IN A NORTHERN FLORIDA
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
It is my pleasure to acknowledge the many friends and
colleagues who have aided me in this work.
The members of
Feinsinger, Jack Kaufmann, and Walter Judd, provided helpful
research and editorial suggestions.
this study is the result of a discus
Feinsinger's community ecology class.
I should mention that
sion question from Peter
I was also very
fortunate to have the assistance
e of Edmund W.
in the nutrient analysis of my fruits.
I have also drawn heavily from Stiles' work on bird-fruit
dispersal systems of North America.
To Ted I extend my
I also thank Doug White,
University, for providing me with invaluable manuscripts.
A number of
associates and graduate
students have helped me throughout my studies.
like to thank Peter G. May, Nathaniel T.
Wheelwright, and E.E.
Spears for their help and suggestions.
I also acknowledge the Department of Zoology and the
Chapman Memorial Fund for their generous
go to the Florida Division of Natural
Felasco Hammock is truly a Florida jewel; I hope that this
study adds to the understanding of this unique community.
I would like to separately thank my wife, Lea, for her
field help and companionship.
the hammock with her. It wil
It was a joy for me to share
.1 be a time and place we will
always remember and cherish.
I must also thank my canine companion,
the miles and mil
of walking, my wee westie was always
there beside me (except when he was off
I would like to thank everyone I've been
associated with at the Department of Zoology at UF for
making my six year stay in Florida a wonderful experience.
TABLE OF CONTENTS
OF FIGURES..... .................................vi
Temperate Bird-Fruit D
SEASONAL PATTERNS OF FRUGIVORE
AVAILABILITY AND FRUIT RIPENING..................9
Study Site and Methods...........................9
Results and Discussion..........................14
Conclusions ..................................... 56
THREE SEASONAL PATTERNS OF BIRD-DISPERSED
FRUIT CHARACTERISTICS AND FRUIT REMOVAL.........57
Results and Discussion..........................59
OF CALCULATING RIPENING SYNCHRONY.......92
FLOWERING AND FRUITING SCHEDULES OF BIRD-
DISPERSED PLANTS OF SAN FELASCO HAMMOCK.........94
III FRUIT TRAITS ANALYZED FOR BIRD-DISPERSED
PLANTS OF SAN FELASCO HAMMOCK.....................98
IV BIRD-DISPERSED PLANT SPECIES AND THE BIRDS THAT
TAB LES V
LIST OF TABLES
Fruit types of vascular plants found
in San Felasco Hammock State Preserve, Florida.........11
observed at San
Hammock during cenususes.......................15
Observed frugivory by major and
minor frugivores in
Resident status of avian frugivores
at San Felasco Hammock.................................19
Spearman rank correlation
for the relationship between frugivo
diversity and abundance and the number
Relationship of mammal-bird fruits
and bird-fruits to
seasonal fruiting patterns..........35
seasonal fruiting patterns for bird-dispersed
plants of San Felasco Hammock..
Seasonal relationships of ripening
characters for bird-dispersed plants of San
rg for ripening characters.......
fruit traits to seasonal
fruiting patterns of bird-dispersed plants of
San Felasco Hammock....................................61
LIST OF FIGURES
Number of frugivorous bird
observed along census route from April 1982
San Felasco Hammock................21
Number of frugivore individuals
through May 19
census route from April 1982
San Felasco Hammock................ 24
Standard deviation of frugivore
numbers for birds observed along
from April 1982 through May 1984
in San Felasco
Number of bird-dispersed plant
with ripe fruit in San Fela
Fall fruiting species.
of the initial total number of fruits removed
within a single week for Cornus foemina,
Callicarpa americana, and Aralia spinosa...............71
Fall-winter fruiting species.
of the initial total number of fruits removed
within a single week for Viburnum rufidulum,
Viburnum obovatum, Symplocos tinctoria,
Ilex decidua, Smilax bona-nox, and Smilax
Winter fruiting sp
of the initial total number of fruits removed
within a single week for
Abstract of Dissertation Presented to the Graduate
School of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy
MUTUALISTIC INTERACTIONS BETWEEN BIRDS AND FRUITS
IN A NORTHERN FLORIDA HAMMOCK COMMUNITY
Stewart T. Skeate
John W. Hardy
The ecological relationships between frugivorous birds
and bird-dispersed plants have been unexplored at lover
latitudes in North America.
This paper examines
bird species and 45 species of bird-
dispersed plants in a northern Florida hardwood hammock
community from January
1982 to May
of frugivore availability, fruiting phenology, and fruit
characteristics were determined and compared.
Frugivore species diversity and abundance were highest
during the fall and winter months and lowest in the spring
in frugivore numbers and diversity
attributed to the presence of migrant thrushes in the
Spring and summer months showed the lowest
number of fruiting species and fall and winter the highest,
with a peak of 28 species in fruit in December.
Fruiting patterns of plant species fell into four
summer fruiting, fall fruiting, fall-winter
and winter fruiting.
Nine species showed summer
bearing ripe fruit between early spring and late
Summer fruits were
characterized by high water and
large fruit mass, and low persistence.
The 12 fall fruiting species ripened fruit at the peak of
frugivore migration through northern Florida.
species were common or abundant in the community and
species with high lipid fruits.
The 20 fall-
winter fruiting species produced highly persistent fruits
that were available to migrant frugivores through fall and
These species were typically
evergreen and included
species with high lipid fruits.
The 4 winter fruiting
species matured fruit in December and depended on
overwintering birds for seed dispersal.
The bird-fruit dispersal system of the lower temperate
latitude hammock community shows similarity
patterns and fruit characteristics to middle latitudes of
the presence of evergreen
The importance of birds
many plant speci
seed dispersal agents for
has long been recognized (Grinnell 1897,
1930, McAtee 1947, Krefting and Roe
only within the last ten years have
researchers examined in depth the relationships between
frugivorous birds and bird-dispersed plants.
of bird-fruit interactions typically involved species
records of birds that were observed visiting fruiting trees
(Petrides 1942, Sutton 1951, Land 1963, Willis 1966, Leck
and did not discuss possible mechanisms of
Snow (1971) and McKey (1975)
first speculated on possible
lutionary trends between
fruit-eating birds and plants,
and presented a theoretical
framework of bird-fruit relationships for future workers to
Bird-fruit interactions have also been of considerable
interest to botanists from evolutionary,
Botanists agree that zoochorous
seed dispersal has an ancient evolutionary history, probably
n'r~i ainat*i nc in -t-rnni 1 anv4~ranmnn-t-a Flrt a vV n
edible portion, an outer protection against
an inner protection of the seed against digestion,
colors that signal maturity,
lack of odor, permanent
attachment, absence of hard rind, and seeds exposed or
dangling in hard fruits.
These adaptations are highly
variable among bird-dispersed plants.
surprising considering the widespread occurrence both
geographically and taxonomically of fruits adapted for bird
Wheelwright et al.
Recent field studies have examined various aspects of
including interactions between
individual plant species and their dispersers
1981, McDiarmid et al. 1977, Howe and DeSteven 1979,
Howe and Vande Kerckhove 1980,
1981, Herrera 1981b, Herrera
and Jordano 1981); nutrient quality of fruit (Foster
Stiles 1980, Sorensen 1981, Herrera 1981a,
White and Stiles unpub.
importance of fruit
color and morphology to dispersal
Turcek 1963, Stil
Willson and Thompson 198
Denslow and Moermond 198
Morden-Moore and Willson 198
, Janson 1983, Willson and Melampy
, Moermond and Dens low 1984)
to succession (Smith 197
Willson 1978, Debussche et al. 1982, McDonnell and Stiles
- a -
LI-- 4..~. It nnh
of bird-dispersed plants (Snow
Thompson and Willson
1979, Herrera 1982b,
have spawned a "second generation" of
on the coevolution of seed dispersal systems
Wheelwright and Orians 198
These papers have contributed
to the theoretical framework of bird-fruit research, but
have also questioned existing the
Wheelwright and Orians 1982).
The field of
other related fields such
in its infan
cy and lags behind
as pollination biology.
more detailed field studies are needed in both tropical and
Temperate Bird-Fruit Dispersal Systems
The study of coevolved relationships between bird-
dispersed plants and frugivorous birds in temperate regions
has received considerable attention since Snow (1971) first
discussed possible temperate
e fruiting patterns (Thompson and
1980, Herrera 1981a,b
1984a,b, Herrera and Jordano
1981, Stapanian 1982, Johnson et
relationship between the time of fruiting by bird-dispersed
patterns of frugivore presence may be the primary selecting
force influencing mid-latitude fruiting patterns in North
They recognized three mid-latitude fruiting
summer, fall, and winter fruiting,
in which the
respective fruiting patterns showed distinct adaptations to
differing levels of frugivore availability.
was the most common pattern,
coinciding with the peak of
avian frugivore migration.
To date, field studies of temperate bird-fruit
dispersal systems in North America have been made at middle
north of the usual wintering grounds of most
major avian frugivores (Sherburne 1972
, Thompson and Willson
1979, Baird 1980, Johnson et al.
between frugivorous birds and bird-dispersed plants at
temperate latitudes in North America has been unexplored.
Lower temperate latitudes differ from higher
their bird-dispersed flora and in seasonal diversity and
abundance of frugivorous birds.
This suggests that
differences in fruiting patterns may be evident between
middle and lower temperate latitudes in North America.
Thompson and Willson (1979) suggest that
both summer and
winter fruiting patterns should be more profitable at
temperate latitudes due to the greater availability of
The majority of bird-fruit studies have involved
dispersal strategies of individual plant species (e.g. Howe
1981, McDiarmid et al. 1977, Howe and Vande Kerchove
1980, 1981, Herrera and Jordano 1981, Jordano 1982).
dealing with single plant species may mis
factors in the evolution or ecology of bird-fruit systems,
factors that may be evident only through research at the
In this study
I examine the bird-fruit system,
involving 45 species of bird-dispersed plants and
frugivorous bird species,
hammock forest community.
in a northern Florida hardwood
Four seasonal fruiting patterns
are analyzed in the context of their relationship to
frugivore availability, ripening synchrony, fruit abundance,
habitat, and evergreenness.
I also compare the bird-fruit
systems of middle and lower temperate latitudes of North
America in an attempt to derive a more complete picture of
in eastern North America.
Nutritional Properties of Temperate Fruits
The nutritional composition of fruits has been a major
focal point in the study of temperate zone bird-fruit
dispersal systems (Sherburne 1972, Stiles 1980,
content and that this
ripening patterns. H
is roughly related to seasonal
igh reward fruits are produced during
peak periods of frugivore availability, and the nutritional
properties of the fruits
correlated with the seasonally
changing demands of their major dispersers
White and Stiles unpub.
Bird-dispersed plants are therefore under some
selection pressure to produce fruits that are nutritionally
attractive to frugivorous birds.
By inducing birds to eat
their fruits and then to regurgitate or defeat
plants obtain an effective method of
seed dispersal away from the parent plant.
may result in lower seed and seedling mortality (Janzen
1970, Hove and Primack 1975), reduced seedling competition
(Connell 1971, Harper
1977), increased gene flow (Levin and
to new habitats
The nutritional reward offered by
a plant in its fruit may
also be related to the quality of
dispersal by birds
, Howe and Estabrook 1977, Howe and
Vande Kerckhove 1980,
Wheelwright and Orians 1982).
The actual nutritional reward in a fruit
by its dry mass content of carbohyd
The latter two elements, howe
appear to be the
in addition to the
(Herrera 1982a) of
a fruit may be
important to a fruit-eating bird and may affect rates of
fruit removal and digestive processing (White and Stiles,
fruit traits that may be important
to birds include pulp moisture (Herrera 1982a), seed size
and number (Sorensen 1984),
1983), fruit mass (Moermond and Denslow
diameter (Wheelwright 1985
, presence of secondary compounds
(Herrera 1982b), and fruit color (Turcek 1963, Morden-Moore
and Willson 1982, Willson and Thompson 1982, Willson and
the attractiveness and
profitability of any fruit to a bird is directly related to
the combination of
chemical and structural components of the
The design and chemistry of fruits, by themselves,
reveal relatively little about the evolution of bird-fruit
Only when this information is examined
in the context of other community parameters can
evolutionary patterns be discerned.
Here, I report on fruit
of 43 species of bird-dispersed plants in a
northern Florida hammock forest community.
characteristics analyzed are fruit mass. seed mass.
nutritional standpoint and have been analysed in other
studies (Stiles 1980, Sorensen 1981. Herrera 1982a, White
and Stiles unpub.
I discuss the relationship of the
nutritional content of fruits to seasonal ripening,
removal rates, frugivore availability, and fruit
I also compare these result
s with the fruiting
patterns described by Stiles (1980) and White and Stiles
ms.) for bird-dispersed plants of
These studies have developed a framework for the
study of bird-fruit dispersal systems for eastern North
America and allow comparisons of fruit characteristics
between middle and lower
SEASONAL PATTERNS OF FRUGIVORE
AVAILABILITY AND FRUIT RIPENING
Study Site and Methods
San Felasco Hammock State Preserve (29045'N, 8230'W),
in Alachua Co., Florida,
covers 2306 hectares and includes
a variety of plant communities, including the largest
protected stand of climax mesic hammock in the state of
Florida (Dunn 1982).
This area is characterized by moderate
winters (15-20C average temperatures) and hot summers
(300C average temperature).
Rainfall is variable in its
seasonal distribution, although more than half the annual
rainfall (average 1370 mm) occurs between June and September
This study focused on the mesic hammock community,
which consists of 129h hectares of rolling hills, plateaus,
and stream valleys within San Felasco Hammock.
mesic hammock refers to
a mixed hardwood forest situated
within a region where the predominant vegetation
marsh, or pine forest (Harper 1905).
community may be found in Laessle (1958),
Delcourt and Delcourt (1977),
while analysis of the plant
communities and flora of San Felasco Hammock may be found in
Ansley (1952) and Dunn (1982).
Dominant canopy trees within
this community include Quercus
tomemtosa, Magnolia grandiflora, Persea
borbonia, and Liquidamber styraciflua.
The mixed deciduous-
evergreen nature of this community is especially evident
during the winter months,
when the forest becomes a mosai
of evergreen and leafless broad-leaved trees.
bird-dispersed fruits make up approximately one-half of the
vascular flora of San Felasco's mesic han
They are especially evident among shrubs,
mock (Table 2-1).
I censused avian frugivore abundance and species
diversity from January
1982 to May
I walked three separate
, one-half hour after s
1984 in the mesic hammock
700 m line transects
sunrise, during which all
or heard for 40 m to
side of the transect
line were recorded.
I determined 40 m to be an effective
maximum distance for all
, particularly in the
Fruit types of vascular plants found in San
Felasco Hammock State Preserve, Alachua Co., Florida.
Life Form Fleshy Fruits (%) Dry Fruits (%) Total Spp
Canopy Trees 4 (25) 12 (75) 16
Understory Trees 10 (67) 5 (33) 15
Shrubs 13 (93) 1 (7) 14
Vines 10 (63) 6 (37) 16
Herbs 4 (21) 15 (79) 19
Total Species 41 (51) 39 (49) 80
the birds during the summer months permitted an accurate
census of their numbers.
I combined data from the three
transects to obtain a single daily count.
Mean values of
four daily transects over a two-week period were calculated,
resulting in two values per month for abundance and
determined to be
observation during bird and fruit censusing.
that swallowed entire fruits and that voided seeds intact
The term frugivore refers
only to birds that disperse
1971, Morton 1973).
Birds that chewed fruits,
swallowing only the fruit pulp and
below the parent
plant, were considered
they did not disperse
plant (Howe and Estabrook 1977).
seeds away from the
major frugivores from minor frugivores by the quantity of
the number of plant
visited, and the
observed frequency of each species during weekly censuses.
I collected phonological data on 4
in a 332
of mesic hammock by
monitoring ten individual plants per
of first ripe fruit and disappearance of all fruit for
I collected fruit ripening data for 27
fruits and 250 plants) in 198
fruits and 180 plants) in 1983-84.
Weekly counts of ripe
and unripe fruits were taken from tagged branches of ten
plants per species when possible.
I tagged approximately
equal numbers of fruits for each individual of a particular
The number of ripe fruits was corrected for fruit
removed; I assumed that removed fruit were ripe.
, I made fruit crop counts for each individual
by direct count or extrapolation from fruit on a few
branches and the total number of branches on the tree.
I analyzed ripening data by four measures: 1) between-
the number of days for all marked
individuals to show 90% ripe fruit
2) population ripening
number of days for the total fruit crop to reach 90%
individual plant ripening
the mean number
days for the marked individual plants of each species to
reach 90% ripe fruit;
4) population synchrony index
determined from a composite measure of individual synchrony.
Individual synchrony measures the overlap of a given
(90% ripe fruit
with those of
This measure of
population ripening synchrony
is a modification of Augspurger's method
I measured the abundance of the bird-dispersed plant
using a belt transect technique.
number of individual plants
I counted the
of each species within
transects measuring 100 m X 10 m along a
through the study area.
km line transect
Only individuals judged to be
capable of fruit production were counted.
were then categorized as abundant,
, based on their censused frequency,
and observed percentage of individual
s fruiting within the
In addition to fruiting records, I recorded flowering
dates for the 45 bird-dispersed species by observation of
plants within the study area.
The flowering to fruiting
duration was the time between the date of first flowering
and the appearance of bhe first ripe fruit for each species.
Results and Discussion
The frugivorous birds I observed in San Felasco Hammock
Avian frugivores observed at San Felasco
Hammock during censuses.
Major frugivores are marked
Red-bellied Woodpecker (Melanerpes carolinus)*
Northern Flicker (Colapt
Yellow-bellied Sapsucker (Sphyrapicus various *
Pileated Woodpecker (Drycocopus pileatus)
Acadian Flycatcher (Empidonax virescens)
Eastern Phoebe (Sayornis phoebe)
Brown Thrasher (Toxostoma rufum)
Gray Catbird (Dumetella carolinensis)*
American Robin (Turdus migratorius)*
Veery (Catharus fuscescens)*
Hermit Thrush (Catharus guttatus)*
Swainson's Thrush (Catharus ustulatus)*
Gray-cheeked Thrush (Catharus minimus)*
Cedar Waxwing (Bombycilla cedrorum)*
Red-eyed Vireo (Vireo olivaceous)
White-eyed Vireo (Vireo griseus)
Solitary Vireo (Vireo solitarius)
Yellow-throated Vireo (Vireo flavifrons)
Yellow-rumped Warbler (Dendroica coronata)*
by these birds and by their observed frequency during
the weekly cenuses
Thus the Eastern Bluebird
while showing a relatively high number of
was not considered a major frugivore in the
community as it was observed only twice in the study area
during the two-year censusing period.
accounted for 96.9% of the total observations of frugivory
(one observation=one feeding bout) and 97.4% of the total
visits (one visit=one fruit
Fruit thieves, birds that ate only the pulp of the
fruit, dropping the seeds, included the Northern Cardinal
Summer Tanager (Piranga rubra),
Rose-breasted Grosbeak (Pheucticus
American Robins (Turdus migratorius) showed the highest
number of fruit visitations among the major frugivores.
Wintering robins were responsible for 57.2% of the total
while Cedar Waxwings (Bombycilla cedrorum) accounted
19.8% of the total
are derived from the total number of visits over a two year
period and may not reflect the importance of any
frugivore at one time of the year.
while accounting for only
7.6% of the
were the most important frugivores within the
community in the early fall orior to the arrival of
Observed frugivory by major and minor frugivores
in San Felas
Visits (%) Spec
The number of frugivores in residence during each
shown in Table 2-4.
Only two major frugivores,
the Red-bellied Woodpecker (Melanerpes carolinus) and the
Wood Thrush (Hylocichla mustelina),
were present in San
Felasco Hammock during the summer.
rare summer resident.
The Wood Thrush is a
The number of major frugivore species
increased in the fall and winter;
and five were winter residents.
four were fall transients
Migrants played a dominant
role within the frugivore guild of thi
constituted 82% of the total number of frugivore species and
92% of the major frugivore species.
responsible for 98.2% of the total
plants, involving 30 plant species.
visits to fruiting
Eastern Bluebird (Sialia sialis) and Northern Flicker
while year-round residents in northern
appeared in the hammock community only during the
fall and winter and were classified
Frugivore Diversity and Abundance
Frugivore and major frugivore species diversity,
remained low throughout the summer months,
increased in late
thing a peak in mid-October in both
This increase was due primarily to the
TABLE 2-4. Resident status of avian frugivores at San
Felasco Hammock, Florida.
Year-round Summer Winter Fall
Residents Residents Residents Transients Total
Frugivores 4 4 10 4 22
Frugivores 1 1 6 4 12
remained in San
Hammock through October and departed at
the end of the month.
At this time,
Yellow-bellied Sapsuckers (Sphyrapicus various ,
and Yellow-rumped Warblers (Dendroica coronata)
arrived in the area.
high throughout the winter,
decreasing from late March to
when the wintering frugivores left the hammock.
Avian migration through San Felasco Hammock during the
spring months was much
pronounced than fall migration,
and frugivore diversity increased only slightly in the
Frugivore abundance showed a pattern similar to that
of frugivore species diversity: the number of frugivore
individuals remained low during the summer months,
, and reached a first peak in mid-October of
A second peak was evident in both
years, in mid-November in 1982 and mid-December in 1983.
The first pulse of frugivore migration was due primarily to
the arrival of migrant thrushes,
while the second was due to
the arrival of wintering robins, Cedar Waxwings, Hermit
, and Yellow-rumped Warblers.
In 1982, robins
arrived in large numbers in early November,
while in 1983,
flocks of robins were not observed until mid-
0 (U *H
$4 :,.$.4 0)
$4 04 -p 4)
^ ^ ^
increased in early
values in the late fall and
while the summer months showed the lowest
The high standard deviations evident in the late
fall and winter months reflect the unpredictable nature of
as a res
ource for fruiting plants
the hammock community (Stapanian
There was no
significant difference in major frugivore
Wilcoxon's Matched Pairs Signed-Ranks
P>0.05) or frugivore
Matched Pairs Signed-Ranks Test,
There was also no significant difference in major
frugivore abundance (Wilcoxon's Matched Pairs Signed-Ranks
P>0.05) or frugivore abundance (Wilcoxon's
P>0.05) between the two
a significant difference in major
frugivore and frugivore abundance from September through
March for the two
s Matched Pairs Signed-
due to the higher number of frugivo
observed in the fall and winter of
compared to 1
due to the small fruit
crop of Cornus florida in
robins to the study
the mean flock
to available fruit crops (Speirs 1946,
Thompson and Willson
The number of plant sp
with ripe fruits from April
1982 through March 1984 is shown in Fig. 2-4.
number of species with ripe fruit occurred during the spring
The number of species in fruit began to
increase in late August and continued to increase through
the fall, reaching a peak of
8 species in fruit in
The number of fruiting species remained high
and began to decrease
There was no significant difference in the number of
in fruit per census period between 19
significant correlation between the number of fruiting
species per census and major frugivore species diversity,
frugivore species diversity, major frugivore abundan
frugivore abundance (Table
There was a significant
difference in the flowering to fruiting interval between
summer fruiting and fall and winter fruiting species (Mann
patterns are independent of seasonal flowering patterns,
iii t~irnt~s timinn nf fri4 +4 na c nnva, ~ n~aan
for the rela
Spearman rank correlation coefficients, r ,
tionship between frugivore species diversity
and abundance and the numb
in fruit per
two-week census period in San Felasco Hammock.
78.4 d (N=9) for summer fruiting
and 142.1 d
(N=36) for fall and winter fruiting species.
The majority of shrub and vine
whereas a number of tree species only
fruited heavily in alternate years or every three years.
latter fruiting behavior was evident in Cornus florida,
Persea borbonia, and Nyssa sylvatica.
Cornus florida had
large fruit crops in 1982 but small crops in 1983,
isolated branches of individual trees produced flowers and
Persea borbonia had large fruit crops in 1981, but
very small fruit crops in 19
individuals produced fruit.
and 1983, when only a few
Most Nyssa sylvatica
individuals produced ripe fruits in 1981 and again in 1983,
but only a few individuals bore fruit in 1982.
fruit crop of Cornus florida in 1983 appeared to be a major
factor in the lower number of frugivores
in the fall of
1983 compared to 1982,
when Cornus florida was a major
source of fruit for a number of frugivore species.
borbonia may also affect frugivore abundance due to its
large fruit crops and lipid rich fruits.
that showed high annual varis
included Symplocos tinctoria,
Ltion in fruit crop
decidua, and Crataegus
'S J -
important factor in the proximate timing of fruiting (see
Symplocos tinctoria showed the greatest
annual variation in ripening times,
as this species ripened
fruit in late July of 1982, but not until early October in
I have classified the 45 species of bird-dispersed
plants in the hammock community into four seasonal fruiting
patterns based on observed seasonal fruiting phenologies
(see Appendix II).
These include summer fruiting, fall
fruiting, fall-winter fruiting, and winter fruiting.
Nine species produced ripe fruit from early spring to
late summer, before the arrival of fall migrants.
subgroups were evident; Vaccinium mrysinites, Smilax
smallii, Rubus cuneifolius, and Morus rubra had ripe fruit
in the spring and early summer, while Chionanthus
virginiana, Prunus angustifolia, P. serotina, Vitis
aestivalis, and V. rotundifolia had ripe fruit in the late
The average fruiting duration for summer fruiting
species was 64.9 d.
There is a decrease in the number of
"mammal-bird" fruits from summer to fall
to winter (Table
The predominance of mammal-bird fruits during the
summer suggests that birds are unreliable seed dispersers at
this time, and mammal-bird fruits may have evolved to
increase the dispersal coterie of these plants.
unreliable nature of birds
summer is most likely due to t]
dispersal agents during the
heir low abundance and
this time and to their preference for abundant,
protein-rich insects as a food source
e (Morton 1973).
Summer fruiting species consisted primarily of uncommon
and rare fruiting species, with only two, Rubus cuneifolius
and Vitis rotundifolia considered common in the study area.
of the 36 fall and winter fruiting species
were either abundant or common in the study area (see
Summer fruiting species typically occurred along second
growth forest edge and in gaps within the forest.
summer species, Smilax smallii and Chionanthus virginiana,
were found within the closed forest,
whereas 28 of the
fall and winter fruiting
were forest inhabitants.
Few frugivores visited second growth habitats in the hammock
during the summer.
The two major frugivores present during
ths ~iimmsr ths Pnd~hnl 1innWnnefn nn T'ru ir
Relationship of mammal-bird fruits and
bird-fruits to seasonal fruiting patterns in San Felasco
No. of Species
Mammal-Bird Fruits (%)
Bird Fruits (%)
Total 9 36
erythrophthalmus), which may be considered fruit-thieves or
I have observed cardinals
on these fruits and dropping the seeds intact,
while feeding only on the pulp.
If a cardinal moves the
seed away from the parent plant, even a short distance,
effective dispersal may result.
Thompson and Willson
(1979) have found that while cardinals crush the large seeds
of Lindera benzoin,
the small seeds of Sambucus canadensis
the tendency of plants inhabiting disturbed
habitats to fruit early may result from the importance of
rapid reproduction and dispersal due to the colonizing
nature of these plants.
reaching maturity, rega
Such plants may reproduce upon
.rdless of the time of year (Janzen
Forest plants may not be under such
may delay fruiting until the fall or winter.
The 12 fall fruiting specie
s (see Appendix II) yielded
ripe fruits during September and October, at the peak of
fall bird migration through northern Florida. The average
duration of fall fruiting
disappearance typically occurred by January.
important than the second in the evolution of fruiting
due its greater predictability and higher
Certain fall fruiting
Amelopsis arborea, Aralia spinosa, Cornus foemina, Magnolia
grandiflora, Parthenocissus quinquefolia, and Phytolacca
americana have their fruit dispersed almost exclusively by
the first pulse,
their fruit crops are exhausted by the
time the second wave arrives.
Other fall fruiting species,
including Cornus florida, Crataegus marshall,
uniflora, Callicarpa americana, and Arisaema dracontium,
fruit available for the second pulse.
Fall fruiting species are predominately "bird" fruits
), although Crataegus uniflora and Crataegus
viridis, a fall-winter fruiter,
appear to be adapted for
primarily by mammals.
These fruits exhibited poor
often falling to the ground while still green.
The ripe fruits are probably
located by smell by ground
The fall-winter fruiting pattern was the most prevalent
fruiting pattern in the hammock,
occurring in twenty
produced ripe fruits during the fall
This fruiting pattern
synchronized with the first
and second pulse of migrant frugivores.
persistent nature of these fruits may be an adaptation to
the unpredictable nature of wintering frugivores in time and
space (Stapanian 1982). Persist
fruit quality and removal rates,
ence may also be related to
low quality fall-winter
low removal rates (see Chapter
these species have ripe fruit in the fall, but do not have
their fruits removed until January or February.
Fall-winter fruiting species consist of a higher
proportion of evergreen sp
than summer or fall
The fruits of most evergreen
species persist or ripen after the deciduous trees in the
community have lost their
leaves in late November.
The possession of persistent fruits or delayed fruit
ripening in evergreen
species may have a physiological
in that the evergreen condition allows for
persistent fruits or continued photosynthate
buildup for maturing fruits.
The abundant fall-winter fruit
production in the scrublands of southern Spain has been
attributed in part to the predominance of evergreen
within this community (Herrera
with persistent fruits are also
I............ *r a a
r r r r r
Relationship of evergreenness to seasonal
fruiting patterns for bird-dispersed plants of San
Felasco Hammock, Florida.
No. of Species
Summer; Fall Fruiting
Fall-Winter; Winter Fruiting
Total 25 20
contingency: X2=12.3, P<0.005
leaves of these species may act
(Stiles 1982), and a
to trees with leaves
foliar fruit flags
vian dispersal agents may be attracted
possible sources of fruit.
The winter fruiting pattern was shown by four evergreen
(see Appendix II).
In these species,
did not occur until December, at which time the majority of
wintering frugivores had arrived in Florida.
fruiting duration of winter fruiting species was 87.8 d.
These fruits showed high persistence and low spoilage.
Important winter fruiting species included Prunus
caroliniana and Phoradendron serotinum.
a common tree along forest edges,
provided an abundant food
supply for wintering robins and woodpeckers, whereas
Phoradendron serotinum constituted 70.5% of the total
to fruiting plants by wintering waxwings.
There was a close
relationship between fruit ripening in Phoradendron
serotinum and the arrival
of waxwings at San Felasco
In 1982 and 1983,
Phoradendron serotinum fruits
ripened in mid-December,
at which time waxwings were first
Winter fruiting specie
large ripe fruit crops at
provide wintering frugivores with low quality (see Chapter
III) but abundant,
concentrated fruits at a time when fruit
is diminishing and frugivore metabolic needs are
increasing due to colder temperatures.
Delayed fruiting in
winter fruiting species may also be an adaptation for the
of competition for dispersal agents in the fall,
when up to
have ripe fruit.
There were no significant differences among the four
seasonal fruiting patterns in between-plant ripening,
population synchrony (Table
individual plant ripening, and
There was also no
significant difference in these characteristics between the
ripening, population ripening, and individual plant ripening
showed a significant positive correlation,
of population synchrony, Z, showed a signi
correlation to between plant ripening, pop
while the index
plant ripening (Table
values (low population synchrony) were associated with long
asynchronous ripening periods (high between-plant ripening,
population ripening, and individual plant ripening values).
'6 0) 0
to- o in
CO *a~' 0
P4 1.. *
.1w' *~44~.f4 $4
o cn LA\0
C0 C\O0 CM
- '.-f H
CM 4a CA *
in- ~n- -~
0 (vi CM 0
H O\0 0
I-f r-*$ 0
*- H -~
UN IA- CO
-a* -t -~ (vi
-~ %- 'Sn- -~ H
o uA-' 04
ir1 U" *
M^ 01 ^
o en LAto
O SW Ovo
Spearman rank correlation coefficients, r
for ripening characters
Correlation coefficients were
calculated from 1983-84 data from Table
100/90% 90% x90% Z
100/90% 1.0000 -
90% 0.9705* 1.0000 -
x/90% 0.8662* 0.8524* 1.0000 -
Z -0.6994* -0.6936* -0.5006* 1.0000
ripening rates were 19.5 d (1982-83) and 14.8 d
(1983-84); and average population synchrony values were
Individual species often differed greatly in the above
Between plant ripening rates ranged from 7
d (Smilax bona-nox) to
Individual ripening rates
varied from 7
d (Smilax bona-nox)
d (Symplocos tinctoria).
Population synchrony ranged
The seasonal relationship of ripening synchrony
suggested by Thompson and Willson (1979), in which summer
fruiting species exhibit asynchronous between-plant ripening
and fall fruiting species show synchronous between-plant
was not evident for the northern Florida fruiting
Herrera (1984a) also has found no obvious
relationship between seasonality and ripening patterns in
the bird-dispersed plants of the scrubland community in
southern Spain. The extended availability of migrant
fruigivores at lower latitudes (September-November) may have
resulted in less selection pressure for synchronous fall
compared to more northern latitudes in North
would be especially true for those
tendency for synchronized ripening.
Also, pronounced annual
variation in ripening rates for a number of species suggests
that proximate factors related to the physiology of fruit
ripening may play an important role in the length of the
Middle and Lower Temperate Comparisons
The phenological patterns of northern Florida bird-
s and seasonal patterns of frugivore
availability show a logical relationship with tho
patterns described by Thompson and Willson (1979) in
Frugivore diversity and abundance were highest in
September in Illinois,
occurred in October.
while in Florida, highest values
The month-long difference would
reflect the movement of migrant frugivores from middle to
eptember peaks in frugivore
diversity and abundance in Illinois correspond to the two
peaks evident in northern Florida in October and November.
The first peak in both areas consisted primarily of thrush
species, while the second pulse consisted of robins, Hermit
Thrushes, and Yellow-rumped Warbler
The number of
in fruit in Illinois begins to
increase in August, reaching peaks in late August and
both areas correspond to periods of high frugivore
from upper to middle
likely that, as migrant frugivores pass
to lover latitudes in North America,
they encounter peaks of fruit production at each latitude.
As predicted by Thompson and Willson
fruiting strategy at lower temperate latitudes in North
America is more pronounced than at middle or upper
Twenty-four species of plants have ripe fruit
during the winter months in the hammock community,
to only four speci
availability, mild winters, and the presence of evergreen
species have made winter fruiting possible and profitable in
Low frugivore abundance,
and the dominance of
deciduous species restrict winter
fruiting to only a few species at northern latitudes.
The second prediction of Thompson and Willson (1979),
that summer fruiting should be more profitable at lower
temperate latitudes, is not supported by this study.
absence of major frugivores, abundance of insects, and
potential problems with microbial spoilage apparently have
restricted summer fruiting to a relatively few species,
which may depend more on mammals for dispersal
that the summer fruiting strategy should
be more profitable at more northern latitudes,
accounts of adult thrushes feeding fruits to their young do
Lower Temperate Latitude Dispersal Systems
Possibly the most thorough single community analysis of
bird-fruit coevolution to dat
e has been made in the
scrublands of southern Spain (Herrera 1981a,b,
1984a,b, Herrera and Jordano 1981, Jordano 1982, Jordano and
imilarities between Florida and southern
Spain in seasonal frugivore availability make a comparison
of the bird-fruit system in these two areas worthwhile.
Southern Spain is a major flyway in the fall for transient
migrating south to wintering grounds in
Africa, while during the winter it is home to a number of
overwintering frugivores (Herrera 1982a,
an analgous situation to that in Florida,
As this is
it is logical
similar phenological strategies related to dispersal should
be evident for the bird-dispersed plants in these two
Fruit-eating by birds in the scrubland community of
southern Spain appears to be limited to passerine species
whereas in the Florida hammock community
four woodpecker species are significant consumers of fruit.
the passerine species in the hammock community.
predators are also more prevalent in the scrublands,
69% of the fruit-eating birds are considered fruit
while only three species of fruit predator were
observed in the hammock community.
does the hammock community,
the scrublands has a
subset of major frugivores, consisting primarily of fall
migrants and overwintering species,
that is responsible for
most of the frugivory within the community (Herrera 1984a).
The major frugivores in the Spanish scrublands tend to be
12-18g body mass (Herrera 1984a,b) than those in
Florida (40-100g body mass
and other North American forests
(Sherburne 1972; Stiles 1980; Rybczynski and Riker
Herrera (1984a) has attributed the absence of large
frugivores in this community to possible difficulties in
foraging for fruit in the dense scrub and to the prevalence
of fruiting displays involving thin stems and erect
The Spanish scrubland community shows a proportion of
bird-dispersed plant species,
49-66% (Herrera 1984a)
similar to that in the hammock community (approximately
The presence of
overwintering frugivores in southern Spain and Florida has
resulted in the availability of fruit during every month of
attributed the prevalence of fall-winter fruiting in the
Spanish scrublands to the high degree of evergreenness and
mild winters that are characteristic of this community.
This is a similar situation to that in the hammock
where mild winters have allowed the establishment
of such subtropical evergreen species
Persea borbonia, Symplocos tinctoria, Osmanthus
americanus, and Prunus caroliniana.
Herrera (1984a) maintains that the scrubland bird-plant
dispersal system during fall and winter is driven by
relatively few pairs of strong reciprocal primary plant-bird
Pistacia lentiscus-Sylvia atricapilla;
Viburnum tinus-Erithacus rubecula).
These highly nutritious
primary plants indirectly favor seed dispersal of low-reward
or rare coexisting plants due to the varied diets exhibited
by the primary frugivores.
Thus the plant species interact
via diet-sharing and form an interdependent dispersal guild
There is also good evidence for the presence of primary
plant-bird interactions in the northern Florida hammock
In the early fall,
when the Veery is the primary
visitor to these plants, Parthenocissus quinquefolia and
Aralia spinosa are the most frequently visited plants in the
n ~'n*' ~ O .b.In a- n ,. ni -n *. a I- 4.4
Parthenocissus and Aralia fruit crops are exhausted,
is a close relationship between Veerys and Cornus florida,
another high lipid fruit.
In November, after Veerys leave
the area, robins become the primary dispersal agent in the
During late fall and winter, robins
sequentially become the primary dispersal agent for a
of fruiting plants,
including Cornus florida, Persea
borbonia, and Prunus caroliniana.
At these times,
majority of visits by robins are to these primary plants,
they also visit 15 other fruiting plant species.
A similar situation exists for the Cedar Waxwing and
The presence of
s principal food within this
attracts the highly frugivorous waxwing to the
It does appear
where it feeds on
, therefore, that
plants of the hammock community form a dispersal guild,
dependent on a relatively few species of lipid-rich or
abundant fruiting species. Th
dependent on a few species of
Veerys and robins).
Lis dispersal guild is in turn
abundant frugivores (e.g.
The temporal sequence of primary plants
ensures frugivore abundance throughout the fall and winter
months and sugg
possible temporal partitioning in
fruiting among the primary plant
The importance of
lower than in 1982,
when Cornus florida showed
a large fruit crop.
Despite the similarities in fruiting patterns between
the hammock and scrubland communities,
limit comparisons between these two
The Spanish scrublands receive approximately
one-half the annual rainfall of the hammock community and
have a drier summer.
the scrublands have a
continuous distribution of fruit-producing low shrubs
the hammock community has a patchy
distribution of fruiting tr
Community analysis of bird-fruit interactions in the
southern California chaparral community of North America
should show patterns more similar to those found in the
Spanish scrublands, due to the similarity in the climate and
vegetation of the two areas.
likely that mesic hardwood communities were
present in northern Florida throughout the Pleistocene,
when this area acted
as a possible glacial refuge for
northern deciduous trees (Delcourt and Delcourt
Glaciation over the past 80,000 years appears to have had a
glacial periods compared to the boreal
vegetation of the
middle and upper temperate latitudes (Delcourt and Delcourt
The Quaternary glaciations also have had a major
influence on avian migration patterns and are responsible
for the current geography
of avian migrations (Dorst
likely that northern Florida, with its mild climate
and mesic hardwood forests, served as a refugium for many
North American bird sp
during the glaciations.
flyway migration route (Lincoln 1939
channels avian migrants from most of
eastern North America
through northern Florida on their way to wintering grounds
in Central America,
and the West Indies.
predictable pulse of fall migrants,
the mild winters of
and the established southern mesic
hardwood forest community have provided the raw materials
necessary for the evolution of mutualistic interactions
between fruiting plants and fruit-eating birds in the
southern mixed-hardwood community.
No discussion on the evolution of eastern North
America forest community structure would be complete without
mention of the extinct Passenger Pigeon,
It was once extremely abundant in deciduous
forests from the great plains to the Atlanti
coast of New
chestnuts, but they were also known to feed upon a variety
including Prunus spp., Ilex spp., Nyssa
Cornus florida, Morus spp., Rhus spp.,
repens, Phytolacca americana,
and Aralia spinosa (Schorger
It is likely that
Passenger Pigeons were effective dispersal agents for the
majority of these fruits,
voiding the seeds intact.
we do not know the relative importance of fruit
in the Passenger Pigeon's diet.
The southernmost limit of the Passenger Pigeon's
wintering range in Florida has been listed
where it was reported to be abundant in the
century (Howell 1932).
The primary ecological role of the
Passenger Pigeon was probably that of a seed predator of mast-
However, due to the inclusion of fruit
within its diet and its large numbers, it is possible that
the Passenger Pigeon played a significant role in the
evolution of the bird-fruit dispersal system in eastern
its presence in northern Florida
during the winter may have in part
selected for the
predominance of fall-winter fruiting in this area.
Numbers of the American Robin,
frugivore in northern Florida today,
the primary wintering
seem to have increased
due to the clearing of forests in is breeding grounds
This would be especially true in the selection for
as the primary wintering ground for
flocks of up to 50,000 have been
sed flora of San Felasco Hammock shows
strong taxonomic affinities with the mid-latitude flora of
eastern North America.
Of 22 plant families represented in
San Felasco Hammock,
Palmae and Symplocaceae, are
confined to the southeastern United States
in their U.S.
At the generic level,
represented in San Felasco Hammock, only 8, Amelopsis,
Callicarpa, Chionanthus, Osmanthus,
and Symplocos, are restricted primarily to the
At the species
of the 45
species present at San Felasco Hammock (64.h%) are limited
to the southeast.
Twenty-four of these species bear ripe
fruit in the fall or winter.
Therefore, at the species
, the bird-dispersed flora of lower latitudes
distinct from the flora of mid-latitudes of eastern North
Yet many of the
species confined to the southeast
have congeners at more northern latitudes that show similar
-erotinum, Persea borbonia, Magnolia grandiflora, and
Symplocos tinctoria, show strong tropical affiliations.
This contrasts with the bird-dispersed flora of southern
where tropical elements comprise approximately 87%
bird-dispersed plant species (Tomlinson 1980).
Certain taxa show wide variation in fruit types and
while others are quite uniform.
angustifolia, P. serotina, and P. caroliniana,
flowering in early spring,
fruiting phenology. Prun
show pronounced differences in
us angustifolia and P. serotina
bear ripe fruit in the summer,
while P. caroliniana ripens
fruit in December.
while P. angustifolia produces a
sweet mammal-bird fruit that shows
low persistence, P.
caroliniana produces a bitter bird-dispersed fruit that is
Evergreenness is a good indicator of
the deciduous species, P.
angustifolia and P. serotina, are summer fruiters,
evergreen, P. caroliniana,
is a winter fruiter.
situation also exists for two
within the Oleaceae,
Chionanthus virginianus and Osmanthus americanus.
virginianus shows summer fruiting,
a winter fruiter.
Within certain genera such
as Smilax and
different species tend to exhibit similar fruiting patterns,
taxonomic constraints or the result of similar selective
pressures acting on each sp
The relationship between
systematics and ecological adaptation
is complex and may be
and the effects of evolutionary canalization must
be considered in analyzing the biological interactions of
The strong similarity in the bird-dispersed floras of
lower and mid-temperate latitudes in eastern North America
has probably played an important role in the evolution of
the bird-fruit seed dispersal system that we
such as Parthenocissus quinquefolia
and Cornus florida, supply migrating frugivores with a
familiar food supply in time and space along the
while congeners in the south show
displays to migrating frugivores similar to those of their
relatives in the north.
Thus throughout eastern North
America there is a continuity in both the floral and
frugivorous components of the bird-fruit dispersal system.
Add to this the highly predictable nature of the annual fall
migration of avian frugivores from upper to lower
not surprising that we
phenological patterns of
SEASONAL PATTERNS OF BIRD-DISPERSED
AND FRUIT REMOVAL
I collected ripe fruits from
in San Felasco Hammock.
pyrene) mass, pulp yield, and water content were determined
from an aggregate sample of
3474 undamaged fruits (mean
Fruits were counted,
weighed, and separated into
pulp and seed components by dissection with forceps and by
Pulp and cleaned
were weighed, and the pulp was
oven-dried to constant mass at 60oC.
measured to the nearest 0.1mg on
a Mettler AK 160
defined pulp moisture
separated pulp. Tota
from the mass of wet pulp
as percent mass lost on drying
1 dry pulp per fruit was determined
fruit mass minus the mass of
clean seeds) and its moisture content.
mass from wet
load per fruit and mean
per fruit for each sampl
-- 1 -
r 1 I I n in r' I C *~ a a p a F. -J
r- f jJ
petroleum ether in Micro-Soxhlet apparatus.
was estimated as
times nitrogen found by micro-Kheldahl
extraction (Association of Official Agricultural Chemists
carbohydrate content was determined by
the anthrone method (Yemm and Willis 1954;
Results for each chemical test a
from analysis of two subsamples.
re mean percent of dry mass
In addition to the 38
species of fruits collected from San Felasco Hammock,
included in my results analysis of
species (Morus rubra,
Rubus cuneifolius, Prunus serotina, Ilex glabra, and
Toxicodendron radicans) from sample
collected in New Jersey
and analysed by White and Stiles.
San Felasco Hammock,
These species occur in
but I was unable to collect sufficient
numbers of fruits for analysis due to their scarcity or
The nutritional composition of species of
Morus, Rubus, Prunus, and Ilex have been shown to be
consistently similar (see Stiles 1980,
White and Stiles
I have therefore included figures
from the New
Jersey samples in my seasonal analyses.
I have omitted
Toxicodendrom radicans from my statistical comparisons as
0 nnrn *1r
patterns are not evident for this
compared to the other genera.
and 13,902 fruits in 1982-83, and for 8 species, involving
74 individuals and 6,334 fruits, in 1983-84.
of fruits on
tagged branches of marked plants were made.
also tagged additional branches of marked plants each week
to replace removed fruits.
This was done in an attempt to
maintain a constant number of marked fruit
s for each
individual for each week. I defined removal rate
percentage of the total marked population of fruit removed
within a seven-day period.
I obtained additional information on fruit removal by
direct observations of fruit-eating birds. I
of fruit-eating while walking a 2100m transect
each week during bird censuses (
fruit censusing from January 1982 to April 1984.
Results and Discussion
The 43 species of bird-dispersed plants of San Felasco
Hammock fall into four fruiting patterns: summer fruiting,
fall fruiting, fall-winter fruiting, and winter fruiting
These fruiting patterns
on seasonality of fruiting, a single fruiting pattern may
preference, and fruit abundance are related to fruiting
Summer fruits (9 species) of the hammock community are
characterized by a large fruit mass,
low seed load,
water content, low
lipid and protein content, and high
carbohydrate content (Table 3-1).
Summer fruits include
that depend on both mammals and birds for seed
1. These "mammal-bird" fruits are characterized by
low persistence (fall
to the ground upon ripening), high
sugar content, and a sweet taste (Stiles 1980).
(but not P.
caroliniana) are eaten on a regular basis
by a number of mammals,
The production of fruits that are
attractive to mammals may be advantageous due to the low
availability of avian frugivores during the summer months.
Fruit use by breeding birds is probably also reduced due to
the abundance of insects at this time and the importance of
a high-protein food for breeding birds (Morton
has suggested that a dichotomy in seed
through the guts of small mammals undamaged,
large-seeded species may produce seeds
discourage seed predation by small mammals.
bird fruits separate into small-seeded species (mean seed
The small-seeded species also
low seed loads (mean seed load
Two summer fruiting species, Prunus angustifolia and
Vitis rotundifolia, produce large fruits (> 2000mg) that are
clearly too large to be swallowed by any resident bird.
Mammal-dispersed fruits have been characterized by a large
and an orange or yellow color (Janson 1983,
Stiles unpub. ms.).
Prunus anugustifolia and
"chewed" by cardinals.
produce purple fruits that are
If the cardinal flies away with the
fruit, even without swallowing the seeds, it may still
deliver short distance dispersal.
Nonetheless, it appears
that mammals may be the primary dispersal agents for these
The presence of a red preripe fruit color is also
summer mammal-bird fruit
display of unripe red and ripe blue-black fruit may increase
the conspicuousness of the fruiting display (Thompson and
Willson 1979, Stiles 1980. 1982. Stananian 1982. White and
of the 11 fall fruiting
of the 19 fall
winter fruiting species.
The 11 fall fruiting species in the hammock community
, Cornus foemina,
Cornus florida, Euonymus
americanus, Parthenocissus quinquefolia, and Magnolia
grandiflora, with high lipid content (>
10% dry mass,
High lipid content has typically been
correlated with high fruit quality
lipids yield about
twice the energy content
As a group, fall fruiting species show the
highest mean percentage in lipid content compared to the
other fruiting patterns (Table
fruits become available in the early fall,
frugivore diversity and abundance are at a peak due to the
presence of migrant frugivores.
also found a strong correlation between the timing of
production of lipid-rich fruits by bird-dispersed plants in
the fall and frugivore abundance in the scrublands of
High quality fruits,
because of their high lipid
content, may also attract insects and microbes.
five fall high-lipid fruits showed some fruit damage
resulting from invertebrate and microbe attack.
According to White and Stiles
of lipid-rich fruits in eastern North America may be limited
by season and plant growth form.
These fruits tend to be
produced in the fall when flocks of vagile,
migrant frugivores are available for rapid fruit removal,
and appear also to be limited to woody shrubs and trees,
which can produce moderate to large fruit crops.
fruit crops of herbs may limit repeat visits by birds and
thus decrease selection for high-lipid fruits (Stiles and
Based on the analysis of 77
mid-latitude bird-dispersed fruits from eastern North
White and Stiles (unpub. ms.) have found that
fruits with lipid contents greater than 10% are limited to
fall fruiting woody plants.
A similar situation occur
latitude hammock community.
No summer fruiting
species produces a high-lipid fruit,
of the 11 fall
fruiting species show fruits rich in lipids.
fruits are also found among fall-winter fruiting species
High lipid fruits in the hammock community are
limited to woody plants,
no herbaceous plants produced high
Stiles (1980) has nronosed two dispersal strateales for
thus permitting slower removal rates.
' dichotomy, Callicarpa americana, Crataegus uniflora,
Aralia spinosa, Arisaema dracontium, Phytolacca americana,
and Amelopsis arborea would be considered low quality fruits
in the hammock community.
dracontium and Crataegus uniflora,
With the exception of Arisaema
these species produce low
< 3.0% dry mass) with small seed masses (mean
= 9.2mg) and low seed loads (mean
Crataegus uniflora fruits show characteristics of
mammal-bird fruits (low persistence, sweet taste,
fruit mass) and appear to be adapted primarily for mammal
The herb Arisaema dracontium produces low-reward
for birds due to their high seed load (40.4%),
and low yield in dry matter per fruit
Stiles and Devito
certain low quality,
rare herbs may rely on generalized
mimicry of high-quality fruits for seed dispersal.
dracontium fruits resemble fallen Cornus florida fruits (a
high-lipid fruit) and may be taken accidentally by ground
foraging thrushes and
lowest mean fruit mass (251.8mg) and lowest mean
water content (55.4%) of the four seasonal
The low water content of these fruits may be
an adaptation for persistence (Stapanian 1982,
summer fruits show a
high mean water content
(82.7%) and are
Fall-winter fruiting species include
borbonia, Symplocos tinctoria, Myrica cerifera, Rhus
coppalina and Toxicodendron radicans,
that show high lipid
White and Stiles
three distinct fruit types for
2) dry (<
) moist (
> 40% water),
oily fruits of the
fleshy fruits of the
high-quality fruiting pattern.
Toxicodendron radicans and
Myrica cerifera would be considered waxy-pulped fruit
will be discussed below.
Rhus coppalina (12.0% lipid) would
be considered a dry, oily
, lipid-rich fruit.
The fruits of
although fed on by
six bird species (Appendix IV), were
Over a two-year period,
I observed only
12 visits to this species,
Sumacs (Rhus spp.) general
in which 190 fruits were removed.
ly are considered to be
unattractive to birds and
used primarily as an emergency
(White and Stiles,
Persea borbonia and SymDlocos tinctoria would he
duration for these
of North America.
had a highly asynchronous
fruit ripening pattern (Chapter
II), resulting in a small
standing crop of ripe fruits at any one time. This resulted
in high removal rates of ripe fruits (see below), thus
limiting the number of ripe fruits available for insects and
The fruits of Symplocos also turned from purple
and moist to brown and dry.
Persea borbonia fruits showed
little evidence of extensive damage from insects or
This species may possess effective toxins that
discourage insect and microbe attack.
The 4 winter fruiting species produced ripe fruits in
January when the majority of fall-fruiting
depleted fruit crops.
The fruits of Prunus caroliniana,
Osmanthus americanus, and Ilex opaca showed a nutrient
content similar to that of fall-winter low-quality fruits
while Phoradendron serotinum
exhibited a distinct fruit type and dispersal syndrome from
the other fall-winter and winter fruiting species (see
Prunus caroliniana, Osmanthus americana, and Ilex
opaca produced large fruits
(mean fruit mass
Specialized Dispersal Systems
Several specialized dispersal systems, involving single
that are dependent upon a small set of
occur in the hammock community.
associations are rare in bird-fruit dispersal systems
(Wheelwright and Orians 1982) and typically involve fruits
with distinct fruit structure or chemistry (White and
Toxicodendron radicans and Myrica
cerifera produce waxy-pulped fruits unlike the fleshy-pulped
fruits exhibited by most species
in the hammock.
These species produce high lipid fruits,
which may be a
necessity for plants that have a small dispersal coterie
Removal of Toxicodendron fruits in the
hammock is limited to woodpeckers and Yellow-rumped Warblers
Based on gut content
records of fruit-eating birds in North America,
Stiles (unpub. ms.) have found woodpeckers and warblers to
significantly more frequently than
other fruit-eating birds.
also found a
close relationship between warblers and the waxy fruits of
Lindakeria laurina in Panama.
Removal of Myrica cerifera
fruits in the hammock
limited primarily to Yellow-rumped
These warblers accounted for 91% of the total
Cedar Waxwings accounted for 80.7% of the total
visits to Phoradendron; mistletoe constituted 70.5% of the
visits to fruiting plants by waxwings.
fruits show a distinct fruit structure typical of many
(Gill and Hawksworth 1961), while vaxwings
probably show similar morphological fruit-eating adaptations
to those of other mistletoe birds (Wetmore 1914,
Phoradendron berries also have a high protein
which may contribute to its prevalence in
The high lipid fruiting species, Cornus foemina and
Symplocos tinctoria, had consistently high fruit removal
for Cornus foemina,
a fall fruiting
, increased in early September upon the arrival of
migrant frugivores (Fig.
Fruit removal remained high
throughout September until its fruit crops were exhausted in
a fall-winter fruiting
had high weekly removal rates from August through
asynchronous ripening pattern (Chapter
in a relatively small standing crop of ripe fruit
s that was
Fall fruiting species.
Percent of the
initial total number of fruits removed within a single
week for Cornus foemina, Callicarpa americana, and
Solid line represents 1982 season,
dashed line 1983 season.
the initial total number of fruits removed within a
single week for Viburnum rufidulum,
Symplocos tinctoria, Ilex decidua, Smilax bona-nox, and
Smilax auriculata. Solid line represents 1982-83
season, dashed line 1983-84 season.
Vib urunum rufidulum
Viburnum obo vatum
- aI a
plant was visited regularly by small flocks of migrant
thrushes (Appendix IV) and was considered a primary plant
species for the local frugiv
display, abundant fruit crop
seed load of its fruits (14.
The showy fruiting
, clumped distribution, and low
9%), may account for the
attractiveness of this species to fruit-eating birds. High
removal rates appear to be advantageous for Aralia as its
fruits are highly susceptible to drying out and rotting.
Callicarpa americana, another fall, low lipid fruiting
species, showed a fruit removal pattern similar to Aralia's.
Removal rates for Callicarpa were high in September and
October until its
fruit crops were exhausted
which was visited regularly by migrant
thrushes (Appendix IV),
has a stunning fruiting display
and was highly visible in the hammock.
The popularity of
Callicarpa may also be attributed to its individual
abundance, large fruit crops, and the low seed load of its
Fall-winter low lipid fruits typically showed low
removal rates throughout the fall until January or February,
when removal rates increased.
These increases occurred
after the fruit crops of the fall and fall-winter high lipid
fruiting species were all removed.
This pattern is shown by
TinY iri11 mlyhn nn S in nirrimt.Vhi m
V i hii rnlr rn
result of the inclusion of a specific fruit in the diet of
the most abundant wintering frugivore in the
Once a fruit was
included in the robin's diet,
its fruit crop was regularly visited until it was exhausted.
The winter fruiting Prunus caroliniana showed a
removal pattern similar to those of fall-winter
Removal rates of Prunus were moderate until it
became the primary fruit for robins.
At this time,
removal rates for this species increased sharply until its
crop was completely removed (Fig.
attractiveness of Prunus caroliniana to robins may be due to
Its fruits, while having a low lipid
content and high seed mass, h
of fruits of its congener, P.
ave three times the dry pulp mass
serotina, a summer fruiting
Also, the large seed of Prunus caroliniana is
regurgitated, which may negate the disadvantages of a high
Seasonal Patterns of Fruit Removal
Based on weekly observations of fruit-eating birds from
1982-84 and fruit removal rates of marked individuals, th
fruit removal picture in the hammock community appears as
In the early fall,
when frugivore availability
Fig. 3-3. Winter fruiting species. Percent of the
initial total of fruits removed within a single week
for Ilex oaca and Prunus caroliniana. Solid line
represents 192-83 season, dashed line 1983-84 season.
i I I --
Fall high lipid fruiting species often had small
fruit crops (e.g. Symplocos tinctoria,
or may be more difficult to
quinquefolia) than many
lipid-rich fruits may offer a higher
caloric reward to
the search time for these fruits may be higher than
for some low-quality species.
Therefore it may be
practical, energetically, for a transient frugivore to visit
low lipid fruits (e.g. Aralia,
lipid-rich fruits are not readily visible
Cornus florida, with its lipid-rich fruit, is the
primary fruiting plant in the mid-fall,
by small flocks of migrant thrushes and later, robins
(Appendix IV). In late fall,
when the fruit crop of Cornus
florida is completely removed, Persea borbonia, another high
lipid species, becomes the primary fruit for wintering
The size of the fruit crop of Cornus florida
and Persea borbonia varies greatly in different years and
of their crops appears to influence frugivore
while Cornus florida and Persea borbonia
were primary food plants for
local frugivores during late
fall and early winter, other fruiting species were also
On one day in mid-November.
appears to be a common practice among many frugivorous birds
1977, Wheelwright 1983, Herrera 1984a) and may be
related to dietary requirements or "optimal foraging".
varied feeding behavior of frugivorous birds may be
essential to low-reward or rare species that by themsel
would be unable to attract dispersers.
Therefore, at the
there is an interdependence, in relation to
in which low-reward species may benefit from
coexisting with primary plant
In late January
species (Herrera 1984a).
, when the fruit crop of Persea borbonia
is gone, the hammock frugivores shift to persistent, low
lipid fruits such as Prunus caroliniana, Osmanthus
americanus, Celtis laevigata, Ilex spp., and Vaccinium
The last fruits to be taken include low
Smilax spp., Mitchella repens, and Viburnum obovatum.
this time wintering frugivores are beginning to lay down fat
deposits for the northward migration to their breeding
I found no absolute relationship between high and low-
quality fruits and fruit removal rates.
The majority of
high-quality fruits did show higher fruit removal rates than
species, however, certain low-quality
(e.g. Callicarpa americanus, Aralia spinosa)
. n a a --
fruiting display, seed load,
and plant distribution, may be equally
content in relation to removal
the dichotomy gf high and low-quality fruits,
, Hove and Estabrook 1977), may not
always be correlated with fruit removal
Eastern North American Comparisons
The bird-dispersed flora of San Felasco Hammock shows a
close taxonomic affinity to the mid-latitude flora of
eastern North America.
especially evident at the
of the 31 genera present in the
hammock are found at mid-latitudes.
ms.) have analyzed the nutritional content of
of mid-latitude bird-dispersed plants of eastern
At the generic level,
13 genera showed similar nutritional composition to
congeners from New Jersey.
common to San
Felasco Hammock and New Jersey,
had marked nutritional
Parthenocissus quinquefolia fruits
were found to have a lower lipid content than
also found higher lipid values for Parthenocissus fruits in
for the congener,
(31.2%) has also
been recorded (Johnson et. al.
a high-quality fruit as determined by White and
showed a much lower
lipid content in Florida than in
1.4% vrs. 14.8%).
The bird-dispersed flora of San Felasco hammock has a
high proportion of lipid-rich fruits
more northern latitudes.
compared to floras of
s is due in part to the
inclusion of such tropical-affiliated species as Magnolia
grandiflora, Persea borbonia,
and Symplocos tinctoria.
These species occur in families that have characteristically
high lipid fruits
The mild Florida
winters have permitted the establishment of these
subtropical species and have also resulted in large
populations of wintering frugivores.
Furthermore, a number
of these species show extended fall-winter fruiting,
wintering frugivores for seed dispersal.
At more northern
small populations of wintering birds should
select against the
evolution of lipid-rich fall-winter
high lipid fruits are limited to fall
fruiting species at middle and upper temperate latitudes
Thompson and Willson 1979
, White and Stiles unpub. ms.).
The remaining high lipid fruits from the hammock
community belong to
the Vitaceae shows both high and low lipid fruiting
species in the hammock.
The fall fruiting Parthenocissus
quinquefolia has a high lipid fruit,
while the summer
fruiting Vitis rotundifolia and V. aestivalis have low lipid
The variability in fruit quality in temperate
families suggests an independent evolution of lipid-rich
for these families
(White and Stiles,
species of bird-dispersed plants in the hammock
community represent a heterogeneous mixture of fruit types,
which exhibit a wide range in nutrient composition.
clear nutritional patterns related to dispersal
strategies and seasonality are evident.
into categories is thus possible
Summer fruiting species
form a distinct set, in which fruit morphology and nutrient
composition show characteristics
persistence) that may favor mixed mammal-bird seed
This strategy appears to be an adaptation to the
low abundance of frugivores during the summer months and to
the preference of breeding birds for insects
primary food supply.
Fall fruiting sp
fall into two category
low-quality, may offer migrants
an abundance of fruits with low seed loads.
fruiting species show variable removal rates,
species showing low removal rates, while others exhibit
removal rates comparable to high lipid species.
species also consist of high and
low lipid species, in which the high lipid species typically
have their fruit crops removed before the low lipid species.
lipid fruits persist until late winter
- early spring,
when they are taken sequentially by flocks of robins.
Winter fruiting species have fruits with low lipid
content and exhibit fruit removal patterns similar to low
lipid fall-winter fruiting
The delayed ripening
pattern of these species may be an adaptation for the
avoidance of competition for dispersal agents from high
Within the hammock community are a series of primary
These plants receive the majority of
visits from local frugivores and may be instrumental in
attracting flocks of migrant frugivores into the community.
preferred fruiting species may thus
coexistence with these primary plant species.
included both high and low lipid species.
Characteristics of primary plants that may account for their
patterns of bird-dispersed fruits in
northern Florida show strong similarities to those patterns
described by Stiles
1980) and White and Stiles
for mid-latitude bird-dispersed plants of
the result of the close
taxonomic relationship in the bird-dispersed floras of the
two regions and to similar
selection pressures related to
the seasonal availability of frugivores.
difference in the bird-fruit system between middle and lower
temperate latitudes in eastern North America is the
prevalence of the fall-winter fruiting pattern at lower
latitudes compared to middle latitudes.
winters have permitted the establishment of high lipid,
bird-dispersed species of tropical affinity and large
populations of wintering frugivores at lower
which have selected for fall-winter fruiting for many
The presence of these subtropical species has
resulted in a greater proportion of high lipid fruits in the
bird-dispersed flora at lower
latitudes compared to middle
latitudes in eastern North America.
The bird-fruit dispersal system of San Felasco Hammock
22 species of frugivorous birds and 45 species of
The frugivorous birds include
members from eight families and range in size from
Twelve of the twenty-two bird
species are considered major frugivores as they were
involved in the majority of fruit visitations.
dispersed flora consists of plants representing
These plants show high variability in fruit type and
The fruiting phenology of the bird-dispersed plants of
San Felasco Hammock show similarities to both tropical and
middle latitude temperate bird-dispersed plant fruiting
of the yea
The hammock community shows ripe fruit every month
typical of many tropical communities
Frankie et. al.
plants of middle temperate latitudes tend to fruit only
during the summer and fall,
s bearing ripe fruit
during the winter months (Thompson and Willson 1979). The
situation to that observed at middle temperate latitudes in
North America (Thompson and Willson 1979, Stiles 1980, White
Most bird-dispersed plants in the hammock
bear ripe fruit in the fall,
when the community is invaded
by a high diversity of migrant frugivores.
diversity and abundance remain high throughout the fall and
winter months as does the number of fruiting species.
number of fruiting species decreases in the spring and
remains low throughout the summer months.
diversity and abundance is also low at this time.
The 44 species of bird-dispersed plants fall into four
distinct seasonal fruiting patterns:
fall-winter fruiting, and winter fruiting.
summer fruiting species have fruit characterized by high
sweet taste, a
large fruit mass,
Lnd low persistence.
high carbohydrate content,
These species produce
"mammal-bird fruits that may depend on mammals and birds
Summer fruiting species typically are rare
or uncommon in the community and occur in disturbed
Fall fruiting species produce ripe fruit
s during the
peak of fall migration of frugivorous birds through northern
Fall fruiting species include five species that
Fall fruiting species usually occur in
the closed forest and are common or abundant in the
The fall-winter fruiting pattern is the most prevalent
fruiting pattern in the hammock, being
shown by 20 species.
Fall-winter fruiting species produce highly persistent fruit
that become ripe in the fall, persisting into January and
This fruiting pattern is thus synchronized with
the first wave of transient frugivores and also the second
wave of overwintering frugivores.
species are typically evergreen, and their fruits have the
lowest mean fruit mass and water content of the four
Six fall-winter fruiting species have
high lipid fruits, including two with waxy-pulped fruit, one
with dry fruits, and three with moist fleshy fruits.
remaining fall-winter fruits produce low lipid fruits.
The winter fruiting pattern
shown by 4 evergreen
species, in which fruit maturation does not occur until
These species use only overwintering frugivores
for seed dispersal.
Winter fruiting species exhibit low
lipid fruits, although Phoradendron serotinum produces a
relatively high protein fruit and
a unique dispersal
syndrome compared to the other winter specie
T1 h ta h-i vrl-fT'll-i-k ^i r1 rnf t n hcwi ^ lr
instrumental in attracting frugvorous birds into the
These primary plants may also indirectly favor
seed dispersal of non-primary plants due to the varied diets
exhibited by fruit-eating birds.
The primary plants in this
community consist primarily of high lipid fruiting species
(e.g. Parthenocissus quinquefolia,
but also include several low
species (e.g. Aralia spinosa,
lipid species produce large fruit crops and high
with small seeds and low seed loads.
Primary plants show high removal rates,
low quality fruiting species exhibit low removal rates until
January or February when the fruit crops of Cornus florida
and Persea borbonia are exhausted.
At this time,
of these highly persistent species are taken regularly until
their fruit crops are gone.
Primary frugivore bird species in the hammock include
migrant thrushes in the early fall and robins in the late
fall and winter.
Veerys are the primary dispersal
agent in the early fall when they form small flocks and
visit a variety of fruiting species.
When Veerys leave the
hammock in the late fall,
due to their
and highly frugivorous diet become the primary dispersal
agent in the community.
Robins become the primary
responsible for the bulk of the fruit removal of the fall-
low quality fruiting species in January and February.
Several specialized dispersal systems, involving single
plant species that are dependent upon a single or few
dispersal agents, are evident in the hammock community.
These include a close relationship between the mistletoe,
Phoradendron serotinum and the Cedar Waxwing,
radicans and woodpeckers and Yellow-rumped Warblers, and
Myrica cerifera and the Yellow-rumped Warbler.
species produce fruits with a distinct fruit structure or
chemistry related to their dispersal syndromes.
The bird-dispersed plants of the lower temperate
latitude hammock community show similar phenological
patterns and fruit characteristics to middle latitude
temperate bird-dispersed plants of North America.
fruiting is the most prevalent fruiting pattern shown at
middle latitudes of North America,
where the timing of
fruiting is closely correlated with the availability of
migrant frugivores (Thompson and Willson 1979).
the fall fruiting pattern is also common
the fall-winter fruiting pattern is shown by th
largest number of
dispersed plants have ripe fruit during the winter months in
the hammock community,
only 4 species in Illinois have fruit
fruiting possible and profitable in northern Florida
compared to the middle latitudes of North America.
Similar selection pressures related to frugivore
availability and close taxonomic affinities have resulted in
close similarities in the fruit characteristics of the bird-
dispersed flora between lower and middle temperate latitudes
of eastern North America.
Nonetheless, the bird-dispersed
flora of the hammock community shows a higher proportion of
lipid-rich fruits than at more northern latitudes.
is due in part to the inclusion within its flora of such
lipid-rich tropical-affiliated species
grandiflora, Persea borbonia, and Symplocos tinctoria.
The bird-dispersed flora of the hammock community also
shows fruiting patterns and fruit characteristics similar to
the lower temperate latitude scrubland community of southern
Spain (Herrera 1981, 1982a, 1984a).
Both communities show a
predominance of fall-winter fruiting species, in which
lipid-rich fruits are produced in the fall.
can be attributed to similar patterns of frugivore
availability, in which the presence of overwintering
frugivores in both communities has selected for fall-winter
fruiting and also to the high degree of evergreenness
characteristic of both communities.
METHODS OF CALCULATING RIPENING
SYNCHRONY (MODIFIED FROM AUGSPURGER 1983)
A. Synchrony of a given individual with its
: Xi, the index of synchrony for
. is defined as:
ej = number of days both individuals
i and J are ripening 90% of fruit crop,
jl; fi = number of days individual
i is ripening 90% of fruit crop;
n = number of individuals in population.
ripening period of individual i overlaps with
ripening period of each other individual, j#i
in the population.
= 0.0, no synchrony
overlap occurs in the ripening period of individual
i and any other individual, jV1 in the population.
Synchrony of the population:
of population synchrony,
where Xi is synchrony of individual i with its
conspecifics from part A. (above).