Epizoochorous seed dispersal by an Afroalpine savanna primate

Abstract

Primates are prolific dispersers of seeds via endozoochory (i.e., defecation and spitting). In contrast, epizoochorous seed dispersal (i.e., via adhesion to fur) has rarely been observed in primates. On the Guassa Plateau in north‐central Ethiopia, grass‐eating geladas ( Theropithecus gelada ) regularly carry on their fur the barbed seeds of a commonly eaten plant, a low‐lying herbaceous forb called Agrocharis melanantha [Apiaceae]. Here, we describe the basic ecology of this plant–primate relationship. For 24 months (November 2017–December 2019), we monitored the number and location of A . melanantha seeds on the fur of geladas ( n = 225 individuals) from four age‐sex classes: adult males, adult females, juveniles, and infants. Seed accumulation ( n = 12649 seeds in total) was seasonal and closely tracked patterns of landscape vegetation phenology, peaking in September near the end of the rainy season. During seasonal periods of heavy seed accumulation, larger animals carried more seeds, which accumulated most often on the hindlimbs and on the long‐haired “cape” (a secondary sexual characteristic) of adult males. Geladas almost never removed seeds during self‐ or social grooming. Rather, data on seed gain and loss from focal follows indicate that geladas gain and lose seeds every few minutes as they walk and sit in an upright feeding position amidst terrestrial vegetation. We estimate that, on average, geladas disperse seeds roughly 80 m from their parent plants. Geladas appear to exert negative and positive fitness impacts on A . melanantha by regularly consuming its herbaceous and underground tissues and dispersing its seeds.
Abstract in Amharic is available with online material.

Full text

358
|
Biotropica. 2022;54:358–369.wileyonlinelibrary.com/journal/btp
Received: 5 July 2021
|
Revised: 16 November 2021
|
Accepted: 23 November 2021
DOI : 10.1111/btp.1305 7
ORIGINAL ARTICLE
Epizoochorous seed dispersal by an Afroalpine savanna primate
Vivek V. Venkataraman1,2 | Carrie Miller2,3 | Iris Foxfoot2 | Bing Lin2,4 |
Zack L. Petrie2 | Ruth A. Simberloff2 | Odin Bernardo2 | Nathan Redon2 |
Triana I. Hohn5 | Jeffrey T. Kerby2,6 | Nga Nguyen2,7 | Peter J. Fashing2 , 7, 8
© 2021 The Association for Tropical Biology and Conservation.
1Department of Anthropology and
Archaeology, University of Calgary,
Calgary, AB, Canada
2Guassa Gelada Research Project, Mehal
Meda, Ethiopia
3Department of Anthropology, University
of Minnesota, Minneapolis, Minnesota,
USA
4Princeton School of Public and
International Affairs, Princeton University,
Princeton, New Jersey, USA
5Department of Biological Sciences,
University of Calgar y, Calgary, AB, Canada
6Aarhus Institute of Advanced Studies,
Section for Ecoinformatics and
Biodiversity, Department of Biology,
Aarhus University, Aarhus, Denmark
7Department of Anthropology &
Environmental Studies Program, California
State Universit y, Fullerton, California, USA
8Department of Biosciences, Centre for
Ecological and Evolutionary Synthesis
(CEES), University of Oslo, Oslo, Norway
Correspondence
Vivek V. Venkataraman, Department of
Anthropology and Archaeology, University
of Calgary, Calgary, AB, Canada.
Email: vivek.venkataraman@ucalgary.ca
Funding information
Primate Conser vation International; San
Diego Zoo Institute for Conservation
Research
Associate Editor: Jennifer Powers
Handling Editor: Kim McConkey
Abstract
Primates are prolific dispersers of seeds via endozoochory (i.e., defecation and spit-
ting). In contrast, epizoochorous seed dispersal (i.e., via adhesion to fur) has rarely
been observed in primates. On the Guassa Plateau in north- central Ethiopia, grass-
eating geladas (Theropithecus gelada) regularly carry on their fur the barbed seeds of
a commonly eaten plant, a low- lying herbaceous forb called Agrocharis melanantha
[Apiaceae]. Here, we describe the basic ecology of this plant– primate relationship.
For 24 months (November 2017– December 2019), we monitored the number and
location of A. melanantha seeds on the fur of geladas (n = 225 individuals) from four
age- sex classes: adult males, adult females, juveniles, and infants. Seed accumulation
(n = 12649 seeds in total) was seasonal and closely tracked patterns of landscape
vegetation phenology, peaking in September near the end of the rainy season. During
seasonal periods of heavy seed accumulation, larger animals carried more seeds,
which accumulated most often on the hindlimbs and on the long- haired “cape” (a sec-
ondary sexual characteristic) of adult males. Geladas almost never removed seeds
during self- or social grooming. Rather, data on seed gain and loss from focal follows
indicate that geladas gain and lose seeds every few minutes as they walk and sit in
an upright feeding position amidst terrestrial vegetation. We estimate that, on aver-
age, geladas disperse seeds roughly 80 m from their parent plants. Geladas appear to
exert negative and positive fitness impacts on A. melanantha by regularly consuming
its herbaceous and underground tissues and dispersing its seeds.
Abstract in Amharic is available with online material.
KEYWORDS
Afroalpine, Agrocharis melanantha, epizoochory, Ethiopia, gelada, primate, seed dispersal,
Theropithecus gelada

|
359
VENK ATARAMAN ET A l.
1 | INTRODUCTION
Seed dispersal plays a critical role in the dynamics of plant com-
munities, ranging from population genetics and demography to
ecosystem function and conservation (Howe & Smallwood, 1982).
Among mammals, primates are prolific dispersers and predators of
seeds, especially in forest habitats (Andresen et al., 2018; Chapman,
1995). Documented cases of seed dispersal by primates typically
involve endozoochory, in which animals spit or defecate out seeds
(Chapman, 1995). In contrast, research on epizoochorous seed dis-
persal, in which seeds adhere to an animal's fur until they drop off,
has heavily focused on ungulates (Albert et al., 2015; Baltzinger
et al., 2019) but has rarely been addressed among primates (Chen
et al., 2018). Given that epizoochorous seed dispersal in other an-
imals primarily occurs among low- lying shrubs in open habitats
(Sorensen, 1986), one would expect certain cercopithecine monkeys
and great apes to also disperse seeds via adhesion. For instance,
baboons (Papio spp.), patas monkeys (Erythorocebus patas), and ver-
vet monkeys (Chlorocebus spp.) are moderately- sized, terrestrial or
semi- terrestrial, and feed omnivorously in savanna woodland habi-
tats, and mountain gorillas (Gorilla beringei) in Bwindi, Uganda often
inadvertently carry seeds and/or flowers of their food items on their
fur as they forage and travel in dense vegetation (J. Rothman, pers.
comm.). However, despite decades of ecological research on these
taxa, we are not aware of any published reports of epizoochorous
seed dispersal among them.
Here, we describe a case of epizoochorous seed dispersal among
another taxon of cercopithecines, geladas (Theropithecus gelada),
which are endemic to the highlands of Ethiopia and unique among
the primates in being largely confined to alpine grassland habitats
(Bergman & Beehner, 2013). We carried out our research on the
Guassa Plateau (hereafter Guassa), an ecologically intact tall- grass
ecosystem in north- central Ethiopia (Fashing et al., 2014; Welch
et al., 2017; Lin et al., 2020). Geladas are exclusively terrestrial and
forage mostly on graminoids and forbs in dense herds of several
hundred individuals (Figure 1a,b; Fashing et al., 2014; Jarvey et al.,
2018). During fieldwork, we noticed that the fur of geladas was
frequently adorned by small yellow- green seeds (Figure 1c– h). The
seeds derived exclusively from Agrocharis melanantha Hochst (fam-
ily Apiaceae) (Figure 1i– k), an herbaceous forb whose aboveground
leaves and underground storage tissues are frequently consumed by
geladas (Fashing et al., 2014).
Successful epizoochorous seed dispersal relies on emigration
(the process of diaspores being loaded by an animal vector), transfer
(the distance and time elapsed between attachment and detach-
ment), and immigration (the subsequent ability of a loaded diaspore
to become an adult plant) (Baltzinger et al., 2019). Such processes
depend on plant (e.g., seed and plant morphology, phenology) and
animal vector traits (e.g., body size, body area, fur characteristics,
behavior), and also on other environmental factors such as the na-
ture of the neighboring vegetation (Baltzinger et al., 2019; Sorensen,
1986). This paper focuses on circumstances in which geladas accu-
mulate (emigration) and lose (transfer) seeds on their fur.
A. melanantha is a low- lying annual or perennial forb, occur-
ring widely across Africa up to the alpine belt around 4200 m
a.s.l., which has fruits (hereafter seeds) that are setose achenes,
oblong- ellipsoid in shape, and roughly 5 mm long, arising from a
compound umbel and having short, rigid, barbed styles (Hedberg
et al., 2003). Agrocharis may be both epizoochorously and endozo-
ochorously dispersed, like its closely related genus Daucus, which
includes carrots (Kleyer et al., 2008). To our knowledge, there are
no published studies on the natural history of A. melanantha; how-
ever, the plant is apparently used to treat the “evil eye” among two
traditional societies (Maale and Ari) in southern Ethiopia (Kidane
et al., 2014).
Given that A. melanantha exhibits functional traits consistent
with an animal dispersal syndrome (but see Green et al., 2021) and
that we have regularly observed these seeds on geladas, we sought
to address five questions (with associated predictions) about the re-
lationship between geladas and A. melanantha at Guassa.
Question 1. What are the distribution and functional traits of
A. melanantha at Gu assa? Bec aus e of the de arth of pu blish ed ecolog-
ical information, we examined several traits of A. melanantha. During
a two- month period of high seed availability (September– October),
we performed vegetation transects to survey the abundance of
A. melanantha across different microhabitats at Guassa (short grass,
roadside, shrub, or tall grass); we also examined mean plant height,
vertical position relative to the surrounding vegetation, number of
seeds per umbel, and total number of seeds per plant.
Question 2. What is the seasonal pat te rn of seed accu mu lation
of A. melanantha on the fur of geladas? In high- elevation tropical
grasslands such as Guassa, rainfall seasonality tends to dictate the
availability of graminoids and herbs, and most plants time their
growth and subsequent reproductive phenology to coincide with
these periods of reduced water limitation (Fashing et al., 2014;
Jarvey et al., 2018). Given the pronounced rainfall seasonality at
Guassa during the warmer summer months (July– August) and that
the arid dry season occurs during colder months with frequent
frosts (November– January; Fashing et al., 2014), we expected to
observe clear seasonality in seed accumulation patterns on gela-
das’ fur across the landscape. Finally, geladas live in multi- level
societies, with the basic social unit being stable one- male units
(OMUs) consisting of a harem male and 2– 10 females (Snyder-
Mackler et al., 2012). Because OMUs generally move throughout
the environment in close spatial proximity (<10 m), we expected
that members of a given OMU would exhibit similar seed accumu-
lation patterns.
Question 3. Do larger geladas tend to carry more A. melanan-
tha seeds? Adult male geladas are roughly twice the body mass of
adult females, and juveniles and infants are substantially smaller,
perhaps less than half the mass of an adult female (Venkataraman
et al., 2014). While a larger animal should provide a larger surface for
seeds to accumulate, animal behavior could contradict this pattern
because juveniles feed on herbaceous vegetation more than adults
do at Guassa (Fashing et al., 2014), and also run, jump, and play fre-
quently, facilitating more opportunities for adhesion.

360
|
VENK ATARAMAN ET A l.
Question 4. Which body parts of geladas attach more seeds? We
expected that the body regions that most frequently contact veg-
etation and that are most conducive to adhesion (e.g., areas with
layers of long hair) should bear the highest number of seeds. Geladas
engage in a stereotypical sitting posture while feeding and shuffling
within food patches (Figure 1b; Wrangham, 1980). In this position,
we expected that the caudal part of the body (lower thorax and
hindlimbs; Figure 2) should attach more seeds. We also expected
the long and flowing hair of the adult male “cape,” a secondary sexual
characteristic (Figure 1e– g), to accumulate many seeds.
Question 5. How long does it take for geladas to gain and lose
A. melanantha seeds, what determines this, and how does this trans-
late into dispersal distance? Grooming often results in barbed seed
removal from fur among mammals (Albert et al., 2015; Baltzinger
et al., 2019; Liehrmann et al., 2018; Sorensen, 1986). As prolific
groomers (Iwamoto & Dunbar, 1983), geladas are expected to
FIGURE 1 (a) An aggregation of gelada
monkeys at Guassa containing adult
males (larger, with prominent cape and
white hair around the red chest patch)
and females. Note that adult males are
nearly twice the body mass of adult
females. Photo by Vivek V. Venkataraman.
(b) Geladas foraging on graminoids and
forbs in dense herds at Guassa. Photo
by Vivek V. Venkataraman. (c) Two adult
females grooming. Note numerous A.
melanantha seeds on the right individual's
leg and thorax. Photo by Jeffrey T. Kerby.
(d) Adult female walking with infant on
back. Numerous seeds are visible. Photo
by Vivek V. Venkataraman. (e, f) Adult
males with dozens of seeds in their capes.
Photos by Vivek V. Venkataraman. (g)
A. melanantha seeds attached to the cape
of an adult male and being ignored by
a female groomer. Photo by Jeffrey T.
Kerby. (h) Close- up view of A. melanantha
seeds adhering to gelada fur. Photo by
Jeffrey T. Kerby. (i) A. melanantha in the
tall- grass ecosystem of Guassa, Ethiopia.
Photo by Peter J. Fashing. (j) The seeds
of A. melanantha with reddish flowers.
Photo by Zack L. Petrie. (k) A. melanantha
exhibits several fruits per umbel and
has short styles on the schizocarp (adult
male fingers for scale). Photo by Zack
L. Petrie
(a) (b)
(c) (d) (e)
(f) (g) (h)
(i) (j) (k)
FIGURE 2 Schematic representation
of (a) female and (b) male gelada monkeys
with labeled body region locations.
Inset shows how each body region
was classified into broader categories.
Illustrations by Samantha Shields

|
361
VENK ATARAMAN ET A l.
remove seeds during self- and social grooming, in addition to losing
seeds from regular habitual movement in their environment.
2 | METHODS
2.1 | Study site
We conducted our study of geladas on the Guassa Plateau, a
large (111 km2) and unusually intact Afroalpine tall- grass eco-
system located along the western edge of the Great Rift Valley
(10°15′– 10°27′N; 39°45′– 39°49′E) at elevations between 3200 and
3600 m a.s.l. (Fashing et al., 2014). Guassa consists of rolling hills, flat
plateau areas, cliffs along its eastern border (on which the geladas
sleep at night), and farms along its other borders, with one road bi-
secting the area. The mammal community at Guassa is intact and in-
cludes klipspringers (Oreotragus oreotragus), grey duikers (Sylvicapra
grimmia), Ethiopian wolves (Canis simensis), African wolves (Canis au-
reus lupaster), spotted hyenas (Crocuta crocuta), servals (Leptailurus
serval), and leopards (Panthera pardus), in addition to rodents (six
species of rats and mice), Abyssinian hares (Lepus habessinicus), rock
hyraxes (Procavia capensis), Egyptian mongoose (Herpestes ichneu-
mon), and zorillas (Ictonyx striatus) (Ashenafi, 2001; Gutema et al.,
2019; Lin et al., 2020). During 2007– 2012, the mean monthly tem-
perature at Guassa was 11.0 ± 1.2 (SD) °C (Fashing et al., 2014).
Mean monthly low and high temperatures were 4.3 ± 0.5 (SE) and
17. 8 ± 0.3 (SE) °C , respec tively. An nua l mean rainf all was 1650 ± 243
(SD) mm (Fashing et al., 2014). Rainfall is strongly seasonal at Guassa,
typically exhibiting a unimodal peak during July and August when
more than two- thirds of the annual rainfall occurs. Rainfall drives the
temporal variation in graminoid availability at Guassa, and Fashing
et al. (2014) found cumulative rainfall three months prior to the end
of each study month to be the best indicator of monthly graminoid
availability.
2.2 | Data collection and analysis
We quantified the long- term annual phenology of landscape green-
ness using a LOESS curve fit to twice- monthly measures of the
Normalized Difference Vegetation Index (NDVI) (Tucker, 1979)— a
proxy for vegetation greenness that correlates with aboveground
productivity (Pettorelli et al., 2005). Using Google Earth Engine
(Gorelick et al., 2017), we extracted the mean MOD13A1 v6 NDVI
product (Didan et al., 2015) from 2000 to 2021 across a core re-
gion of the focal band's range, and filtered the data to only include
observations with the highest quality atmospheric conditions (i.e.,
QA score of 0, indicating no clouds and limited aerosols). By includ-
ing data from multiple decades, we highlight the long- term pattern
of vegetation seasonality at this site, while also reducing the sub-
stantial impacts of cloud artifacts on the satellite data in any given
year. The growth patterns identified correspond to the “wet season”
(July– October) and “dry season” (November– June) rainfall thresh-
olds used in Venkataraman et al. (2014).
2.2.1 | Question 1 (Plant functional traits)
During September– October 2018, we conducted systematic sur-
veys of A. melanantha across the main microhabitats (tall grass, short
grass, shrub, or roadside) at Guassa by walking 20 randomly cho-
sen non- overlapping 400- m transects within the home range of the
gelada study band. Within the transects, we placed a 0.5 × 0.5 m
wooden square plot directly on the transect line every 50 m. Within
these plots, we collected the following information: the number of
A. melanantha plants and whether they were flowering, the absolute
height of the plant from the ground and the height relative to the
local vegetation, the number of umbels on each plant, and the num-
ber of seeds on each umbel. We also noted the types of microhabitat
(e.g., short grass, tall grass, mixed forb) that characterized the imme-
diate vicinity of the transect. From these data, we calculated means
and standard deviations on the abundance of A. melanantha plants
across the Guassa landscape.
2.2.2 | Question 2 (Seed accumulation patterns on
geladas)
We conducted a behavioral study between November 2017 and
December 2019 and focused on the Steelers band of geladas at
Guassa that has been investigated since December 2005 and followed
on a near- daily basis since January 2007 (Nguyen et al., 2015). We
collected data from a total of 225 individually- recognized geladas dur-
ing the 24 months. Seven trained observers (BL, CM, IF, RS, ZP, NR,
and OB) counted A. melanantha seeds on the fur of geladas with the
aid of binoculars prior or subsequent to focal behavioral follows that
are regularly conducted as part of our long- term research protocol.
Repeated sampling of the same individuals on a single day occasion-
ally occurred, and in these cases, the counts were averaged into a sin-
gle daily observation for that individual. Observational conditions are
generally excellent for geladas, with fieldworkers often able to stand
as close as 1. 5 m from focal individuals and to observe the animal from
multiple angles. Counts were recorded when the observer was satis-
fied that all body regions had been clearly observed for long enough
to enable accurate seed counts. This protocol minimizes the chance
for bias in body region representation in the dataset. Gelada fur does
occasionally carry unidentifiable debris and seeds of Rumex nepalensis,
as well as Rumex nervosus. However, this is rare compared with the
presence of A. melanantha. Moreover, confusion is unlikely because
the seeds of Rumex sp. are red, whereas the seeds of A. melanantha
are green. Consequently, we are confident that there are virtually no
errors of commission (i.e., items wrongfully identified as A. melanan-
tha) in the dataset. To assess the annual pattern of seed dispersal, we
plot ted seed cou nts across time (ever y individual with a correspon din g

362
|
VENK ATARAMAN ET A l.
count of seeds on their fur on a given day), with the expectation of a
unimodal curve centered around the wet season.
2.2.3 | Question 3 (Seed accumulation as a
function of body size)
We surveyed four size classes of geladas: adult males, adult fe-
males, juveniles, and infants. The juvenile category includes small
and medium juveniles of either sex. The infant category includes
both brown (slightly older) and black (younger) infants of either sex.
Both brown and black infants are considered unweaned and spend
most of their time in close proximity or directly on the bodies of
their mothers. Large juveniles, regardless of sex, were placed in the
category “adult female” because they are most similar in body size
to adult females. Subadult males were assigned to the “adult male”
category because they appear most similar in size to adult males
and have partially- developed capes. These body size designations
are similar to those employed in Fashing et al. (2014), Nguyen et al.
(2015), and Venkataraman et al. (2014). To avoid confusion due to
these mixed- sex classifications, we hereafter refer to these size
classes as large (adult and subadult males), medium (adult females
and large juveniles), small (juveniles), and very small (infants). The
mean number of observations on a monthly basis was highest in
the medium class (n = 208), followed by small (n = 128), very small
(n = 85), and large individuals (n = 56). For regression analyses, we
converted these body size categories to estimates of body mass in
kilo grams. Adult males weigh rough ly 19 kg and adult femal es 10 kg
(Bergman & Beehne r, 2013). We are not aware of published data on
the size of juveniles in geladas, so we estimated animals in the very
small cate gor y as weig hing 3.5 kg and th ose in the small categ ory as
weighing 7 kg, wit h the under standing th at there is variation within
these categories. There is undoubtedly error introduced to the
analysis with these methodological decisions; however, we believe
assigning relatively accurate body masses is preferable to conduct-
ing an analysis with ordered categorical variables.
To determine how time of year and body mass are associ-
ated with seed accumulation through time, we compared several
generalized linear mixed models using an information theoretic
approach (Table 1). The dependent variable was seed counts, which
was modeled with a Poisson distribution due to significant zero-
inflation in the dataset (60% zeros). The global model included the
variables month, year, and body size as fixed effects, and the in-
dividual identity of geladas (nested within OMU membership) as
a random effect. Month and year were treated as nominal cate-
gorical variables. Data from March and April were removed from
the dataset for this portion of the analysis as no seeds were ob-
served during this time, and during some years, no data were col-
lected during these months. The other models consisted of various
combinations of the fixed effects. To compare model fit, we used
the corrected Akaike information criterion (AICc) and calculated
model weights (wi) (Anderson, 2008). Lower AICc values indicate
a better- fit model to the data among the candidate set (Burnham &
Anderson, 2002; Burnham et al., 2010). To assess goodness of fit,
we also calculated marginal and conditional R2 values. We used the
package lme4 (Bates et al., 2015) for the mixed modeling portion
of the analysis, AICcmodavg (Mazerolle, 2013) to compute ΔAICc
and Akaike weights for all candidate models, and MuMIn (Barton,
2011) to estimate R2 values. All analyses were conducted in the R
programming langauge (version R 4.0.3).
2.2.4 | Question 4 (Patterns of seed accumulation
across body regions)
We assig ned se eds to 20 body regi ons , which were fur ther catego-
rized into five main body regions (Figure 2). Gelada fur is relatively
uniform across the body, with the tuft on the tail and the cape
of the males having longer and bushier hair. Although the adult
male cape overlaps with other body regions (e.g., shoulder, back,
and flank), we coded it as a distinct body region because of its
prominence. We distinguished between the cape and the nearest
body location on the basis of proximity of the seed to the skin.
When seeds were attached to the shorter hair closer to the skin,
we coded these seeds by location rather than as belonging to the
cape. Seeds coded as “cape” were those on the outer, shaggy por-
tion of the coat. Because we did not double- count seeds, this cod-
ing scheme necessarily results in elevated counts for the cape and
Fixed effects KAICc ΔAICc wiLogL R2
BodyMass + Year + Month 15 29804.72 — 1−14887.31 .71
Yea r + Month 14 29903.89 9 9.1 72 2 0−149 37.9 0 .63
BodyMass + Month 13 30 612 .81 808.09 0−15293.36 .66
Month 12 306 45.67 840.95 0−153 10 . 8 0 .60
BodyMass + Year 741067. 76 11263.04 0−205 27.87 .22
Yea r 441271.14 11466.42 0−20630.57 .02
BodyMass 5416 6 7. 24 11862 .52 0−20 829.62 .11
None (random intercept only) 2417 5 7. 39 11952. 67 −20875 .69 .00
Note: The individual identity of geladas (nested within OMU membership) was included as a
random effect in all models.
TAB LE 1 Results of generalized
linear mixed model selection procedure
(K = number of estimated parameters;
LogL = log- likelihood; wi = Akaike weights;
R2 = marginal R2 values) predicting the
number of A . melanantha seeds attached
to geladas

|
363
VENK ATARAMAN ET A l.
lower counts for other parts. To assess whether particular body
regions accumulated more seeds, we restricted the analysis to the
months of June– December, when seeds were most abundant on
the animals. For each day, we averaged the number of seeds at a
given body location across all individuals of a given size class, then
averaged these values across months. We did not conduct statisti-
cal tests for this portion of the analysis.
2.2.5 | Question 5 (Context of gain and loss of
seeds)
The circumstances of seed gain and loss were observed on an a d
libitum basis. We provide anecdotal descriptions of the behavioral
and/or ecological context of such events. In addition, for 32 days
between August and November 2019, we recorded short- term
changes in the number of seeds attached to specific body regions
during focal follows. During these focal follows, which ranged
from seven to 21 min, an observer randomly chose a body region
and conducted seed counts at intervals of approximately 3 min. In
total, 489 focal follows were recorded. To compare the potential
for contribution to seed dispersal of each body region, we calcu-
lated the mean rate of gain and loss for five different body regions
(cape, hindlimb, forearm, back, and ventrum). It is important to
note that these body regions are not the exact same as those in
Figure 2. To obtain an estimate of how fa r geladas can carr y seeds,
we performed a basic calculation incorporating the mean rate of
seed loss across each of the five body regions and the mean speed
and distance that geladas travel in a day at Guassa. Finally, we esti-
mated the flux of seed acquisition and loss at a fine timescale dur-
ing focal follows to test the expectation that seeds are acquired in
clumps and then gradually lost.
3 | RESULTS
3.1 | Question 1 (Plant functional traits)
Agrocharis melanantha plants were patchily distributed at Guassa:
75% (135/180) of plots on the transects had no A. melanantha
plants, and those that did contained between two and eight indi-
vidual plants. 87% (118/135) of individual A. melanantha plants en-
countered were in mixed short- grass habitats, most of which were
complex mixes of graminoids, forbs, and shrubs. Only 9% were
found in tall- grass habitats. The remaining 4% occurred along the
roadside or in shrub- dominated habitats. In September– October
2018, we closely examined 33 A. melanantha plants to characterize
their functional traits. Mean plant height was 16.1 cm (min = 5 cm,
max = 37 cm, SD = 7.1 cm), and mean height above the surround-
ing vegetation was roughly 2.0 cm (min = −7.0 cm, max = 13.8 cm,
SD = 4.8 cm). A. melanantha had a mean of 5.4 seeds per umbel
(n = 153 umbels; min = 1, max = 16, SD = 2.9) and 25 se eds per plant
(min = 1, max = 95, SD = 24).
3.2 | Question 2 (Seed accumulation patterns on
geladas)
We counted a total of 12649 A. melanantha seeds attached to geladas
(n = 225 individuals) over the course of the study. The prevalence of
seeds on the geladas was highly seasonal and mirrored patterns of land-
scape phenology driven by rainfall at Guassa (Figure 3). Seeds were most
abundant on the fur of geladas during the end of the wet season in late
September when hundreds of seeds were routinely observed on indi-
viduals (Figure S1). In one extraordinary case, we counted 511 seeds on
an adult female (note: In this particular case, the observer counted one
body region and extrapolated to the rest of the body to avoid double-
counting; thus, the precision of this number should be treated with cau-
tion). Between January and May, few seeds were observed (Figure 3).
3.3 | Question 3 (Seed accumulation as a
function of body size)
The patterns of A. melanantha attachment to geladas’ fur generally fol-
lowed our expectations on the role of seasonality and body size (Figure 4,
Table 1). The best- ranking model predicting seed counts was the global
model, which included body size, month, and year as fixed effects, with
an AICc weight equal to 1 and accounting for 71% of the variance in the
dataset (Table 1). The magnitude of the estimates for the fixed effect of
body size (Table 2) was moderate, indicating an increase of ~0.1× seeds
per kilogram of gelada body mass. Model output from the random effect
of individual identity (nested within OMU) showed moderate variation be-
tween individuals and OMUs, with a standard deviation of ~0.8. Such ef-
fects, however, were minor in comparison with the effect of month, which
appears to explain most of the variation in patterns of seed accumulation:
the regression coefficients for month during periods of high seed avail-
ability (September– October) ranged between 4.8 and 5.5. Notably, we
did observe some variation in seed accumulation patterns between years
(Table 2), which is likely due to interannual rainfall patterns.
3.4 | Question 4 (Patterns of seed accumulation
across body regions)
Agrocharis melanantha seeds were distributed across multiple body
regions in each size class of geladas (Figure 5). In general, seeds
were most abundant on the hindlimb, followed by the mid- upper
body, then the forelimb, and face and head (Figure 5). The male
cape attracted more seeds, on average, than any other body region;
however, we note that this result stems in part from our sampling
methodology (see Section 2). In some cases, we observed that seeds
buried in the cape remained for long periods of time (e.g., upwards
of a week). Aside from the cape, males accumulated the most seeds
on their lower back, followed by the thigh and knee and lower legs
(Figure S2). Adult females carried more seeds on their lower legs
than on their thighs and knees (Figure S2). Juveniles accumulated
seeds primarily on their thighs, knees, and lower legs (Figure S2).

364
|
VENK ATARAMAN ET A l.
3.5 | Question 5 (Context of gain and loss of seeds)
We observed the context of seed acquisition or removal in 17 in-
stances (Table 3). Instances of acquisition occurred when walking,
sitting, or playing on A. melanantha patches. In one set of cases, two
adult females removed seeds on their arms and ate them. There were
also cases in which infants removed seeds from their fur and test- bit
them before rejecting them. We observed only one case of seeds
being removed during grooming. Some seeds were observed to fall
off when geladas walked or ran through the grass. Our data based
on focal follows indicate that instances of acquisition and loss could
involve five or more seeds, suggesting that seeds can be gained or
lost relatively rapidly. Seeds were observed to be gained in clumps
irregularly, while seeds were observed to be lost in a more gradual
pattern (Figure S3). There was a gain of seeds during only 6.8% of
the focal follow intervals. In contrast, seed loss was observed during
14% of the focal follow intervals in which seed loss was possible (i.e.,
seeds were present to be lost). Seeds were more frequently gained
than lost in clumps of a given size. For example, 68% of instances of
seed acquisition involved clumps of more than one seed, while 55%
of instances of seed loss involved clumps of more than one seed.
The maximum clump size lost at a given time was 12 seeds, while
the maximum clump size gained at one time was 28 seeds. The cape
region had the fastest mean rate of gain (roughly one seed every
5 min) but a slow rate of loss (roughly one seed every 27 min) in
comparison with rates of gain and loss observed in the other body
regions (Table 4). The ventrum had substantially lower rates of both
gain and loss than the other body regions (Table 4).
To estimate how far geladas might be dispersing seeds, we per-
formed a simple calculation that combined estimates of the mean dis-
tance geladas travel with our findings on the time it takes for a seed
to be lost from a gelada's fur. The mean daily path length of the gelada
study band at Guassa is ~3.5 km, the annual home range is ~9 km2, and
the sleeping cliff site changes at least every few days (Moua, 2015).
Over several years of data collection, we have found that the study
band at Guassa is away from their sleeping cliff(s) for approximately ten
hours throughout the day (PJ Fashing, unpubl. data). From these values
of daily ranging and day length, we calculated that geladas move with
a mean speed of 0.35 km/h. Assuming one seed is lost approximately
every 13 min 32 s (Table 4), we estimated that geladas have the poten-
tial to disperse seeds roughly 80 m from their parent plant.
4 | DISCUSSION
The functional traits of A . melanantha at Guassa are consistent with
an animal- based dispersal syndrome. The plants vary in height ac-
cording to microhabitat but are generally low- lying and extend just
FIGURE 3 (a) Seasonal variation in landscape vegetation NDVI,
i.e., “greenness,” (b) total daily counts (n = 4050) of A. melanantha
seeds on geladas, and (c) the relationship between NDVI and seed
counts (highest 200 counts in a given month). The years of the
study period (2017– 2019) are coded by color in (a) and (b), with
month shown on the x- axis. In (a), landscape greenness is measured
by the Normalized Difference Vegetation Index, which is unitless.
In (b), each point represents daily data from one individual. Seed-
carrying of A. melanantha by geladas occurs primarily during and
after the wet season months when the landscape is greenest, a
period when A. melanantha is flowering and fruiting. There is a
strong positive relationship between NDVI and peak seed counts

|
365
VENK ATARAMAN ET A l.
above the surrounding dense vegetation where the barbed achenes
are likely to be picked up by passing animals. Indeed, most A. mel-
anantha plants were found in the short- grass habitat that geladas
generally prefer to feed in— and where they shuffle at least twice
as frequently— relative to other habitats (Eriksen, 2017). While
A. melanantha is patchily distributed (occurring in only 25% of plots),
patches often contain many plants (up to eight), and individual plants
often hold many seeds (up to 95). Thus, when geladas encounter
patches of A. melanantha while traveling or feeding, many seeds can
be accumulated at once. The observed patterns of seed carrying
by geladas were highly seasonal, as expected based on the strong
rainfall and landscape greenness seasonality at Guassa. We demon-
strate that, in accordance with expectations, both time of year and
body mass were good predictors of seed counts on geladas (Tables
1 and 2, Figure 3). Adult geladas carried more seeds than infants
and juveniles, and adult males carried marginally more seeds than
adult females (Table 2). Aside from the cape of adult males, geladas
accumulated seeds most often on their hindlimbs (Figure 5), suggest-
ing that geladas incidentally accumulate A. melanantha seeds as they
move across the landscape and habitually feed while sitting on their
haunches. Our estimates suggest that seeds adhered to geladas have
the potential to travel, on average, 80 m from the parent plant.
In studies of epizoochorous dispersal, seed acquisition is often
considered to be inadvertent and random on the part of the animal
(Sorensen, 1986). While geladas do randomly encounter A. melanan-
tha as they travel, they are also actively drawn to these patches,
as A. melanantha is a commonly consumed edible forb by geladas
at Guassa. After all, edible forbs, including A. melanantha, represent
<7% of land cover at Guassa yet comprise almost 30% of the geladas’
diet, suggesting a high level of selectivity for A. melanantha and other
forbs (Fashing et al., 2014). During the dry season (December– May),
when geladas shift their diet to include more underground foods,
the roots of A. melanantha are also heavily exploited (Venkataraman
et al., 2014; Figure S4).
4.1 | Geladas as dispersers of A. melanantha
Geladas may be one of the most effective animal dispersers of A.
melanantha at Guassa and more generally in the Ethiopian highlands,
although additional data are needed to assess this possibility. In
FIGURE 4 Mean number of A.
melanantha seeds on geladas, color- coded
by size class, during June– December,
averaged across years 2017– 2019. Data
represent mean seed counts per individual
during a given month. Bars and whiskers
represent means and standard errors,
respectively
TAB LE 2 Model output for the best- fitting model (the global
model in Table 1) predicting seed counts on geladas
Fixed effect bSE Z- value p- Value
Intercept −4.68 0.53 −8.82 <.005
BodyMass 0.11 0.01 11.0 5 <.005
Yea r
2017
2018 0.07 0.042 1.70 .09
2019 − 0. 61 0.045 −13.55 <.005
Month
January
February −0.37 0.62 −0.59 . 55
May −0.64 0.64 −1 . 0 0 .32
June 2.24 0.50 4.45 <.005
July 3.36 0.50 6.73 <.005
August 3.79 0.50 7. 6 0 <.005
September 5.45 0.50 10.96 <.005
October 4.84 0.50 9.7 2 <.005
November 4.19 0.50 8.42 <.0 05
December 2.82 0.50 5.63 <.005
Notes: Model coefficients (b) and standard errors (SE), as well as test
statistic Z- and p- values, are given for fixed effects. January and
2017 were reference levels for the fixed effects of month and year,
respectively. Bolded p- values indicate significance at the α = 0.05 level.

366
|
VENK ATARAMAN ET A l.
Afroalpine habitats in Ethiopia, geladas have at least five times the
biomass of other wild large mammals such as klipspringer (O. ore-
otragus), ibex (Capra walie), and bushbuck (Tragelaphus sylvaticus)
(Dunbar, 1978). Geladas also form some of the largest aggregations
of any primate (up to ~800 individuals in herds at Guassa), and their
herds move relatively cohesively through the landscape (Figure 1b).
It is currently unknown whether and to what extent other Guassa
mammals— including domestic livestock (e.g., sheep, goats, donkeys,
horses, and cows)— consume and/or disperse A. melanantha.
The effectiveness of A. melanantha dispersal via gelada fur relies
not only on the emigration and transfer phases detailed in this paper,
but also on the immigration phase in which seeds fall to the ground
and begin the process of growing into a new plant. At present, we
do not know how long A. melanantha seeds are retained on gelada
fur because we were not able to track the acquisition and loss of
individual seeds. We did document numerous cases in which large
numbers of seeds were acquired and lost relatively quickly (within
hours or days), as has been found among ungulates (Bullock et al.,
FIGURE 5 Counts of Agrocharis
melanantha seeds on geladas across body
regions. The x- axis shows gelada body size
classes. Data are from the months of June
through December and represent the
mean of daily seed counts. Data averaged
across years by month
TAB LE 3 Descriptions of opportunistic observations of Agrocharis melanantha seeds being acquired or lost on gelada fur.
Event Sex Size class Acquisition/loss Quantity Body location Notes
1MSmall juvenile A4Hindquarters Sat on flower
2MSmall juvenile A32 thigh; 1 hindquarters Sat on flower
3MMedium juvenile A8Hindquarters Sat on flower
4FSmall juvenile A2Forearm Feeding
5M Brown infant A 53 thigh; 2 lower leg Sat on flower
6M Large juvenile A 6Ventru m Entire umbel attached to stomach
during feeding
7FSmall juvenile A5Thigh Entire umbel attached to thigh after
sitting on plant
8F Brown infant A 41 foot; 1 lower leg; 1
mid- tail; 1 thigh
Playing
9FAdult A2Lower leg Walking
10 MAdult A1Thigh Walking
11 F Brown infant L 1Thigh Running
12 FSmall juvenile L2Forearm Removed and eaten
13 FBlack infant L1Upper arm Removed and eaten
14 FAdult L1Hindquarters Epizoochory by transfer; removed
by individual A while grooming B;
attempted to eat but interrupted by
individual C; transferred to C’s arm
15 F Brown infant L 1Lower leg Removed from fur, inspected, eaten,
then spit out
16 FAdult L5Hindquarters Sat on plant and acquired entire umbel
17 FAdult L1Proximal tail Walking

|
367
VENK ATARAMAN ET A l.
2011; Liehrmann et al., 2018). In general, seeds appear to be sporad-
ically gained in large clumps and then gradually lost.
The variation in time of seed gain and loss reported for the five
observed body regions further indicates that some body regions are
more involved in epizoochorous seed dispersal than others. This is not
surprising, as variation in fur qualities has been shown to drive epizo-
ochorous seed dispersal patterns in some ungulates (Liehrmann et al.,
2018). In geladas, although the male cape does gain seeds the quick-
est, its relatively low loss rate indicates that any seeds gained by the
cape are typically accumulated. This pattern of accumulation comple-
ments the high number of seeds observed on the cape throughout the
entirety of this study. In contrast, the hindlimbs, forearms, and back
evinced more dynamic seed fluxes. These body regions are in close
proximity to plants, while geladas are moving and feeding. Finally,
seeds were rarely gained or lost by the ventrum, which does not regu-
larly contact vegetation with the full weight of the individual.
Geladas not only carry seeds of A. melanantha on their fur but also
consume its foliage, seeds, and underground storage organs (Fashing
et al., 2014). Thus, we suspect that geladas may also be endozoochor-
ously dispersing A. melanantha. Indeed, several ungulates (e.g., wild
pigs, bison, and red deer) disperse seeds epizoochorously via their
fur, and also consume, then endozoochorously disperse, the seeds
of those same plants (Baltzinger et al., 2019). Although the seeds of
A. melanantha are only occasionally consumed, we have observed that
geladas do target individual seeds of A. melanantha on the umbels as
food items, indicating that the seeds are likely to be chewed. To better
understand the fitness impact that geladas might have on A. melanan-
tha, it would be useful to know whether gelada feces contain intact
(i.e., unchewed) A. melanantha seeds, and how this process influences
seed viability. Overall, these considerations suggest that geladas exert
conflicting fitness impacts on A. melanantha. The leaves of A. melanan-
tha are regularly consumed during both the wet and dry season, and
the roots of A. melanantha are highly sought after during the season,
particularly by larger animals with the strength to extract them from
the ground (Fashing et al., 2014; Figure S4). Thus, while geladas can
be predators of A. melanantha in both the dry and rainy season, the
present study also underscores the unique dispersal contributions of
geladas to A. melanantha seeds during many months of the year.
4.2 | Why don't geladas remove seeds?
Sorensen (1986) argued that epizoochorous dispersal is more suc-
cessful if animals are unaware of the seeds, as animals may remove
them during grooming. Given that geladas do consume A. melanantha
seeds, their behavior toward the seeds of A. melanantha on the fur
is puzzling. In only one instance have we observed geladas removing
seeds while engaging in social grooming. In only two instances did we
observe seed removal during self- grooming, suggesting that the seeds
cause little discomfort. Given that grooming is carefully observed dur-
ing focal follows, the absence of seed removal behaviors during such
instances of self- or social grooming is notable. Perhaps seed removal
during grooming is not attempted because the seeds are not worth
the time investment to remove, as geladas are highly time- constrained
(Dunbar, 1992); because the seeds that are deeply entangled in clumps
could cause painful hair- pulling, potentially leading to an agonistic in-
teraction, and/or because seed removal does not result in the hand-
to- skin contact that appears to characterize effective social grooming.
Another possibility is that the individual burrs provide fewer calories
and/or nutrients than the insects and scabs the geladas usually pick off
each other's bodies while grooming.
A defining aspect of epizoochorously dispersed fruits is that
there is no reward for dispersal (Sorensen, 1986). Janzen’s (1984)
“foliage is the fruit” hypothesis argued that selection has shaped
some plants to attract large herbivores to eat their leaves and inci-
dentally ingest, then disperse, their seeds. Based on observations of
donkey feeding behavior and seed dispersal, Couvreur et al. (2005)
argued that epizoochory and endozoochory may exist in a single
system and be complementary. We may be observing a similar phe-
nomenon with A. melanantha and geladas: numerous parts of the
plant are consumed and attractive to geladas, epizoochory is clearly
occurring (this study), and endozoochory is plausible. We suspect
that A. melanantha, which has a wide distribution across Africa (Lee
& Downie, 1999), has been shaped by evolution to appeal to large
mammals— or at least to attach to them. Our study suggests that, at
Guassa, it seems to be geladas in particular that help A. melanantha
to flourish. Further information on the dispersal ecology of A. mela-
nantha is necessary to explore coevolution between this plant and
other mammal communities.
5 | CONCLUSION
Here, we have described a case of epizoochorous seed dispersal in
a savanna primate. Geladas regularly and effectively transport the
seeds of a frequently consumed forb, A. melanantha, in their fur
in large quantities, and they do not appear to actively remove the
seeds during self- or social grooming. These circumstances contrast
with the only other description of epizoochorous seed dispersal in a
TAB LE 4 Mean rate of seed gain and loss by body region
Body region Time to gain one seeda
Time to lose
one seedb
Cape 5 min 6 s 27 min 35 s
Hindlimbs 7 min 15 s 10 min 33 s
Forearms 20 min 30 s 6 min 52 s
Back 29 min 59 s 22 min 51 s
Vent ru m 1 h 53 min 17 s 44 min 32 s
All Regions 14 min 14 s 13 min 32 s
aSample size for gains includes only intervals where there could have
been an observed gain of seeds (i.e., it was not the first interval of the
focal), n = 2 16 9.
bSample size for losses includes only intervals where there could have
been an observed loss of seeds (i.e., there was a non- zero number of
seeds in the preceding interval and it was not the first interval of the
focal), n = 555 .

368
|
VENK ATARAMAN ET A l.
primate (Chen et al., 2018), in which free- ranging golden snub- nosed
monkeys (Rhinopithecus roxellana) in the temperate forests of cen-
tral China carried haired, hooked, or awned seeds on their fur de-
rived primarily from three herbaceous plant s (Geum aleppicum, Torilis
japonica, and Agrimonia pilosa), which they did not consume (Chen
et al., 2018; Yao et al., 2021). The relationship between geladas and
A. melanantha seems distinctive in the extent to which geladas rely
on A. melanantha as a food source and potentially exert both positive
and negative fitness impacts on this plant species.
ACKNOWLEDGMENTS
The authors thank the Ethiopian Wildlife Conservation Authority,
Amhara Regional Government, and Mehal Meda Woreda for permis-
sion to conduct this research. We thank Dean Gibson and San Diego
Zoo, California State University Fullerton, Primate Conservation Inc.,
Gisela and Norman Fashing, Joseph and Patricia Healey, Anita and
Hans- Peter Profunser, and the many contributors to the Christopher
Schroen Memorial Fund for their financial support of the long- term
gelada research at Guassa. Metikay Basasen, Badiloo Muluyee,
Shoafera Tessema, Bantilka Tessema, and Tasso Wudimagegn pro-
vided crucial logistical support in the field. Kim McConkey and
Christophe Baltzinger provided valuable comments that greatly im-
proved the manuscript. This paper is dedicated to the communities
living near the Guassa Plateau who have contributed to the conser-
vat ion of the area and are now affected by the civil wa r in the region.
CONFLICT OF INTEREST
The corresponding author confirms on behalf of all authors that
there have been no involvements that raise the question of bias in
the work reported or in the conclusions, implications, or opinions
stated.
AUTHOR CONTRIBUTIONS
VV and CM conceived the study. VV wrote the paper. All authors
contributed to editing the manuscript, and contributed to data col-
lection and analysis.
DATA AVAIL AB ILI T Y STAT EME N T
The data used in this study, along with computer code to replicate
the data analyses, have been posted at https://github.com/vivek
vasi/Agroc haris.
ORCID
Vivek V. Venkataraman https://orcid.org/0000-0001-5016-4423
Bing Lin https://orcid.org/0000-0002-5905-9512
REFERENCES
Albert, A., Auffret, A. G., Cosyns, E., Cousins, S. A. O., D'hondt, B.,
Eichberg, C., Eycott, A. E., Heinken, T., Hoffmann, M., Jaroszewicz,
B., Malo, J. E., Mårell, A., Mouissie, M., Pakeman, R. J., Picard, M.,
Plue, J., Poschlod, P., Provoost, S., Schulze, K. A., & Baltzinger,
C. (2015). Seed dispersal by ungulates as an ecological filter: A
trait- based meta- analysis. Oikos, 124, 110 9– 1120. ht tps://doi.
org /10.1111/oik.02512
Anderson, D. R. (2008). Model based inference in the life sciences: A primer
on evidence. Springer.
Andresen, E., Arroyo- Rodriguez, V., & Ramos- Robles, M. (2018). Primate
seed dispersal: Old and new challenges. International Journal
of Primatology, 39, 44 3– 465. https://doi.org/10.10 07/s1076
4 - 0 1 8 - 0 0 2 4 - z
Ashenafi, Z. T. (2001). Common property resource management of an Afro-
alpine habitat supporting a population of the critically endangered
Ethiopian Wolf (Canis simensis). PhD thesis, University of Kent.
Baltzinger, C., Karimi, S., & Shukla, U. (2019). Plants on the move: Hitch-
hiking with ungulates distributes diaspores across landscapes.
Frontiers in Ecology and Evolution, 7, 38. https://doi.org/10.3389/
fevo.2019.00038
Barton, K. (2011). MuMIn: Multi- model inference. The R Project for
Statistical Computing.
Bates, D., Machler, M., Bolker, B., & Walker, S. (2015). Fitting linear
mixed- effects models using lme4. Journal of Statistical Software, 67,
1– 4 8 .
Bergman, T. J., & Beehner, J. C. (2013). Theropithecus gelada (Gelada
Baboon). In T. But ynski, J. Kingdon, & J. Kalina (Eds.), Mammals of
Africa: Volume II: Primates (pp. 240– 244). Bloomsbury Publishing.
Bullock, J. M., Galsworthy, S. J., Manzano, P., Poschlod, P., Eichberg, C.,
Walker, K., & Wichmann, M. C. (2011). Process- based functions for
seed retention on animals: A test of improved descriptions of dis-
persal using multiple data sets. Oikos, 120, 1201– 1208. https://doi.
org /10.1111/j.1600 - 0706.2010.19092.x
Burnham, K. P., & Anderson, D. R. (2002). Model selection and multi- model
inference: A practical information- theoretic approach. Springer.
Burnham, K. P., Anderson, D. R., & Huyvaert, K. P. (2010). AIC model selection
and multimodel inference in behavioral ecology: Some background,
observations, and comparisons. Behavioral Ecology and Sociobiology,
65, 23– 35. https://doi.org/10.1007/s0026 5- 010- 1029- 6
Chapman, C. A. (1995). Primate seed dispersal: Coevolution and conser-
vation implications. Evolutionary Anthropology, 4, 74– 82. https://
doi.org/10.1002/evan.13600 40303
Chen, Y., Chen, H., Zhang, Y., Yao, H., Yang, W., Zhao, Y., Ruan, X., &
Xiang, Z. (2018). First evidence of epizoochorous seed dispersal by
golden snub- nosed monkeys (Rhinopithecus roxellana) in temperate
forest. Plant Ecology, 219, 417– 427. https://doi.org/10.1007/s1125
8- 018- 0805- 7
Couvreur, M., Cosyns, E., Hermy, M., & Hoffmann, M. (2005).
Complementarity of epi- and endozoochory of plant seeds by free
ranging donkeys. Ecography, 28, 37– 48. ht tps://doi.or g/10.1111/
j.0906- 7590.2005.04159.x
Didan, K., Munoz, A. B., Solano, R., & Huete, A. (2015). MODIS vegetation
index user’s guide (MOD13 series). University of Arizona: Vegetation
Index and Phenology Lab.
Dunbar, R. I. M. (1978). Competition and niche separation in a high-
altitude herbivore community in Ethiopia. African Journal of
Ecology, 16, 183– 199. https://doi.org/10.1111/j.1365- 2028 .1978 .
t b 0 0 4 3 9 . x
Dunbar, R. I. (1992). Time: A hidden constraint on the behavioural ecol-
ogy of baboons. Behavioral Ecology and Sociobiology, 31, 35– 49.
https://doi.org/10.10 07/BF001 67814
Eriksen, G. A. (2017). Foraging strategies of wild gelada monkeys
(Theropithecus gelada) in an Afroalpine grassland at Guassa, Ethiopia.
MSc Thesis, University of Oslo.
Fashing, P. J., Nguyen, N., Venkataraman, V. V., & Kerby, J. T. (2014).
Gelada feeding ecology in an intact ecosystem at Guassa, Ethiopia:
Variability over time and implications for theropith and hominin di-
etary evolution. American Journal of Physical Anthropology, 155, 1–
16. ht tps://doi.org/10.1002/ajpa.22559
Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., & Moore,
R. (2017). Google Earth Engine: Planetary- scale geospatial analysis
for everyone. Remote Sensing of Environment, 202, 18– 27. https://
doi.org/10.1016/j.rse.2017.06.031

|
369
VENK ATARAMAN ET A l.
Green, A. J., Baltzinger, C., & Lovas- Kiss, Á. (2021). Plant dispersal syn-
dromes are unreliable, especially for predicting zoochory and long-
distance dispersal. Oikos, 1– 11. https://doi.or g/10.1111/oik.08327
Gutema, T. M., Atickem, A., Tsegaye, D., Bekele, A., Sillero- Zubiri, C.,
Ma rin o, J., Kas so, M., Venk ata raman , V. V., Fas hin g, P. J., & Stenseth ,
N. C. (2019). Fora ging ecolog y of Afr ica n wo lves (Canis lupaster) and
its implications for the conservation of Ethiopian wolves (Canis si-
mensis). Royal Society Open Science, 6, 190772.
Hedberg, I., Edwards, S., & Nemomissa, S. (2003). Flora of Ethiopia and
Eritrea. Vol. 4. Part 1, Apiaceae to Dipsacaceae. National Herbarium,
Addis Ababa University.
Howe, H. F., & Smallwood, J. (1982). Ecology of seed dispersal. Annual
Review of Ecology and Systematics, 13, 201– 228. https://doi.
org/10.1146/annur ev.es.13.110182.001221
Iwamoto, T., & Dunbar, R. I. M. (1983). Thermoregulation, habitat quality
and the behavioural ecology of gelada baboons. Journal of Animal
Ecology, 52, 357– 366. https://doi.org/10.2307/4559
Janzen, D. H. (1984). Dispersal of small seeds by big herbivores: Foliage
is the fruit. The American Naturalist, 123, 338– 353. https://doi.
org/10.1086/284208
Jarvey, J. C., Low, B. S., Pappano, D. J., Bergman, T. J., & Beehner, J. C.
(2018). Graminivor y and fallback foods: Annual diet profile of gela-
das (Theropithecus gelada) living in the Simien Mountains National
Park, Ethiopia. International Journal of Primatology, 39, 105– 126.
h t t p s : // d o i . o r g / 1 0 . 1 0 0 7 / s 1 0 76 4 - 0 1 8 - 0 0 1 8 - x
Kidane, B., van Andel, T., van der Maesen, L. J. G., & Asfaw, Z. (2014). Use
and management of traditional medicinal plants by Maale and Ari
ethnic communities in southern Ethiopia. Journal of Ethnobio logy and
Ethnomedicine, 10, 46. https://doi.org/10.1186/1746- 4269- 10- 46
Kleyer, M., Bekker, R. M., Knevel, I. C., Bakker, J. P., Thompson, K.,
Sonnenschein, M., Poschlod, P., van Groenendael, J. M., Klimeš,
L., Klimešová, J., Klotz, S., Rusch, G. M., Hermy, M., Adriaens, D.,
Boedeltje, G., Bossuyt, B., Dannemann, A., Endels, P., Götzenberger,
L., … Peco, B. (2008). The LEDA Traitbase: A database of life- history
traits of the Northwest European flora. Journal of Eco logy, 96, 1266–
1274. https://doi.org/10.1111/j.1365- 2745.2008.01430.x
Lee, B. Y., & Downie, S. R. (1999). A molecular phylogeny of Apiaceae
tribe Caucalideae and related taxa: Inferences based on ITS
sequence data. Systematic Botany, 24, 461– 479. https://doi.
org/10.2307/2419700
Liehrmann, O., Jégoux, F., Guilbert, M.- A., Isselin- Nondedeu, F., Saïd, S.,
Locatelli, Y., & Baltzinger, C. (2018). Epizo ochorous dispersal by un-
gulates depends on fur, grooming and social interactions. Ecolog y
and Evolution, 8, 1582– 1594. https://doi.org/10.1002/ece3.3768
Lin, B., Fox fo ot, I. R., Miller, C. M. , Venkatar a m a n , V. V., Ker by, J. T., Be cht ol d ,
E. K., Kellogg, B. S., Nguyen, N., & Fashing, P. J. (2020). Leopard pre-
dation on gelada monkeys at Guassa. Ethiopia. American Journal of
Primatology, 82, e23098. https://doi.org/10.1002/ajp.23098
Mazerolle, M. J. (2013). Package ‘AICcmodavg’. The R Project for Statistical
Computing.
Moua, C. (2015). Long- term ranging patterns of wild gelada monkeys
(Theropithecus gelada) on an intact afro- alpine grassland at Guassa,
Ethiopia. MSc, California State University, Fullerton.
Nguyen, N., Fashing, P. J., Boyd, D. A ., Barry, T. S., Burke, R. J., Goodale,
C. B., Jones, S. C. Z., Kerby, J. T., Kellogg, B. S., Lee, L. M., Miller,
C. M., Nurmi, N. O., Ramsay, M. S., Reynolds, J. D., Stewart, K. M.,
Turner, T. J., Venkataraman, V. V., Knauf, Y., Roos, C., & Knauf, S.
(2015). Fitness impacts of tapeworm parasitism on wild gelada
monkeys at Guassa, Ethiopia. American Journal of Primatology, 77,
579– 594. https://doi.org/10.1002/ajp.22379
Pettorelli, N., Vik, J. O., Mysterud, A., Gaillard, J. M., Tucker, C. J., &
Stenseth, N. C . (2005). Using the satellite- derived NDVI to assess
ecological responses to environmental change. Trends in Ecology a nd
Evolution, 20, 503– 510. https://doi.org/10.1016/j.tree.2005.05.011
Snyder- Mackler, N., Beehner, J. C., & Bergman, T. J. (2012). Defining
higher levels in the multilevel societies of geladas (Theropithecus
gelada). International Journal of Primatology, 33, 1054– 1068. https://
d o i . o r g / 1 0 . 1 0 0 7/ s 1 0 7 6 4 - 0 1 2 - 9 5 8 4 - 5
Sorensen, A. E. (1986). Seed dispersal by adhesion. Annual Review of
Ecology and Systematics, 17, 443– 463. https://doi.org/10.1146/
annur ev.es.17.110186.002303
Tucker, C. J. (1979). Red and photographic infrared linear combinations
for monitoring vegetation. Remote Sensing of Environment, 8, 127–
150. https://doi.org/10.1016/0034- 4257(79)90013 - 0
Venkataraman, V. V., Glowacka, H., Fritz, J., Clauss, M., Seyoum, C.,
Nguyen, N., & Fashing, P. J. (2014). Effects of dietary fracture
toughness and dental wear on chewing efficiency in geladas
(Theropithecus gelada). American Journal of Physical Anthropology,
155, 17– 32.
Welch, C., Frost, T. B., & Kerby, J. T. (2017). Where the world’s only grass-
eating monkeys thrive. National Geographic.
Wrangham, R. W. (1980). Bipedal locomotion as a feeding adapta-
tion in gelada baboons, and its implications for Hominid evo-
lution. Journal of Human Evolution, 9, 329– 331. https://doi.
o r g / 1 0 . 1 0 1 6 / 0 0 4 7 - 2 4 8 4 ( 8 0 ) 9 0 0 5 9 - 7
Yao, H., Bai, Y., Chen, Y., Chen, H., Yang, W., Ruan, X., & Xiang, Z. (2021).
Endozoochorous seed dispersal by golden snub- nosed monkeys in
a temperate forest. Integrative Zoology, 16, 120– 127. https://doi.org
/10.1111/1749- 4 877.12488
SUPPORTING INFORMATION
Additional supporting information may be found in the online
version of the article at the publisher’s website.
How to cite this article: Venkataraman, V. V., Miller, C.,
Foxfoot, I., Lin, B., Petrie, Z. L., Simberloff, R. A., Bernardo, O.,
Redon, N., Hohn, T. I., Kerby, J. T., Nguyen, N., & Fashing, P. J.
(2022). Epizoochorous seed dispersal by an Afroalpine savanna
primate. Biotropica, 54, 358– 369. https://doi.org/10.1111/
btp.13057