
processing times.
Murray (1987) found that the proportion of plants
receiving at least some visits from frugivorous birds
on a given day increased with increasing crop size
in three species of pioneer plants (Phytolacca rivi-
noides, Witheringia meiantha [previously named W.
solanacea], and W. coccoloboides). The absolute num-
ber of fruits removed per day increased over the whole
range of crop sizes in all three species, but the pro-
portion removed decreased (i.e., removal from plants
with large fruit crops did not increase in direct pro-
portion to crop size). There was no indication of a
peak in visitation or removal rate at intermediate crop
sizes, and because the "waste" of uneaten fruits from
individuals with large fruit crops is unlikely to result
in lower survival of adult plants in short-lived pio-
neers, the higher visitation and number of seeds dis-
persed at high crop sizes implied selection for syn-
chronous fruit ripening (Murray 1987). Studies in
Monteverde thus support the general prediction that
fruit crop size influences the behavior of dispersers
in a manner beneficial to plants although they pro-
vide only equivocal support in some particulars.
Chemical control of seed passage rates Plants
may enhance their reproductive success by manipu-
lating the treatment that their seeds receive in dispers-
ers' guts. Using artificial fruits made with agar, sugar,
and natural seeds and pulp extracts from Witheringia
meiantha fruits, Murray et al. (1994) showed that
some unidentified chemical(s) in the pulp increases
the rate at which seeds pass through the gut of its
major disperser, the Black-faced Solitaire. This chemi-
cal may serve to balance the consequences of seed
passage through solitaire guts: rapidly passed seeds
emerged in viable condition more frequently than did
those spending longer periods of time in the gut, but
they were also deposited nearer to the parent plant.
Shorter dispersal distances are disadvantageous for
pioneer plants because the probability of encounter-
ing a gap increases with dispersal distance (Murray
1988). This study was the first empirical evidence of
a "laxative" agent in a wild fruit, other than high fiber
content. Results might be explained by slight differ-
ences in the sugar concentrations of the experimen-
tal fruits rather than to laxative chemicals (Witmer
1996), but other data indicate that differences in sugar
concentration are too small to produce the observed
differences in seed passage rates (K. G. Murray, un-
publ. data). Recent work by Wahaj et al. (1998) has
shown the existence of a laxative chemical in the
fruits of a related species (Solanum americanum), but
the chemical itself was not identified. More research
is needed to identify the existence of laxative chemi-
cals in Witheringia and other species, and to eluci-
date their roles in dispersal ecology.Spatial and temporal patterns in plant-frugivore inter-
actions at Monteverde. Plant-frugivore interactions
vary in space and time; studies in Monteverde have
focused on the causes and consequences of this
variation.
Fruiting phenologies Most plant species at Monte-
verde have distinct fruiting seasons, but in a few cases
individual plants are synchronous within crowns but
highly asynchronous as a population (Wheelwright
1985a, Murray 1986a; see Bronstein, "Fig Pollina-
tion," pp. 271-273). In understory shrubs and treelets,
community-level fruiting (the number of species fruit-
ing, not the number of ripe fruits) is less distinctly sea-
sonal than is flowering. Flowering peaks during the
dry-wet season interface (April-June) each year,
whereas fruiting tends to decrease generally and peak
only weakly; small fruiting peaks occur at different
times each year. The lower seasonality of fruiting
may be because plants differ in the amount of time
between flowering and fruit ripening. For example,
flowering is highly synchronous in Meliosma subcor-
data (Sabiaceae) and most members of the Araliaceae.
Meliosma ripens fruits over most of the following
year, whereas most Araliaceae do so within a few
months of flowering (Koptur et al. 1988). Fruiting at
Monteverde is less seasonal than in the dry forests of
Guanacaste, but more seasonal than in wet forest at
La Selva, which may be related to precipitation. Drier
sites tend to have more seasonal flowering and fruit-
ing than wetter ones, and Monteverde is intermedi-
ate between Guanacaste and La Selva in annual rain-
fall (Frankie et al. 1974, Koptur et al. 1988).
Koptur et al. (1988) documented differences in
fruiting phenology between the two years of their
study. Heavy rains in the second year caused the flow-
ers of many species to rot; fewer species set fruit. Some
species of frugivores may have been adversely af-
fected, as were the fruit-eating birds and mammals of
Barro Colorado Island, Panama, following a weather-
related fruiting failure there (Foster 1982). Many ani-
mal seed dispersers migrate altitudinally in response
to food abundance, so understanding the periodicity
and severity of fruit failures at Monteverde is a high
research priority.
Wheelwright (see "A Hypothesis," pp. 281-282)
hypothesized that flowering phenologies in laurace-
ous trees may be affected by competition for pollina-
tion, whereas fruiting phenologies may be constrained
more by abiotic factors (e.g., rainfall). Similarly,
Murray (I986a; see Murray, "Fruiting Phenologies,"262 Plant-Animal Interactions