ArticlePDF Available

Fruit size, crop mass, and plant height explain differential fruit choice of primates and birds

  • Fundación Jocotoco, Quito, Ecuador

Abstract and Figures

Seed dispersal by animals is an important ecological process shaping plant regeneration. In general, seed dispersers are highly variable and often opportunistic in their fruit choice. Despite much research, the factors that can explain patterns of fruit consumption among different animal groups remain contentious. Here, we analysed the interactions between 81 animal species feeding on the fruits of 30 plant species in Kakamega Forest, Kenya, during 840 h of observations. Our aim was to determine whether plant characteristics, fruit morphology, fruit colours and/or fruit compounds such as water, sugar, phenols and tannins explained the relative importance of fruit consumption by the two most important consumer groups, primates and birds. We found significant differences in fruit choice between both groups. Primates fed on larger fruits and on higher trees that had larger fruit crops, whereas birds were observed feeding on smaller fruits and on smaller plants producing fewer fruits. Fruit colours did not differ between fruits consumed by primates and those consumed by birds. However, differences in the fruit choice among frugivorous birds were associated with differences in fruit colours. Smaller plants with smaller fruits produced red fruits which contrasted strongly with the background; these fruits were dispersed by a distinct set of bird species. The contents of water, sugar, phenols and tannins did not differ between fruits eaten by primates and those eaten by birds. Some phylogenetic patterns were apparent; primates fed preferentially on a phylogenetically restricted subsample of large plants with large fruits of the subclass Rosidae. We discuss why the observed primate dispersal syndrome is most likely explained by a process of ecological fitting.
Content may be subject to copyright.
Oecologia (2010) 164:151–161
DOI 10.1007/s00442-010-1655-8
Fruit size, crop mass, and plant height explain diVerential fruit
choice of primates and birds
Martina Flörchinger · Julius Braun ·
Katrin Böhning-Gaese · H. Martin Schaefer
Received: 20 November 2009 / Accepted: 26 April 2010 / Published online: 19 May 2010
© Springer-Verlag 2010
Abstract Seed dispersal by animals is an important eco-
logical process shaping plant regeneration. In general, seed
dispersers are highly variable and often opportunistic in
their fruit choice. Despite much research, the factors that
can explain patterns of fruit consumption among diVerent
animal groups remain contentious. Here, we analysed the
interactions between 81 animal species feeding on the fruits
of 30 plant species in Kakamega Forest, Kenya, during
840 h of observations. Our aim was to determine whether
plant characteristics, fruit morphology, fruit colours and/or
fruit compounds such as water, sugar, phenols and tannins
explained the relative importance of fruit consumption by
the two most important consumer groups, primates and
birds. We found signiWcant diVerences in fruit choice
between both groups. Primates fed on larger fruits and on
higher trees that had larger fruit crops, whereas birds were
observed feeding on smaller fruits and on smaller plants
producing fewer fruits. Fruit colours did not diVer between
fruits consumed by primates and those consumed by birds.
However, diVerences in the fruit choice among frugivorous
birds were associated with diVerences in fruit colours.
Smaller plants with smaller fruits produced red fruits which
contrasted strongly with the background; these fruits were
dispersed by a distinct set of bird species. The contents of
water, sugar, phenols and tannins did not diVer between
fruits eaten by primates and those eaten by birds. Some
phylogenetic patterns were apparent; primates fed preferen-
tially on a phylogenetically restricted subsample of large
plants with large fruits of the subclass Rosidae. We discuss
why the observed primate dispersal syndrome is most likely
explained by a process of ecological Wtting.
Keywords Seed dispersal syndrome · Fruit colour ·
Nutrients · Frugivore · Food selection
Seed dispersal by animals is an important mutualistic inter-
action between Xeshy-fruited plants and frugivorous ani-
mals (Howe and Smallwood 1982; Bascompte and Jordano
2007). This interaction compensates for the limited mobility
of plants and determines the population dynamics and gene
Xow of plant populations (Howe and Miriti 2000).
Dispersal by animals is particularly important in the tropics
where up to 90% of tropical tree species have Xeshy, verte-
brate-dispersed fruits (Jordano 1992). Animals in turn ben-
eWt from fruit consumption by obtaining nutrients such as
carbohydrates and lipids.
In general, the interactions between fruit-bearing plants
and frugivorous animals are asymmetric and characterised
by a low degree of specialization (Bascompte and Jordano
2007). Most plant species interact with many animal
species (and vice versa), and those plant species that are
Communicated by Christopher Johnson.
Electronic supplementary material The online version of this
article (doi:10.1007/s00442-010-1655-8) contains supplementary
material, which is available to authorized users.
M. Flörchinger · K. Böhning-Gaese
Institut für Zoologie, Abteilung V,
Johannes Gutenberg-Universität, Mainz,
Becherweg 13, 55099 Mainz, Germany
J. Braun · H. M. Schaefer (&)
Department of Evolutionary Biology and Animal Ecology,
Faculty of Biology, University of Freiburg,
Hauptstrasse 1, 79104 Freiburg, Germany
152 Oecologia (2010) 164:151–161
dispersed by few animals tend to be dispersed by abundant
and unspecialized animals. Recent studies on the structure
of mutualistic networks suggest that these networks might
be generated by two processes: ecological Wtting and
coevolutionary convergence (Bascompte and Jordano
2007). Interactions could arise by ecological Wtting where
animals disperse those plants they can most easily exploit
without coevolution taking place. Alternatively, as soon as
an interaction evolves and is successful, more species can
attach to this interaction by convergent evolution of their
traits (Bascompte and Jordano 2007). The consequence
would be the evolution of dispersal syndromes, i.e. non-
random suites of plant traits that are associated with spe-
ciWc disperser groups (Bascompte and Jordano 2007).
Although many studies report the existence of seed dis-
persal syndromes (e.g. van der Pijl 1969; Janson 1983;
Gautier-Hion et al. 1985; Voigt et al. 2004), the dispersal
syndrome hypothesis has also been rejected by a number of
phylogenetically controlled studies (e.g. Fischer and Chapman
1993; but see Lomáscolo et al. 2008). As such, the exis-
tence and the speciWcity of seed dispersal syndromes remain
contentious. For example, Gautier-Hion et al. (1985) and
Dowsett-Lemaire (1988) contrast a combined “bird–monkey-
syndrome” with an “elephant–rodent–ruminant-syndrome”,
whereas others describe distinct dispersal syndromes for birds
and primates (Janson 1983; Poulsen et al. 2002). According
to the latter view, fruits pertaining to a primate dispersal
syndrome are large, have a thick husk and are yellow,
green, brown, orange or red (Janson 1983; Dew and Wright
1998), whereas fruits consumed by birds are small, with a
thin husk and are black, blue, violet, or red (Janson 1983;
Lomáscolo and Schaefer 2010).
The existence of distinct dispersal syndromes of birds
and primates is not only of academic, but also of conserva-
tion interest because mammals and birds can diVer in their
eVects on seed-mediated gene Xow (Jordano et al. 2007).
Moreover, primates are generally hunted more intensely
than birds and are often rare or locally extinct in disturbed
and degraded forests (Johns and Skorupa 1987; Corlett
2007; Nuñez-Iturri and Howe 2007). Therefore, it is impor-
tant to assess if certain plant species are dispersed primarily
by primates because regeneration of these species might
decline with the local extinction of their dispersers.
Although dispersal syndromes are important from an
evolutionary and conservation-oriented perspective, stan-
dardized, quantitative observations of the fruit choice of
birds and primates are often lacking on a community level.
Moreover, few studies have used a phylogenetically con-
trolled approach to examine simultaneously which fruit
traits or combination thereof deWne fruit syndromes. This
gap in the literature is surprising given that fruit morphol-
ogy, biochemistry, and colour can strongly Wlter fruit con-
sumers (Dennis et al. 2007).
It was our aim to study fruit consumption of diurnal ver-
tebrates in a tropical forest by standardised, quantitative
observations while accounting for fruit consumption by
nocturnal animals qualitatively. We speciWcally asked: are
there plant species whose fruits are mainly dispersed by
birds or primates? We then investigated whether plant char-
acteristics (e.g. plant height, habitat), fruit morphology,
fruit colour, or speciWc fruit compounds inXuenced the fruit
choice of primates and birds diVerentially, thereby deWning
seed dispersal syndromes. Because colour perception is in
the eye of the beholder, we modelled fruit colour according
to the sensory sensitivities of birds and primates, respec-
tively. We Wnally explored whether some bird species over-
lapped in their fruit choice with primates.
Materials and methods
Study area
From December 2007 until March 2008, the season of high
fruit availability, we observed frugivores in Kakamega
Forest, a mid-altitudinal tropical rainforest in western
Kenya (0°14–0°21N, 34°47–34°48E). The forest com-
prises 12,000 ha at 1,500–1,700 m a.s.l. (Bleher et al.
2006). The rainy season is from April to November; the
mean monthly temperature lies between 15 and 27°C. The
forest is an eastern remnant of the Congo basin rainforests,
and it is known for its high diversity of plant and animal
species (KIFCON 1994; Bleher et al. 2006). Our study area
covered approximately 250 ha in the northern part of the
forest, in the region of Buyangu. This area is marked by a
heterogeneous habitat structure including all successional
forest stages from grassland, forest edge, to secondary and
primary forest. Very large, typical seed dispersers such as
elephants are now missing although some plants with very
large fruits remain potentially pertaining to a syndrome of
megafaunal fruits (see Guimarães et al. 2008), although
none of these plants fruited during our study period.
Plant species
We observed fruit consumption of 32 plant species (mainly
trees, but also climbers, shrubs and herbs) with Xeshy fruits.
No frugivorous animals were observed on Leea guineensis
and we observed only a single blue-headed coucal (Centropus
monachus) on hill raspberry (Rubus niveus). We thus
excluded both species from further analyses. We observed
four diVerent individuals of each plant species except for
three rare species of which we observed the same individu-
als repeatedly on diVerent days (Table 1). A mean species
accumulation curve calculated over all 32 plant species
shows that we did not observe the full visitor spectrum in
Oecologia (2010) 164:151–161 153
most plants during our 28-h observation period for each
species (Fig. 1). As such, our analyses are based on the rel-
ative attractiveness of plant species for primates and birds
during the peak of fruit availability and not on the entire
fruit consumer set that may visit a plant.
Plant characteristics
For each plant we measured plant height with a clinometer
and categorised its habitat as 1 = undisturbed primary
forest, 2 = disturbed primary forest, 3 = old secondary
forest, 4 = young secondary forest, and 5 = grassland. We
estimated crop size by counting the number of ripe and
unripe fruits on 5% of the plant and extrapolating this esti-
mate of fruit density separately to each of the four quarters
of a plant because fruit numbers were often not evenly dis-
tributed within a plant. We then summed fruit numbers of
each of the quarters to obtain a total for each plant individ-
ual. All estimates were done by a single observer. At the
beginning of the observation period, estimates were
repeated by a second observer until a high repeatability was
ensured. The variable fruit crop size was log10-transformed
for further analysis to achieve heteroscedascity and to
reduce errors in fruit estimates. Finally, we estimated the
Table 1 Observed plant species, number of observed plant individuals (ind.), and number of primate, bird and fruit bat visitors consuming fruits
over 4 observation days
The number of visitors swallowing or pecking at fruits is indicated for birds and primates. Relative visitation rate by primates (Visits), relative time
spent in the trees by primates (Time), and relative fruit consumption by primates (Fruits) were calculated as the mean of the daily ratios over the
4 observation days. All variables are scaled so that a value of 0 indicates that only birds visited, spent time or consumed fruits, whereas a value of
1 indicates that these actions were only carried out by primates. All species were observed for 28 h
Species Plant ind. Primates Birds Swallow
Peck (primate) Peck
Fruit bats Visits Time Fruits
Allophylus abyssinicus 409060 10000
Antiaris toxicaria 4 36 118 15 1 2 50 1 0.25 0.36 0.46
Bridelia micrantha 4 59 294 16 164 0 38 0 0.16 0.43 0.55
Carissa edulis 4 0 33 0 21 0 8 0 0 0 0
Casearia battiscombei 2 6 54 2 0 1 44 0 0.08 0.15 0.3
Cissus oliveri 3 0 23 0 15 0 1 0 0 0 0
Clerodendrum capitatum 4 0 18 0 8 0 10 0 0 0 0
CoVea eugenioides 408030 50000
Culcasia falcifolia 4014090 50000
Cyphostemma cyphopetalum 40550190 283000
Draecaena fragrans 405010 40000
Ficus exasperata 4 50 49 27 7 2 20 2 0.36 0.62 0.83
Ficus cf. lutea 4 30 602 21 192 0 329 0 0.06 0.13 0.25
Ficus sur 4 130 108 61 13 48 65 4 0.57 0.83 0.93
Ficus cf thonningii 4 68 470 40 113 0 31 0 0.12 0.21 0.27
Harungana madagascariensis 4 0 68 0 46 0 2 0 0 0 0
Jasminum Xuminense 40330150 150000
Maesa lanceolata 3 0 21 0 12 0 4 0 0 0 0
Manilkara butugi 4 77 72 35 9 10 23 0 0.44 0.59 0.74
Morus mesozygia 4 82 32 35 3 0 19 0 0.76 0.88 0.95
Peponium vogelii 4130007 00111
Piper guineense 4010060 20000
Pittosporum viridiXorum 4 0 19 0 12 0 3 0 0 0 0
Polyscias fulva 4 1 97 1 57 0 13 0 0.01 0.01 0.01
Prunus africana 4 3 407 3 45 0 97 0 0.03 0.03 0.08
Psychotria peduncularis 408030 50000
Rhamnus prinoides 4 0 19 0 13 0 0 0 0 0 0
Solanum mauritianum 4 0 35 0 1 0 22 0 0 0 0
Toddalia asiatica 4 1 15 1 10 0 3 0 0.05 0.06 0.16
Trilepisium madagascariense 4 94 6 33 3 0 3 0 0.87 0.91 0.9
154 Oecologia (2010) 164:151–161
ratio of ripe fruits to all fruits (ripeness) and arcsine-square
root transformed it for analysis.
Fruit morphology
To obtain fresh fruits we collected fruits directly from the
plants which also required climbing trees that fruited in the
canopy. For logistical reasons we thus measured fruit char-
acteristics mainly within individuals on one or two individ-
uals per species and up to 20 fruits from each individual.
We measured fruit length, diameter, and fresh mass. These
variables correlated strongly; we performed a principal
component (PC) analysis (PCA) of them. PC1 explained
more than 95% of total variance and was included as fruit
size index (FSI) to account for variation in fruit size. We
calculated crop mass of ripe fruits per plant by multiplying
the estimated number of ripe fruits (crop size) with the
mean fresh fruit mass. This variable was log10-transformed
to achieve heteroscedascity. We categorised husk thickness
as 1 = no protection, 2 = thin husk, 3 = thick husk/capsule
and counted the number of seeds. We calculated a seed size
index (SSI) analogue to FSI from seed length and seed
diameter. Again PC1 explained 95% of the variance. We
categorised infructescences as 1 = single fruit, 2 = one to
ten fruits, 3 = ten to 20 fruits, 4 = >20 fruits and recorded
the presence of a peduncle.
Fruit colour and colour perception
The reXectance of 20 ripe fruits, ten leaves, and associated
structures such as non-green bracts and capsules were mea-
sured for each species with a spectrometer (AvaSpec2048
with a DH 2000 Deuterium-Halogen lamp, Avantes).
ReXectance was measured from 300 to 700 nm as the pro-
portion of a standard white reference tile (WS-2). We cal-
culated a PCA on the spectra of ripe fruits (for clarity, we
name the fruit colours as those seen by humans). The Wrst
three PCs explained more than 95% of the total variance
and were included as quantitative variables describing fruit
reXectance in the analyses independent of consumer vision
(PC1, PC2, PC3). High PC1 values indicated relatively
high UV reXection of fruits; low values show low relative
UV reXection. Black and dark purple fruits scored high on
PC2, whereas red fruits scored low on PC2 but had high
values on PC3. Low values on PC3 indicated yellow and
green fruits.
To estimate how birds perceive fruit colours, we mod-
elled the probability of photon capture of the four cone
types that birds use for colour vision. The avian eye model
we used assumes that colour discrimination is mediated by
noise originating in the cones (Vorobyev and Osorio 1998).
The probability of photon capture by each cone can be
transformed into a three-dimensional colour space that is
characterised by the Cartesian coordinates x, y and z (for
calculations, see Endler and Mielke 2005). In general, avian
colour vision is relatively conservative (Ödeen and Håstad
2003). Given that the exact spectral sensitivities of most
birds are unknown and that most frugivorous birds
belonged to the passerines, we used the spectral sensitivi-
ties of the blue tit (Cyanistes caeruleus) to model colour
perception (our conclusions do not change qualitatively if
we used the other violet-sensitive avian visual system).
In contrast to birds, Old World primates have three
cones used for colour vision. We modelled primate percep-
tion of fruit colours using catarrhine spectral sensitivities
determined by Bowmaker et al. (1991) and Osorio and
Vorobyev (1996). We then projected the resulting values
onto a two–dimensional plane with variables x and y fol-
lowing Kelber et al. (2003). We calculated the perception
of brightness (that is variation in light intensity) by analys-
ing the stimulation of the avian double cone and of the
combined input of the middle- and long-wavelength cones
in primates according to Kelber et al. (2003) and Siddiqi
et al. (2004). Finally, we calculated the maximal chromatic
and achromatic (luminance) contrast between fruits and
their background (leaves or any of the associated struc-
tures) for birds and primates as the perceived diVerence in
colour or luminance between fruits and background
(Vorobyev and Osorio 1998).
Chemical fruit traits
After reXectance measurements fruits were dried at a tem-
perature of <50°C to avoid mould. We calculated the water
content by subtracting dry fruit mass from fresh fruit mass.
Fig. 1 Mean accumulation curve (and SE) of the fruit consumer
species of 32 plant species over the observation period of 28 h. We
observed diVerent plant individuals, each for 7 h. Thus, diVerences in
microhabitats among plant individuals and diVerences in ripening
stages among fruits likely explain the more or less pronounced increases
at 8, 15, and 22 h
Oecologia (2010) 164:151–161 155
In contrast to primates, passerine birds cannot digest
sucrose (Martinez del Rio and Stevens 1989). Glucose and
fructose are thus the predominant sugars of bird-dispersed
fruits but not of mammal-dispersed fruits (Baker et al.
1998). Owing to the strong prediction that glucose and
fructose pertain to a bird dispersal syndrome, we focused
on these two sugars and analysed the contents of glucose
and fructose photometrically with enzymatic D-glucose/
D-fructose Boehringer analysis kits (r-biopharm). This
method is robust to the presence of anthocyanins. It does
not measure sucrose, a sugar that occurs in some fruits dis-
persed by mammals such as Old World bats. Phenol con-
centration in fruits was measured using the Prussion blue
method using gallic acid as standard (Price and Butler
1977). Tannin concentration was measured using the pro-
anthocyanidin method using catechin as a standard (Porter
et al. 1986). A limitation of both methods is that not all
phenolic compounds show identical responses to oxidation.
As such, the results of these tests should not be treated as
yielding an estimate of the contents of phenols and tannins
on a per mass basis but rather as equivalents of gallic acid
and catechin, respectively. Nevertheless, these assays cor-
respond well to behavioural data on fruit consumption both
of avian seed dispersers as well as of fruit pests such as
microbes and fungi in two diVerent plant communities
(Schaefer et al. 2003; Cazetta et al. 2008).
Frugivore observations
Plant individuals were observed continuously from 7 a.m.
to 2 p.m. Observations were conducted from camouXaged
hides with binoculars (10 £50) to reduce biases caused by
our presence. We determined the species’ identity and the
number of individuals of all fruit consumers as well as the
duration of their visits. If species identiWcation was diYcult
(e.g. diVerent greenbul species) we recorded the genus or
weight classes of fruit consumers. Weight classes were cat-
egorised following Kingdon (1997) and Dunning (1993):
class 1, small birds (body mass < 18.5 g); class 2, e.g. bul-
buls and weavers (body mass 18.5–40 g); class 3, e.g. bar-
bets and thrushes (body mass 40–100 g); class 4, large birds
(body mass 100–400 g); class 5, squirrels; class 6, horn-
bills; class 7, guenons; and class 8, colobus monkeys
(Appendix 1 in Supplementary material). Finally, we
recorded whether animals fed on ripe or unripe fruits and
classiWed their fruit handling behaviour. That is we
recorded whether animals swallowed fruits or carried them
away, pecked at them or whether they dropped the fruit.
We conducted night observations to record the presence
of nocturnal vertebrates feeding on fruits. At least one indi-
vidual from each plant species was observed with infrared
night vision glasses (5 £52 and 2 £24) for 2 h from night-
fall. We integrated the number of nocturnal fruit consumers
in our calculations as an independent variable to account
for their possible eVect in deWning fruit syndromes.
Response variables
We restricted analyses to visits of animals that fed on fruits.
We evaluated the ratio of primates relative to bird visitors
on three levels; the number of visits per plant individual,
the duration of visits by each consumer group, and Wnally
the estimated number of consumed fruits by each animal
group. For the number of visits per plant, we calculated the
ratio of primate visits in relation to the total number of bird
and primate visits (V) such that a ratio of 0% indicated that
a plant species was only visited by birds, whereas a ratio of
100% indicated that it was only visited by primates.
We calculated the sum of duration of visits of all bird
and primate individuals per plant. If the duration of a visit
could not be determined with certainty (21% of total obser-
vations), we substituted it by the mean duration of the same
animal species or weight class in the same plant. A ratio of
duration of time spent by primates in relation to time spent
by primates and birds (T) was calculated.
Finally, we estimated the ratio of the number of fruits
consumed by primates and birds, respectively. Often, we
could not record exactly the number of consumed fruits
because several animals fed concomitantly on the plant. We
therefore used the mean number of fruits consumed per
minute for each class of fruit consumers from Farwig et al.
(2006) and Kirika et al. (2008a, b; Appendix 1 in Supple-
mentary material). To estimate fruit consumption we calcu-
lated the mean fruit consumption per minute for each
weight class and multiplied it by the total time (in minutes)
spent by each animal species on each plant. We then calcu-
lated the ratio of the estimated number of fruits consumed
by primates in relation to the estimated total number of
fruits consumed by primates and birds for each plant (F).
We calculated the mean of the daily ratio values. If a tree
was not visited during one day, the mean was calculated of
the remaining days. All ratio data were arcsine-square root
transformed. As expected, all three ratio values (V, T, F)
were highly correlated (Pearson correlations all > r= 0.97,
P< 0.001, n= 30). We therefore restricted the following
analyses to the relative importance of primate visits (V), the
variable we had directly observed.
Statistical analyses
To visualise the complex patterns of similarity among
plants (in their visitor spectra) and animals (in their fruit
choice) we used non-metric multidimensional scaling
(NMDS). This method is an iterative search procedure that
places objects from a distance matrix of visitor spectrum
(or fruit choice) in a space of minimized dimensionality
156 Oecologia (2010) 164:151–161
while it conserves the rank order of species as far as possi-
ble. NMDS is a robust, multivariate analysis that illustrates
the relationship among species characterised by multiple
variables. All categorical variables with more than two cat-
egories were ranked and used as continuous variables; vari-
ables with only two categories were used as dummy
variables. We Wrst calculated pairwise correlations between
all plant and fruit variables. We then applied NMDS to ana-
lyse diVerences in visitor spectra of plant species (Kruskal
1964). NMDS produces a two-dimensional description of
similarity among species (based on the Bray index). This
method illustrates the increasing similarity among plant
species in their consumer spectrum by decreasing distances
in two-dimensional space (Fig. 2). Vector calculation then
identiWes those plant traits that most reliably associate with
the observed proximity among plant species. DiVerences
between mathematical similarity and graphic distances are
calculated as “stress” and are minimised by iterative pro-
As a basis for the NMDS, we developed a matrix with
plants in rows and animal species in columns. Each cell
contained the number of interactions (number of animals
feeding on fruits). We only included animal species that
paid more than 20 visits to fruiting plants to avoid biases
introduced by rare observations. This criterion was used to
assure the robustness of our results. Psychotria peduncu-
laris was only visited by very rare animal species and was
thus excluded from the matrix, which Wnally contained 29
plant and 22 animal species. The matrix was square root
and Wisconsin transformed to minimize outliers. We then
calculated the similarity of the visitor spectra of plants with
the Bray index. The Kruskal stress 1 was 18% and in the
normal range of ecological studies (Kruskal 1964). We
introduced plant and fruit variables as vectors in the NMDS
plane and tested their inXuence on the multidimensional
species composition with correlations. We tested for sig-
niWcance by drawing 1,000 randomized permutations of the
While the NMDS explores the similarity among species,
it is not a predictive statistical analysis. To determine which
plant and fruit variables predicted the visitation rate of pri-
mates, we calculated a multiple regression model using
only those seven variables that correlated with the dimen-
sions of the NMDS space because the fruit choices of pri-
mates and birds segregated on NMDS 1. To assess the
relative importance of each variable we used multi model
comparisons based upon the summed Akaike weight of
each variable as recommended by Burnham and Anderson
(2002). We calculated seven models with three independent
variables each so that each variable was present in three
models and tested once with each of the other variables.
The Akaike weight of each model represents the likelihood
that this model is the best available model. The summed
Akaike weight for each variable thus represents the relative
importance of each variable in contributing to the best
model; higher values indicate more important variables. We
tested models for phylogenetic eVects and these were
corrected using phylogenetic eigenvector regression (see
Phylogenetic eVects
Species cannot be regarded as independent data points in
regressions because they diVer in relatedness. We compen-
sated for such phylogenetic eVects using phylogenetic
eigenvector regression (Diniz-Filho et al. 1998). The phylo-
genetic relationship between species (family to class nomi-
nation) was taken from Stevens (2008). We generated a
phylogenetic distance matrix D with the plant species in
rows and in columns. The matrix cells contain an index of
the phylogenetic relatedness of two plant species. If two
species belonged to the same genus, they scored 1, if they
belonged to diVerent genera but to the same family, they
scored 2, and so on for order (3), subsubclass (4), subclass
(5), class (6), monocotyl or dicotyl (7), or angiosperms (8).
Next, this phylogenetic distance matrix D was double cen-
tred, i.e. each cell was transformed to ¡0.5 D2ij, then the
row and the column mean was subtracted and the grand
mean was added (Gower 1966). Finally, we calculated
eigenvectors from the transformed matrix.
Fig. 2 Position of all plant species (black type) and selected animal
species (grey italics) in the non-metric multidimensional scaling
(Nmds) plane. Animal species are only displayed if located >0.7 scale
values distant from the centre on at least one axis. Additionally, plant
and fruit characteristics are displayed that correlated signiWcantly with
one or two of the NMDS axis with the two-dimensional NMDS scores
(*P< 0.05, **P< 0.01, ***P< 0.001). Length of the vectors is pro-
portional to the correlation coeYcient
Oecologia (2010) 164:151–161 157
We tested for phylogenetic eVects in the residuals of all
the multiple regressions. We calculated Moran’s I autocor-
relation coeYcients for all phylogenetic distance classes
(Gittleman and Kot 1990). All regressions with signiWcant
Moran’s I values were corrected by phylogenetic eigenvector
regression (Diniz-Filho et al. 1998), including the phyloge-
netic eigenvector with the highest eigenvalue in the regres-
sion model. Residuals of the regression were again tested
for phylogenetic autocorrelation. After including the phylo-
genetic eigenvector, none of Moran’s I values were signiW-
cantly diVerent from zero. Including the phylogenetic
eigenvector in the regressions did not change any of the
All calculations were conducted in R 2.6.0 and 2.10
(R Development Core Team 2005). Packages “vegan”
(Oksanen et al. 2008) and “MASS” (Venables and Ripley
2002) were used for NMDS. Package “ape” (Paradis et al.
2004) was used for the calculation of Moran’s I and package
“bblme” for the calculation of Akaike weights.
Fruits and fruit consumers
In total, 3,375 visits by fruit consumers were observed. We
identiWed 76 bird species, four primate species and one
squirrel species (Appendix 1 in Supplementary material).
Primates fed on 45% of all plant species totalling 19% of all
visits (Table 1; Appendices 1 and 2 in Supplementary
material). Birds fed on all but one plant species, Peponium
vogelii, which was consumed only by primates (Table 1).
Frugivorous birds mostly swallowed (50.3%) and pecked
(49.9%) at fruits, rarely dropped them (0.04%) or carried
them away in their beaks (0.002%; the cumulative percent-
age is >100% because some individuals behaved in more
than one way). Primates mostly swallowed fruits entirely
(79%), dropped them (41%) or ate them piecemeal (25%).
Birds pecked at large fruits more frequently than at small
fruits (linear regression: pecking birds versus FSI: =65.1,
t= 2.4, P= 0.02, r2=0.18, n= 30). We included in the fol-
lowing analysis all interactions in which animals fed on
fruits. However, the results do not change qualitatively if
we only include interactions where fruits were swallowed.
Primates fed on fruits of plants with a mean height of
27.7 m, birds fed on fruits of plants with a mean height
of 15.7 m. Fruits eaten by primates had a mean length of
21 mm (range 10–75 mm) and a mean weight of 5.8 g
(range 0.4–49 g). Fruits eaten by birds were on average
13 mm long (range 4–29 mm) and weighed 1.4 g (range
0.02–8.3 g). Both groups fed on red, violet, dark violet,
yellow and orange fruits. Black, white and blue fruits were
only consumed by birds, whereas green fruits were only
consumed by primates (Appendix 3 in Supplementary
material). Fruit bats of undetermined species were observed
feeding on Ficus sur, Ficus exasperata, Antiaris toxicaria
(all Moraceae) and Cyphostema cyphopetalum (Vitaceae;
Table 1).
Correlation of plant and fruit characteristics
We found multiple correlations among plant and fruit char-
acteristics (all Pearson’s correlations). Small fruits stood in
larger infructescenses than large fruits (FSI vs. infructescence:
r=¡0.63, P< 0.001, n= 30 for all correlations), and fewer
small fruits had a peduncle (FSI versus peduncle: r= 0.52,
P< 0.01). Plant height correlated with fruit size (r= 0.46,
P< 0.01), fruit number (r=0.60, P< 0.001), and inversely
with the proportion of simultaneously ripe fruits
(r=¡0.38, P= 0.04). Plants with fewer fruits and plants
with a lower crop mass produced fruits with a higher chro-
matic contrast to the background according to a model of
avian vision (crop size, r=¡0.45, P=0.01; crop mass,
r=¡0.39, P= 0.03). Plants producing many ripe fruits at
the same time produced fruits with higher fructose and
glucose contents than plants with a more staggered fruit
production (proportion ripeness versus sugar: r= 0.42,
P= 0.02). The contents of glucose and fructose were higher
in plants from open habitats compared to those in undis-
turbed forest (r= 0.49, P< 0.01).
The phylogenetic eigenvector with the highest eigen-
value (Ev1) correlated negatively with fruit size (FSI vs.
Ev1: r=¡0.37, P= 0.045) and plant height (plant height
versus Ev1: r=¡0.64, P<0.001).
Visitor spectra and fruit choice
The NMDS placed the plants and animals in a two-dimen-
sional space according to their similarity in visitor spectra
and fruit choice, respectively. The amount of variance
explained by each axis is indicated by the range of values
(Fig. 2).
In general, the range of plants was larger than that of ani-
mals, which shows that some plants were more specialized
on few dispersers than animals on few plants. The fruit
choice of primates diVered from that of birds. All primate
species fed on similar fruit types (Peponium vogelii,
Trilepisium madagascariense, all Ficus spp., Antiaris
toxicaria), all located between 1.1 and 1.7 on NMDS 1 and
between ¡0.2 and 0.2 on NMDS 2. Birds were more catho-
lic in their fruit choice, which ranged from ¡0.65 on
NMDS 1 (Cameroon sombre greenbul, Andropadus curvi-
ostris) to 1.1 on NMDS 1 (black and white casqued horn-
bill, Bycanistes subcylindricus). This hornbill, the largest
bird species, had the highest score on NMDS 1 and was
therefore most similar in fruit choice to the primates. In
158 Oecologia (2010) 164:151–161
general, the body mass of birds was positively correlated
with NMDS 1 score (r=0.55, P= 0.017, n=18). Birds
also covered a wide range of values on NMDS 2,
from ¡1.2 (violet-backed starling, Cinnyricinclus leucogaster)
to 0.7 (yellow-rumped tinkerbird, Pogoniulus subsulphu-
reus). NMDS 2 scores of birds correlated negatively with
their average group size while visiting a plant (r=¡0.50,
P= 0.034, n= 18); there was no correlation with body
To illustrate how plant and fruit characteristics inXu-
enced the fruit choice of primates and birds, we plotted
them as vectors in the NMDS plane. The relative impor-
tance of primates as plant visitors increased with increasing
values on NMDS 1. Increasing height of a plant, increasing
crop mass and increasing fruit size distinguished plants that
were visited by primates from those that were not. As such,
primates fed on tall plants with large fruits and high crop
mass, whereas birds fed primarily on smaller fruits and
fruits growing in large infructescences.
Plant height and crop mass also showed a gradient along
the second axis that distinguished between diVerent bird
species. Fruit colour did not separate primates and birds as
fruit consumers but separated diVerent bird species along
NMDS 2. Small plants with low fruit production produced
fruits with high chromatic contrast for avian vision and
were visited by a distinct subset of bird species, in particular
the yellow-rumped tinkerbird, yellow-whiskered greenbul
(Oriolus auratus) and joyful greenbul (Chlorocichla
laetissima). Red fruits indicated by high values on PC3
were associated with high chromatic contrast against the
background. Plants with such highly contrasting fruits had
a high proportion of ripe fruits.
Phylogenetic eigenvectors 1 and 13 had a signiWcant
inXuence on the distribution of plant species in the NMDS
plane. Eigenvector 1 had a high eigenvalue of 27% and ran
in parallel to NMDS 1 separating plants whose fruits were
consumed by primates from those consumed only by birds.
Primates fed primarily on plants of the subclass Rosidae
with one Asteridae, Manilkara butugii, as an exception.
Eigenvector 13 had a very low eigenvalue of 0.02% and
was consequently neglected as an explanatory variable.
Model of fruit choice
A multiple regression model of relative primate visitation
rates as dependent variable against the plants and fruit char-
acteristics that were signiWcant in the NMDS (plant height,
crop mass, ripeness, FSI, infructescence, PC3, chromatic
contrast detected by birds) showed that fruit size, plant
height, and crop mass were the most important variables
explaining fruit choice by primates (global model:
F7,22 =6.12, r2=0.66, P< 0.001, Akaike’s information
criterion = 5.22; Table 2). Moran’s I values of the residuals
were not signiWcantly diVerent from zero. Thus, no correc-
tion for phylogenetic eVects was necessary in this analysis.
Primates were observed feeding on 45% of all fruiting plant
species, whereas birds fed on all species except Peponium
vogelii. Taken together, the four primate species had a more
specialized fruit diet than the frugivorous birds. Moreover,
the fruits consumed by primates were more similar to each
other than the fruits consumed by birds. This result reXects
the fact that the four species of primates were more similar
to each other in terms of morphology and size than the 18
species of birds included in the NMDS.
Plant and fruit morphology inXuence fruit choice
Plant height, fruit size, and crop size were all correlated.
These correlations complicate causal inferences on the rela-
tive importance of each of the factors. Yet, such non-
random correlations form the functional basis of the seed
dispersal syndrome hypothesis. The interpretation of such
syndromes requires care. Although we found overall little
evidence for phylogenetic signals in our analyses, the phy-
logenetic eigenvector with the highest eigenvalue corre-
lated with the relative importance of primates as fruit
visitors as well as with plant height and fruit size. As such,
our results indicate that primates fed on a phylogenetically
restricted subsample of large plants with large fruits of the
subclass Rosidae. The primate seed dispersal syndrome that
we found can be explained by ecological Wtting without
evolutionary change, that is that primates consume the
fruits of a phylogenetically restricted subsample that they
can exploit most eYciently. Alternatively, the pattern could
Table 2 Global model on plant traits that contribute to diVerential
fruit choice of primates and birds
PC Principal component
aThe relative importance of plant traits for explaining the proportion
of fruits consumed by primates is indicated by high summed Akaike
weights. Large fruit size, tall plants, and large crop mass characterised
plants whose fruits were consumed by primates
Plant traitsaSE Akaike weight
Intercept ¡0.1134 0.2900
Fruit size index 0.3577 0.1038 0.9957
Plant height 0.0104 0.0061 0.8165
Crop mass 0.0242 0.0702 0.8162
Fruit colour (PC 3) ¡0.0472 0.0510 0.1308
Chroma ¡0.0008 0.0041 0.1287
Infructescence 0.0179 0.0403 0.0568
Ripe fruits 0.0026 0.0023 0.0552
Oecologia (2010) 164:151–161 159
be explained by one-sided evolution of one of the partners
or by mutual coevolution where both partners adapt to each
other. We are unable to diVerentiate between these hypoth-
eses directly. However, the covariance between plant and
fruit size likely indicates genetic or developmental con-
straints. Consequently, this covariance likely restricts the
evolutionary potential for plants to respond to selection
upon fruit size imposed by seed dispersers.
The novel aspect of our results is that the seed dispersal
syndromes were correlated to plant height and therefore
likely explicable by functional (morphological and physio-
logical) linkages between complex plant traits. Such link-
ages are termed phenotypic integration and are explicable by
genetic, physiological or developmental constraints. Covari-
ance among reproductive traits and between vegetative and
reproductive plant traits have been noted previously and are
attributable to phenotypic responses to the environment,
genetic responses to natural selection, and the limits of such
responses imposed by developmental and genetic linkage
such as epistasis and pleiotropy (e.g. Herrera 2002; Perèz
et al. 2007). As such the covariance among distinct plant
traits that we observed suggests that ecological Wtting is the
most parsimonious hypothesis to explain trait matching
between fruiting plant and fruit-eating primates.
According to our models, fruit size was the most impor-
tant trait determining the relative importance of primates as
fruit consumers. Primates consume only fruits larger than
10 mm (Dew and Wright 1998), whereas few birds con-
sumed large fruits (up to 29 mm). Importantly, the func-
tional overlap between both groups is less than these
numbers suggest. Birds peck at large fruits which reduces
the likelihood of successful seed dispersal, whereas prima-
tes can swallow larger fruits entirely. Our conclusion that
body size (and gape width) is the major determinant of fruit
choice is corroborated by the observation that the largest
frugivorous bird, the black and white casqued hornbill, was
more similar to primates in fruit choice than to other birds.
Plant height is another important factor shaping the
diVerential fruit choice of primates and birds. Primates con-
sumed fruits from plants with an average height of 27.7 m;
considerably higher than the fruiting plants visited by birds
(mean height of 15.7 m). The higher foraging height of pri-
mates might, at least partly, be explained by the larger fruits
in the canopy rather than by an intrinsic foraging preference
for higher trees. This pattern is probably further exacer-
bated by the fact that larger primates may not be able to
climb small bushes with weak branches.
Finally, crop mass was positively correlated to the visita-
tion rate of primates. This result is consistent with the gen-
erally high attractiveness of plants with a large fruit crop.
Large fruit crops might be especially important for primates
owing to their relatively high body mass coupled with their
social foraging behaviour (Chapman and Chapman 2000).
Yet, given that birds are likewise attracted by large fruit
crops (Ortiz-Pulido and Rico-Gray 2000), our result might
rather be explicable by the correlations between fruit crop
and fruit size. Similarly, our result that primates feed pri-
marily on single fruits might be explicable by the negative
correlation between fruit number within an infructescence
and fruit size.
The only fruit consumed exclusively by primates, Pepo-
nium vogelii, had a very thick husk which birds cannot
open with their bills. As such, this plant is specialized for
seed dispersal by primates; all birds are excluded from its
dispersal, even the large-sized hornbills. The missing inter-
action between P. vogelii and birds is a “forbidden link”
(Jordano et al. 2003). This species is thus expected to be
most severely aVected by the over-hunting of primates.
We emphasise that our observation period was not suY-
ciently long to account for a full visitor spectrum of most
plants. Thus, our analyses capture the relative attractiveness
of plants for primates and birds during diVerent days of the
4-month observation period of peak fruit availability but
not the full visitor spectrum of each plant. We found no
pronounced seasonality in the abundance of frugivores
within our observation period making it unlikely that diVer-
ences in the observed visitor spectrum among plant species
are due to changes in consumer abundances rather than
diVerences in fruit traits. However, it is important to note
that the patterns of fruit choice are likely to change accord-
ing to the pronounced seasonality of fruit availability in
tropical forests. This is because fruit consumers are more
choosy at times of high fruit availability compared to sea-
sons of relative fruit scarcity (Schaefer and Schaefer 2006).
Given that our study period covered the peak of fruit avail-
ability, the diVerences in the foraging behaviour of primates
and birds might be less pronounced in the Kakamega forest
during the rest of the year.
Fruit colour and fruit compounds
Given that fruit colour diVers in a large sample of 402 plant
species whose fruits were either dispersed by primates or
by birds (Lomáscolo and Schaefer 2010), it is perhaps sur-
prising that fruit colour had no inXuence on the relative
number of visits by birds and primates. We suggest that this
discrepancy is due to the considerably lower sample size of
our study. However, fruit colour variation did explain
diVerential fruit choice among bird species. In the NMDS,
the vector of increasing chromatic contrasts of fruits against
their background (as seen by birds) was parallel to the sec-
ond axis of the NMDS. In particular, birds such as yellow-
rumped tinkerbirds and grey-throated barbets fed primarily
on small plants with few, red fruits that contrasted strongly
with the background. The negative correlation between
scores on NMDS 2 and group size of birds is at least partly
160 Oecologia (2010) 164:151–161
explicable by the social behaviour of violet-backed star-
lings that occurred mostly in groups of more than 23 indi-
viduals and that scored low on NMDS 2.
Plants producing relatively few fruits had relatively high
chromatic contrast to the background. We suggest that this
pattern is adaptive because increased chromatic contrast
facilitates long-distance fruit detection by frugivorous birds
(Cazetta et al. 2009). Importantly, fruit detection is particu-
larly critical in plants that produce few fruits. The larger a
fruit crop, the easier it is detected by means other than
visual contrasts (i.e. by size and by the activity of other fru-
givorous animals). At the same time plants with few fruits
had a large proportion of ripe fruits which may increase
fruit detection. On the other hand a large proportion of ripe
fruits may simply reXect the low attractiveness of such
plants to frugivores leading to the accumulation of ripe,
undispersed fruits.
The fruit compounds we measured did not diVerentiate
the fruits consumed primarily by primates and those con-
sumed primarily by birds. Our results show that the content
of glucose and fructose depended mainly on light availabil-
ity in open habitats rather than on disperser identity. We
can obviously not exclude the possibility that protein, lipids
or secondary compounds other than the indices of phenolic
contents that we tested might diVerentially aVect the feed-
ing behaviour of primates and birds. Taken together, the
novel contribution of this study is that fruit syndromes in
this African rainforest are deWned by plant and fruit size
and most likely characterised by a process of ecological
We found a primate plant-syndrome that was characterised
by plant and fruit size as well as crop mass but not by fruit
colour or the contents of phenols, tannins, glucose or fruc-
tose. Our data suggest that plants’ potential to respond to
selection imposed by seed dispersers is constrained by
covariance between fruit and plant morphology. A bird syn-
drome was less apparent because birds fed on a larger vari-
ety of fruit types. Fruit colour explained diVerential fruit
choice of bird species. Small plants with few fruits pro-
duced red fruits that contrasted strongly with the back-
ground and were consumed by a distinct set of bird species.
Few plant species depended strongly on primates as seed
dispersers. It is conceivable that the dispersal of species
like P. vogelii might diminish if primates are overexploited
in this region.
Acknowledgments We would like to thank the people around
Kakamega Forest for their hospitality and support and the Kenya
Wildlife Service for their permission to work in Kakamega Forest.
N. Sajita, D. Shitandayi, J. Andaye, M. Wendy, L. Mutola, M. Nzisa,
T. Ikime, M. Njeri and S. Karimi Murage helped during Weldwork. We
thank I. Malombe for species categorization and D. Berens, M. Schleuning
and L. Crick, Christopher Johnson and two anonymous referees for
useful comments. Financial support was provided by the BMBF
(BIOTA East Africa 01LC0025/E11).
Baker HG, Baker I, Hodges SA (1998) Sugar composition of nectars
and fruits consumed by birds and bats in the tropics and subtropics.
Biotropica 30:559–586
Bascompte J, Jordano P (2007) Plant-animal mutualistic networks: the
architecture of biodiversity. Annu Rev Ecol Evol Syst 38:567–
Bleher B, Uster D, Bergsdorf T (2006) Assessment of threat status and
management eVectiveness in Kakamega forest, Kenya. Biodivers
Conserv 15:1159–1177
Bowmaker JK, Astell S, Hunt DM, Mollon JD (1991) Photosensitive
and photostable pigments in the retinae of Old World monkeys.
J Exp Biol 156:1–19
Burnham KP, Anderson DR (2002) Model selection and multi-model
inference. Springer, New York
Cazetta E, Schaefer HM, Galetti M (2008) Does attraction to frugi-
vores or defense against pathogens shape fruit pulp composition?
Oecologia 155:277–286
Cazetta E, Schaefer HM, Galetti M (2009) Why are fruits colorful? The
relative importance of achromatic and chromatic contrasts for
detection by birds. Evol Ecol 23:233–244
Chapman CA, Chapman LJ (2000) Determinants of group size in pri-
mates: the importance of travel costs. In: Boinski S, Garber P
(eds) On the move: how and why animals travel in groups.
Chicago University Press, Chicago, pp 24–42
Corlett RT (2007) The impact of hunting on the mammalian fauna of
tropical Asian forests. Biotropica 39:292–303
Dennis AJ, Schupp EW, Green RJ, Westcott DA (2007) Seed dis-
persal: theory and its application in a changing world. CABI,
Development Core Team R (2005) R: A language and environment for
statistical computing. R foundation for Statistical Computing,
Dew JL, Wright P (1998) Frugivory and seed dispersal by four species
of primates in Madagascar’s eastern rainforest. Biotropica
Diniz-Filho JAF, Sant’Ana CER, Bini LM (1998) An eigenvector
method for estimating phylogenetic inertia. Evolution 52:1247–
Dowsett-Lemaire F (1988) Fruit choice and seed dissemination by
birds and mammals in the evergreen forests of upland Malawi.
Rev Ecol 43:251–285
Dunning JB Jr (1993) CRC handbook of avian body masses. CRC
Press, Boca Raton
Endler JA, Mielke PW (2005) Comparing entire colour patterns as
birds see them. Biol J Linn Soc 86:405–431
Farwig N, Böhning-Gaese K, Bleher B (2006) Enhanced seed dispersal
in fragmented and disturbed forests? Oecologia 147:238–252
Fischer KE, Chapman CA (1993) Frugivores and fruit syndromes:
diVerences in patterns at the genus and species level. Oikos
Gautier-Hion A, Duplantier JM, Quris R, Feer F, Sourd C, Decoux JP,
Dubost G, Emmons L, Erard C, Hecketsweiler P, Moungazi A,
Roussilhon C, Thiollay JM (1985) Fruit characters as a basis of
fruit choice and seed dispersal in a tropical forest vertebrate com-
munity. Oecologia 65:324–337
Gittleman JL, Kot M (1990) Adaptation: statistics and a null model for
estimating phylogenetic eVects. Syst Zool 39:227–241
Oecologia (2010) 164:151–161 161
Gower JC (1966) Some distance properties of latent root and vector
methods used in multivariate analysis. Biometrika 53:325–338
Guimarães PR, Galetti M, Jordano P (2008) Seed dispersal anachronisms:
rethinking the fruits extinct megafauna ate. PLoS One 3:e1745
Herrera CM (2002) Correlated evolution of fruit and leaf size in bird-
dispersed plants: species-level variance in fruit traits explained a
bit further. Oikos 97:426–432
Howe HF, Miriti MN (2000) No question: seed dispersal matters.
Trends Ecol Evol 15:434–436
Howe HF, Smallwood J (1982) Ecology of seed dispersal. Annu Rev
Ecol Syst 13:201–228
Janson JH (1983) Adaptation of fruit morphology to dispersal agents
in a neotropical forest. Science 219:187–188
Johns AD, Skorupa JP (1987) Responses of rainforest primates to hab-
itat disturbance: a review. Int J Prim 8:157–191
Jordano P (1992) Fruits and frugivory. In: Fenner M (ed) Seeds: the
ecology of regeneration in plant communities. CAB International,
Wallingford, pp 105–156
Jordano P, Bascompte J, Olesen JM (2003) Invariant properties in
coevolutionary networks of plant-animal interactions. Ecol Lett
Jordano P, García C, Godoy JA, García-Castaño JL (2007) DiVerential
contribution of frugivores to complex seed dispersal patterns.
Proc Natl Acad Sci USA 104:3278–3282
Kelber A, Vorobyev M, Osorio D (2003) Animal colour vision–behav-
ioural tests and physiological concepts. Biol Rev 78:81–118
Kingdon J (1997) The Kingdon Weld guide to African mammals. Aca-
demic Press, London
Kirika JM, Bleher B, Böhning-Gaese K, Chira R, Farwig N (2008a)
Fragmentation and local disturbance of forests reduce frugivore
diversity and fruit removal in Ficus thonningii trees. Basic Appl
Ecol 9:663–672
Kirika JM, Farwig N, Böhning-Gaese K (2008b) EVects of local dis-
turbance of tropical forests on frugivores and seed removal of a
small-seeded afrotropical tree. Conserv Biol 22:318–328
Kruskal JB (1964) Non-metric multidimensional scaling: a numerical
method. Psychometrika 29:115–129
Lomáscolo SB, Schaefer HM (2010) Signal convergence in fruits:
a result of selection by frugivores? J Evol Biol 23:614–624
Lomáscolo SB, Speranza P, Kimball R (2008) Correlated evolution of
Wg size and color supports the dispersal syndromes hypothesis.
Oecologia 156:783–796
Martinez del Rio C, Stevens BR (1989) Physiological constraint on
feeding behavior: intestinal membrane disaccharidases of the
starling. Science 243:794–796
Nuñez-Iturri G, Howe HF (2007) Bushmeat and the fate of trees with
seeds dispersed by large primates in a lowland rainforest in west-
ern Amazonia. Biotropica 39:348–354
Ödeen A, Håstad O (2003) Complex distribution of avian color vision
systems revealed by sequencing the SWS1 opsin from total DNA.
Mol Biol Evol 20:855–861
Oksanen J, Kindt R, Legendre P, O’Hara B, Simpson GL, Stevens
MHH, Wagner H (2008) Vegan: community ecology package,
R package version 1.13-1.
ON KIFC (1994) Kakamega Forest. The oYcial guide. Kenya Indige-
nous Forest Conservation Programme, Nairobi
Ortiz-Pulido R, Rico-Gray V (2000) The eVect of spatio-temporal var-
iation in understanding the fruit crop size hypothesis. Oikos
Osorio D, Vorobyev M (1996) Colour vision as an adaptation to frugi-
vory in primates. Proc R Soc B 263:593–599
Paradis E, Claude J, Strimmer K (2004) APE: analyses of phylogenet-
ics and evolution in R language. Bioinformatics 20:289–290
Pérez F, Arroyo MTK, Medel R (2007) Phylogenetic analysis of Xoral
integration in Schizanthus (Solanaceae): does pollination truly
integrate corolla traits? J Evol Biol 20:1730–1738
Porter L, Hrstich L, Chan B (1986) The conversion of procyanidins and
prodelphinidins to cyanidin and delphinidin. Phytochem 25:223–
Poulsen JR, Clark CJ, Connor EF, Smith TB (2002) DiVerential
resource use by primates and hornbills: implications for seed
dispersal. Ecology 83:228–240
Price ML, Butler LG (1977) Rapid visual estimation and spectropho-
tometric determination of tannin content of sorghum grain.
J Agric Food Chem 25:1268
Schaefer HM, Schaefer V (2006) The fruits of selectivity: how birds
forage on Goupia glabra fruits of diVerent ripeness. J Ornithol
Schaefer HM, Schmidt V, Winkler H (2003) Testing the defence trade-
oV hypothesis: how contents of nutrients and secondary com-
pounds aVect fruit removal. Oikos 102:18–328
Siddiqi A, Cronin TW, Loew ER, Vorobyev M, Summers K (2004)
InterspeciWc and intraspeciWc views of color signals in the straw-
berry poison frog Dendrobates pumilio. J Exp Biol 207:2471–2485
Stevens PF (2008) Angiosperm phylogeny website, version 9. http:// Accessed 13 Sept 2008
van der Pijl L (1969) Principles of dispersal in higher plants. Springer,
New York
Venables WN, Ripley BD (2002) MASS: modern applied statistics with
S. Springer, New York.
Voigt FA, Bleher B, Fietz J, Ganzhorn JU, Schwab D, Böhning-Gaese
K (2004) A comparison of morphological and chemical fruit traits
between two sites with diVerent frugivore assemblages. Oecolo-
gia 141:94–104
Vorobyev M, Osorio D (1998) Receptor noise as a determinant of
colour thresholds. Proc R Soc Lond B 265:351–358
... There are many factors to help seed dispersion such as gravity, water, wind or animal but animals are the main factor. Plant species have evolved different fruit traits to attract seed dispersion factors, especially specific frugivores (Van der Pjil 1982), with a range of fruit traits reported to influence fruit choice of frugivores including size (Chen et al. 2004;Flörchinger and Braun 2010;Galetti et al. 2011;Kitamura et al. 2002), type of fruits (Chen et al. 2004;Stiles 1989), colour (Chen et al. 2004;Galetti et al. 2011;Gautier-Hion et al. 1985;Kitamura et al. 2002;Wheelwright & Janson 1985), physical protection of fruits (Chen et al. 2004;Gautier-Hion et al. 1985), nutrients (Flörchinger & Braun 2010;Kitamura et al. 2002) and phenology (Chen et al. 2004), as well as seed traits such as seed size (Chen et al. 2004;Galetti et al. 2011;Kitamura et al. 2002), seed protection (Chen et al. 2004;Gautier-Hion et al. 1985) and seed number (Chen et al. 2004;Gautier-Hion et al. 1985;Kitamura et al. 2002). Fruit size is an important trait that affects seed dispersal because fruit size relates to body size of frugivores. ...
... There are many factors to help seed dispersion such as gravity, water, wind or animal but animals are the main factor. Plant species have evolved different fruit traits to attract seed dispersion factors, especially specific frugivores (Van der Pjil 1982), with a range of fruit traits reported to influence fruit choice of frugivores including size (Chen et al. 2004;Flörchinger and Braun 2010;Galetti et al. 2011;Kitamura et al. 2002), type of fruits (Chen et al. 2004;Stiles 1989), colour (Chen et al. 2004;Galetti et al. 2011;Gautier-Hion et al. 1985;Kitamura et al. 2002;Wheelwright & Janson 1985), physical protection of fruits (Chen et al. 2004;Gautier-Hion et al. 1985), nutrients (Flörchinger & Braun 2010;Kitamura et al. 2002) and phenology (Chen et al. 2004), as well as seed traits such as seed size (Chen et al. 2004;Galetti et al. 2011;Kitamura et al. 2002), seed protection (Chen et al. 2004;Gautier-Hion et al. 1985) and seed number (Chen et al. 2004;Gautier-Hion et al. 1985;Kitamura et al. 2002). Fruit size is an important trait that affects seed dispersal because fruit size relates to body size of frugivores. ...
... Fruit size is an important trait in fruit selection by frugivores, (Peña et al. 2020& Valenta et al. 2020, with selected fruit size related to mouth size (Valido et al. 2011) -i.e. large fruits consumed more often by large frugivores and small fruits by small frugivores (Flörchinger & Braun 2010). Fruit colour also affects the fruit visitation of frugivores. ...
Fruits are a major food resource for wildlifes and have evolved different traits which attract specific frugivores and facilitate seed dispersal. This study examines the quantity of the frequency of fruit tree species, distribution amongst fruit traits and estimates the potential availability of the fruit resource for frugivores in a 16-ha permanent forest plot at Doi Suthep–Pui, Thailand. The similarity amongst traits for fleshy fruited species was explored using Principal Component Analysis (PCA). Fleshy fruited species comprised 122 of 208 tree species >2 cm diameter at breast height (DBH) recorded in the permanent plot. Amongst fleshy fruited species, small fruits (length <20 mm) were most common (63.16% of species) while large fruits were rare (4.1%). Black was the most common fruit colour (43.4%). Principal Component Analysis of fruit traits explained 57% of total variance on the first three axes, and allowed identification of three species groups. Litsea martabanica and Persea gamblei are the greatest density and represented the major PCA group; black, small-sized and thin husk indehiscent fruits. These fruiting trees scattered throughout the permanent plot and were of good regeneration status. Indicating fleshy fruit can be a food resource for frugivores especially small-sized fruit. Furthermore, large-fruited species such as Madhuca floribunda is low density but important to preserve for food resource of large frugivores. This finding is very important not only for forest protection policy but also for wildlife conservation as food resources.
... Thus, they plan their visit appropriately, when the fruit is ripe Cunningham & Janson, 2007;Janmaat et al., 2012;Noser & Byrne, 2010). In addition, primates can be picky: they tend to select the trees that produce the most food (Flörchinger et al., 2010), or those that produce the sweetest fruit . They are therefore one of the main actors in the selection of plant characteristics (Lomáscolo & Schaefer, 2010). ...
... Naturally, the conclusions, built upon a single study (this of Manuscript 8), deserve stronger evaluation, in particular because dietary guilds are generally gross, only distinguishing folivorous species and frugivorous species for instance. Still, frugivores might be very selective in what they eat, if only due to their own size and the size of the fruit (Flörchinger et al., 2010). Consequently, dietary overlap between sympatric frugivores can still be limited. ...
À l’ombre du feuillage dense de la forêt tropicale d’Afrique centrale, un gorille de l’Ouest (Gorilla gorilla) passe presque le tiers de sa journée à manger. Chaque jour, il a besoin de plus de 5 000 kcal et parcourt plusieurs kilomètres pour trouver sa nourriture en quantité suffisante. De l’efficacité de sa recherche peut ainsi dépendre sa valeur sélective. Si la sélection naturelle a donc pu jouer sur les capacités facilitant la digestion et la perception des ressources, ou la locomotion, elle a aussi pu influencer la capacité de ce grand singe à traiter, mémoriser et réutiliser un savoir de long terme permettant de connaître quand et où se trouve quelle ressource. Cette thèse propose de mieux comprendre l’implication de la cognition dans la recherche de nourriture chez le gorille de l’Ouest, et plus largement chez les primates. Nous comblons d’abord un vide théorique et illustrons par jeu de simulations agent-centrées comment une mémoire temporelle permet d’améliorer l’efficacité de la recherche de nourriture dans un environnement variable mais prédictible, comme la forêt tropicale et saisonnière où vivent les gorilles de l'Ouest. Par suite, nous prenons appui sur un suivi de long terme de cinq groupes de gorilles de l’Ouest d’Afrique centrale, habitués à la présence humaine, ainsi que sur un inventaire botanique et phénologique de leurs ressources alimentaires afin de montrer l’existence d’une mémoire spatio-temporelle. Pour ce faire, nous nous concentrons sur deux comportements : la restriction du déplacement au sein d’un domaine vital et les choix alimentaires effectués. Par suite, nous mettons en exergue une partie des mécanismes cognitifs spatiaux sous-jacents aux déplacements en nous focalisant sur l’utilisation par les gorilles de marécages largement dispersés mais aux ressources uniques et essentielles, et celle d’un réseau de sentiers d’éléphants de forêt. Ces études soulignent l’existence d’une mémoire spatiale précise jusqu’à de larges échelles, avec un encodage topologique de l’information. Nous montrons ensuite comment le savoir spatio-temporel est utilisé et insensible aux variations des ressources par analyse des patrons de déplacement et de revisite aux sites d’alimentation. Pour finir, nous élargissons notre champ de vision à travers plusieurs études multi-espèces se dessinant autour (a) d’une approche expérimentale, afin de relier stratégie de recherche alimentaire, cognition et degré de frugivorie chez les primates; (b) d’une revue d’opinion proposant une méthodologie standardisée pour comparer la cognition liée au déplacement des primates en milieu sauvage; (c) d’une approche phylogénétique, afin de comprendre l’influence de la sympatrie entre primates frugivores sur l’histoire évolutive de leur cognition et leur diversification. Cette thèse contribue ainsi à lever le voile sur les mécanismes proximaux et ultimes façonnant les stratégies de recherche de nourriture et les capacités cognitives chez les primates.
... The diet of frugivorous animals is determined by intrinsic characteristics, such as their physiology (i.e., digestive capacity and energy requirements) and behavior (i.e., feeding habits, dispersal, and social systems), and extrinsic characteristics, such as food availability (i.e., the abundance of a given species in the environment) and biochemical contents (e.g., nutrients and metabolites) (Bergman et al., 2005;Searle et al., 2007). Likewise, the consumption of fleshy fruits may be stimulated by conspicuous features of the vegetative structures, such as size (Dias da Silva et al., 2020;Flörchinger et al., 2010;, color (Regan et al., 2001;Osorio et al., 2004), odor (Nevo and Heymann, 2015;Nevo et al., 2016;Nevo and Valenta, 2018), hardness (Kinzey and Norconk, 1990), and specific chemical compounds (e.g., ethanol; Dominy, 2004). Additionally, fruits have different shapes (berries, drupes, pommel), high energy content (sugars and lipids), and low fibrosity (structural fibers) that make them part of the diet of a wide array of frugivores (Fleming and Kress, 2011;Rothman et al., 2011;Seymour et al., 2013). ...
... For example, guerezas (Colobus guereza) feed principally on trees from the Moraceae family (Fashing, 2001). Likewise, cercopithecid monkeys (Cercopithecus ascanius, C. mitis, and Papio anubis) in a tropical forest in Kakamega, Africa, feed mainly on large fruits of plants belonging to the subclass Rosidae (Flörchinger et al., 2010). Our results demonstrate that the subclass Rosidae plays an important role in the diet of night monkeys inhabiting tropical forests. ...
Full-text available
The study of diet and food selection is foundational to understanding how primates interact with their environment. Due to the potential evolutionary relationship between fruit physical traits and frugivory in animals, our goal was to understand the composition and the evolutionary relationships of plant species in the Andean night monkey’s diet (Aotus lemurinus) in Pijao-Quindío, Colombia. We hypothesize that phylogenetically related plants share fruit physical traits, such as size, which make them more likely to be selected. We estimated the diet of two groups of monkeys through feeding time (4,431 min) and surveyed fruit-bearing trees available (from 40 spp.) through five phenological transects from July 2018 to July 2019. To evaluate whether feeding time is influenced by fruit size, controlling for phylogeny, we mapped the fruit size of each species onto the phylogeny and tested for the phylogenetic signal using comparative phylogenetic methods. We found that the Andean night monkey selected some fruit species (of 29 spp. selected) disproportionately more often relative to their abundance and spent more time feeding on fruit-bearing trees belonging to the Rosidae subclass than other subclasses (pattern also found in other populations in Colombia). Although fruit size had a strong phylogenetic signal, it was not significantly related to time spent feeding on fruits, even after controlling for the phylogenetic relationships among fruit species. Our results suggest that fruit size does not influence fruit selection and that we need to assess other fruit traits (physical and phytochemical) to better understand the coevolutionary dynamics in this tropical forest.
... These two tree species are abundant in the study area and seem to adapt to varying environmental conditions of the successional gradient, including disturbed areas [17,47]. In fact, fruit size has been identified as a trait that drives fruit removal and seed dispersal [76][77][78], particularly of several Bursera species [12,13,79]. Thus, the interaction between the fruit size of these trees and potential legitimate dispersers increases the probability of germination and the establishment of seedlings in the TDF of the study area. ...
... For example, small-bodied pulp consumers, such as Haemorhous mexicanus, Passerina versicolor, and P. leclancherii, removed small-sized fruits in B. aptera and B. schlechtendalii [78], while large-bodied legitimate dispersers such as Calocitta formosa, Melanerpes chrysogenys, Ortalis poliocephala, and T. verticalis consumed large-sized fruits in Sideroxylon capiri, B. longipes, and Capparis spp. This shows that the selection of zoochorous trees by fruit-eating birds is highly dependent on plant fruit size, as previously described [77,85]. Although this is not surprising, it has important ecological implications because the effectiveness of seed dispersal by fruit-eating birds depends, to a great extent, on their body mass [74,76], which highlights the importance of morphological traits in the mutualistic frugivory networks in tropical environments [85]. ...
Full-text available
Tropical dry forests (TDFs) are affected by land-use changes. These modifications impact their composition and arboreal structure, as well as the availability of food for several bird groups. In this study, we evaluated the foraging preferences in zoochorous trees of fruit-eating birds during the dry season of the year in three successional stages (early, intermediate, and mature) of TDFs in southern Mexico. The fruits of these trees are important in the diet of several birds during the dry season, a period during which food resources are significantly reduced in TDFs. We estimated foliar cover (FC) and foliage height diversity (FHD) of zoochorous trees in 123 circular plots. These variables were recognized as proxies of food availability and tree productivity. Foraging preferences were evaluated at the community level, by frugivore type, and by bird species. We evaluated the effect of the structural variables and the fruit size of zoochorous plants on fruit removal by birds and related the bird body mass and fruit size removed in the successional gradient. A total of 14 zoochorous tree species and 23 fruit-eating bird species were recorded along the successional gradient. Intermediate and mature stages showed greater fruit removal. The birds removed mainly B. longipes fruits across the three successional stages. The FHD and fruit size were important drivers in the selection of zoochorous trees and fruit removal by fruit-eating birds. Fruit size and bird body mass were positively related along the successional gradient. The results suggest that fruit removal by fruit-eating birds in the successional gradient can promote the demographic dynamics of several zoochorous tree species, especially of Bursera spp. along the TDFs.
... We selected 100 diaspores from the fruit pool of each plant species to determine fresh diaspore mass using a precision balance (0.0001 g) in the laboratory. It has been previously documented that fruit variables (i.e., fruit size and fruit mass) are highly correlated (Flörchinger et al. 2010). Additionally, for all species we counted the number of seeds per fruit from 100 fruits. ...
Full-text available
Invasive alien plant species (IAPS) are severely changing ecosystems on earth. Studying the interactions that allow IAPS to establish and spread in the new regions is crucial. Ants can disperse exotic fleshy fruits. We asked the following questions at three different sites of Chaco Mountain Forest (Córdoba, Argentina): (1) Do ants disperse diaspores of native, neonative and IAPS differentially? (2) Which is the ant species assemblage and their role in the secondary dispersal of each of the selected plant species? (3) Do ants interact in different ways with intact and manipulated fruits, and these interactions vary within plant species origin? and (4) Are diaspore traits different among the plant species considered? We selected four plant species: Celtis ehrenbergiana (native), Lantana camara (neonative), Pyracantha angustifolia and Ligustrum lucidum (IAPS). Two experiments were performed: (1) To disentangle the contribution of ants to the secondary dispersal process, and (2) To investigate the ant behavior of ground-foraging ant species when they encountered the fruits. Additionally, we measured fruit mass and the number of seeds per fruit. Ants were the main diaspore dispersers on the Chaco Mountain Forest ground. Twelve ant species interacted with the fruits; the native presented the higher number, followed by the neonative, and the two IAPS. Only Acromyrmex crassispinus and Pheidole cordyceps removed diaspores. Furthermore, the fruits differed in their mass and also in the number of seeds. Our results highlight the importance of ants and also diaspores traits in these diffuse mutualisms, and enhance their role in plant-invasive processes in subtropical ecosystems.
... Habitat heterogeneity is a pivotal driver of niche differentiation in plant and animal communities and may increase biodiversity by providing ecological opportunities for species to coexist (Flesch et al. 2015;Stroud & Losos 2016;Machado-de-Souza et al. 2019). Variation in tree heights, for example, turns tropical forests into complex ecosystems with unique microenvironmental conditions (humidity, light, temperature, foliage density, wind, etc.) distributed unevenly across the forest strata (Flörchinger et al. 2010;Purificação et al. 2020). Many animals exploit this complexity and use different strata nonrandomly: a widespread phenomenon called vertical stratification (Allee et al. 1949). ...
Animals may exploit habitat complexity in a myriad of ways, and a widespread pattern of habitat use is vertical stratification. In tropical forests, it often increases niche diversity for frugivorous birds, and it strongly influences seed dispersal and community persistence. Yet, most studies disregard intraspecific variations in the use of forest layers and their ecological and evolutionary consequences. In this study, we asked whether adult males and green individuals (females and juveniles) of a dimorphic frugivorous bird, the helmeted manakin (Antilophia galeata, Pipridae) foraged at different strata in a forest from Southeastern Brazil. We also investigated how the seasonal dynamics of fruit availability and distribution within the forest affect between-and within-phenotype use of forest layers. We conducted field observations of the foraging activity of both phenotypes and found that they were mainly midstory foragers flying at similar heights in both seasons. However, adult males increased their stratum breadth and their foraging height during the wet season, when fruits were also available at higher strata, while the foraging behaviour of green individuals remained seasonally invariant. The different effects of seasonality on the phenotypes of the helmeted manakin may be a consequence of sexual competition among adult males in the wet season to secure fruits in taller plants. Altogether, our findings show that seasonal-ity affects differently the spatial use behaviours of males and green individuals, and understanding these intraspecific differences may help us predict how species respond to environmental disturbances affecting habitat structure and frugivores assemblages.
... As frugivores move through the environment, they may use information on intrinsic qualities of available fruiting plants, assessed at the moment or inferred from previous visits, to make subsequent foraging and movement decisions (Plante et al. 2014, Suarez 2014, Trapanese et al. 2019. Frugivores may direct their foraging towards preferred plant species, as well as toward individual plants with favorable traits, such as large fruit crops or heights (Flörchinger et al. 2010, Palacio and Ordano 2018, Crestani et al. 2019, Schupp et al. 2019. Intraspecific variation in plant traits has been associated with variation in frugivore visitation rates (Murray 1987), visiting frugivore diversity (Guerra et al. 2017), the number of fruits removed by frugivores (Davidar and Morton 1986), and the quality of seed-handling (Palacio et al. 2017). ...
While plant–animal interactions occur fundamentally at the individual level, the bulk of research examining the mechanisms that drive interaction patterns has focused on species or population levels. In seed-dispersal mutualisms between frugivores and plants, little is known about the role of space and individual-level variation among plants in structuring patterns of frugivory and seed dispersal in a plant community. Here we use a zoocentric approach to examine how space and variation between individual plants affect movement and visitation by frugivores foraging on individual fruiting plants. To do this, we used a spatially explicit network approach informed by observations of the movement and foraging of a frugivorous lemur species Eulemur rubriventer among individual plants in a diverse plant community in Madagascar. The resulting hierarchical networks show a few individual plants receiving the bulk of the interactions, demonstrating that a generalist frugivore species could act as an individual-plant specialist within a plant community. The few individual plants that dominated interactions with lemurs shaped the modular spatial structure of frugivory interactions in the community and facilitated visitation to near neighbors. This interaction structure was primarily driven by extrinsic factors, as lemur movements among plants were significantly influenced by the individual plant's spatial position and the fruiting plant richness in its immediate neighborhood. Individual plants in central spatial locations with large fruit crops received the most visits. The observed inequality in the interactions of a generalist frugivore within a highly diverse plant community highlights the importance of considering individual-level variation for essential ecosystem processes such as seed dispersal.
... Fruits and seeds vary in size (Michaels et al., 1988) both within and between species. The selection of fruits by primates and other frugivorous involves nutritional and sensory factors such as color (Melin et al., 2019), smell (Nevo et al., 2015), touch (Wrangham, 1975), level of toxins, hardness (Ayres, 1986;Barnett et al., 2016;Norconk & Veres, 2011), nutritional content (Felton et al., 2009;Rothman, Raubenheimer, & Chapman, 2011), and size (Corlett & Lucas, 1990;Flörchinger, Braun, Böhning-Gaese, & Schaefer, 2010;Stevenson, Pineda, & Samper, 2005). The size of the fruit, however, is considered as the primary selection criterion for many frugivorous species (Jordano, 1995a(Jordano, , 1995b(Jordano, , 2014Martin, 1985;Mello, Leiner, Guimarães, & Jordano, 2005). ...
Full-text available
Optimal foraging theory predicts that animals will seek simultaneously to minimize food processing time and maximize energetic gain. To test this hypothesis, we evaluated whether a specialist seed‐predator primate forages optimally when choosing among variable‐sized thick‐husked fruits. Our objects of study were the golden‐backed uacari (Cacajao ouakary, Pitheciidae) and single‐seeded pods of the macucu tree (Aldina latifolia, Fabaceae). We predict that golden‐backed uacari will consume fruits of the size class that requires the least time to obtain, handle, and ingest. We used scan sampling, ad libitum to record feeding observations, and measured fruits, their penetrability, and the size of taxidermized C. ouakary hands. To test whether uacaris selected for optimal characteristics, we compared 8 metrics from 75 eaten and 105 uneaten seeds/fruits collected. Uacaris selected fruits of medium size and weight disproportionately to their abundance. Processing large fruits took six times longer than did medium‐sized fruits, but seeds were only four times as large, that is, for energetic yield per unit time, thus choosing medium‐sized pods was optimal. Disproportionate selection by C. ouakary of fruits of medium size and mass in relation to their abundance suggests active sub‐sampling of the available weight–size continuum. This selectivity probably maximizes trade‐offs between the energy derived from a seed, and time and energy expended in processing fruit to access this, so following optimal foraging theory predictions. The greater time spent processing large pods is attributed to difficulties manipulating objects five to seven times the size of the animal's palm and one‐sixth its own body weight.
Land-use change is one of the greatest threats to biodiversity. It is predicted that conversion of land and habitats will increase rapidly over the next few decades in Africa. Over the years, these changes potentially reduced the capacity of ecosystems to sustain food production for vertebrates. Ficus species (Moraceae), commonly known as figs, occupy diverse habitats and typically produce large numbers of nutritional fleshy fruits that are important to frugivores. However, a decline in Ficus spp. distribution because of land-use changes may negatively affect frugivores and their ecosystems (e.g. via seed dispersal). We systematically searched for studies on the distribution of Ficus spp. in Africa and their frugivore interactions together with the effects of land-use changes up until 2021. Our search resulted in 70 eligible papers. A total of 124 Ficus spp. were recorded across 30 African countries representing approximately 56% of the African countries. Cameroon had the highest record of 63 species, while Benin, Burundi, Ghana, and Rwanda had two, the least number of Ficus spp. recorded. East Africa had the highest Ficus spp. richness recorded (96 species), followed by southern Africa (74 species), Central and Northern Africa (72 species), and West Africa with the least (31 species) recorded. Information about the effect exerted by anthropogenic land-use changes on Ficus-frugivore interaction in Africa was limited. However, research has been conducted on the impact of anthropogenic land-use changes on plant-frugivore and frugivore feeding ecology. Ficus spp. fruit were identified as significant in the diets of various frugivores across Africa, as it is found globally. However, it is essential to understand the impacts of anthropogenic land-use changes on the mutual interaction between frugivores and Ficus spp. and the attendant consequences for ecosystem service provision.
We propose a new method to estimate and correct for phylogenetic inertia in comparative data analysis. The method, called phylogenetic eigenvector regression (PVR) starts by performing a principal coordinate analysis on a pairwise phylogenetic distance matrix between species. Traits under analysis are regressed on eigenvectors retained by a broken-stick model in such a way that estimated values express phylogenetic trends in data and residuals express independent evolution of each species. This partitioning is similar to that realized by the spatial autoregressive method, but the method proposed here overcomes the problem of low statistical performance that occurs with autoregressive method when phylogenetic correlation is low or when sample size is too small to detect it. Also, PVR is easier to perform with large samples because it is based on well-known techniques of multivariate and regression analyses. We evaluated the performance of PVR and compared it with the autoregressive method using real datasets and simulations. A detailed worked example using body size evolution of Carnivora mammals indicated that phylogenetic inertia in this trait is elevated and similarly estimated by both methods. In this example, Type I error at α = 0.05 of PVR was equal to 0.048, but an increase in the number of eigenvectors used in the regression increases the error. Also, similarity between PVR and the autoregressive method, defined by correlation between their residuals, decreased by overestimating the number of eigenvalues necessary to express the phylogenetic distance matrix. To evaluate the influence of cladogram topology on the distribution of eigenvalues extracted from the double-centered phylogenetic distance matrix, we analyzed 100 randomly generated cladograms (up to 100 species). Multiple linear regression of log transformed variables indicated that the number of eigenvalues extracted by the broken-stick model can be fully explained by cladogram topology. Therefore, the broken-stick model is an adequate criterion for determining the correct number of eigenvectors to be used by PVR. We also simulated distinct levels of phylogenetic inertia by producing a trend across 10, 25, and 50 species arranged in "comblike" cladograms and then adding random vectors with increased residual variances around this trend. In doing so, we provide an evaluation of the performance of both methods with data generated under different evolutionary models than tested previously. The results showed that both PVR and autoregressive method are efficient in detecting inertia in data when sample size is relatively high (more than 25 species) and when phylogenetic inertia is high. However, PVR is more efficient at smaller sample sizes and when level of phylogenetic inertia is low. These conclusions were also supported by the analysis of 10 real datasets regarding body size evolution in different animal clades. We concluded that PVR can be a useful alternative to an autoregressive method in comparative data analysis.
A guide to using S environments to perform statistical analyses providing both an introduction to the use of S and a course in modern statistical methods. The emphasis is on presenting practical problems and full analyses of real data sets.
Inferences about mechanisms at one particular stage of a visual pathway may be made from psychophysical thresholds only if the noise at the stage in question dominates that in the others. Spectral sensitivities, measured under bright conditions, for di-, tri-, and tetrachromatic eyes from a range of animals can be modelled by assuming that thresholds are set by colour opponency mechanisms whose performance is limited by photoreceptor noise, the achromatic signal being disregarded, Noise in the opponency channels themselves is therefore not statistically independent, and it is not possible to infer anything more about the channels from psychophysical thresholds. As well as giving insight into mechanisms of vision, the model predicts the performance of colour vision in animals where physiological and anatomical data on the eye are available, but there are no direct measurements of perceptual thresholds. The model, therefore, is widely applicable to comparative studies of eye design and visual ecology.
This chapter gives results from some illustrative exploration of the performance of information-theoretic criteria for model selection and methods to quantify precision when there is model selection uncertainty. The methods given in Chapter 4 are illustrated and additional insights are provided based on simulation and real data. Section 5.2 utilizes a chain binomial survival model for some Monte Carlo evaluation of unconditional sampling variance estimation, confidence intervals, and model averaging. For this simulation the generating process is known and can be of relatively high dimension. The generating model and the models used for data analysis in this chain binomial simulation are easy to understand and have no nuisance parameters. We give some comparisons of AIC versus BIC selection and use achieved confidence interval coverage as an integrating metric to judge the success of various approaches to inference.