Citation
Behavioral and demographic responses to habitat change by the Tana River crested mangabey (Cercocebus galeritus galeritus)

Material Information

Title:
Behavioral and demographic responses to habitat change by the Tana River crested mangabey (Cercocebus galeritus galeritus)
Creator:
Kinnaird, M
Publication Date:
Language:
English
Physical Description:
xxii, 277 leaves : ill., maps ; 29 cm.

Subjects

Subjects / Keywords:
Food ( jstor )
Foraging ( jstor )
Forests ( jstor )
Fruiting ( jstor )
Fruits ( jstor )
Infants ( jstor )
Population estimates ( jstor )
Population size ( jstor )
Primates ( jstor )
Species ( jstor )
Dissertations, Academic -- Forest Resources and Conservation -- UF
Forest Resources and Conservation thesis Ph. D
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1990.
Bibliography:
Includes bibliographical references (leaves 261-276).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Margaret F. Kinnaird.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
027771640 ( ALEPH )
26498450 ( OCLC )

Downloads

This item has the following downloads:


Full Text












BEHAVIORAL AND DEMOGRAPHIC RESPONSES TO HABITAT CHANGE
BY THE TANA RIVER CRESTED MANGABEY
(Cercocebus qaleritus qaleritus)



















By

MARGARET F. KINNAIRD


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1990






























To Ratilli Kawa

and the women of rural East Africa















ACKNOWLEDGMENTS


My research was funded by Wildlife Conservation

International, New York Zoological Society. I was sponsored

in Kenya by the National Museums of Kenya (NMK) Biological

Resources Program and the Institute of Primate Research

(IPR). The National Council for Science and Technology and

The Office of the President granted me permission to work

within Kenya (Permit No. OP. 13/001/160-282/12). While at

the University of Florida, I was supported by assistantships

from the Center of African Studies and Dr. John Eisenberg. I

warmly thank these people and organizations for the

opportunities they provided.

Many people have contributed, both knowingly and

unknowingly, to my study and to each one I owe a great debt.

Drs. John Eisenberg, John Robinson, Jack Putz, Michael

Collopy and Tom O'Shea, members of my doctoral committee,

helped shape my ideas both before and during the field work,

and have been constructive editors throughout the writing

phase. I extend special thanks to Dr. John Eisenberg, my

major professor, whose advice to remain ever "flexible and

patient" proved to be the most valuable and who, in his own

way, demanded my best.


iii









I owe perhaps my greatest debt to Dr. Tom Struhsaker.

He helped in the study design, pushed for funding, made

several well-needed trips to my field site and has been my

most critical editor. Tom encouraged and reassured me when I

most needed it and pushed me back on course when I was

straying. He has been teacher, college, interrogator,

project supervisor, and, most importantly, friend. For all

of this, I extend my deepest thanks.

Many of my ideas were formed through conversations with

fellow students and colleagues. I thank Jeff Hardesty, Dr.

Susan Jacobson, Lysa Leland, Jay Malcolm and Peter Polshek

for their input, but in no way hold them responsible for the

final product. Gainesville friends Lane Devereaux, Mark

Edmiston, Nina Hoffer, Susan Jacobson, Jim and Katherine

Reid and Bob and Deb Wallace helped logistically and/or

emotionally. I also thank Laurie Wilkens and Dr. Steve

Humphrey for providing me with large and friendly working

spaces during the writing phase. Dr. Andy Dobson and the

students and staff of the Ecology and Evolutionary Biology

Department of Princeton University also are thanked for

providing me with a stimulating atmosphere and comfortable

office in the final month of writing.

The previous and extensive studies by Dr. Katherine

Homewood were an invaluable resource. Dr. Homewood also

furnished unpublished data, supplemental information and

inspiration.








While in Kenya, my life was made enormously easier by

the NMK and IPR. Dr. Jim Else dealt with every millimeter of

red tape in Nairobi while I was isolated in the bush. Dr.

Mohammed Isahakia, Dr. Steven Njuguna, Richard Leakey,

Margaret Omoto, and Mary Sefu also played critical roles in

securing the necessary visas, permits and affilliations.

Mary Matibo, Dorothy McCullough and Debbie Snelson of the

African Wildlife Foundation kindly dealt with licenses and

insurance for my jeep and saw that I was legal on the road.

My stays in Nairobi were memorable and warm due to the

hospitality and generosity of Jim, Margaret and Jessie Else.

In Malindi, Elsa Friman provided me with a luxurious home

base. She, Romolo, and their employee Sammi, renewed me in

ways they will never know. I also thank Ann and Ian

Robertson for the many wonderful afternoons and discussions

on their Malindi porch.

My research was made much easier by the comfortable

accommodations of the Tana River Research Camp. Fred Decker

designed, built and kept the camp in working order for most

of my stay; the fact that all ran smoothly in his absence is

testimony to his skill. I also thank Alex Njui for managing

the camp after Fred's departure. To our camp staff Mustafa

Lordis, Bwana Mzee, Jumaa Galana, and Abio I give thanks for

their help and friendship. I also owe very special thanks to

my field assistants Galole Galana and Bakari Kawa. They

taught me much about jungle life and without their help, the








process of habituating elusive mangabeys would have been a

much harder and more arduous task.

Kim Medley, Barbara Decker, Odhiambo Ochiago, Paula

Kahumbu, Francis Muli, and Alex Njue shared the camp with me

as fellow students and bush companions for various periods

of time. Their company was greatly appreciated, especially

that of Kim Medley, who shared the first year and many first

experiences with me.

My experience living on the Tana River was made rich

and rewarding by a host of kind and generous people. At the

top of the list comes the Kawa family, who allowed me to

construct a small, auxiliary campsite on their shamba. I

could never repay them for their overwhelming hospitality.

I thank Father Spiro, Sisters Romona and Evelyn, and

the late Sister Agnes of the Wema Catholic Mission for

opening their home to me on numerous occasions when I was

surveying the forests of the lower Tana. Their strength will

always be remembered as an inspiration to me.

I single out Tim O'Brien for very special and sincere

thanks. This product would not have been the same without

his input. He has helped in literally all stages--from

assisting in the field to advising on statistical analysis,

writing numerous computer programs, and reading every

chapter. He suffered through all phases of the project and,

in spite of it all, stuck beside me.








I also thank my parents who have endured much, including

my long research stints out of the country. They have

encouraged and supported me in numerous ways throughout all

my endeavors, even when they may not have understood fully

just what drove me.

Finally, I thank the "Makarau Mgogo", those rare and

elusive primates who allowed me, albeit with insouciance, to

investigate their lives. Without their cooperation, this

study would not have been feasible.


vii















TABLE OF CONTENTS



ACKNOWLEDGMENTS . . . . . . . . . . iii

LIST OF TABLES . . . . . . . . . .. x

LIST OF FIGURES . . . . . . . . . .. xiv

ABSTRACT . . . . . . . . . . . .. XX

CHAPTER ONE INTRODUCTION . . . . . . . 1

CHAPTER TWO ABUNDANCE AND DISTRIBUTION OF FRUIT
RESOURCES IN THREE TANA RIVERINE FORESTS ... . 12
Introduction . . . . . . . . .. 12
Methods . . . . . . . . . .. .14
Study Sites . . . . . . . .14
Abundance and Distribution of Resources . 17
Environmental Correlates with Flowering and
Fruiting . . . . . . . .. 24
Comparisons with 1973-74 . . . . .. 25
Results . . . . . . . . ... * * .26
Abundance and Spatial Distribution of
Resources . . . . ................. 26
Temporal Distribution of Resources . . 27
Environmental Correlates with Flowering and
Fruiting . . . . . . . . .. .60
Comparisons with 1973-74 . . . . .. 72
Discussion . . . . . . . . . .. 73

CHAPTER THREE ACTIVITY PATTERNS AND DIET OF THE TANA RIVER
CRESTED MANGABEY: EFFECTS OF SEASONAL AND LONG-TERM
HABITAT CHANGE . . . . . . . . .. 79
Introduction . . . . . . . . .. 79
Methods ............ ........................ ..82
Data Collection . . . . . . .. 82
Data Analysis . . . . . . . .. 85
Results . . . ... o... . . . . .86
Variation in Activities 1988-89 . . . 86
Comparison with 1973-74 Activity Budgets . 101
Diet . . . . . . . . . .. ill
Discussion . . . . . . . . . .. 124


viii









CHAPTER FOUR COMPETING USES OF A FOREST PALM, PHOENIX
RECLINATA N. J. JACQUIN, BY HUMANS AND TANA RIVER
CRESTED MANGABEYS . . . . . . .. 136
Introduction . . . . . . . . .. 136
Methods . . . . . . . . . .. 138
Study Species . . . . . . . .. 138
Data Collection . . . . . . .. 139
Results . . . . . . . ... . . . 141
Discussion . . . . . . . . . . 149

CHAPTER FIVE TANA RIVER CRESTED MANGABEY HOME RANGE
AND SPATIAL DEFENSE: EFFECTS OF FRUIT ABUNDANCE
AND DISTRIBUTION . . . . . . . .. .153
Introduction . . . . . . . . .. 153
Methods .......... ........................ 155
Data Collection . . . . . . .. 155
Data Analysis . . . . . . . .. 157
Results . . . . . . . . . .. 160
Use of Space . . . . . . . .. 160
Movements . . . . . . . . . 179
Intergroup Behavior . . . . . .. 186
Discussion . . . . . . . . . .. 205

CHAPTER SIX POPULATION VIABILITY OF THE TANA RIVER
CRESTED MANGABEY . . . . . . . .. 214
Introduction . . . . . . . . .. 214
Methods . . . . . . . . . . 216
Mangabey Data Set . . . . . . .. 216
Population Genetics Model . . . . .. .219
Demographic Extinction Model . . . .. 226
Results . . ..... ........................ 227
Estimates of Ne . . . . . . .. 227
Population Subdivision . . . . .. 224
Demographic Extinction Model . . . .. 236
Discussion ... . . . . . . . . . 242
Effective Population Size . . . . .. .242
Demographic Extinction Model . . . .. 246

CHAPTER SEVEN CONCLUSION . . . . . . .. 249

APPENDIX A ESTIMATING TREE CANOPY SURFACE AREA AND
VOLUME . . . . . . . . . . . . 255

APPENDIX B NUMBER OF FEEDING RECORDS AND PERCENT OF TOTAL
RECORDS BY PLANT TAXON AND MANGABEY GROUP . . .. .258

REFERENCES . . . . . . . . . . .. 261

BIOGRAPHICAL SKETCH . . . . . . . . .. 277















LIST OF TABLES


Table 2.1. Composition of monthly phenology sample by
species for Mchelelo, Nkano and Mnazini forests. 20

Table 2.2. Density of reproductive trees and lianas
enumerated in 16.32 ha of Mchelelo forest and
13.45 ha in Nkano forest, and the degree of
clumping of each species using Morisita's index
(I) If I, is significantly greater or less than
unity using the F-test, the species is considered
clumped or evenly distributed in space,
respectively. If I, is not significantly
different from unity than the species is
considered randomly distributed . . . .... 27

Table 2.3. Estimated canopy surface area (S), canopy
volume (V), mean fruit production (FN) and mean
fruit weight (FW) for 16 species. Species are
categorized into 4 groups: those that bear their
fruits on the canopy surface (C) or along the
branches of the tree (R), lianas (L) and palms (P)
. . . . . . . . . . o . 54

Table 2.4. Spearman's rank correlation coefficients
for mean monthly environmental variables and mean
monthly fruiting scores for Mchelelo forest.
MINTEMP=minimum monthly temperature; RAIN=total
monthly rainfall (mm); RAINLAG1, RAINLAG2 AND
RAINLAG3=monthly rainfall lagged by 1,2, and 3
months, respectively; RAINADV=previous months'
rainfall; FLOW=mean monthly river flow (mcm);
FLOWLAG1, FLOWLAG2, FLOWLAG3=mean monthly river
flow lagged by 1,2 and 3 months, respectively;
FLOWADV=previous months' mean river flow. . . 68

Table 2.5. Spearman's rank correlation coefficients
for mean monthly environmental variables and mean
monthly fruiting scores for Nkano forest.
Variables defined in Table 2.4 . . . . .. 69

Table 2.6. Spearman's rank correlation coefficients
for mean monthly environmental variables and mean
monthly fruiting scores for Mnazini forest.
Variables defined in Table 2.4 . . . .. . 70








Table 2.7. Spearman's rank correlation coefficients
for mean monthly environmental variables and mean
monthly fruiting scores for all forests combined.
Variables defined in Table 2.4 . . . . .. 71

Table 3.1. Mean percent time spent in each activity by
each month by mangabeys in the 1988-89 N.
Mchelelo, S. Mchelelo, and Nkano study groups.
Standard errors are in parentheses. See text for
definition of activities . . . . . .. 90

Table 3.2. MANOVA results for group membership,
season, and time-of-day differences in 1988-89
mangabey activity budgets . . . . . .. 91

Table 3.3. ANOVA results for group membership, season,
and time-of-day differences in 7 behaviors.
Significance values are calculated from Type III
partial sums of squares . . . . . .... 92

Table 3.4. ANOVA results for study (Homewood's 1973-74
study and the present 1988-89 study) and age
differences in 3 measures of mother/infant
association: a) amount of time infant is on its'
mother, b) amount of time infant has mother as
nearest neighbor and c) amount of time infant is
separate from mother. Data were nested by infants
within studies . . . . . . . . .. . 110

Table 3.5. Rank order of the five most common plant
species eaten and the percentage of the total
feeding records for identified plants (N) by group
and month. Complete scientific names of plant
species listed in Appendix B . . . . .. .113

Table 3.6. Percentage of diet items eaten by month and
group. Unidentified items are not included. . 115

Table 3.7. Frequency of plant parts eaten for common
diet species of the 3 study groups . . .... 117

Table 4.1. Human uses of P. reclinata grouped into 6
broad categories. Plant parts utilized and
harvesting methods employed are outlined . . 142

Table 5.1. Mean diversity of quadrat use (H'), mean
number of unique quadrats entered, mean distances
travelled, mean half-hour step distances, mean
turning angles, and mean number of path crossings
by month for the 3 study groups . . . ... 178








Table 5.2. Spearman's Rank correlation coefficients
(rY) between frequency of quadrat use by mangabeys
and frequency of occurrence of top 5 plant diet
species each month for the North Mchelelo, South
Mchelelo, and Nkano groups. (-) indicates that
ranging data or plant frequency data were not
available for all top 5 diet species that month. 180

Table 5.3. ANOVA results for effect of South Mchelelo
group presence and month on distance traveled and
measures of travel pattern and use of space by
North Mchelelo group. F-statistics are calculated
from Type III partial sums of squares . . .. .190

Table 5.4. Number of adult male long-calls by study
group and month. North refers to long-calls given
by the North Mchelelo group males. South M refers
to long-calls given by the South Mchelelo group
males while present in Mchelelo forest and South C
refers to long-calls given by the South Mchelelo
group males while in the neighboring Congolani
forest. Alien refers to long-calls given by males
of a second, unhabituated group in Nkano forest,
and Unk represents long-calls of unidentified
males in the Nkano forest . . . . . .... 192

Table 5.5. Intergroup behavior categories and measures
of proximity and fighting by month. F, M, and A
designate fights, merge, and avoid months,
respectively. N includes all 12-hr observation
days and days for which only presence or absence
of South group was determined . . . . .. 194

Table 6.1. Minimum and maximum numbers of mangabey
groups censused in forests on the east and west
banks of the lower Tana River in 1988-89 and 1973-
74. Forests not censused indicated with (-);
those no longer standing indicated with (x). Data
from 1973-74 taken from Homewood (1976) and Marsh
(1976, 1978) . . . . . . . . .. 217

Table 6.2. Age-sex composition of 7 mangabey groups
with comparative means from 1973-74 (Homewood
1976). AM = adult male (>4-5 yr); SubM = subadult
male (3-5 yr); AF = adult female (>4 yr); SubF =
subadult female (2.5-3 yr); JM = juvenile male (8
mon-3 yr); JF = juvenile female (8 mon-2.5 yr); IM
and IF = infant male and infant female,
respectively (< 7 mon) . . . . . . .. 218

Table 6.3. Reproductive parameters for male and female
mangabeys . . . . . . . . . .. 220


xii








Table 6.4. Minimum and maximum parameter values and
estimates of effective population size. N =
census number of males or females; K = estimate of
lifetime reproductive success of males or females;
K = variance in lifetime reproductive success;
and Ne = effective population size. Subscripts
denote parameters for males vs females. . . 229

Table 6.5. Parameter values and estimates of the
genetic fixation index, F,,, for forests on east
and west banks and for west bank and east bank
subpopulations. East vs west bank figures are
calculated for group migration from one bank to
the other for two estimates of high flood
frequency . . . . . . . . . ..235


xiii














LIST OF FIGURES


Figure 1.1. Adult male Tana River crested mangabey,
Cercocebus galeritus qaleritus, seated in a
Hypheane compressa palm . . . . . . . 3

Figure 1.2. Map of Kenya and location of the Tana
River National Primate Reserve (TRNPR; map
modified from Hughes 1985) . . . . . . 8

Figure 2.1. Map of the study area showing the
Mchelelo, Mnazini, and Nkano forests within the
Tana River National Primate Reserve . . .. 16

Figure 2.2. Number of individuals/0.25 ha of 4 common
species within Mchelelo forest showing clumped
spatial distributions a) Hyphaene compressa;
b)Alanqium salviifolium; c)Phoenix reclinata; d)
Oncoba spinosa . . . . . . . . .. 30

Figure 2.3. Number of individuals/0.25 ha of 4 common
species within Nkano forest showing clumped
spatial distributions, a) Ficus sycomorus; b)
Alanqium salviifolium; c) Phoenix reclinata; d)
Pachystela msolo . . . . . . . . 32

Figure 2.4. Examples of randomly distributed Ficus
spp. a) Ficus natalensis in Mchelelo forest; b)
Ficus sycomorus in Mchelelo forest; c) Ficus
natalensis in Nkano forest . . . . . .. 34

Figure 2.5. Mean flower scores averaged over all
species by month, a) Mchelelo forest; b) Nkano
forest; c) Mnazini forest . . . . . . 36

Figure 2.6. Mean fruiting scores averaged over all
species by month, a) Mchelelo forest; b) Nkano
forest; c) Mnazini forest . . . . . . 38


xiv








Figure 2.7. Mean monthly fruit scores from Mchelelo
and Mnazini forests for Mimusops fruticosa and
Acacia robusta, seasonally fruiting species with 1
peak in fruiting per year. a) Mimusops fruiticosa
in Mchelelo forest; b) Acacia robusta in Mchelelo
forest; c) Mimusops fruticosa in Mnazini forest;
d) Acacia robusta in Mnazini forest . . .. 41

Figure 2.8. Mean monthly fruit scores from Mchelelo
and Mnazini forests for Garcinia livingstonei and
Sorindeia madagascariensis, seasonally fruiting
species with 2 fruiting peaks per year. a) Garcinia
livingstonei in Mchelelo forest; b) Sorindeia
madagascariensis in Mchelelo forest; c) Garcinia
livingstonei in Mnazini forest; d) Sorindeia
madagascariensis in Mnazini forest . . .. 43

Figure 2.9. Mean monthly fruit scores for Phoenix
reclinata. a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest . . . . . . . . .. 45

Figure 2.10. Mean monthly fruit scores for Diospyros
mespiliformes. a) Mchelelo forest; b) Nkano forest;
c) Mnazini forest . . . . . . . . 47

Figure 2.11. Mean monthly fruit scores for 3 species
showing patterns of continuous fruit production.
a) continuous fruiting due to asynchrony among
individuals of the species (Ficus natalensis); b)
continuous fruiting due to retention of fruits
over time (Hyphaene compressa); c) continuous
fruiting due to rapid turnover of fruits (Oncoba
spinosa) . . . . . . . . . .. 49

Figure 2.12. Mean monthly fruit scores for Ficus
sycomorus. a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest . . . . . . . . .. 51

Figure 2.13. Seasonality of mean monthly fruit scores
for Saba comorensis in the 3 study forests, a)
seasonal fruiting in Mnazini forest; b) seasonal
but extended fruiting in Nkano forest; c)
continuous fruiting in Mchelelo forest . . .. 53

Figure 2.14. Contribution to monthly fruit biomass
indices by different species categories, a)
Mchelelo forest; b) Nkano forest; c) Mnazini
forest . . . . . . . . . . .56









Figure 2.15. Contribution of ripe and unripe fruits to
monthly fruit biomass index, a) Mchelelo forest; b)
Nkano forest; c) Mnazini forest . . . . ... .59

Figure 2.16. Monthly mean minimum and mean maximum air
temperature in degrees centigrade, a) Mchelelo
forest; b) Nkano/Mnazini forest areas . . .. 62

Figure 2.17. Total monthly rainfall in millimeters.
a) Mchelelo forest; b) Nkano/Mnazini forest area 64

Figure 2.18. Mean monthly river flow for the Tana
River Hola Station . . . . . . . .. 66

Figure 3.1. Mean annual percent time spent among 7
behaviors by three 1988-89 mangabey groups and two
1973-74 mangabey groups, a) 1988-89; b) 1973-74 89

Figure 3.2. Mean percent time spent in 4 major
behaviors (eating, foraging, moving and inactive)
by time-of-day and season for the 1988-89 N.
Mchelelo group. . . . . . . . . 96

Figure 3.3. Mean percent time spent in 4 major
behaviors (eating, foraging, moving and inactive)
by time-of-day and season for the 1988-89 S.
Mchelelo group . . . . . . . . .. 98

Figure 3.4. Mean percent time spent in 4 major
behaviors (eating, foraging, moving and inactive)
by time-of-day and season for the 1988-89 Nkano
group . . . . . . . . . . .. 100

Figure 3.5. Mean percent time spent in 7 social
behaviors by 1973-74 and 1988-89 adult males and
females, a) adult males; b) adult females . . 104

Figure 3.6. Mean percent time spent in 7 social
behaviors by three age-sex classes in 1973-74 and
1988-89. a) juvenile males; b) juvenile females;
c) male and female infants combined . . .. .106

Figure 3.7. Percent time 1988-89 and 1973-74 infants
spent in 3 levels of association with their
mothers. Asterisks denote months when levels of
association were significantly different between
the 2 study periods, a) Association A: the amount
of time an infant was on its mother; b)
Association B: the amount of time an infant had
its mother as nearest neighbor; c) Association C:
the amount of time infants were separate from
their mothers . . . . . . . . .. 109


xvi








Figure 3.8. Percent composition of diet by item for
1973-74 and 1988-89. a) 1988-89; b) 1973-74 . 119

Figure 3.9. Percent ripe and unripe fruit and seed
items inthe diet. a) for three 1988-89 mangabey
groups; b) for 1973-74 mangabey groups . . .. .123

Figure 4.1. The number of harvested and unharvested
palms in Nkano and the number of reproductive and
nonreproductive palms in Mchelelo by size class.
a) harvested and unharvested palms by size class
in Nkano; b) reproductive and nonreproductive
palms in Mchelelo by size class . . . .. .145

Figure 4.2. The number of feeding observations on P.
reclinata by month and item for mangabeys in the
Mchelelo and Nkano forests . . . . . .. .148

Figure 5.1. Cumulative frequency of sightings of one
or more North Mchelelo group mangabeys versus rank
order of 0.25 h quadrats. Quadrats were ranked
with respect to the number of sightings occurring
in each. The areas accounting for 50 and 100% of
all sightings are designated . . . . .. .162

Figure 5.2. Cumulative frequency of sightings of one
or more South Mchelelo group mangabeys versus rank
order of 0.25 h quadrats. Quadrats were ranked
with respect to the number of sightings occurring
in each. The areas accounting for 50 and 100% of
all sightings are designated . . . . .. .164

Figure 5.3. Cumulative frequency of sightings of one
or more Nkano group mangabeys versus rank order of
0.25 ha quadrats. Quadrats were ranked with
respect to the number of sightings occurring in
each. The areas accounting for 50 and 100% of all
sightings are designated . . . . . .. 166

Figure 5.4. Taut-string line (shaded with diagonal
lines) enclosing all ad libitum sightings of South
Mchelelo group when the group ranged outside the
Mchelelo forest . . . . . . . .. 168

Figure 5.5. Frequency distributions of intensity of
quadrat use. Asterisks indicate the values
expected if quadrat occupancy were random
(following a Poisson distribution).a) Frequency
distribution for N. Mchelelo group; b) Frequency
distribution for S. Mchelelo group; c) Frequency
distribution for Nkano group . . . . .. . 171


xvii








Figure 5.6. Occupancy of each 0.25 ha quadrat by North
and South Mchelelo groups; percent occupancy is
divided into 1 of 5 intensities of use . . .. .173

Figure 5.7. Occupancy of each 0.25 ha quadrat by Nkano
group; percent occupancy is divided into 1 of 5
intensities of use . . . . . . . .. 175

Figure 5.8. Frequency distributions of half-hour
distance steps in intervals of 20 m. Asterisks
indicate values expected if distance steps were
distributed randomly (following a Poisson
distribution), a) Frequency distribution for N.
Mchelelo group; b) Frequency distribution for S.
Mchelelo group; c) Frequency distribution for
Nkano group . . . . . . . . .. 182

Figure 5.9. Frequency distributions of half-hour
turning angles in 10 degree intervals, a)
Frequency distribution for N. Mchelelo group; b)
Frequency distribution for S. Mchelelo group; c)
Frequency distribution for Nkano group . . .. .185

Figure 5.10. Degree of monthly overlap by North Mchelelo
group (solid line) and South Mchelelo group (broken
line) and the area of overlap in hectares (bars) 189

Figure 5.11. Daily travel paths by North and South
Mchelelo groups during one day of a month of
intergroup avoidance. Timing and location of
adult male long-calls are plotted for North
(ovals) and South (rectangles) groups . . .. .196

Figure 5.12. Daily travel paths by North and South
Mchelelo groups during one day of a month of
intergroup merging. Timing and location of adult
male long-calls are plotted for North (ovals) and
South (rectangles) groups . . . . . .. .198

Figure 5.13. Daily travel paths by North and South
Mchelelo groups during one day of a month of
intergroup fighting. Timing and location of adult
male long-calls are plotted for North (ovals) and
South (rectangles) groups . . . . . .. .201

Figure 5.14. Location and outcome of 34 fights
observed between North and South Mchelelo groups.
N and S indicate that North and South group won
fights, respectively. D indicates that no clear
winner was determined . . . . . . .. .204


xviii









Figure 5.15. Percent time North and South Mchelelo
groups spent feeding on foods with different
distributions as a function of intergroup
behavior, a) Percent time feeding on foods with
uniform distributions; b) Percent time feeding on
foods with patchy distributions . . . .. .207

Figure 6.1. Distribution of estimated female lifetime
reproductive success. Infant values are rounded
to the nearest whole number . . . . .. 231

Figure 6.2. Distribution of male lifetime reproductive
success estimated for males with 1, 2 and 3 tenures as
a dominant, breeding group male. a) 1973-74 male
lifetime reproductive success calculated for 1 male:
9.75 females; b) 1988-89 male lifetime reproductive
success calculated for 1 male:5.85 females . . 233

Figure 6.3. Persistence time in years as a function of
population size (N). Dashed lines connect 1973-74
and 1988-89 population sizes to estimated mean
persistence times . . . . . . . ... 238

Figure 6.4. Probability of extinction p(E) within 100,
50, 25, and 10 years (denoted by labeled curves)
for populations with mean persistence times (T) of
200 to 2000 years . . . . . . . .. 241

Figure 6.5. Effective population size (Nj) as a
function of number of breeders (Nb) and variance
of male lifetime reproductive success (Vm) . . 244


xix














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

BEHAVIORAL AND DEMOGRAPHIC RESPONSES TO HABITAT CHANGE
BY THE TANA RIVER CRESTED MANGABEY
(CERCOCEBUS GALERITUS GALERITUS)

By

Margaret F. Kinnaird

December 1990

Chairperson: Dr. John F. Eisenberg
Major Department: Forest Resources and Conservation
(Wildlife and Range Sciences)


The Tana River crested mangabey (Cercocebus qaleritus

qaleritus) is an endangered primate found only in small

forests of the Tana River in Kenya. The mangabey is afforded

some protection in the Tana River National Primate Reserve.

Since the establishment of the reserve in 1976 the mangabey

population has decreased by 45%. The reduction in habitat

size and quality is believed to have been a major factor

contributing to the population decline. I compare data on

mangabey behavior and demography with a study conducted in

1973-74 to investigate the effects of habitat loss and

reductions in important diet species on mangabey behavior

and risk of extinction.









The Tana mangabey's behavioral and dietary flexibility

enables it to buffer fluctuations in resource availability

in a highly seasonal habitat and to adapt to large-scale

habitat change. Mangabeys forage and eat more, and show

extended infant dependency in response to changes in fruit

distributions and lower fruit availability. A dietary shift

between 1973-74 and 1988-89 from primarily ripe fruits to

unripe seeds and invertebrates may constitute the best

alternative strategy when ripe fruits are less available.

Mangabey ranging patterns are influenced by a complex

interaction of the spatial and temporal distribution of

foods, forest patch size, group size, and intergroup

interactions. Seasonal variation in food availability and

distribution result in facultative defense of space by

mangabeys.

The Tana mangabey's endangered status is primarily the

result of the highly restricted distribution of Tana forest

habitat and the rapid destruction of this habitat.

Degradation of the remaining forests may result from non-

sustainable human exploitation of forest species such as the

palm, Phoenix reclinata. Results of demographic and genetic

models suggest that the mangabey is at risk of losing

genetic variation and has a high probability of extinction

over the next 50-100 years given its present population

status. A conservation strategy should incorporate measures

for habitat preservation and enrichment. Because the


xxi









mangabey is a generalist species that easily exploits new

habitat, it should respond well to habitat management.


xxii















CHAPTER ONE

INTRODUCTION


The Tana River crested mangabey, or Tana mangabey,

(Cercocebus galeritus galeritus Peters 1879) is a rare and

highly endangered cercopithecine monkey (IUCN 1976, Lee et

al. 1988) found only in small forest patches of the Tana

floodplain in north-eastern Kenya (Homewood 1976; Figure

1.1). Mittermeier (1981) ranked C. g. qaleritus among the

most endangered primates in the world, and the IUCN/SSC

Primate Specialist Group listed it in the highest priority

category for conservation (Oates 1985). These rankings were

based on the small population size of C. g. galeritus, its

highly restricted range, and the present threats to its

habitat. C. g. galeritus experiences many of the classic

factors contributing to endangerment: it has a small

population size, is highly restricted in range, occupies a

naturally patchy habitat that is becoming increasingly

fragmented by human exploitation, and has experienced a

population decline despite the creation of a reserve for its

protection (Eisenberg 1980, Terborgh and Winter 1980).

The Tana mangabey, is thought to be a relict population

of a once widely distributed species (Homewood 1975, 1976);






























Figure 1.1. Adult male Tana River crested mangabey,
Cercocebus galeritus galeritus, seated in a
Hypheane compressa palm.













other subspecies, C. g. chrysoqaster and C. g. aqilis

inhabit the Congolese forest block from Cameroon through

Gabon, Congo (Brazzaville), and Zaire (Dandelot 1974, Lee et

al. 1988). A fourth subspecies, C. g. sanlei, occurs in the

Uzungwa Mountains of Tanzania (Homewood and Rodgers 1981,

Lee et al. 1988). All subspecies of C. galeritus are semi-

terrestrial and broadly similar in ecology and social

behavior (Homewood 1975, Quris 1975) but vary in morphology.

Dobroruka and Badalec (1966) and Groves (1978) have

classified the Tana mangabey as a separate species based on

differences in skull morphology. Irrespective of the

taxonomy used, the important point is that the Tana mangabey

is a geographically, and perhaps genetically, distinct

population and should be considered separately from its

conspecifics in the conservation arena.

The Tana mangabey occurs only within a 60 km stretch of

the lower Tana River (Andrews et al. 1975, Homewood 1976)

where ground-water seepage from the river is sufficient to

support patches of riverine forest in a region of otherwise

semi-arid bushland. Within the mangabeys' range there are an

estimated 65 forest patches, averaging 35 ha in size (range

<10-625 ha; Marsh 1978a, 1986), separated by stretches of

deciduous woodland, bush, and grass. Forest patches, or

fragments, result in part from changes in the river's course

and patterns of flooding. Changes in the river's course

during high floods leaves some forests far from the life-









5

supporting ground-water supply; others are cut and eroded by

the scouring action of the river.

In recent years, natural forest fragmentation has been

exacerbated by increasing human exploitation and land use

(Homewood 1975, 1976; Hughes 1985, 1987, Kinnaird et al.

1990, Marsh 1986). An increasing local population (e.g., the

number of families in one village, Nkano, increased from 1

to 7 between 1974 and 1986; personal observation T.T.

Struhsaker), dependent on slash-and-burn agriculture, has

resulted in increases in forest clearing and tree felling

for houses and dugout canoes (Kinnaird et al. 1990). Medley

(1990) showed a 56% reduction in forest area and

fragmentation of 5 forest areas into 15 patches between 1960

and 1975. This loss was attributed, in part, to changes in

the rivers course (Medley 1990), and to the destruction of

forest for agriculture (Decker 1989).

Forests also have been cleared for large-scale

irrigation schemes (Medley et al. 1989), and possibly have

experienced a reduction in regeneration due to alterations

in the rivers flow by upstream dams (Hughes 1985, 1990, Lee

et al. 1988, Marsh 1986). Although changes in the flooding

regime due to upstream dams have not been verified, Hughes

(1985, 1990) has speculated that a reduction in peak floods

due to damming will reduce the sediment/nutrient load and

the frequency of meander formation and cut-off, thereby

reducing the dynamic nature of the floodplain and the











occurrence of favorable conditions for seed germination and

seedling establishment. Data from Medley (1990) indicate

that upstream dams may be having an effect; the Tana River

has shortened its course in recent years as a result of less

meandering and sediments loads delivered during floods have

declined. Agricultural practices also influence forest

regeneration. Sites of active forest regeneration (i.e.

point-bars) are selected for farming because much of the

crop production by the local people is dependent on biannual

flooding and the nutrient rich soils near the river.

The Tana River National Primate Reserve (TRNPR; Figure

1.2) was gazetted in 1976 by the government of Kenya to

protect the best remaining stands of riverine forest and

populations of the Tana River crested mangabey and red

colobus monkey (Colobus badius rufomitratus). At the time of

establishment, the reserve occupied approximately 171 km2 of

forest, dry woodland and savanna habitat on the east and

west banks of the river (Marsh 1976). The forest habitat

covered an estimated 17.5 km2, broken into approximately 40

patches. In the 15 years since the establishment of the

reserve, there has been a decline in forest area and

increased fragmentation of the remaining forest stands

(Kinnaird et al. 1990, Medley 1990). Primate populations

also have declined by an estimated 83% for the red

colobus (Marsh 1986, Decker 1989) and 25-45% for the

mangabey (Chapter 6, Marsh 1986). The mangabey population






























Figure 1.2. Map of Kenya and location of the Tana River
National Primate Reserve (TRNPR; map modified
from Hughes 1985).

















f ^ Lake '
LI Turkana




IUganda


r-

)Lake
aaa Victoria




ai`


Tanzani


T


200


Kenya


I Somalia


Mt. 00
Kenya I

Tana River



Ns TRNPR .-

a^ ^ .t A . A .I *f J
a. Garsen
aaa'aal f ndi a flaK"
.i~ .. .. ... .. ...
^ ,/2*l*>X'^--Ocea'.l;" "
A A M A MA A A

AA AA A.AA.. ^ ...: : .: .: .2
km .. A.. ..-....
}!:iii! A . A A . A . A .ii~


or
<


/
/


Sudan


Ethiopia










declined through a loss in the total number of groups, but

the remaining groups showed no significant changes in size

(Decker and Kinnaird 1990).

Kinnaird et al. (1990) speculated that the extensive

loss of forest in the late 1960s led to a compression of the

primate populations and unusually high population densities.

In the subsequent 15-20 years these populations were

believed to have declined and stabilized at levels in

equilibrium with the new carrying capacity of the reduced

forest area. Loss of forest area likely was exacerbated by

reduction in habitat quality due to forest senescence

(Hughes 1985, Marsh 1986) and possibly, by severe drought

conditions in the early 1980s (Decker and Kinnaird 1990,

Marsh 1986).

The present study examines the effects of habitat change

on mangabey behavior and demography. The underlying theme of

the dissertation is comparative, using baseline data on

behavior and demography collected during a 1973-74 study by

Homewood (1976), the only other study to date of the Tana

mangabey. Comparison of my data with that of Homewood's

(1976) allowed me to address questions concerning the loss

of mangabey groups and effects of habitat change (e.g. loss

of forest area and reductions in important mangabey diet

species) on mangabey activity budgets, diet, ranging

behavior, and risk of extinction. Detailed analysis of

ecological and behavioral data collected during 1988-89 also










provided measures of variance in mangabey behaviors due to

seasonal habitat change alone.

I begin in Chapter 2 by describing the abundance and

distribution of important mangabey fruit resources and

examining associations between the temporal availability of

these resources and various environmental parameters. I also

examine changes over the 15-year period in the temporal

pattern of fruit availability. In Chapter 3, I investigate

the effect of changes in forest structure and the food

resource base on activity budgets. I also examine the

temporal distribution (diurnal and seasonal) of activity

patterns and the influence of diet on time budgets. In

Chapter 4, I present an example of the impact of human

exploitation of forest products on mangabeys. Specifically,

I present data on competing uses of an important mangabey

diet species, Phoenix reclinata, by mangabeys and humans. I

quantify mangabey ranging patterns and defense of space in

Chapter 5 and incorporate data from Chapters 2 and 3 to test

the influence of seasonal changes in food availability and

distribution on variation in defense of space by mangabeys.

In Chapter 6, I present 2 approaches, demographic and

genetic, for estimating population sizes necessary for the

long-term persistence of the Tana mangabey. In this chapter,

I evaluate the mangabey's risk of extinction by comparing

derived population sizes with present population estimates.

I summarize in Chapter 7 the major results of the body of









i1
the dissertation and broadly discuss management

recommendations.
















CHAPTER TWO

ABUNDANCE AND DISTRIBUTION OF FRUIT RESOURCES
IN THREE TANA RIVERINE FORESTS



Introduction

Temporal patterns of resource abundance and scarcity,

coupled with the spatial distribution of fruit trees, are

considered crucial to molding the ranging and foraging

behaviors of frugivorous birds and primates (Chivers and

Raemaekers 1980, Leighton and Leighton 1983, Terborgh 1983,

Levey 1988). In most tropical trees, flowering and fruiting

is episodic and seasonal peaks in fruit abundance have been

recorded for many forests of the Neotropics (Foster 1974,

Terborgh 1983, Levy 1988) and Old World tropics (Koelmeyer

1960, Medway 1972, Struhsaker 1975, Leighton and Leighton

1983, Gautier-Hion et al. 1985, van Schaik 1986). The timing

of flowering and fruiting in tropical trees has been

ascribed to climatic, edaphic and biotic factors (Rathche

and Lacey 1985). In most tropical forests, variation in

rainfall appears to be the most significant climatic factor

influencing the phenologies of flowering and fruiting

(Foster 1974, Hilty 1980, Borchert 1983), although the

relationship between rainfall and plant reproduction is










variable in less seasonal forests (Hilty 1980, Heideman

1989). The phenological patterns of seasonal, ground-water

dependent forests are not well documented and may be

influenced more strongly by variation in river water level

than other climatic factors.

The Tana River forests occur in a semi-arid environment

with low annual rainfall (about 400 mm/year) and are

dependent on ground-water from the river for their existence

(Homewood 1976, Marsh 1976, Hughes, 1985). The occurrence of

ground-water forest communities along the Tana appears to be

related to the forest height above flood levels and duration

of flooding, soil texture and therefore ability to hold

water, soil chemistry, age of substrate and the history of

human activities within the forest (Hughes 1985).

Little information exists on the influence of forest

fragmentation, isolation or habitat change on tree

phenologies. Increased leaf fall has been noted in Brazilian

forests after fragmentation (Lovejoy et al. 1986) but even

qualitative changes in plant reproduction have not been

investigated. Varying degrees of habitat degradation,

fragmentation and isolation within the Tana River forests

over the last 15 years provide a basis for examining the

effects of habitat change on plant reproduction.

In this chapter I describe the phenological and spatial

patterns of fruit species eaten by Tana mangabeys in 3

forests. I derive 2 indices of monthly fruit production in










each forest and examine the associations between flowering

and fruiting phenology, river flow, temperature and

rainfall. I also make comparisons with similar data

collected during 1973-74 (Homewood 1976) to determine if

fruiting patterns differ between the 2 time periods.

Methods

Study Sites

Three forests were chosen for intensive study:

Mchelelo, Nkano and Mnazini (Figure 2.1). Baseline data were

available for these forests from Homewood's (1976) 1973-74

study. The forests vary in size, vegetation, degree of

flooding, human disturbance, and mangabey density. Mchelelo

is a small forest (17 ha) with a diverse plant species

composition. It is on a raised river levee and experiences

minimal flooding. Mchelelo presently is occupied by 1-2

groups of mangabeys, depending on the season and is not

disturbed by local people. Plant species composition has

changed since the mid-1970s (Marsh 1986, Medley 1990) and

several important mangabey food species have declined in

abundance.

Nkano is a larger forest (38 ha) with a much less

diverse plant species composition than Mchelelo (Homewood

1976, Medley 1990). Nkano presently is occupied by 2

mangabey groups and experiences a high degree of flooding

and human exploitation. Although forest area in Nkano has

decreased by an estimated 24% due to a fire in 1981 (Marsh





























Figure 2.1. Map of the study area showing the Mchelelo,
Mnazini, and Nkano forests within the Tana
River National Primate Reserve.

































-




'6
1
I
I
I


I,










1986, Decker 1989), plant species composition has not

changed significantly since the mid-1970s (Medley 1990).

The 35 ha Mnazini forest is less than 0.5 km south of

the Nkano forest, being separated from it by an old river

course and an area of grassland. Mnazini is a low diversity

forest (Homewood 1976, Medley 1990) that experiences high

levels of flooding and human exploitation. The forest has

decreased in size since the mid-1970s by approximately 50%

due to floods and fires in 1982 and 1983, respectively. The

number of mangabey groups in Mnazini has declined from 2 to

1 since the mid-1970s.

The climate is strongly seasonal. Rainfall averages 400

mm per year, most of which falls during the long rains

(March-May) and the short rains (November-January), with

intervening dry periods generally occurring during February

(inter-rains) and June-October (dry season) (Hughes 1984).

Floods are highly variable in intensity and timing but in

general, water levels peak during the long and short rains

(Hughes 1984).

Abundance and Distribution of Resources

Density. I enumerated selected tree species in 2 of the

3 study sites where mangabey groups were intensively studied

(Mchelelo and Nkano) to calculate the abundance and describe

the dispersion patterns of mangabey foods. Species were

selected based on their importance in the diet during 1973-

74 (Homewood 1976) or the present study. Only those trees of










reproductive size for each species were enumerated. Foot

trails cut along compass bearings at 50m intervals were used

to delineate quadrats of 0.25 ha. I enumerated 23 plant

species within 65 quadrats (16.25 ha) for Mchelelo and 14

species within 54 quadrats (13.5 ha) for Nkano. Nine species

present in Mchelelo were not found in Nkano. Average

density was calculated by summing the number of individuals

for a particular species in all quadrats and dividing by the

total number of hectares sampled.

Spatial Distribution. I used Morisita's (1959) index of

spatial pattern (I) to measure the distribution of each

species in the two forests. I chose Morisita's index over

the more commonly used variance to mean ratio because many

of the species I considered had calculated means less than

1.0 individual/ha. Greig-Smith (1983) has shown the variance

to mean ratio to behave erratically when the mean is less

than 1.0. Morisita's index provides a measure of the extent

to which individuals of a particular species are nonrandomly

distributed among identically-sized quadrats. The index is

calculated as

I = [ni(ni l)/n(n-l)] N
where ni is the number of individuals in the ith quadrat, n

is the total number of individuals in all quadrats, and N is

the number of quadrats. Morisita's I ranges from 1/N to N;

values approaching 1.0 indicate a random distribution and

those less than or greater than 1.0 indicate regular and










patchy distributions, respectively. Deviation of the index

from unity is tested with the F statistic.

Although Morisita's Index is relatively independent of

quadrat size, Poole (1974) suggests that quadrats be large

enough to contain at least one individual of one of the more

common species. A quadrat size of 0.25 ha meets this

requirement: the first and third most abundant species in

Mchelelo forest, Phoenix reclinata and Hyphaene compressa,

occur at least once in over 90% and 70% of the quadrats,

respectively; at least one individual of one of the 3 most

abundant species occurs in over 90% of the quadrats. The 3

most common species in Nkano forest, Alangium salviifolium,

Phoenix reclinata and Pachystela msolo, all occur in over

90% of the quadrats and all quadrats contain at least 1

individual of 1 of these species.

Temporal Distribution. Temporal fluctuations in the

availability of flowers and fruits were tabulated for 240

individuals in 16 species within the 3 forest areas: 126

individuals in 15 species for Mchelelo forest, 49

individuals in 9 species for Nkano forest, and 65

individuals in 8 species for Mnazini South forest (Table

2.1). I chose the species based on their presence in each

study site and their importance as mangabey foods (Homewood

1976). The more common species were represented by at least

10 individuals; species with fewer than 10 individuals in

the study areas were represented by the total number of











Table 2.1. Composition of monthly phenology sample by
species for Mchelelo, Nkano and Mnazini forests.


Number of Individuals

Taxon Mchelelo Nkano Mnazini


ALANGIACEAE
Alangium salvifolium 10 5 -

ANACARDIACEAE
Sorindeia madagascariensis 10 8

APOCYNACEAE
Saba comorensis 7 3 8

EBENACEAE
Diospyros mespiliformis 13 5 9

FLACOURTIACEAE
Oncoba spinosa 10

GUTTIFERAE
Garcinia livingstonei 9 10

MIMOSACEAE
Acacia robusta 10 10
Albizia gummifera 10 -

MORACEAE
Ficus bubu 1 3 0
Ficus bussei 1 -
Ficus natalensis 6 7
Ficus sycomorus 10 8 7

PALMAE
Hvyphaene compressa 11 -
Phoenix reclinata 10 7 3

SAPOTACEAE
Mimusops fruticosa 8 10
Pachystela msolo 10








21
individuals present. I chose only those individuals believed

to be of a species' reproductive size and individuals whose

canopies were relatively easy to view.

The phenological state of each individual was assessed

monthly for Mchelelo (n=19 months), Nkano (n=17 months) and

Mnazini (n=13 months) beginning October 1987. I used a 0-5

scale developed and used by Homewood (1976) in 1973-74 to

visually assess the presence of flowers and ripe and unripe

fruits on each individual. The scale is expressed as a

proportion of the total canopy covered by flowers or fruits,

where 0=no flowers or fruits, 1=1-20% coverage, 2=21-40%

coverage, 3=41-60% coverage, 4=61-80% coverage and 5=81-100%

coverage. Flowers and fruits were examined separately, each

receiving a potential maximum score of 5.

Two indices of flower and fruit availability were

calculated for each study forest using the 0-5 fruit scale.

The first index is comparable to that used by Homewood

(1976) and provides a relative measure of variation in

flowering and fruiting. This index is calculated by pooling

monthly flower or fruit scores for individuals of each

species and averaging to give a mean species score for each

month. An index of total monthly flowering or fruiting is

calculated by summing monthly means for all species and

dividing by the number of species in the sample.

The second index, used only for fruits, considers

species differences in abundance and productivity and










provides an indication of monthly fruit biomass production.

The index incorporates estimated numbers of fruits produced

by a species during maximum fruit production, the estimated

weight of fruits or fruit parts eaten by mangabeys, and the

abundance of reproductive individuals. Fruit biomass indices

were calculated for 4 categories of species, depending on

how fruits were distributed throughout the canopy. Species

bearing fruits primarily on the outer surface of the canopy

(n=ll) were categorized as surface fruiters. Species bearing

fruits along their branches, either in clusters or evenly,

were classified as ramiflorous fruiters (n=4). Palms (n=2)

and lianas (n=l) each were classified separately because

they do not have identifiable canopies.

I counted fruits only when individuals received the

maximum phenological score of 5; this limited my sample size

for some species because certain individuals were never

scored as a 5. I estimated the number of fruits for all

categories except the palms by counting the number of fruits

visible within an approximately 1 m3 field of vision of

binoculars. One count in the top, middle and lower sections

of the canopy or liana "tangle" was made for each individual

and an average was taken. Fruits of the 2 palm species were

counted directly for a minimum of 3 infructescenses. The

numbers of fruits for the 3 infructescenses were averaged

and multiplied by the total number of infrutescenses

present. For Hyphaene compressa, a palm that branches and










has multiple crowns, the above value was multiplied by the

number of crowns per individual.

Fruit counts for all species, excluding palms, were

multiplied by an estimate of relative food-producing area to

calculate the total number of fruits potentially produced by

an average individual of a particular species. Fruit

producing area was calculated using surface area and volume

formulas for an open-bottomed cylinder, a shape that best

approximates a variety of differing tree shapes (Appendix

A). Surface areas were calculated for the category of

species bearing their fruits primarily on the surface of the

canopy and volumes were calculated for ramiflorous species.

Branches of ramiflorous species were subjectively estimated

to occupy little over 1/4 of the entire volume of the

canopy; volume estimates, therefore, were divided by 4. The

food-producing area of the liana species, which grows over

the top of tree canopies, was calculated by estimating the

amount of surface area of the tree canopy covered by the

liana. I did not correct for canopy size in palms because a

more direct count of fruit numbers was possible.

Samples of 10 or more fruits were collected and weighed

from a maximum of 10 individuals per species. Only those

fruit parts consumed by mangabeys were weighed and parts

were weighed only during the stage of ripeness when

mangabeys were feeding on them. Weights were averaged for

individuals and individual averages were pooled to calculate










a species mean. The density of potentially reproductive

individuals in each species for Mchelelo and Nkano study

sites were taken from the species enumerations described

above. Species densities for Mnazini were taken from Medley

(1990).

A final monthly biomass (kg/ha) index of fruits of

species sampled was calculated as:

FB = [{(Xsi*Dsi*SAsi*FPsi)/Isi}] +

[5{(Xri*Dri*Vri*FPri)/Iri}] +

[ {(Xpi*Dpi*FP p i)/Ipi] +

[4 {(Xli*Dli*SA li *FPi)/I1i)] where
X = monthly fruit score for species i / 5 (to scale the

scores to maximum production);

I = number of individuals scored for species i;

D = density of species i per hectare;

SA = surface area of species i;

V = branch volume of species i;

FP = total fruit production measured as mean fruit count *

mean fruit weight of species i;

and subscripts s, r, p and 1 denote surface, ramiflorous,

palm and liana categories, respectively.

Environmental Correlates with Flowering and Fruiting

I examined associations between 3 environmental

parameters and the temporal patterns of flowering and

fruiting in the 3 study sites. A Spearman's rank correlation

analysis (Siegel 1956) was conducted to examine possible










associations between rainfall, riverflow and temperature,

and flower and fruit scores for trees and lianas for each

forest and all forests combined. I conducted separate

analyses for flowers and fruits for all species combined.

For fruits, I examined seasonal and continuous fruiting

species separately, and ripe versus, unripe fruits. Because

flowering occurred as much as 3 months prior to fruiting,

monthly flowering, rainfall and river flow data were lagged

up to 3 months in each analysis. Associations between

flowering and fruiting and environmental conditions during

the months after flowering or fruiting were examined by

advancing rain and river data by one month. I also conducted

a multiple regression analysis to test whether the

environmental parameters chosen explain a significant amount

of variability in fruiting. A square root transformation was

performed on monthly scores because they were discrete data

(Sokal and Rohlf 1981). Monthly rainfall levels were

recorded for Mchelelo and Nkano/Mnazini areas, and Tana

River discharge levels for the Hola Station were obtained

from the Kenya Water Department, Hydrology Section.

Comparisons with 1973-74

I compared monthly fruit scores for 9 species examined

during this study and by Homewood in 1973-74 to determine if

fruiting patterns varied between the 2 study periods. Seven

species from Mnazini and 6 species from Mchelelo forest were

compared. I subsampled from my 1987-89 data to correspond










with Homewood's 12 month data from Mnazini forest and 7

month data from Mchelelo forest. Kolmogrov-Smirnov two-

sample tests (Hollander and Wolfe 1973) were performed by

species to test for differences in the distributions of

fruiting between the 2 study periods. Wilcoxon signed-ranks

tests also were performed by species to test whether fruit

scores were consistently higher during 1 study period than

the other.

Results
Abundance and Spatial Distribution of Resources

Density values ranged widely for enumerated species

both within and between forests (Table 2.2). Over 50% of the

species enumerated in Mchelelo (n=13 of 25) and Nkano (n=8

of 15) were rare, occurring at densities of less than 1

individual per ha. The palm, Phoenix reclinata, was one of

the most common species, occurring at the highest and second

highest densities (193.6 and 101.6 individuals/ha) of all

species in Mchelelo and Nkano, respectively. Alanqium

salviifolium, an understory species, also occurred at very

high densities (33.6-112.3 individuals/ha) in both forests.

Five species present in both forests (Pachystela solo,

Ficus sycomorus, Alanqium salviifolium, Aporrhiza paniculata

and Saba comorensis) occurred at much higher densities in

Nkano than Mchelelo; 4 additional species (Sorindeia













Table 2.2. Density of reproductive trees and lianas
enumerated in 16.32 ha of Mchelelo forest and 13.45 ha in
Nkano forest, and the degree of clumping of each species
using Morisita's index (I) If I is significantly
greater or less than unity using the F-test, the species is
considered clumped or evenly distributed in space,
respectively. If I. is not significantly different from
unity than the species is considered randomly
distributed.



Mchelelo Nkano


Taxon Code Density Ig F-test (p) Density Ig F-Test (p)


ALANGIACEAE
Atangiwi salvifotium AS
ANACARDIACEAE
Lannea stuhtmannii LS
Sorindeia madagascariensis SO
APOCYNACEAE
Rauvolfia mnombasiana RH
Saba comorensis SF
BORAGINACEAE
Cordia goetzei CG
CAESALPINACEAE
Cynometra webberi CW
Tamarindus indica TI
EBENACEAE
Diospyros mespitiformis DM
FLACOURTIACEAE
Oncoba spinosa OS
GUTTIFERAE
Garcinia tivingstonei GL
MIMOSACEAE
Acacia robusta AR
Atbizia guivnifera AG
ALbizia gtaberrima AL
MORACEAE
Ficus bubu FU
Ficus bussei FB
Ficus natatensis FN
Ficuss sycomorus FS
PALMAE
Hyphaene compressa HC
Phoenix rectinata PR
SAPINDACEAE
Aporrhiza panicutata AP
BIighia unijugata BU
SAPOTACEAE
Mimusops fruticosa MF
Pachysteta msoLo PB
STERCULIACEAE
Stercutia appendicutata SA


33.56 2.71 <0.001

0.55 1.81 ns
6.74 2.16 <0.001

52.01 3.68 <0.001
0.86 3.57 <0.005

4.59 1.52 <0.001

0.43 9.29 <0.001
0.49 4.64 <0.001

7.72 1.51 <0.001

6.49 2.67 <0.001

1.72 8.60 <0.001

0.67 3.55 <0.001
1.65 2.78 <0.001
0.55 3.61 ns


0.06
0.06
0.55
1.16


0.00
1.14


17.45 2.48 <0.001
193.57 1.39 <0.001

1.64 4.94 <0.001
0.86 17.86 <0.001

0.86 1.43 ns
0.31 0.00 ns

0.89 5.91 <0.001


112.26 1.27 <0.001

0.30 27.00 <0.001
0.15 54.00 <0.001


6.77 1.31


<0.005


0.00

0.00
0.00

1.19 1.35 ns

0.45 14.40 <0.001

0.37 5.40 ns

0.00
0.00
0.00


0.89
0.00
0.45
4.68

0.00
101.64


2.45

0.00
3.98


ns

ns
<0.001


1.90 <0.001


8.70 1.90 <0.001
0.15 0.00 ns


0.00
35.99

0.07


1.35 <0.001










madagascariensis, Diospyros mespiliformis, Oncoba spinosa

and Garcinia livingstonei) occurred at much higher densities

in Mchelelo than Nkano (Table 2.2).

The majority of species in both forests were patchily

or randomly distributed in space and no species showed a

uniform spatial pattern (Table 2.1, Figures 2.2-2.4). The

tendency towards patchy distributions of species was

consistent on a larger spatial scale: quadrats with high

numbers of individuals of a species tended to be clumped

throughout the forests (Figures 2.2 and 2.3). Five species

showed different patterns of spatial distribution between

the 2 forests. Diospyros mespiliformis, Garcinia

livingstonei, and Blighia unijugata occurred in clumps in

Mchelelo but were randomly distributed in Nkano. Pachystela

msolo and Ficus sycomorus, on the other hand, occurred in

clumps in Nkano but were randomly distributed in Mchelelo.

These apparently random distributions, however, may be an

artifact of small sample size.

Temporal Distribution of Resources

Patterns of flowering and fruiting. Flower and fruit

scores averaged over all species by month suggest

seasonality of flowering and fruiting in the 3 forests

(Figures 2.5 and 2.6). Increased flowering occurred from

August to December with minor increases in February and

March. Fruiting was high from January to May of 1988, and

November to March of 1988-89; low levels of fruiting






























Figure 2.2.


Number of individuals/0.25 ha quadrat of 4
common species within Mchelelo forest showing
clumped spatial distributions.
a) Hyphaene compressa; b) Alangium
salviifolium; c) Phoenix reclinata; d) Oncoba
spinosa.







Alangium salviifolium


Phoenix reclinata


Oncoba spinosa


I w 0 m m
1 2-5 6-10 11-20 20+


Hypheene compressa





























Figure 2.3.


Number of individuals/0.25 ha quadrat of 4
common species within Nkano forest showing
clumped spatial distributions.
a) Ficus sycomorus; b) Alangium salviifolium;
c) Phoenix reclinata; d) Pachystela msolo.








Alangium salviifol


Phoenix reclinata Pac


hystela msolo


ED
1


El
2-5


m
11-20


20+
20+


Ficus sycomorus


E3
6-10






























Figure 2.4.


Examples of randomly distributed Ficus spp.
a) Ficus natalensis in Mchelelo forest; b)
Ficus sycomorus in Mchelelo forest; c) Ficus
natalensis in Nkano forest.







Ficus natalensis


Ficus natalensis


1
1


E m
6-10 11-20


2-5E
2-5


20+
20+


/00




Hf


Ficus sycomorus































Figure 2.5. Mean flower scores averaged over all species by
month.
a) Mchelelo forest; b) Nkano forest; c) Mnazini
forest.















Flower score
6-

4-

3-

2-


1-im.


._lm~i=._. _


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9


Nkano


Flower score
5-

4-

3-

2-


N D J F M A M J J A S O N D J F M AM
8 8 8
7 8 9


Mnazini


Flower score
6-

4-

3-

2-


* m_-


-.IiiN


Mchelelo


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9
Month






























Figure 2.6. Mean fruiting scores averaged over all species
by month.
a) Mchelelo forest; b) Nkano forest; c) Mnazini
forest.


















Mchelelo

Fruit score
3-












0
N D J F M A M J J A S O N D J F M A M
8 8 6
7 8 9




Nkano
Fruit score
3-



2








0-
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9



Mnazlnl

Fruit score
3-



2-








N D J F M A M J J A S O N D J F M A M

8 8M8
7 8 9
Month








39

occurred from June to October in Mchelelo and Nkano and from

July through November in Mnazini.

Several species showed seasonal patterns of fruiting

(Figures 2.7-2.10). The majority of seasonal fruiters bore

fruits only once in a year; however, individuals of 2

species (Sorindeia madagascariensis and Garcinia

livingstonei) produced fruits twice within a year (Figure

2.8). Of the seasonal fruiters, only Phoenix reclinata and

Diospyros mespiliformes produced fruit during the cool, dry

months between June and September (Figures 2.9 and 2.10).

Three species appeared to fruit continuously (Figures 2.11

and 2.12) due to 1) continuous production of fruit by all

individuals (Oncoba spinosa and Ficus sycomorus); 2)

retention of fruits over long periods of time (Hyphaene

compressa); and 3) unsynchronized bouts of fruiting (Ficus

natalensis).

Species sampled in more than 1 of the forests tended to

show similar patterns of fruiting among forests. One

exception was the liana, Saba comorensis. Saba comorensis

exhibited highly seasonal fruiting in Mnazini, seasonal but

extended fruiting in Nkano, and continuous fruiting in

Mchelelo (Figure 2.13).

Estimates of fruit biomass. Estimated maximum fruit

numbers, fruit weight and canopy size varied widely between

species and shape categories (Table 2.3), and strongly

influenced monthly biomass indices of fruit (Figure 2.14).



















Figure 2.7. Mean monthly fruit scores from Mchelelo and Mnazini forests for Mimusops
fruticosa and Acacia robusta, seasonally fruiting species with 1 peak in
fruiting per year.
a) Mimusops fruiticosa in Mchelelo forest; b) Acacia robusta in Mchelelo
forest; c) Mimusops fruticosa in Mnazini forest; d) Acacia robusta in Mnazini
forest.









Mcheleo


m MImusovs







.Il .mo


Fruit score
6-
5

4 -

3

2-


II, -


I Acacia


Fruit score
5

4 -

3-

2

1
0n11


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9






Mnazlnl


Fruit score
6-
4

3-

2

1I


I MImusoos


II


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9





Mnazlnl
Fruit score
51 Acacia


rnM =


N D J F M A M d J A S O N D J F M A M
8 8 8
7 8 9
Month


.1lll.


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9
Month


IM-


v


Mchelelo


wJ



















Figure 2.8.


Mean monthly fruit scores from Mchelelo and Mnazini forests for Garcinia
livingstonei and Sorindeia madagascariensis, seasonally fruiting species
with 2 fruiting peaks per year.
a) Garcinia livingstonei in Mchelelo forest; b) Sorindeia madagascariensis
in Mchelelo forest; c) Garcinia livingstonei in Mnazini forest; d) Sorindeia
madagascariensis in Mnazini forest.











Mchelelo


Fruit score


Fruit score


= Garclnla


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9







Mnazlnl


Fruit score
6-

4-

3-

2-

1 -


- Garclnla


if I


= Sorlndele


__ n


mR a


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9







Mnazlnl
Fruit score


- Sorlndele


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9
Month
L.


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9
Month


Mchelelo


n -






























Figure 2.9.


Mean monthly fruit scores for Phoenix
reclinata.
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.


















Mchelelo

Fruit score
56-

4


3-

2




0-
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 0




Nkano

Fruit score
5"

4


3

2-

1


0-
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8



Mnazini

Fruit score
6-
5


4

3-

2




0
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 g
Month






























Figure 2.10. Mean monthly fruit scores for Diospyros
mespiliformes.
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.

















Mchelelo

Fruit score
6

4-

3-

2-



0
1 iI I

N D J F M A M J A S O N D J F M A M
8 8 8
7 8 9




Nkano

Fruit score
6-

4-

3-

2-

1

0
N D J F MAM J J A S O N D J F MA M
8 8 8
7 8 9



Mnazini
Fruit score
6-

4-

3-

2-



0-
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9
Month





























Figure 2.11.


Mean monthly fruit scores for 3 species
showing patterns of continuous fruit
production.
a) continuous fruiting due to asynchrony among
individuals of the species (Ficus natalensis);
b) continuous fruiting due to retention of
fruits over time (Hyphaene compressa); c)
continuous fruiting due to rapid turnover of
fruits (Oncoba spinosa).












49





Ficus natalensis

Fruit score
6

4

3


2

1



N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9




Hyphaene compressa

Fruit score
6-

4-


3

2-





N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9



Oncoba spinosa

Fruit score
6-

4

3

2




0-
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9
Month





























Figure 2.12. Mean monthly fruit scores for Ficus sycomorus.
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.















Mchelelo
Fruit score
6-

4-

3

2-

1

04..ll6h
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9




Nkano
Fruit score
5-

4-

3-

2-

1

0
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9



Mnazini
Fruit score
6-

4-

3-

2-

1 H IN


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9
Month





























Figure 2.13. Seasonality of mean monthly fruit scores for
Saba comorensis in the 3 study forests.
a) seasonal fruiting in Mnazini forest; b)
seasonal but extended fruiting in Nkano forest
c) continuous fruiting in Mchelelo forest.















Mchelelo


Fruit score


N D J F M A M J J A S O N D
8 8
7 8


J F M A M
8
9


Nkano


Fruit score


N D J F M A M J J A S O N D J F
8 8 8
7 8 9


M A M


Mnazini


Fruit score
6-


III..


m m


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9
Month


'I..Ilmlllmmmm.









Table 2.3. Estimated canopy surface area (S), canopy volume (V), mean fruit production
(FN) and mean fruit weight (FW) for 16 species. Species are categorized into 4 groups:
those that bear their fruits on the canopy surface (C) or along the branches of the tree
(R), lianas (L) and palms (P).




Taxon Group S(m3 ) V(m3) FN(m3 ) FW(gr)


Alanqium salvifolium C 125.9 48 0.86
Sorindeia madagascariensis R 313.0 142 0.65
Saba comorensis L 232.0 6 119.15
Diospyros mespiliformis C 1016.3 47 1.27
Oncoba spinosa C 74.3 12 29.07
Garcinia livingstonei C 567.0 47 4.87
Acacia robusta C 931.2 48 0.96
Albizia gummifera C 731.6 26 0.70
Ficus bubu R 412.0 122 7.75
Ficus bussei C 1965.0 55 7.80
Ficus natalensis C 1004.8 176 0.80
Ficus sycomorus R 1345.0 90 7.86
Hyphaene compressa p 980 6.88
Phoenix reclinata P 11502 0.81
Mimusops fruticosa C 695.8 67 6.18
Pachystela msolo R 627.3 121 0.94





























Figure 2.14. Contribution to monthly fruit biomass indices
by different species categories.
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.


















Mchelelo

Blomass Index of fruits (Kg)/ha


.... 1 1 l ..i ~ i 1
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9


Nkano


Biomass Index of fruits (kg)/he
6000-
4600-
4000 -


3600-*J^^^^ l^^
2500-

1600
1000
600"
0
N D J F M A M J J A S O N D J F
8 8 8
7 8 8


M A M


Mnazini


Index of fruits (kg)/ha


lana
- palm
sBurface
ramlflorous


6000
4600
4000
3600
3000
2600
2000
1600
1000
600


U


N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9
Month








57

Palms, particularly Phoenix reclinata, often accounted for a

large proportion of the monthly biomass indices in Mchelelo

because of their potential for large fruit crops and their

high density in the forest. Although ramiflorous species

were not abundant in Mchelelo, their capacity for large

fruit crops results in a relatively large contribution to

the monthly biomass indices. In Nkano, ramiflorous species

such as Pachystela msolo and Ficus sycomorus were abundant

and they dominated the monthly biomass index.

Although fewer species were sampled in Nkano, the

forest produced a much greater biomass of fruits than either

Mchelelo or Mnazini for all months of the year with the

exception of August. The high fruit biomass production by

Nkano was primarily the result of high densities of Ficus

sycomorus, a species that produces large fruit crops

throughout the year, with a decline in production during

June and July (Figure 2.12). Mchelelo had a more consistent

production of fruits across the months due to the

availability of palm fruits between May and October, a

period of relative scarcity in the other 2 forests. Mnazini,

with a low abundance of figs and palms, was more seasonal in

fruit production than Mchelelo or Nkano and experienced a 4

month period (May-Aug) of fruit scarcity.

Unripe fruits contributed the most to the biomass index

over all months in all study forests; ripe fruits were

entirely unavailable during several months (Figure 2.15).






























Figure 2.15. Contribution of ripe and unripe fruits to
monthly fruit biomass index.
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.
















Mchelelo


Blomass Indclex of fruits (kg)/ha
6000-
4600-
4000-
3600-
3000-
2500-
2000-
1500-

500 -
0
N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9


Nkano


6000-
4500-
4000-











8 8 8
7 8 9


Mnazlni

6000-
4600 -RpFru









4000 ESUnrip* Fruit
253600 -









3000-
2000-










1600
1000










600-

N D J F M A M J J A S O N D J F M A M










8 8 8
7 8 9
MnaziMonthni
4500- = Pipe Fruit
4000- M Unripe Fruit
3500 -
3000 -
2600 -
20O00

1500"
1000'

0
N D J F M A M J J A S 0 N D J F M AM
8 8 8
7 8
Month










The predominance of unripe fruits may be the result of the

time necessary to ripen a fruit crop and/or asynchronous

ripening within individuals, and fruits being selectively

chosen by animals as soon as they ripen. Nkano and Mnazini

generally had larger biomasses of ripe fruits relative to

Mchelelo, possibly due to the higher density of figs in

Nkano and species such as Mimusops fruticosa and Garcinia

livingstonei in Mnazini, that tend to ripen more

synchronously.

Environmental Correlates with Flowering and Fruiting

Temperature, rainfall and river flow showed strong

seasonality. The hottest months occurred between November

and April and the coolest months occurred between June and

October (Figure 2.16). Rainfall was high during November and

December and again from March-June (Figure 2.17). Rainfall

in Nkano generally was higher than Mchelelo although there

was a strong correlation between monthly totals for the 2

areas (r=0.85, n=ll, p<0.05). River flow was highest April-

June, 1988 and January-April, 1989 (Figure 2.18), and was

correlated with the previous months rainfall in both areas

(r=0.72 and 0.76 for Nkano and Mchelelo, respectively, n=ll,

p<0.05).

Flowering was negatively correlated with rainfall in

Nkano (r=-0.57, n=14, p<0.05) and with river flow in

Mchelelo (r=-0.51, n=16, p<0.05) and Mnazini (r=-0.77, n=10,

p<0.05). When all 3 forests are combined, there were































Figure 2.16. Monthly mean minimum and mean maximum
air temperature in degrees centigrade.
a) Mchelelo forest; b) Nkano/Mnazini forest
areas.








62





Mchelelo

Temperature (C)


30-
10 "^^\""-- ^^

20 \


10-



S 0 N D J F M A M J J A S 0 N D J F M A
8 8 8
7 8 9






Nkano
Temperature (C)
40-

30 -- '

20) -

10 Mean maximum
SMean minimum

S O N D J F M A M J J A S O N D J F M A
8 8 8
7 8 9
Month
































Figure 2.17. Total monthly rainfall in millimeters.
a) Mchelelo forest; b) Nkano/Mnazini forest
area.

















Mchelelo


300

250

200

150

100

50

0


Rainfall (mm)


A S O N D J F M A M J J A S O N D J F M A M
8 8 8
7 8 9


Rainfall (mm)
300

250

200

150-

100-o!.1




O N D J F M A M J J A S O N D
8 8
7 8
Month


Nkano


J F M
8
9






























Figure 2.18. Mean monthly river flow for the Tana River
Hola Station.















Million Cubic Meters
250

200

150

100

50-
l lll lld __l_ !


A S O N D J F M A M J J A S O N D J F M A
8 8 8
7 8 9
Month








67
significant negative correlations between flowering and both

rainfall and river flow (r=0.-41 and -0.35 for rainfall and

river flow, respectively, n=41, p<0.05). Correlations

between flowering and fruiting were significant when flower

scores were lagged by 3 months in Mchelelo, Nkano and for

all 3 forests combined (r=0.76, p<0.05 for Mchelelo, r=0.55,

p<0.05 for Nkano and r=.45, n=41, p<0.05 for all forests

combined).

There were no consistent associations between monthly

fruit scores and either rainfall or river flow for the 3

forests (Tables 2.4, 2.5 and 2.6). Minimum monthly

temperature was positively correlated with monthly fruit

scores for all species combined in the 3 separate forests.

There were also positive correlations between minimum

monthly temperature and scores from seasonal fruiting

species in Nkano and Mnazini. When the palm, P. reclinata,

was removed from the analysis, there were significant

negative correlations between fruiting of seasonal species

and river flow lagged by 3 months. Phoenix reclinata showed

significant positive correlations with river flow lagged

only by 1 or 2 months.

When data for all forests were combined, many of the

results were consistent with the analysis performed on the

forests separately (Table 2.7). There were significant

positive correlations between fruit scores for continuous

species and rainfall. Minimum monthly temperature showed









Table 2.4. Spearman's rank correlation coefficients for mean monthly environmental
variables and mean monthly fruiting scores for Mchelelo forest. MINTEMP=minimum monthly
temperature; RAIN=total monthly rainfall (mm); RAINLAG1, RAINLAG2 AND RAINLAG3=monthly
rainfall lagged by 1,2, and 3 months, respectively; RAINADV=previous months' rainfall;
FLOW=mean monthly river flow (mcm); FLOWLAG1, FLOWLAG2, FLOWLAG3=mean monthly river flow
lagged by 1,2 and 3 months, respectively; FLOWADV=previous months' mean river flow.


RIPE AND UNRIPE


Total Continuous Seasonal


RIPE ONLY


TotaL Continuous Seasonal


RIPE AND UNRIPE


Seasonal
(w/o fP. rectinata)


RIPE AND UNRIPE


P. reclinata


MINTEMP 0.54 ns ns ns ns ns n ns
RAINTEMP 0.54ns ns ns ns ns ns ns ns
RAINLAG ns ns ns ns ns ns ns ns
RAINLAG2 ns ns ns ns ns ns ns ns
RAINLAG2 ns ns ns ns ns ns ns ns
RAINLAG3 ns ns, ns ns ns ns ns ns
RAINADV ns .74 ns ns ns ns ns ns
FLOW ns ns ns ns ns ns ns ns
FLOWLAG1 ns ns ns ns 0.64 ns ns 0.78**
FLOWLAG2 ns ns ns ns ns ns ns 0.83**
FLOWLAG3 ns ns ns 0.48 ns ns -0.59 0.79
FLOWADV ns ns ns ns ns ns ns ns


** p-O.05
p<0.O01









Table 2.5. Spearman's rank correlation coefficients for mean monthly environmental
variables and mean monthly fruiting scores for Nkano forest. Variables defined in Table
2.4.


RIPE AND UNRIPE


Total Continuous Seasonal


RIPE ONLY


Total Continuous Seasonal


RIPE AND UNRIPE


Seasonal
(w/o P. rectinata)


RIPE AND UNRIPE


P. reclinata


MINTEMP 0.72 0.63 0.61 ns na na 0.74 0.75
RAIN ns ns ns ns ns ns ns ns
RAINLAG1 ns ns ns ns 0.60 ns ns ns
RAINLAG2 ns ns ns ns ns ns ns ns
RAINLAG3 ns ns ns ns ns ns ns ns
RAINADV ns ns ns ns ns ns ns ns
FLOW ns ns ns ns ns ns ns ns
FLOULAG1 ns ns ns ns ns ns ns 0.783*
FLOULAG2 ns ns nse ns ns ns ns 0.83**
FLOWLAG3 ns ns ns ns ns ns -0.59 0.79
FLOWADV ns ns ns ns ns ns ns ns

*
** p p<0.001









Table 2.6. Spearman's rank correlation coefficients for mean monthly environmental
variables and mean monthly fruiting scores for Mnazini forest. Variables defined in Table
2.4.


RIPE AND UNRIPE


Total Continuous Seasonal


RIPE ONLY


Total Continuous Seasonal


RIPE AND UNRIPE


Seasonal
(w/o P. rectinata)


RIPE AND UNRIPE


P. rectinata


NTEP08* ** 0*
MINTEMP 0.58 ns 0.58 0.72 0.60 0.68 0.58 0.75
RAIN ns ns ns ns ns* ns ns ns
RAINLAG1 ns ns ns ns 0.60 ns ns ns
RAINLAG2 ns ns ns ns ns ns ns ns
RAINLAG3 ns ns ns ns ns ns ns ns
RAINADV ns ns __ ns ns ns ns ns ns
FLOW ns 0.78 ns ns 0.57 ns ns ns
FLOWLAGI ns ns ns 0.59 ns ns ns 0.78*
FLOWLAG2 ns ns ns ns ns ns ns 0.83*
FLOWLAG3 ns ns -0.59 -0.63 ns ns -0.71 0.79
FLOWADV ns ns ns ns ns ns ns ns


* p<-0.05
p<0.001











Table 2.7. Spearman's rank correlation coefficients for mean monthly environmental
variables and mean monthly fruiting scores for all forests combined. Variables defined in
Table 2.4.


RIPE AND UNRIPE


Total Continuous Seasonal


RIPE ONLY


Total Continuous Seasonal


RIPE AND UNRIPE


Seasonal
(w/o P. rectinata)


RIPE AND UNRIPE


P. rectlinata


*
MINTEMP ns ns ns ns ns 0.39 0.53 ns
RAIN ns ns ns ns ns ns ns ns
RAINLAG1 ns ns ns ns 0.37 ns ns ns
RAINLAG2 ns ns ns ns ns ns ns ns
RAINLAG3 ns ns ns ns ns ns ns 0.46
RAINADV ns ns ns ns ns ns ns ns
FLOW ns ns ns ns ns ns ns ns
FLOWLAG1 ns ns ns ns ns ns ns 0.39**
FLOWLAG2 ns ns ns ns ns ns -0.41k 0.51*
FLOWLAG3 ns ns -0.58 ns ns ns -0.59 0.34
FLOWADV ns ns ns ns ns ns ns ns


** p p







72

significant positive correlations, and river flow lagged by

2 and 3 months showed significant negative correlations with

seasonal fruiting species excluding P. reclinata. Combined

P. reclinata scores for the 3 sites were significantly

correlated with rainfall lagged by 2 months and river flow

lagged by at least 1 month.

A multiple regression model incorporating minimum

temperatures and river flow lagged by 3 months, the 2

variables most highly correlated with fruit scores,

explained 25.7% of the variance in monthly fruit scores for

seasonal species excluding P_. reclinata (F37=7.75, p=0.004).

Temperature explained 8.9% and river flow 16.7% of the total

variance.

Comparisons with 1973-74

Distributions of fruit scores differed significantly

between 1973-74 and 1988-89 for 5 of the 9 species examined.

Acacia robusta, Pachystela msolo and Ficus sycomorus of

Mnazini and Diospyros mespiliformes and Ficus natalensis of

Mchelelo had significantly different distributions of fruit

scores between 1973-74 and 1988-89 (Kolmogrov-Smirnov J'= 7,

9, 11, 7 and 7 for A. robusta, P. msolo, F. sycomorus, D.

mespiliformes and F. natalensis, respectively, p<0.05).

Acacia robusta had an extended fruiting season in 1973-74

relative to 1988-89 and Pachystela msolo had an extended

fruiting season in 1988-89 relative to 1973-74. Diospyros

mespiliformes, Ficus sycomorus and F. natalensis had more










uniform distributions of fruiting in 1973-74 than in 1988-

89.

There were significant differences in monthly fruit

scores between 1973-74 and 1988-89 for 3 of the 7 species in

Mnazini and for 2 of the 6 species in Mchelelo. Ficus

sycomorus and Acacia robusta in Mnazini received

significantly higher mean monthly fruiting scores in 1973-74

than 1988-89 (T=2.94 and 2.75 for F. sycomorus and A.

robusta, respectively, p<0.05), whereas individuals of

Mimusops fruticosa received significantly higher fruiting

scores in 1988-89 than 1973-74 (T=2.28, p<0.05). The same

species in Mchelelo showed no significant differences in

fruiting scores. Diospyros mespiliformes and the palm,

Hyphaene compressa, had significantly higher mean fruiting

scores in 1988-89 than 1973-74 in Mchelelo (T=2.28 and 2.28

for D. mespiliformes and H. compressa, respectively,

p<0.05). Individuals of D. mespiliformes in Mnazini showed

no significant differences in monthly fruiting scores

between the 2 time periods.

Discussion
Mangabey foods were highly variable in their temporal

abundance and spatial distribution. The majority of food

species examined were rare and spatially clumped. Rarity and

spatial aggregation are characteristic of many tropical tree

species (Hubbel and Foster 1986, Robinson 1986, Struhsaker

1975) and may be the result of chance or particular










microhabitat requirements (Hubbel and Foster 1986). Hughes

(1985) showed that several species occurring in the Tana

River forests have very particular soil and water

requirements for germination and growth. Because the

frequency and depth of flooding, soil type and soil moisture

holding capacity vary widely between forests and often

within a single forest stand, different vegetative

assemblages occur among forests and certain plant species

may occur only in particular areas within a forest (Hughes

1985, Marsh 1978a, Homewood 1976).

The spatial distribution of mangabey food species

should affect ranging behavior throughout the year, but

because of the small size of the forests and the mangabeys'

ability to traverse the entire area, spatial distribution

should not affect whether the fruits of a particular tree

species will be included in the diet. The relationship

between plant abundance and temporal availability of fruits

likely is more important. The abundance of a particular

species limits the absolute availability of fruits to

mangabeys; as plants become rarer, total fruit availability

decreases and the amount of time when no fruits are

available to mangabeys may increase.

Mangabey fruit resources show periods of relative

abundance (November-May) and relative scarcity (June-

October). The abundance of continuous, asynchronously

fruiting species (e.g. Ficus sycomorus and Ficus










natalensis), however, can affect the temporal variation in

fruit availability by dampening the severity of periods of

scarcity. A critical number of individuals, however, is

required to ensure that fruits are available each month. As

these species decline in abundance, the probability that the

species population has an individual bearing fruits during

bottlenecks decreases. This is apparent in the monthly

phenological scores for F. sycomorus individuals from

Mnazini (n=7) and Mchelelo (n=10) forests (Fig. 12). Figs

were unavailable during 3 months in Mnazini but at least 1

fig was fruiting in all months sampled in Mchelelo.

Species such as Phoenix reclinata and Diospyros

mespiliformes that fruit at times when other species are not

in fruit also can form important and critical food resources

for mangabeys during months of otherwise low fruit

production. The large fruit crop of P. reclinata, coupled

with its extremely high density in the Mchelelo forest,

resulted in abundant fruit during months of relative food

scarcity in Nkano and Mnazini, forests with lower densities

of reproductive palms (Fig. 13). The abundance of species

such as P. reclinata and Ficus spp. may be a key factor in

explaining the presence or absence of mangabeys in the

various forest patches along the Tana River.

Many tropical forest species show adaptations toward

fruit-drop and/or seed germination near the beginning of

heavy rains (Rathche and Lacey 1985). Flowering occurs










primarily during periods of drought and low soil moisture

(Ayres 1986, Monasterio and Sarmiento 1976, Croat 1975,

Foster 1974, Koelmeyer 1960). This phenological pattern does

not appear to prevail within the Tana River study sites.

Although flowering occurs during times of low river water

and rainfall, when soil moisture is low, there are few

consistent associations between fruiting and environmental

conditions. Fruit production by species that produce fruit

continuously throughout the year, particularly Ficus spp.,

appear to be associated with the onset of the rains but the

more seasonal species show no associations with rain. River

flow and temperature also explained very little of the

variance (25.6%) in a multiple regression model. It is

likely that fruiting patterns of the Tana forests are

influenced by a complex interaction of the variables

examined here and perhaps more importantly, the position of

the forest relative to the water table, the occurrence of

floods, the depth and length of inundation during flooding

and soil type. Fruiting time also may be more strongly

correlated with elapsed time from germination than with

environmental variables. Environmental cues seldom stimulate

the onset of fruit ripening; onset is determined primarily

by internal factors that control the rate of fruit

development (e.g., production of sugars; Rathche and Lacey

1985).










Pronounced year-to-year variation in the onset and

termination of the dry season could have important

consequences for the flowering and fruiting phenologies of

tree species (e.g. Chivers and Raemaekers 1980). The Tana

riverine forests experience high year-round temperature with

some seasonal variation, twice yearly rains and irregular

flooding, with the severity and timing of rains and floods

varying appreciably between years (Hughes 1985, Homewood

1976, Marsh 1976). The extreme unpredictability and annual

variation in rainfall, temperature and flooding, overlaid

with habitat degradation and fragmentation, make it

difficult to determine causes of significant differences in

fruiting patterns between 1973-74 and 1988-89. Wheelwright

(1986) states that even 7 years of phenological data on

species occurring in a Costa Rican montane forest, a habitat

far less variable than the Tana River, were too few to

determine supra-annual cycles of fruit production. The

critical point however, is that total fruit availability

will decline with decreasing abundances of reproductive

trees. The abundances of several important food species

(e.g. Ficus sycomorus, Ficus natalensis, Acacia robusta,

Albizia glaberrima, and Mimusops fruticosa) declined between

1973-74 and 1988-89 in at least 1 or more of the study sites

(Decker 1989, Medley 1990), effectively decreasing overall

fruit production. Such long-term changes in the fruit









78
resource base may be shaping the present day feeding ecology

of the Tana mangabey.




Full Text
11
the dissertation and broadly discuss management
recommendations.


Short Rains
Long Rains
Morning
Mid-day
Afternoon
Inter-rains
Dry Season
Morning
Mid-day
Afternoon


81
Therefore, when animals must go through a long search or
handling time per unit energy, major shifts in activity
budgets will occur.
Although seasonal and daily variation in activity
patterns are well documented for several primate species,
very few studies have investigated long-term variation in
activity budgets (Oates 1986). Furthermore, none has
addressed the question of whether large-scale habitat change
(e.g., forest loss and fragmentation) and subsequent change
in the food resource base (e.g., lower resource
availability), causes primates to reallocate time spent in
different activities. The Tana River crested mangabey
(Cercocebus galeritus) is an ideal species for addressing
this question. Tana mangabeys have experienced severe
habitat alteration over the last 15 years due to forest
senescence, forest cutting for agriculture, and selective
felling and pruning of species for building materials (Marsh
1986). These habitat changes, which involved the loss of
important mangabey diet species, have resulted in an overall
reduction in the food resource base.
In this paper, I examine the effects of these 15-year
changes in the forest structure and food resource base on
long-term variation in activity budgets. I also examine the
temporal distribution (diurnal and seasonal) of activity
patterns within and among 3 mangabey groups observed in
1988-89 and the influence of diet on time budgets. Finally,


17
1986, Decker 1989), plant species composition has not
changed significantly since the mid-1970s (Medley 1990).
The 35 ha Mnazini forest is less than 0.5 km south of
the Nkano forest, being separated from it by an old river
course and an area of grassland. Mnazini is a low diversity
forest (Homewood 1976, Medley 1990) that experiences high
levels of flooding and human exploitation. The forest has
decreased in size since the mid-1970s by approximately 50%
due to floods and fires in 1982 and 1983, respectively. The
number of mangabey groups in Mnazini has declined from 2 to
1 since the mid-1970s.
The climate is strongly seasonal. Rainfall averages 400
mm per year, most of which falls during the long rains
(March-May) and the short rains (November-January), with
intervening dry periods generally occurring during February
(inter-rains) and June-October (dry season) (Hughes 1984).
Floods are highly variable in intensity and timing but in
general, water levels peak during the long and short rains
(Hughes 1984).
Abundance and Distribution of Resources
Density. I enumerated selected tree species in 2 of the
3 study sites where mangabey groups were intensively studied
(Mchelelo and Nkano) to calculate the abundance and describe
the dispersion patterns of mangabey foods. Species were
selected based on their importance in the diet during 1973-
74 (Homewood 1976) or the present study. Only those trees of


180
Table 5.2. Spearman's Rank correlation coefficients (r )
between frequency of quadrat use by mangabeys and frequency
of occurrence of top 5 plant diet species each month for the
North Mchelelo, South Mchelelo, and Nkano groups. (-)
indicates
available
that ranging data
for all top 5 diet
or plant frequency
species that month
data were not
Month
North Mchelelo
rxy
South Mchelelo
rxy
Nkano
rxy
Feb 8 8
0.38**
-0.01
_
Mar
0.29*
0.35
0.12
Apr
-0.10
May
23**
-
-
Jun
0.55
0.05
0.01
Jul
0.35

0.44
0.19
Aug
-
-
*08**
0-28
Sep
i
o
o
u>
Oct
0.22
-
0.37
Nov
-0.02^
-0.08
0.28
Dec
0.59
-0.04
0.28**
Jan 89
0.25
k k
0.43
k k
0.50
Feb
0.10
-
0.45
Mar
0.50
*
*
p<0.05, n=58
p<0.1, n=58


I owe perhaps my greatest debt to Dr. Tom Struhsaker.
He helped in the study design, pushed for funding, made
several well-needed trips to my field site and has been my
most critical editor. Tom encouraged and reassured me when I
most needed it and pushed me back on course when I was
straying. He has been teacher, collegue, interrogator,
project supervisor, and, most importantly, friend. For all
of this, I extend my deepest thanks.
Many of my ideas were formed through conversations with
fellow students and colleagues. I thank Jeff Hardesty, Dr.
Susan Jacobson, Lysa Leland, Jay Malcolm and Peter Polshek
for their input, but in no way hold them responsible for the
final product. Gainesville friends Lane Devereaux, Mark
Edmiston, Nina Hoffer, Susan Jacobson, Jim and Katherine
Reid and Bob and Deb Wallace helped logistically and/or
emotionally. I also thank Laurie Wilkens and Dr. Steve
Humphrey for providing me with large and friendly working
spaces during the writing phase. Dr. Andy Dobson and the
students and staff of the Ecology and Evolutionary Biology
Department of Princeton University also are thanked for
providing me with a stimulating atmosphere and comfortable
office in the final month of writing.
The previous and extensive studies by Dr. Katherine
Homewood were an invaluable resource. Dr. Homewood also
furnished unpublished data, supplemental information and
inspiration.
IV


Figure 2.15. Contribution of ripe and unripe fruits to
monthly fruit biomass index, a) Mchelelo forest; b)
Nkano forest; c) Mnazini forest 59
Figure 2.16. Monthly mean minimum and mean maximum air
temperature in degrees centigrade, a) Mchelelo
forest; b) Nkano/Mnazini forest areas 62
Figure 2.17. Total monthly rainfall in millimeters.
a) Mchelelo forest; b) Nkano/Mnazini forest area 64
Figure 2.18. Mean monthly river flow for the Tana
River Hola Station 66
Figure 3.1. Mean annual percent time spent among 7
behaviors by three 1988-89 mangabey groups and two
1973-74 mangabey groups, a) 1988-89; b) 1973-74 89
Figure 3.2. Mean percent time spent in 4 major
behaviors (eating, foraging, moving and inactive)
by time-of-day and season for the 1988-89 N.
Mchelelo group 96
Figure 3.3. Mean percent time spent in 4 major
behaviors (eating, foraging, moving and inactive)
by time-of-day and season for the 1988-89 S.
Mchelelo group 98
Figure 3.4. Mean percent time spent in 4 major
behaviors (eating, foraging, moving and inactive)
by time-of-day and season for the 1988-89 Nkano
group 100
Figure 3.5. Mean percent time spent in 7 social
behaviors by 1973-74 and 1988-89 adult males and
females, a) adult males; b) adult females .... 104
Figure 3.6. Mean percent time spent in 7 social
behaviors by three age-sex classes in 1973-74 and
1988-89. a) juvenile males; b) juvenile females;
c) male and female infants combined 106
Figure 3.7. Percent time 1988-89 and 1973-74 infants
spent in 3 levels of association with their
mothers. Asterisks denote months when levels of
association were significantly different between
the 2 study periods, a) Association A: the amount
of time an infant was on its mother; b)
Association B: the amount of time an infant had
its mother as nearest neighbor; c) Association C:
the amount of time infants were separate from
their mothers 109
xvi


Figure 2.14. Contribution to monthly fruit biomass indices
by different species categories,
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.


120
Associations between activity budget and diet. Monthly
activity budgets varied with diet composition and food
availability. The amount of time spent feeding (eating +
foraging) for the 3 1988-89 groups was negatively correlated
with percent ripe fruit in the diet (r=-0.48, n=38,p<0.002)
and was positively correlated with percent unripe seed in
the diet (r=0.36, n=38, p<0.02). When eating and foraging
were treated separately, eating, but not foraging, was
negatively correlated with the amount of fruit in the diet
(r=-0.40, n=38, p<0.01). There were no significant
associations between foods other than fruit in the diet
(e.g. invertebrates) and amount of time spent in the
different activities. Although levels of inactivity showed
no relationships with diet, inactivity was strongly
correlated with feeding (r=-0.49, n=38, p<0.002); as the
amount of time spent foraging and eating decreased, levels
of inactivity increased. Plant species dietary diversity
(Shannon-Weiner H', Pielou 1969) also was negatively
correlated (r=-0.49, n=30, p<0.005) with measures of fruit
availability (see Chapter 2), particularly ripe fruit. Thus,
as fruit resources became scarce, the diet broadened and the
amount of time spent in feeding behaviors increased.
Comparison with 1973-74 diet composition. The
contribution of fruit, seed and other items to the diet
differed between 1973-74 and 1988-89 (Figure 3.8). MANOVA
showed a highly significant effect of group on the overall


32
20+


84
2) Foraging. This category included both manipulatory
and/or searching activity, excluding ingestion. Manipulatory
foraging was any movement which involved manipulating a
potential food object. Searching was defined as closely
inspecting or sifting through leaf litter or foliage.
3) Moving. Movement included all locomotion except short
movements during foraging, and locomotion during activities
such as playing.
4) Inactive. This category included all stationary,
nonsocial activity. Animals sleeping, sitting, lying, or
standing were considered inactive.
5) Autogrooming. Any animal scratching or otherwise self
cleaning was autogrooming.
6) Allogrooming. Allogrooming included animals grooming or
being groomed by another animal.
7) Social/Sexual. This was a composite category including:
a) approach individuals giving or receiving friendly or
sexual presentations or embraces; b) aggression -
individuals supplanting another from an occupied position,
or lunging at, chasing or biting another, and individuals
behaving submissively to another by posturing or retreating
from a potential conflict; c) play individuals engaged in
mock-chasing or mock-fighting; d) sexual behaviors -
mounting or copulating; e) maternal behaviors nursing,
picking up infant, etc.; and f) vocalizing giving alarm
calls and adult male long-calls.


Figure 5.8. Frequency distributions of half-hour distance
steps in intervals of 20 m. Asterisks indicate
values expected if distance steps were
distributed randomly (following a Poisson
distribution).
a) Frequency distribution for N. Mchelelo
group; b) Frequency distribution for S.
Mchelelo group; c) Frequency distribution for
Nkano group.


247
2000 individuals would be adequate. This estimate is still 3
times greater than the present mangabey population size but
close to the population sizes reported for mangabeys in the
mid 1970s.
Models incorporating details of a species life history
and demography often require larger populations than those
based on genetic grounds alone (Lande 1988). Application of
data for the Tana mangabey to both genetic and demographic
models result in very similar population requirements. Given
the relationship between current population sizes and
calculated Ne's (Ne/N) a census population of 3600-4500
individuals would be required if an Nc of 500 is considered
necessary to maintain genetic variability within the
population. If the effect of isolation between river banks
is considered, as many as 9000 mangabeys may be necessary to
achieve an Nc of 500. These requirements are very similar to
those predicted by the demographic model for a 95%
probability of persistence over the next 100 years. Although
these similarities may be fortuitous, the important point is
that two very different approaches to population viability
indicate that the mangabey is at risk at its present
population size and require populations of the same order of
magnitude to avert loss of genetic variance or probable
extinction.
One of the most important results of viability analysis
to date is that viable populations generally are large, so


234
generations require detailed life-tables which are not
available for the mangabey. Without incorporating the
effects of overlapping generations, the final estimates will
be overestimates, probably by a factor of 2 (Soul 1980).
Population Subdivision
Values for genetic fixation between demes (Fst) based on
estimated levels of migration (Table 6.5), indicate that
forest demes along either bank of the river are not
effectively isolated from one another. Over 95% of the
genetic variance within the population may be due to
differences within the forests and only 4%-5% of the genetic
variance (Fst East=0.036, Fsl West = 0.048) appears to be due
to differentiation among the forests. Because estimates of m
(Table 6.5) do not incorporate secondary transfer by males,
the reported Fst values likely overestimate the amount of
differentiation and therefore are conservative. Estimates of
m used, however, are not unrealistic when m is viewed as an
emigration rate that is influenced more by the number of
potential breeding slots available in a neighboring deme
within 1 generation than by the actual number of migrants
leaving a deme. With an estimated mean male tenure time of
approximately 3 years, 2 males every 6 years (equivalent to
2 males/generation) could on average gain access to a
neighboring deme and breed. Considering each deme in a
linear array has 2 neighbors, a maximum number of 4


274
Soul, M.E. 1980. Thresholds for survival: maintaining
fitness and evolutionary potential. Pages 151-169 in
M.D. Soul and B.A. Wilcox, editors. Conservation
Biology: An Evolutionary-Ecological Perspective.
Sinauer Associates, Sunderland, MA.
Soul, M.E. 1987. Where do we go from here? Pages 175-184
in M.E. Soul, editor. Viable Populations for
conservation. Cambridge University Press, Cambridge,
UK.
Srikosamatara, S. 1986. Group Size in Wedge-capped
Capuchin Monkeys (Cebus olivaceus): Vulnerability to
Predators,Intragroup and Intergroup Feeding Competition.
Unpublished Ph.D. dissertation, University of Florida.
Stephens, D.W. and J.R. Krebs. 1986. Foraging Theory.
Princeton University Press, Princeton, NJ.
Stearns, S.C. 1976. The evolution of life history traits:
A critique of the theory and a review of the data. Ann.
Rev. Ecol. and Syst. 8:145-171.
Struhsaker, T.T. 1975. The Red Colobus Monkey. University
of Chicago Press, Chicago, IL.
Struhsaker, T.T. 1978. Food habits of five monkey species
in the Kibale Forest, Uganda. Pages 225-248 in D.J.
Chivers and J. Herbert, editors. Recent Advances in
Primatology, Vol. 1, Behaviour. Academic Press, New
York, NY.
Struhsaker, T.T. 1980. Comparison of the behaviour and
ecology of red colobus and redtail monkeys in the Kibale
Forest, Uganda. African Journal of Ecology 18:33-51.
Struhsaker, T.T. and L. Leland. 1979. Socioecology of five
sympatric monkey species in the Kibale Forest, Uganda.
Advanced Studies in Behavior 9:159-228.
Terborgh, J. 1983. Five New World Primates: A Study in
Comparative Ecology. Princeton University Press,
Princeton, NJ.
Terborgh, J. and B. Winter. 1980. Some causes of
extinction. Pages 119-134 in M.E. Soule and B.A.
Wilcox, editors. Conservation Biology: An
Evolutionary-Ecological Perspective. Sinauer
Associates, Sunderland, MA.


CHAPTER ONE
INTRODUCTION
The Tana River crested mangabey, or Tana mangabey,
(Cercocebus qaleritus qaleritus Peters 1879) is a rare and
highly endangered cercopithecine monkey (IUCN 1976, Lee et
al. 1988) found only in small forest patches of the Tana
floodplain in north-eastern Kenya (Homewood 1976; Figure
1.1). Mittermeier (1981) ranked C. g. qaleritus among the
most endangered primates in the world, and the IUCN/SSC
Primate Specialist Group listed it in the highest priority
category for conservation (Oates 1985). These rankings were
based on the small population size of C. g. qaleritus. its
highly restricted range, and the present threats to its
habitat. C. g. qaleritus experiences many of the classic
factors contributing to endangerment: it has a small
population size, is highly restricted in range, occupies a
naturally patchy habitat that is becoming increasingly
fragmented by human exploitation, and has experienced a
population decline despite the creation of a reserve for its
protection (Eisenberg 1980, Terborgh and Winter 1980).
The Tana mangabey, is thought to be a relict population
of a once widely distributed species (Homewood 1975, 1976);
1


Figure 3.8. Percent composition of diet by item for 1973-74
and 1988-89.
a) 1988-89; b) 1973-74.


> E
244
Ne
~ Vm 1 Nb


Figure 4.2. The number of feeding observations on P.
reclinata by month and item for mangabeys in
the Mchelelo and Nkano forests,
a) Mchelelo; b) Nkano.


Table 5.2. Spearman's Rank correlation coefficients
(r^) between frequency of quadrat use by mangabeys
and frequency of occurrence of top 5 plant diet
species each month for the North Mchelelo, South
Mchelelo, and Nkano groups. (-) indicates that
ranging data or plant frequency data were not
available for all top 5 diet species that month. 180
Table 5.3. ANOVA results for effect of South Mchelelo
group presence and month on distance traveled and
measures of travel pattern and use of space by
North Mchelelo group. F-statistics are calculated
from Type III partial sums of squares 190
Table 5.4. Number of adult male long-calls by study
group and month. North refers to long-calls given
by the North Mchelelo group males. South M refers
to long-calls given by the South Mchelelo group
males while present in Mchelelo forest and South C
refers to long-calls given by the South Mchelelo
group males while in the neighboring Congolani
forest. Alien refers to long-calls given by males
of a second, unhabituated group in Nkano forest,
and Unk represents long-calls of unidentified
males in the Nkano forest 192
Table 5.5. Intergroup behavior categories and measures
of proximity and fighting by month. F, M, and A
designate fights, merge, and avoid months,
respectively. N includes all 12-hr observation
days and days for which only presence or absence
of South group was determined 194
Table 6.1. Minimum and maximum numbers of mangabey
groups censused in forests on the east and west
banks of the lower Tana River in 1988-89 and 1973-
74. Forests not censused indicated with (-);
those no longer standing indicated with (x). Data
from 1973-74 taken from Homewood (1976) and Marsh
(1976, 1978) 217
Table 6.2. Age-sex composition of 7 mangabey groups
with comparative means from 1973-74 (Homewood
1976). AM = adult male (>4-5 yr); SubM = subadult
male (3-5 yr); AF = adult female (>4 yr); SubF =
subadult female (2.5-3 yr); JM = juvenile male (8
mon-3 yr); JF = juvenile female (8 mon-2.5 yr); IM
and IF = infant male and infant female,
respectively (< 7 mon) 218
Table 6.3. Reproductive parameters for male and female
mangabeys 220
Xll


Table 3.6
Continued
c) Nkano
Month
Ripe
Fruits
Unripe
Fruits
Ripe
Seeds
Unripe
Seeds
Flowers
Leaves
Sprouts
Shoots
Wood
Bark
Gum
Animal
N
Mar 88
37.2
5.2
2.1
29.7
4.2
1.6
2.6
3.7
0.0
13.6
191
Apr
45.6
12.3
0.0
23.6
0.0
1.5
3.1
2.1
0.0
11.8
195
Jun
53.5
2.9
0.6
7.6
0.0
1.2
6.4
2.9
0.0
25.0
172
Jul
43.1
2.5
13.7
4.1
0.0
1.5
8.6
8.1
0.0
18.3
197
Aug
4.5
3.0
0.0
43.3
25.4
3.0
6.0
6.0
0.0
9.0
67
Sep
19.5
8.8
0.0
31.2
34.4
0.0
0.0
0.0
0.0
6.0
215
Oct
31.2
5.8
12.6
35.5
1.1
1.1
1.1
2.2
1.1
9.0
189
Nov
10.1
9.0
66.0
0.0
1.1
0.0
0.7
5.2
0.0
7.8
268
Dec
58.1
1.3
0.6
0.6
0.0
0.6
21.9
3.9
0.0
12.9
155
Jan 89
45.3
11.3
0.6
5.7
0.6
0.6
8.2
6.3
0.0
21.4
159
Feb
3.3
2.8
0.0
67.0
5.6
1.4
2.3
4.2
0.0
13.5
215
Mar
8.4
5.3
0.0
64.2
3.5
0.9
2.2
1.8
0.0
10.2
226
(-*
CTl


71
Table 2.7. Spearman's rank correlation coefficients
for mean monthly environmental variables and mean
monthly fruiting scores for all forests combined.
Variables defined in Table 2.4
Table 3.1. Mean percent time spent in each activity by
each month by mangabeys in the 1988-89 N.
Mchelelo, S. Mchelelo, and Nkano study groups.
Standard errors are in parentheses. See text for
definition of activities 90
Table 3.2. MANOVA results for group membership,
season, and time-of-day differences in 1988-89
mangabey activity budgets 91
Table 3.3. ANOVA results for group membership, season,
and time-of-day differences in 7 behaviors.
Significance values are calculated from Type III
partial sums of squares 92
Table 3.4. ANOVA results for study (Homewood's 1973-74
study and the present 1988-89 study) and age
differences in 3 measures of mother/infant
association: a) amount of time infant is on its'
mother, b) amount of time infant has mother as
nearest neighbor and c) amount of time infant is
separate from mother. Data were nested by infants
within studies 110
Table 3.5. Rank order of the five most common plant
species eaten and the percentage of the total
feeding records for identified plants (N) by group
and month. Complete scientific names of plant
species listed in Appendix B 113
Table 3.6. Percentage of diet items eaten by month and
group. Unidentified items are not included. . 115
Table 3.7. Frequency of plant parts eaten for common
diet species of the 3 study groups 117
Table 4.1. Human uses of P. reelinata grouped into 6
broad categories. Plant parts utilized and
harvesting methods employed are outlined 142
Table 5.1. Mean diversity of quadrat use (H'), mean
number of unique quadrats entered, mean distances
travelled, mean half-hour step distances, mean
turning angles, and mean number of path crossings
by month for the 3 study groups 178
xi


CHAPTER FIVE
TANA RIVER CRESTED MANGABEY HOME RANGE AND SPATIAL
DEFENSE: EFFECTS OF FRUIT ABUNDANCE AND DISTRIBUTION
Introduction
Variability in primate ranging patterns have been
related primarily to the availability and distribution of
food resources (e.g., Altmann and Altmann 1970, Clutton-
Brock 1977, Robinson 1986, van Schaik et al. 1983, Wrangham
1980). Species that depend on uniformly distributed,
renewable resources tend to show an even use of space while,
at the opposite extreme, species relying on highly patchy
resources shift activities from one location to another as
different patches become productive (Clutton-Brock and
Harvey 1977, Struhsaker 1980, Struhsaker and Leland 1979,
Terborgh 1983). Factors other than food also shape home
range use. Such factors include weather (Chivers 1974, Marsh
1978b), location of traditional sleeping sites (Gittins and
Raemaekers 1980, Rasmussen 1979), reproductive strategies of
troop members (Rasmussen 1983) and intergroup interactions
(Struhsaker 1975, Waser 1976, Isbell 1983, Srikosamatara
1986).
153


4
other subspecies, C. g. chrvsoqaster and C. g. agilis
inhabit the Congolese forest block from Cameroon through
Gabon, Congo (Brazzaville), and Zaire (Dandelot 1974, Lee et
al. 1988). A fourth subspecies, C. g. saniei. occurs in the
Uzungwa Mountains of Tanzania (Homewood and Rodgers 1981,
Lee et al. 1988). All subspecies of C. qaleritus are semi
terrestrial and broadly similar in ecology and social
behavior (Homewood 1975, Quris 1975) but vary in morphology.
Dobroruka and Badalec (1966) and Groves (1978) have
classified the Tana mangabey as a separate species based on
differences in skull morphology. Irrespective of the
taxonomy used, the important point is that the Tana mangabey
is a geographically, and perhaps genetically, distinct
population and should be considered separately from its
conspecifics in the conservation arena.
The Tana mangabey occurs only within a 60 km stretch of
the lower Tana River (Andrews et al. 1975, Homewood 1976)
where ground-water seepage from the river is sufficient to
support patches of riverine forest in a region of otherwise
semi-arid bushland. Within the mangabeys' range there are an
estimated 65 forest patches, averaging 35 ha in size (range
<10-625 ha; Marsh 1978a, 1986), separated by stretches of
deciduous woodland, bush, and grass. Forest patches, or
fragments, result in part from changes in the river's course
and patterns of flooding. Changes in the river's course
during high floods leaves some forests far from the life-


208
resources of wedge-capped capuchin monkeys (Cebus olivaceus)
was due, in part, to their knowledge of resource location
and ability to move accordingly.
Invertebrate consumption also did not appear to
influence movement patterns. Although invertebrates may be
highly renewable, they appeared to be distributed uniformly
through the habitat; mangabeys frequently fed on
invertebrates in the leaf litter when traveling between
fruit trees and no quadrats could be identified as areas of
high invertebrate consumption. A uniform distribution of
invertebrates would decrease the influence of invertebrate
foraging on use of space and movement patterns.
Forest patch size may limit mangabey movements and use
of space. Tana mangabeys are constrained in movement by the
small size and isolation of the forests they occupy. If
groups are unable to travel between neighboring patches,
home range and movements are limited as groups reach forest
boundaries and are forced to turn or backtrack. Although
daily path lengths are similar among Tana mangabeys and
other closely related species (Waser 1984), home range
estimates for mangabeys occupying larger, less fragmented
forests are an order of magnitude greater (Cercocebus
galeritus agilis. 200+ ha, Quris 1975; Cercocebus torauatus.
247 ha, Mitani 1989; Cercocebus albigena. 410 ha, Waser
1975). Only the black mangabey (Cercocebus aterrimus) of the
Zaire Basin has been reported to have home range sizes


176
variances >2.5, df 59,13, p<0.05), indicating that mangabey
behavior was more consistent in high-use quadrats.
I also tested for differences in mean species diversity
and mean densities of the 10 top mangabey diet plant species
(Appendix B) in high- and low-use quadrats to determine if
high-use quadrats differed vegetatively from low-use
quadrats. There were no significant differences in mean
species diversity between high- and low-use quadrats for
North Mchelelo, South Mchelelo, or Nkano (Student's two-
sample t-test, p>0.05). Differences in diet species
densities were significantly higher in high-use quadrats of
the North Mchelelo group for Phoenix reclinata (t=2.1,
df=63, p=0.04) and Oncoba spinosa (t=2.8, df=63, p=0.01).
High-use quadrats of the South Mchelelo group had
significantly higher densities of Alangium salviifolium
(t=2.0, df=63, p=0.04) and Ficus svcomorus (t=2.6, df=63,
p=0.02), and those of the Nkano group had significantly
higher densities of Pachvstela msolo (t=3.9, df=52,
p=0.0004) and Alangium salviifolium (t=2.9, df=52, p=0.006).
Diet species densities in high- versus low-use quadrats,
however, generally did not correspond with differences in
the consumption of these species by mangabeys. There were no
significant differences in the proportion of different diet
species eaten in high- and low-use quadrats for the North
Mchelelo group (Student's two sample t-test, p>0.05). The
South Mchelelo group ate a significantly higher proportion


231


259
APPENDIX B Continued
N. Mchelelo
S. Mchelelo
Nkano
Taxon
No. of
No. of
No. of
records %
records %
records %
EUPHORBIACEAE
Antidesma venosum 19
Drvpetes natalensis 1
FLACOURTIACEAE
Oncoba spinosa 251
GRAMINEAE
grasses (including
Urochloa setiaera) 75
GUTTIFERAE
Garcinia livingstonei 10
MALVACEAE
Abutiln mauritianum 12
Hibiscus micranthus
Thespesia danis
MIMOSACEAE
Acacia robusta 104
Albizia qummifera 16
MORACEAE
Ficus bubu 14
Ficus bussei
MORACEAE
Ficus natalensis 128
Ficus svcomorus 90
PALMAE
Borassus aethiopium
Hyphaene compressa 322
Phoenix reclinata 1012
PAPILIONACEAE
Rhvnchosia viscosa 2
SALICACEAE
Populus illicifolia
0.6
4
0.2
2
0.1
0.03
2
0.1
5
0.3
rH
CO
69
4.0
13
0.7
2.5
85
o
in
54
2.8
0.3
21
1.2
4
0.2
0.4
_
-
-
-
1
0.05
2
0.1

3.2
6
0.4
1
0.05
0.5
4
0.2

0.5

1
0.05

43
2.5
-
-
4.1
63
3.7
_

2.9
43
2.5
342
17.9
1
0.06
1
0.05
1 G. 4
137
8.0
-
-
32.7
602
35.2
449
23.3
0.07
-
-
1
0.05
_
1
0.06


133
colobus mothers provided their infants with a greater
opportunity to play.
A reduction in the overall food supply, and/or a shift
to less nutritious food items also may have increased the
length of dependency on the mother if a) the mother had a
decreased milk production or b) an inexperienced, infant
forager was unable to locate or compete for food resources.
Extended infant dependency should delay the incorporation of
juveniles into the social network of the group. Juveniles
therefore might be expected to spend more time with peers
than with older group members (O'Brien 1990, Walters 1986);
this may be supported by the great amount of time 1988-89
juvenile males spent playing. Increased play, however, would
not be expected during times of nutritional stress. Baldwin
and Baldwin (1971) and Lee (1984) show reduced play in
squirrel monkeys (Saimiri spp.) and vervets (Cercopithecus
aethiops), respectively, in less productive habitats and
during times of reduced food supply. Differences in
estimates of play by 1988-89 and 1973-74 juvenile male
mangabeys may be more a function of group demography than
nutritional status. The juvenile sex ratio in 1988-89 was 8
males: 5 females versus 1 male: 4 females in 1973-74
(Homewood 1976). Because juvenile male mangabeys play more
than any other age/sex class (Homewood 1976), the
differences in sex ratio could have effected rates of play.


190
Table 5.3. ANOVA results for effect of South Mchelelo
group presence and month on distance traveled and measures
of travel pattern and use of space by North Mchelelo group.
F-statistics are calculated from Type III partial sums of
squares.
Source
df
Sum of
Squares
Mean
Squares
F
P>F
Distance
Month
13
996995.2
76691.9
3.39
0.001
Presence
1
1.0
1.0
0.00
0.994
Error
42
949264.3
22601.5
Turning Angle
Month
13
2701.5
207.8
2.01
0.043
Presence
1
832.3
832.3
8.07
0.007
Error
42
4333.0
103.2
Path Crossing
Month
13
11.3
0.9
1.96
0.05
Presence
1
1.9
1.9
4.25
0.04
Error
42
18.6
0.4
Quad Entries
Month
13
271.1
20.9
1.48
0.167
Presence
1
84.7
84.7
6.00
0.019
Error
42
592.8
14.1
Quad Diversity
Month
13
0.7
0.1
1.31
0.247
Presence
1
0.2
0.2
4.93
0.032
Error
42
1.7
0.1


5
supporting ground-water supply; others are cut and eroded by
the scouring action of the river.
In recent years, natural forest fragmentation has been
exacerbated by increasing human exploitation and land use
(Homewood 1975, 1976; Hughes 1985, 1987, Kinnaird et al.
1990, Marsh 1986). An increasing local population (e.g., the
number of families in one village, Nkano, increased from 1
to 7 between 1974 and 1986; personal observation T.T.
Struhsaker), dependent on slash-and-burn agriculture, has
resulted in increases in forest clearing and tree felling
for houses and dugout canoes (Kinnaird et al. 1990). Medley
(1990) showed a 56% reduction in forest area and
fragmentation of 5 forest areas into 15 patches between 1960
and 1975. This loss was attributed, in part, to changes in
the rivers course (Medley 1990), and to the destruction of
forest for agriculture (Decker 1989).
Forests also have been cleared for large-scale
irrigation schemes (Medley et al. 1989), and possibly have
experienced a reduction in regeneration due to alterations
in the rivers flow by upstream dams (Hughes 1985, 1990, Lee
et al. 1988, Marsh 1986). Although changes in the flooding
regime due to upstream dams have not been verified, Hughes
(1985, 1990) has speculated that a reduction in peak floods
due to damming will reduce the sediment/nutrient load and
the frequency of meander formation and cut-off, thereby
reducing the dynamic nature of the floodplain and the


APPENDIX A
ESTIMATING TREE CANOPY SURFACE AREA AND VOLUME
Tree canopies display an abundance of shapes, varying
from the classic "Christmas tree" form to the flat,
spreading, umbrella shape of an Acacia. Some tree canopies
approximate geometric forms such as cones or spheres
but many others, exhibiting odd shapes and canopy
irregularites, do not. Deciding on geometric formulae when
estimating the surface area and volume of tree canopies
therefore is problematic. To determine a geometric form that
could best generalized across various shapes, I generated a
set of dimensions and calculated surface area and volume
based on formulae for different geometric shapes. I then
correlated the areas and volumes generated from all
formulae. The shape giving areas and volumes that correlated
best with all others was considered the most generalizable.
Values varying from 1-3 were used for 3 dimensions:
height, radius 1 and radius 2. All possible combinations for
the 3 dimensions were made, giving a total sample of 27 sets
of dimensions, or tree shapes. I calculated surface area and
volumes for all sets using formulas for 6 geometric shapes:
1) an open-bottomed cylinder, 2) a cone, 3) a 1/2 sphere, 4)
the zone of a spheriod, 5) a 1/2 oblate spheroid and 6) a
255


16

k


3h) o**- o > a> coo. oco
189
O > coa ca i-

80
Monthly variation in activity budgets, particularly in
the time spent feeding, has been demonstrated for several
primate species and shown to covary with measures of food
availability (e.g., Harrison 1985, Homewood 1976, Oates
1977, Robinson, 1986, Terborgh 1983, Waser 1975). In
general, species relying heavily on foliage tend to decrease
their level of activity during months when high-quality food
is scarce, while species that rely heavily on fruits or
insects increase the amount of time spent looking for and
processing food during times of scarcity (Oates 1986). The
spatial distribution of food also influences the amount of
time individuals spend moving; individuals of species that
feed on high density, evenly dispersed food spend less time
moving than those of species that feed on widely dispersed
food (Struhsaker 1980, 1978; Struhsaker and Leland 1979).
Optimal foraging theory predicts that an animal will
act to maximize energy intake per unit time to meet demands
for maintenance, movement, and reproduction (Stephens and
Krebs 1986). Activity budgets therefore should reflect these
actions. As the distribution and availability of food
changes, time spent in energy-accruing behaviors (eating)
versus more costly behaviors of searching for and moving
between patches of food should change. The proportion of
time spent in various activities forms a mutually dependent
set, i.e., time spent in one activity affects the time
available for other activities (Altmann and Altmann 1970).


250
distribution result in facultative defense of space by Tana
mangabeys. As fruit availability increases, the distribution
of plant species in the diet influences the type of
interaction; peaceful, intergroup interactions generally
occur when mangabeys eat evenly distributed species and
aggressive interactions occur when mangabeys eat species
with patchy distributions. The question of whether defense
has increased over the last 15 years as fruit resources
became more limiting and distributions changed could not be
adequately addressed with the available database and poses
interesting questions for future work.
The flexible behavioral attributes of the Tana mangabey
make it well suited for highly dynamic, naturally patchy
riverine forests. There are, however, limitations to the
adaptability of any species? the Tana mangabey is limited by
habitat quality and other specific requirements. My results
show that mangabey groups go extinct as forests become
isolated from the river and senescence. Small forests,
isolated from neighboring patches, support fewer mangabey
groups than larger, less isolated forests (Kinnaird unpubl.
data). The presence of particular food species also may be
critical to mangabey presence and/or persistence in a
forest. For example, Phoenix reclinata and Ficus spp.
provide critical resources when other fruits are unavailable
and may greatly improve patch quality for mangabeys.
Mangabey population size therefore is limited by the number


Figure 5.6. Occupancy of each 0.25 ha quadrat by North and
South Mchelelo groups; percent occupancy is
divided into 1 of 5 intensities of use.


This dissertation was submitted to the Graduate Faculty
of the School of Forest Resources and Conservation in the
College of Agriculture and to the Graduate School and was
accepted as partial fulfillment of the requirements for the
degree of Doctor of Philosophy.
December, 1990
Director, School of
Forest Resources
and Conservation
Dean, Graduate School


21
individuals present. I chose only those individuals believed
to be of a species' reproductive size and individuals whose
canopies were relatively easy to view.
The phenological state of each individual was assessed
monthly for Mchelelo (n=19 months), Nkano (n=17 months) and
Mnazini (n=13 months) beginning October 1987. I used a 0-5
scale developed and used by Homewood (1976) in 1973-74 to
visually assess the presence of flowers and ripe and unripe
fruits on each individual. The scale is expressed as a
proportion of the total canopy covered by flowers or fruits,
where 0=no flowers or fruits, 1=1-20% coverage, 2=21-40%
coverage, 3=41-60% coverage, 4=61-80% coverage and 5=81-100%
coverage. Flowers and fruits were examined separately, each
receiving a potential maximum score of 5.
Two indices of flower and fruit availability were
calculated for each study forest using the 0-5 fruit scale.
The first index is comparable to that used by Homewood
(1976) and provides a relative measure of variation in
flowering and fruiting. This index is calculated by pooling
monthly flower or fruit scores for individuals of each
species and averaging to give a mean species score for each
month. An index of total monthly flowering or fruiting is
calculated by summing monthly means for all species and
dividing by the number of species in the sample.
The second index, used only for fruits, considers
species differences in abundance and productivity and


126
and late afternoon. The Tana mangabeys' temporal pattern of
activities is more consistent with that observed for large
bodied, terrestrial primates. Several species of savanna
baboons display a less marked diurnal variation with
feeding, in particular, occupying a relatively constant
proportion of the daily activity budget (Papio anubis:
Aldrich-Blake et al. 1971, Nagel 1973, Harding 1976, Rose
1977; Papio cvnocephalus; Post 1981; Papio ursinus: Davidge
1978, Hall 1962; Theripithecus pelada; Crook and Aldrich-
Blake 1968, Dunbar and Dunbar 1974, Iwamoto 1975). This may
be due to the poorer quality diets of these species and the
need to feed more consistently throughout the day, and/or
their use of multiple diet items, each requiring different
handling and search times.
Although daily variability in activity patterns may be
influenced by climatic conditions such as temperature and
rainfall (Bernstein 1976, Marsh 1978b, 1981a) or "scheduled
activities" (sensu Altmann 1974) such as the necessity to
climb into sleeping trees before dark, seasonal variability
in activity budgets more likely is affected by changes in
food availability than climatic variation per se (Clutton-
Brock 1974). Post (1981) and Clutton-Brock (1974) found no
correlation between activity and climate for the yellow
baboon (Papio cvnocephalus) and red colobus (Colobus
badius), respectively and both implicated seasonal changes
in food resources. Although Marsh (1978b) found a


66
Million Cubic Meters
250 -i
Month
8
9


Figure 2.4.
Examples of randomly distributed Ficus spp.
a) Ficus natalensis in Mchelelo forest; b)
Ficus svcomorus in Mchelelo forest? c) Ficus
natalensis in Nkano forest.


Figure 3.7. Percent time 1988-89 and 1973-74 infants spent
in 3 levels of association with their mothers.
Asterisks denote months when levels of
association were significantly different
between the 2 study periods.
a) Association A: the amount of time an infant
was on its mother; b) Association B: the amount
of time an infant had its mother as nearest
neighbor; c) Association C: the amount of time
infants were separate from their mothers.


83
hrs, with the exception of the S. Mchelelo group. This group
frequently left the study site and was followed only while
it was present in the Mchelelo forest. The group was
observed for 3 consecutive days per month January 1988-April
1988 and November 1988; all other monthly samples consisted
of non-consecutive 3-day samples or 1 or 2 day samples. No
data were collected during May 1988 for either S. Mchelelo
or Nkano nor during August 1988 for S. Mchelelo. The total
number of contact hours for the 3 study groups was 595, 308
and 374 for N. Mchelelo, S. Mchelelo, and Nkano,
respectively.
I sampled groups every half hour using a 10-minute scan
(Altmann 1974, Robinson 1986), and recorded the following
data for each individual located during a sample: 1) age;
2) sex; 3) identity; 4) height above the ground; 5) first
activity sustained for 5 seconds; 6) species and item of
food if eating (ripe or unripe fruit, ripe or unripe seed,
flower bud or flower, shoots, stems and sprouts, leaves,
gum, bark or dead wood, and animal prey); and 7) age, sex,
and identity of the nearest neighbor within a 5 m radius.
I grouped activities into the same categories used by
Homewood (1976):
1) Eating. This activity included animals ingesting or
chewing food items. Animals scored as eating while moving
were classified as eating.


152
and basket production) could be devastating. Several of the
human uses of the palms clearly are not sustainable.
Individuals of P. reclinata with construction-size trunks
have been removed from unprotected forests. Due to the lack
of tall palms for the construction of ceremonial buildings,
village leaders have set aside one small forest for pole
production. Hyphaene compressa. another palm heavily
utilized for thatching has been severely depleted from
unprotected areas; people now travel more than 10 km to the
reserve to harvest leaves (pers. obs.).
By eliminating palms or decreasing fruit production,
palm harvesting degrades the habitat for the Tana mangabey
and other frugivorous birds and mammals that rely heavily on
the fruits and seeds of the palm. Given the continued
elimination of forest area and increased fragmentation of
forests due to agricultural expansion along the Tana River,
activities further degrading the habitat will not be
sustainable. Continued, but highly controlled, harvesting
outside the reserve and cessation of harvesting within the
reserve are necessary to provide benefits to both the humans
and the endangered Tana River crested mangabey.


121
diet composition (Wilks' Lambda=0.28, df=16,153, F=4.95,
p<0.0001). There were significant group differences in the
representation of fruit (1-way ANOVA, F=7.24, df=4,53,
p<0.001), seed (F=6.99, df=4,53, pcO.001) and invertebrates
(F=15.74, df=4,53, p<0.001) in the diet. Significant least
sguare means comparisons (p<0.05) showed that the 1988-89
groups consumed fewer fruits and greater numbers of seeds
and invertebrates than the 1973-74 groups (Figure 3.8).
Data on consumption of ripe versus unripe fruits and
seeds for one 1973-74 Mchelelo group suggested that the
1988-89 groups consumed less ripe and more unripe fruits and
seeds than the 1973-74 group (Figure 3.9). MANOVA showed a
significant effect of group on the amount of ripe and unripe
fruits and seeds in the diet (Wilks Lambda=0.68, F=3.25,
df=6,92, p<0.006). The 1973-74 group had a significantly
higher intake of ripe material (F=5.03, df=3,47, p<0.004)
and a significantly lower intake of unripe material (F=4.69,
df=3,47, p<0.006) than the 1988-89 groups. The differences
in the intake of ripe and unripe material were due primarily
to an increase in unripe seed consumption (F=5.06,
df=3,p<0.004) and, to a lesser degree, a decrease in ripe
fruit pulp consumption by the 1988-89 groups (Figure 3.9).
Measures of diet diversity were similar between the
1988-89 and 1973-74 Mchelelo groups. Mean monthly diet
diversity for the 1973-74 Mchelelo group (H'=1.90) did not
differ significantly from H' for either the 1988-89 N.


78
resource base may be shaping the present day feeding ecology
of the Tana mangabey.


ACKNOWLEDGMENTS
My research was funded by Wildlife Conservation
International, New York Zoological Society. I was sponsored
in Kenya by the National Museums of Kenya (NMK) Biological
Resources Program and the Institute of Primate Research
(IPR). The National Council for Science and Technology and
The Office of the President granted me permission to work
within Kenya (Permit No. OP. 13/001/160-282/12). While at
the University of Florida, I was supported by assistantships
from the Center of African Studies and Dr. John Eisenberg. I
warmly thank these people and organizations for the
opportunities they provided.
Many people have contributed, both knowingly and
unknowingly, to my study and to each one I owe a great debt.
Drs. John Eisenberg, John Robinson, Jack Putz, Michael
Collopy and Tom O'Shea, members of my doctoral committee,
helped shape my ideas both before and during the field work,
and have been constructive editors throughout the writing
phase. I extend special thanks to Dr. John Eisenberg, my
major professor, whose advice to remain ever "flexible and
patient" proved to be the most valuable and who, in his own
way, demanded my best.
iii


LIST OF TABLES
Table 2.1. Composition of monthly phenology sample by
species for Mchelelo, Nkano and Mnazini forests. 20
Table 2.2. Density of reproductive trees and lianas
enumerated in 16.32 ha of Mchelelo forest and
13.45 ha in Nkano forest, and the degree of
clumping of each species using Morisita's index
(I§) If I5 is significantly greater or less than
unity using the F-test, the species is considered
clumped or evenly distributed in space,
respectively. If I5 is not significantly
different from unity than the species is
considered randomly distributed 27
Table 2.3. Estimated canopy surface area (S), canopy
volume (V), mean fruit production (FN) and mean
fruit weight (FW) for 16 species. Species are
categorized into 4 groups: those that bear their
fruits on the canopy surface (C) or along the
branches of the tree (R), lianas (L) and palms (P)
54
Table 2.4. Spearman's rank correlation coefficients
for mean monthly environmental variables and mean
monthly fruiting scores for Mchelelo forest.
MINTEMP=minimum monthly temperature; RAIN=total
monthly rainfall (mm); RAINLAG1, RAINLAG2 AND
RAINLAG3=monthly rainfall lagged by 1,2, and 3
months, respectively; RAINADV=previous months'
rainfall; FL0W=mean monthly river flow (mcm);
FLOWLAG1, FLOWLAG2, FLOWLAG3=mean monthly river
flow lagged by 1,2 and 3 months, respectively;
FLOWADV=previous months' mean river flow 68
Table 2.5. Spearman's rank correlation coefficients
for mean monthly environmental variables and mean
monthly fruiting scores for Nkano forest.
Variables defined in Table 2.4 69
Table 2.6. Spearman's rank correlation coefficients
for mean monthly environmental variables and mean
monthly fruiting scores for Mnazini forest.
Variables defined in Table 2.4 70
x


102
Social Behavior. I examined social behaviors more
closely because the social/sexual category was a composite
of rare but important behaviors, and 1973-74 social behavior
data were available by age-sex class. I tested for
differences in the 1973-74 and 1988-89 annual means for 7
social behaviors available for 1973-74 adult males, adult
females, juvenile males, juvenile females, and male and
female infants combined. These behaviors included:
aggressing, aggressed, approaching, approached,
allogrooming, allogroomed and playing. Because there were no
significant interactions between groups and age-sex classes
for grooming or social/sexual behaviors in 1988-89 (ANOVA,
p>0.3), I combined data for my 3 groups and calculated
annual means for social behaviors for each age-sex class.
There were no significant differences in the
distribution of social behaviors between 1973-74 and 1988-89
for either adult males (G=3.5, df=6, p>0.05) or females
(G=5.9, df=6, p>0.05, Figure 3.5). Differences exist,
however, between the 2 studies for juvenile males (G=21.5,
df=7, p<0.005), who spent a greater than expected amount of
time playing in 1988-89 than 1973-74 (G=9.9, df=l, p<0.005)
and juvenile females (G=15.4, df=7, p<0.025), who engaged in
grooming more in 1973-74 than 1988-89 (Figure 3.6). There
also was significant heterogeneity in the distributions of
infant social behaviors between 1973-74 and 1988-89 (G=68.7,
df=5, p<0.001). Infants of 1973-74 appeared to be more


151
fruit sources, or both. Species that fruit throughout the
year, such as Ficus svcomorus. tend to dampen the effect of
seasonal troughs in fruit availability and provide mangabeys
with a more dependable resource base. In forests with a low
abundance of figs or other species providing a continuous
fruit resource, P. reclinata may be critical to the presence
or persistence of mangabeys.
Mangabeys appear to play a much greater role in seed
predation than seed dispersal. The occasional consumption
however, of whole ripe seeds and the rejection of seeds
after eating the sweet mesocarp of ripe fruits may aid P.
reclinata seed dispersal and germination. Mangabeys
frequently fill their cheek pouches with palm fruits before
traveling to another food source and occasionally drop seeds
as they move. Seeds extracted from the feces of mangabeys
showed 100% germination success indicating that mangabeys
can be effective seed dispersers.
Will human use of P. reclinata influence the long-term
persistence of the palm, thereby having an impact on both
people and endangered Tana mangabeys? It appears unlikely
that present levels of human exploitation would result in
the extirpation of this abundant palm; however, results
indicate that such use may alter the population structure
and influence reproduction. Although some of the human uses
of P. reclinata may be sustainable at present levels,
development of a local industry using palms (e.g., for mat


Figure 3.6. Mean percent time spent in 7 social behaviors
by three age-sex classes in 1973-74 and 1988-
89.
a) juvenile males; b) juvenile females; c) male
and female infants combined.


253
and palms are necessary elements of an action plan. The
second theme central to the recommendations is habitat
improvement. The establishment of habitat corridors between
forests by planting trees and encouraging natural
regeneration is strongly recommended; corridors effectively
create larger forest areas and encourage movement and gene
flow between isolated primate groups. Increasing forest area
within suitable habitat and enriching existing forests by
tree planting and protection against fire are important
measures for forest areas with primate groups that are far
removed from other forests, and may allow or encourage
recolonization of small forests that have lost all primate
groups.
Manipulation of mangabey groups is another management
consideration. Although translocation of mangabey groups
from "doomed forests" and adult male mangabeys between river
banks should decrease loss of groups and improve gene flow,
several problems must be considered: a) the logistics of
translocating endangered species are not trivial; b) the
probabilities of new males securing groups are not known and
may be low; and c) group take-overs may be associated with
infanticide (Kinnaird 1990a), thereby lowering infant
survival rates.
Kinnaird et al. (1990) state that "the unique biological
attributes of the TRNPR combined with its highly threatened
status, warrant national park status". National park status


107
involved in the overall social network of the group; they
were the recipients of aggression and were involved in
approaches significantly more than 1988-89 infants (G=3.7,
20.6 and 15.9 for aggressed, approach and approached,
respectively, df=l, p<0.05), whereas 1988-89 infants stayed
with their peer groups and played significantly more than
the 1973-74 infants (G=21.3, df=l, p<0.01) (Figure 3.6).
Mother/infant associations. I addressed the guestion of
whether the 1988-89 infants interacted less with group
members and were weaned at a later stage than the 1973-74
infants by examining the patterns of mother/infant
associations for the first 7 months of infancy. The
percentages of time an infant a) clung to or nursed its
mother, b) had its mother as its nearest neighbor, and c)
was separated from its mother were compared for 5 infants
from 1973-74 and 7 infants from 1988-89.
Infants were carried and suckled less, and associated
with individuals other than their mother more as they aged
(Figure 3.7, Table 3.4). Mother/infant associations and
infant development, however, differed between the 2 study
periods. The 1988-89 infants had their mothers as nearest
neighbors more and associated with other group members less,
particularly during the last 3 months of infancy, than 1973-
74 infants (Figure 3.7). There were significant differences
between studies in the amount of time infants spent clinging
or suckling as they aged (Table 3.4, study*age interaction).


179
were available) for the North Mchelelo, South Mchelelo, and
Nkano groups, respectively (Table 5.2). The months for which
correlations were not significant, however, were months in
which grasses occurred in the top 5 diet species and/or in
which invertebrate consumption was high (Tables 3.5 and
3.6).
Movements
Daily path length and patterns of movement. Mean daily
path length was 1138 m (range=588-1728 m, n=42 days), 1278 m
(range=779-1744 m, n=24 days), and 1141 m (range=646-2039 m,
n=34 days) for the North Mchelelo, South Mchelelo, and Nkano
groups, respectively. Rate of movement, or group speed,
measured as the mean half-hour step distance, was 48 m
(range=0-199 m) for North Mchelelo, 52 m (range=0-221 m) for
South Mchelelo, and 49 m (range=0-260 m) for Nkano group.
The distribution of distances moved in half hour periods
deviated significantly from expected Poisson distributions
for all 3 groups (X2=19.2, df=8, P<0.01; X2=28.1, df=8,
p<0.0004; X2=87.9, df=8, p<0.0001 for N. Mchelelo, S.
Mchelelo, and Nkano, respectively). There were more short
and more long movements than would be expected if group
speeds were random (Figure 5.8), indicating that groups
moved rapidly through some areas and spent more time in
others.
Half-hour turning angles indicated that all groups
tended to move ahead (< 90 from the direction moved the


Figure 2.17. Total monthly rainfall in millimeters.
a) Mchelelo forest; b) Nkano/Mnazini forest
area.


500
400
300
200
100
0
time (years)
Population size (N)


Figure 2.10. Mean monthly fruit scores for Diospyros
mespiliformes.
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.


201


Figure 2.6. Mean fruiting scores averaged over all species
by month.
a) Mchelelo forest; b) Nkano forest; c) Mnazini
forest.


137
potentially negative, indirect influences of harvesting on
other animal species are rarely considered. Many animal and
plant species harvested from tropical forests by humans are
important in the diets of other animals; harvesting
therefore may reduce or eliminate the availability of their
food resources.
Palms are among the most frequently harvested plant
species in the tropics (McCurrach 1960). Palms are
harvested for food, building materials, cloth, ornaments,
medicines and ritualistic purposes, and are of primary
importance to many South American, Asian and African
cultures (e.g., Moore 1973, Prance et al. 1987, Bates 1988,
Davis 1988, Peters et al. 1989). Palms also figure
prominently in the life cycles of a variety of insects and
are important food sources to rodents, bats, ungulates and
primates (see Uhl and Dransfield 1987 for review).
Overexploitation is one of the major threats to the survival
of many wild palm species (Uhl and Dransfield 1987).
Although increased use of palms may provide economic
benefits to rural people (Anderson 1988, Peters et al.
1989), some species, particularly high-potential market
species, are being depleted by destructive harvesting
(Vasquez and Gentry 1989).
Palms of the genus Phoenix are harvested throughout
their range in tropical Africa and Asia. The fruit of
Phoenix spp. have been important in Middle Eastern diets for


49
Ficus natalen8¡s
Fruit score
6-
4 -
3-
NDJFMAMJ JASONDJFMAM
8 8 8
7 8 9
Hyphaene compressa
Fruit score
6-
4 -
NDJFMAMJJASONDJFMAM
Oncoba spinosa
Fruit score
6 i
Ulliliiiiiilililii
NDJFMAMJJASONDJFMAM
8
7
8
8
Month
8
9


212
may be greater than for resources typically showing uniform
distributions. A ripe fig for example, offers a locally
superabundant source of fleshy fruits high in sugar and low
in secondary compounds (Waterman 1984). Fruits of more
abundant, uniformly distributed species (e.g., Phoenix
reclinata and Diospyros mespiliformes) are eaten primarily
unripe and may offer a less easily digested, lower quality
reward. Mangabeys also may be less likely to defend
uniformly distributed food resources because they impose
fewer constraints on available feeding sites and decrease
feeding competition. The large, continuous clusters of the
palm Phoenix reclinata afford numerous feeding sites for
mangabeys and were often sites of peaceful, intergroup
interactions. Fruit trees, on the other hand, often could
not contain an entire group of mangabeys.
Waser and Homewood (1979) predicted that Tana mangabeys
should show tendencies toward defense of space based on
their small home ranges, the relatively high productivity of
the Tana forests, and the fact that mangabeys often fed on
tree species that produce fruit continuously over periods of
many months. Field experiments involving long-call playbacks
did not support their expectations; Tana mangabeys showed no
indication of site-specific territorial defense (Waser and
Homewood 1976). My observations however, suggest that
response to vocal playbacks (territorial defense) should
vary according to resource availability and distribution and


9
declined through a loss in the total number of groups, but
the remaining groups showed no significant changes in size
(Decker and Kinnaird 1990).
Kinnaird et al. (1990) speculated that the extensive
loss of forest in the late 1960s led to a compression of the
primate populations and unusually high population densities.
In the subsequent 15-20 years these populations were
believed to have declined and stabilized at levels in
equilibrium with the new carrying capacity of the reduced
forest area. Loss of forest area likely was exacerbated by
reduction in habitat quality due to forest senescence
(Hughes 1985, Marsh 1986) and possibly, by severe drought
conditions in the early 1980s (Decker and Kinnaird 1990,
Marsh 1986).
The present study examines the effects of habitat change
on mangabey behavior and demography. The underlying theme of
the dissertation is comparative, using baseline data on
behavior and demography collected during a 1973-74 study by
Homewood (1976), the only other study to date of the Tana
mangabey. Comparison of my data with that of Homewood's
(1976) allowed me to address questions concerning the loss
of mangabey groups and effects of habitat change (e.g. loss
of forest area and reductions in important mangabey diet
species) on mangabey activity budgets, diet, ranging
behavior, and risk of extinction. Detailed analysis of
ecological and behavioral data collected during 1988-89 also


Mchelelo
Fruit score
Mnazlni
Fruit score
Month
8
9
Mchelelo
5 n
4 -
2 -
Fruit score
1 -
III..
i 1 r~
lilil
NDJFMAMJJASONDJFMAM
8 8
7 8
Mnazlni
Fruit score
8 8 8
7 8 9
Month


Table 6.4. Minimum and maximum parameter values and
estimates of effective population size. N =
census number of males or females; K = estimate of
lifetime reproductive success of males or females;
K, = variance in lifetime reproductive success;
and N = effective population size. Subscripts
denote parameters for males vs females
Table 6.5. Parameter values and estimates of the
genetic fixation index, F, for forests on east
and west banks and for west bank and east bank
subpopulations. East vs west bank figures are
calculated for group migration from one bank to
the other for two estimates of high flood
frequency
229
235
xiii


> 4%
3-4% 2-3% 1-2%
< 1%


Adult Male
Adult Female
100-
80-
60-
Aggress Aggr'd Approach Appr'd Groom Groomed Play


229
Table 6.4. Minimum and maximum parameter values and
estimates of effective population size for differing numbers
of male tenure. N = census number of males or females; K =
estimate of lifetime reproductive success of males or
females? K, = variance in lifetime reproductive success; and
Nc = effective population size. Subscripts denote parameters
for males vs females.
1973-74
1988-89
Parameter
Min Max
Min Max
N
1245
1511
683
725
Nm
47
57
32
35
N(
458
556
193
205
Kf
2.57
2.57
2.57
2.57
K*
1.84
1.84
1.84
1.84
Nc(
514
624
216
230
1
male
tenure
Km
7.41
7.41
4.45
4.45
K
44.13
44.13
15.88
15.88
Nem
28
34
20
22
Nc
107
129
74
80
2
male
tenures
Kra
11.85
11.8
7.13
7.13
K
56.37
56.37
20.41
20.41
Ncm
36
43
26
28
Ne
133
162
93
99
3
male
tenures
Km
15.75
15.75
9.47
9.47
K
60.32
60.32
21.82
21.82
Nera
40
48
28
31
Ne
148
179
99
108


72
significant positive correlations, and river flow lagged by
2 and 3 months showed significant negative correlations with
seasonal fruiting species excluding P. reclinata. Combined
P. reclinata scores for the 3 sites were significantly
correlated with rainfall lagged by 2 months and river flow
lagged by at least 1 month.
A multiple regression model incorporating minimum
temperatures and river flow lagged by 3 months, the 2
variables most highly correlated with fruit scores,
explained 25.7% of the variance in monthly fruit scores for
seasonal species excluding reclinata (F37=7.75, p=0.004).
Temperature explained 8.9% and river flow 16.7% of the total
variance.
Comparisons with 1973-74
Distributions of fruit scores differed significantly
between 1973-74 and 1988-89 for 5 of the 9 species examined.
Acacia robusta. Pachvstela msolo and Ficus sycomorus of
Mnazini and Diospyros mespiliformes and Ficus natalensis of
Mchelelo had significantly different distributions of fruit
scores between 1973-74 and 1988-89 (Kolmogrov-Smirnov J'= 7,
9, 11, 7 and 7 for A. robusta. P. msolo. F. sycomorus. D.
mespiliformes and F. natalensis. respectively, p<0.05).
Acacia robusta had an extended fruiting season in 1973-74
relative to 1988-89 and Pachvstela msolo had an extended
fruiting season in 1988-89 relative to 1973-74. Diospyros
mespiliformes. Ficus sycomorus and F_¡_ natalensis had more


Table 2.7. Spearman's rank correlation coefficients for mean monthly environmental
variables and mean monthly fruiting scores for all forests combined. Variables defined in
Table 2.4.
RIPE AND UNRIPE
RIPE ONLY
RIPE AND UNRIPE
RIPE AND UNRIPE
Total Continuous Seasonal
Total Continuous Seasonal
Seasonal
(u/o P. reclinata)
P. reclinata
*
MINTEMP
ns
ns
ns
ns
ns
0.39
0.53
ns
RAIN
ns
ns
ns
ns
ns
ns
ns
ns
RAINLAG1
ns
ns
ns
ns
0.37
ns
ns
ns
RAINLAG2
ns
ns
ns
ns
ns
ns
ns
ns
RAINLAG3
ns
ns
ns
ns
ns
ns
ns
0.46
RAINADV
ns
ns
ns
ns
ns
ns
ns
ns
FLOW
ns
ns
ns
ns
ns
ns
ns
ns *
FL0WLAG1
ns
ns
ns
ns
ns
ns
ns
FL0WLAG2
ns
ns
ns
ns
ns
ns
0.51
FL0WLAG3
ns
ns
-0.58
ns
ns
ns
-0.59
0.34
FLOWADV
ns
ns
ns
ns
ns
ns
ns
ns
*. P<0-05
p<0.001
-j
H


73
uniform distributions of fruiting in 1973-74 than in 1988-
89.
There were significant differences in monthly fruit
scores between 1973-74 and 1988-89 for 3 of the 7 species in
Mnazini and for 2 of the 6 species in Mchelelo. Ficus
svcomorus and Acacia robusta in Mnazini received
significantly higher mean monthly fruiting scores in 1973-74
than 1988-89 (T=2.94 and 2.75 for F\_ svcomorus and A.
robusta. respectively, p<0.05), whereas individuals of
Mimusops fruticosa received significantly higher fruiting
scores in 1988-89 than 1973-74 (T=2.28, p<0.05). The same
species in Mchelelo showed no significant differences in
fruiting scores. Diospyros mespiliformes and the palm,
Hyphaene compressa. had significantly higher mean fruiting
scores in 1988-89 than 1973-74 in Mchelelo (T=2.28 and 2.28
for D^_ mespiliformes and compressa. respectively,
p<0.05). Individuals of D^_ mespil if ormes in Mnazini showed
no significant differences in monthly fruiting scores
between the 2 time periods.
Discussion
Mangabey foods were highly variable in their temporal
abundance and spatial distribution. The majority of food
species examined were rare and spatially clumped. Rarity and
spatial aggregation are characteristic of many tropical tree
species (Hubbel and Foster 1986, Robinson 1986, Struhsaker
1975) and may be the result of chance or particular


191
Frequency of male long calls. The numbers of long calls
given by adult males of North and South Mchelelo groups were
relatively consistent among months and seasons (Table 5.4);
ANOVA showed no significant effect of month or season on
calling by North or South Mchelelo males, or both groups
combined (p>0.05). South group's presence in the Mchelelo
forest also had no effect on the frequency of calling by
North group males (ANOVA, p>0.05). Month but not season had
significant effects on numbers of long calls given by Nkano
group males (F=4.1, df=ll,22, p<0.002), males of the
neighboring, unhabituated group (F=8.64, df=ll,22,
pcO.OOOl), unidentified callers (F=3.73, df=ll,22, p<0.004),
and all calls combined (F=5.3, df=ll,22, p<0.0004). The
effect of month on calling, however, was due to extensive
counter-calling between several males during November, a
month of an attempted group takeover. When November was
eliminated from the analysis the differences are no longer
significant.
There were no significant correlations between long-
calls and monthly total fruit availability, unripe fruit
availability, or the percent of ripe and unripe fruit in the
diet for all 3 study groups (Spearman's Rank correlation,
p>0.05). Ripe fruit availability was significantly
correlated only with the number of long-calls given by North
group males (r=0.83, pcO.OOOl, n=14).


175
> 4% 3-4% 2-3%
1-2% < 1%


Figure 6.5. Effective population size (Ne) as a function of
number of breeders (Nb) and variance of male
lifetime reproductive success (VJ .


1200
1000
800
600
400
200
0
frequency
80
G\
4^


Figure 5.15. Percent time North and South Mchelelo groups
spent feeding on foods with different
distributions as a function of intergroup
behavior. a) Percent time feeding on foods
with uniform distributions; b) Percent time
feeding on foods with a patchy distributions.


Figure 3.3. Mean percent time spent in 4 major behaviors (eating, foraging, moving and
inactive) by time-of-day and season for the 1988-89 S. Mchelelo group.


226
Sm = probability of an infant male
surviving to reproduce.
The estimated fraction of members exchanged with adjacent
forests each generation (m) is simply M/N, where N is the
number of breeders present in a given deme (Allendorf 1983).
The rate of long-distance migration (m>) is taken to be of
the same order as mutation, i.e. lO"1 (Kimura and Weiss 1964,
Lande and Barrowclough 1987).
Demographic Extinction Model
I chose a demographic extinction model developed by
Goodman (1987b) and approximated by Belovsky (1987) that
calculates average persistence time for a population of a
given size with certain demographic parameters:
T
N
[(N 1)V r][(V+r/V-r)(1/r(V-r)( S
N
- l/r ( s l/y(yV-r))
Y=l
y=l
i/y2)
N N
- 2V/V-r( E l/y2) ( E l/[(y-l)V r] (yV-r) ]
y=l y=l
where T = expected persistence time;
r = mean per-capita population growth rate;
V = variance of r attributed to environmental
fluctuations;
N = maximum population size
Estimating model parameters. Per-capita population
growth rate (r) was calculated for the 3 main study groups


Figure 3.8. Percent composition of diet by item for
1973-74 and 1988-89. a) 1988-89; b) 1973-74 . 119
Figure 3.9. Percent ripe and unripe fruit and seed
items inthe diet, a) for three 1988-89 mangabey
groups; b) for 1973-74 mangabey groups 123
Figure 4.1. The number of harvested and unharvested
palms in Nkano and the number of reproductive and
nonreproductive palms in Mchelelo by size class.
a) harvested and unharvested palms by size class
in Nkano; b) reproductive and nonreproductive
palms in Mchelelo by size class 145
Figure 4.2. The number of feeding observations on P.
reclinata by month and item for mangabeys in the
Mchelelo and Nkano forests 148
Figure 5.1. Cumulative frequency of sightings of one
or more North Mchelelo group mangabeys versus rank
order of 0.25 h quadrats. Quadrats were ranked
with respect to the number of sightings occurring
in each. The areas accounting for 50 and 100% of
all sightings are designated 162
Figure 5.2. Cumulative frequency of sightings of one
or more South Mchelelo group mangabeys versus rank
order of 0.25 h quadrats. Quadrats were ranked
with respect to the number of sightings occurring
in each. The areas accounting for 50 and 100% of
all sightings are designated 164
Figure 5.3. Cumulative frequency of sightings of one
or more Nkano group mangabeys versus rank order of
0.25 ha quadrats. Quadrats were ranked with
respect to the number of sightings occurring in
each. The areas accounting for 50 and 100% of all
sightings are designated 166
Figure 5.4. Taut-string line (shaded with diagonal
lines) enclosing all ad libitum sightings of South
Mchelelo group when the group ranged outside the
Mchelelo forest 168
Figure 5.5. Frequency distributions of intensity of
quadrat use. Asterisks indicate the values
expected if quadrat occupancy were random
(following a Poisson distribution).a) Frequency
distribution for N. Mchelelo group; b) Frequency
distribution for S. Mchelelo group; c) Frequency
distribution for Nkano group 171
xvii


The Tana mangabey's behavioral and dietary flexibility
enables it to buffer fluctuations in resource availability
in a highly seasonal habitat and to adapt to large-scale
habitat change. Mangabeys forage and eat more, and show
extended infant dependency in response to changes in fruit
distributions and lower fruit availability. A dietary shift
between 1973-74 and 1988-89 from primarily ripe fruits to
unripe seeds and invertebrates may constitute the best
alternative strategy when ripe fruits are less available.
Mangabey ranging patterns are influenced by a complex
interaction of the spatial and temporal distribution of
foods, forest patch size, group size, and intergroup
interactions. Seasonal variation in food availability and
distribution result in facultative defense of space by
mangabeys.
The Tana mangabey's endangered status is primarily the
result of the highly restricted distribution of Tana forest
habitat and the rapid destruction of this habitat.
Degradation of the remaining forests may result from non-
sustainable human exploitation of forest species such as the
palm, Phoenix reclinata. Results of demographic and genetic
models suggest that the mangabey is at risk of losing
genetic variation and has a high probability of extinction
over the next 50-100 years given its present population
status. A conservation strategy should incorporate measures
for habitat preservation and enrichment. Because the
xxi


Figure 2.16. Monthly mean minimum and mean maximum
air temperature in degrees centigrade,
a) Mchelelo forest; b) Nkano/Mnazini forest
areas.


242
100 years, a mean persistence time of 2000 years, or a
population of nearly 8000 individuals, is necessary.
Discussion
Effective Population Size
Estimates of effective population size for the Tana
mangabey varied widely (up to 64%) as a function of the
initial size of the breeder population and the values used
for male lifetime reproductive success. Chepko-Sade et al.
(1987) state that for mammals, reproductive variance has the
greatest effect on Ne; Ne declines with increasing
reproductive variance. This is true, however, only when mean
reproductive success is constant. For the mangabey, the
initial number of breeders had as great an effect on Ne as
the variance in male reproductive success (Figure 6.5). Ne
increased with increasing male variance in progeny
production. As male mangabeys were given more opportunities
to breed by increasing numbers of tenures, mean male
reproductive success increased with the variance. Thus, it
is the ratio of the mean to the variance that appears to be
influencing Ne for mangabeys rather than variance alone.
This may be true for primates in general or populations of
other long-lived, polygamous mammals.
The least conservative estimates of Ne (from 99-108)
suggest that the Tana mangabey is at risk of losing genetic
variation. Values of N,. calculated with the highest


74
microhabitat requirements (Hubbel and Foster 1986). Hughes
(1985) showed that several species occurring in the Tana
River forests have very particular soil and water
requirements for germination and growth. Because the
frequency and depth of flooding, soil type and soil moisture
holding capacity vary widely between forests and often
within a single forest stand, different vegetative
assemblages occur among forests and certain plant species
may occur only in particular areas within a forest (Hughes
1985, Marsh 1978a, Homewood 1976).
The spatial distribution of mangabey food species
should affect ranging behavior throughout the year, but
because of the small size of the forests and the mangabeys'
ability to traverse the entire area, spatial distribution
should not affect whether the fruits of a particular tree
species will be included in the diet. The relationship
between plant abundance and temporal availability of fruits
likely is more important. The abundance of a particular
species limits the absolute availability of fruits to
mangabeys; as plants become rarer, total fruit availability
decreases and the amount of time when no fruits are
available to mangabeys may increase.
Mangabey fruit resources show periods of relative
abundance (November-May) and relative scarcity (June-
October). The abundance of continuous, asynchronously
fruiting species (e.g. Ficus sycomorus and Ficus


106
Juvenile Male
Juvenile Female
100 i
80 -
Aooress Aflgr'd Approach Appro Groom Groomed Play


154
Although many primates maintain stable home ranges and
patterns of intergroup relations throughout the year (Mitani
and Rodman 1979), some studies have shown extreme
variability in ranging and intergroup interactions for a
species in different habitats and/or populations (e.g.,
Cercopithecus aethiops. Chapman and Fedigan 1984, Kavanagh
1981; Cercopithecus mitis. Butynski 1990; Colobus badius
Marsh 1981b, Struhsaker and Leland 1979; Colobus quereza.
Oates 1977; Macaca svlvanus. Mehlman 1989; Papio ursinus.
Hamilton et al. 1976; Presbvtis entellus. Yoshiba 1968;
Presbvtis melalophos. Bennett 1986). These studies document
differing strategies between groups or populations of a
species. Few data exist, however, on variability of defense
of space within groups of a single population.
Homewood (1976) and Waser and Homewood (1979) noted for
the Tana River crested mangabey (Cercocebus qaleritus
galeritus) that any generalizations about defense were
precluded by the high intragroup variability in defense. The
extent of variation in mangabey intergroup behaviors is
striking, ranging from descriptions of intergroup fighting
to peaceful, non-aggressive encounters in which groups
sometimes merge. The Tana mangabey, therefore, is an ideal
subject for investigating factors influencing use and
defense of space.
In this chapter, I quantify and examine variation in use
and defense of space by the Tana River crested mangabey.


22
provides an indication of monthly fruit biomass production.
The index incorporates estimated numbers of fruits produced
by a species during maximum fruit production, the estimated
weight of fruits or fruit parts eaten by mangabeys, and the
abundance of reproductive individuals. Fruit biomass indices
were calculated for 4 categories of species, depending on
how fruits were distributed throughout the canopy. Species
bearing fruits primarily on the outer surface of the canopy
(n=ll) were categorized as surface fruiters. Species bearing
fruits along their branches, either in clusters or evenly,
were classified as ramiflorous fruiters (n=4). Palms (n=2)
and lianas (n=l) each were classified separately because
they do not have identifiable canopies.
I counted fruits only when individuals received the
maximum phenological score of 5; this limited my sample size
for some species because certain individuals were never
scored as a 5. I estimated the number of fruits for all
categories except the palms by counting the number of fruits
visible within an approximately 1 m3 field of vision of
binoculars. One count in the top, middle and lower sections
of the canopy or liana "tangle" was made for each individual
and an average was taken. Fruits of the 2 palm species were
counted directly for a minimum of 3 infructescenses. The
numbers of fruits for the 3 infructescenses were averaged
and multiplied by the total number of infrutescenses
present. For Hyphaene compressa. a palm that branches and


27
Table 2.2. Density of reproductive trees and lianas
enumerated in 16.32 ha of Mchelelo forest and 13.45 ha in
Nkano forest, and the degree of clumping of each species
using Morisita's index (I§) If I§ is significantly
greater or less than unity using the F-test, the species is
considered clumped or evenly distributed in space,
respectively. If I§ is not significantly different from
unity than the species is considered randomly
distributed.
Mchelelo
Nkano
Taxon
Code
Density
'§
F-test (p)
Density
1
:-Test (p)
ALANGIACEAE
Alanqium salvifolium
AS
33.56
2.71
<0.001
112.26
1.27
<0.001
ANACARDIACEAE
Lannea stuhlmannii
IS
0.55
1.81
ns
0.30
27.00
<0.001
Sorindeia madaqascariens
is SO
6.74
2.16
<0.001
0.15
54.00
<0.001
APOCYNACEAE
Rauvolfia mombasiana
RM
52.01
3.68
<0.001
.
.
Saba comorensis
SF
0.86
3.57
<0.005
6.77
1.31
<0.005
BORAGINACEAE
Cordia qoetzei
CG
4.59
1.52
<0.001
0.00
CAESALPINACEAE
Cynometra webberi
CW
0.43
9.29
<0.001
0.00
Tamarindus indica
TI
0.49
4.64
<0.001
0.00
-
-
EBENACEAE
Diospvros mesoiliformis
DM
7.72
1.51
<0.001
1.19
1.35
ns
FLACOURTIACEAE
Oncoba spinosa
OS
6.49
2.67
<0.001
0.45
14.40
<0.001
GUTTIFERAE
Garcinia livinqstonei
GL
1.72
8.60
<0.001
0.37
5.40
ns
M1M0SACEAE
Acacia robusta
AR
0.67
3.55
<0.001
0.00
Albizia qummifera
AG
1.65
2.78
<0.001
0.00
-
-
Albizia qlaberrima
AL
0.55
3.61
ns
0.00
-
.
MORACEAE
Ficus bubu
FU
0.06
_
_
0.89
2.45
ns
Ficus bussei
FB
0.06
-
-
0.00
-
-
Ficus nata lens is
FN
0.55
0.00
ns
0.45
0.00
ns
Ficus svcomorus
FS
1.16
1.14
ns
4.68
3.98
<0.001
PALMAE
Hyphaene comoressa
HC
17.45
2.48
<0.001
0.00
Phoenix reclinata
PR
193.57
1.39
<0.001
101.64
1.90
<0.001
SAPINDACEAE
ADorrhiza paniculata
AP
1.64
4.94
<0.001
8.70
1.90
<0.001
Bliqhia uniiuqata
BU
0.86
17.86
<0.001
0.15
0.00
ns
SAPOTACEAE
Mimusoos frut i cosa
MF
0.86
1.43
ns
0.00
Pachvstela msolo
PB
0.31
0.00
ns
35.99
1.35
<0.001
STERCULIACEAE
Sterculia appendiculata
SA
0.89
5.91
<0.001
0.07
-
-


TABLE OF CONTENTS
ACKNOWLEDGMENTS iii
LIST OF TABLES X
LIST OF FIGURES xiv
ABSTRACT XX
CHAPTER ONE INTRODUCTION 1
CHAPTER TWO ABUNDANCE AND DISTRIBUTION OF FRUIT
RESOURCES IN THREE TANA RIVERINE FORESTS 12
Introduction 12
Methods 14
Study Sites 14
Abundance and Distribution of Resources ... 17
Environmental Correlates with Flowering and
Fruiting 24
Comparisons with 1973-74 25
Results 26
Abundance and Spatial Distribution of
Resources 26
Temporal Distribution of Resources 27
Environmental Correlates with Flowering and
Fruiting 60
Comparisons with 1973-74 72
Discussion 73
CHAPTER THREE ACTIVITY PATTERNS AND DIET OF THE TANA RIVER
CRESTED MANGABEY: EFFECTS OF SEASONAL AND LONG-TERM
HABITAT CHANGE 79
Introduction 79
Methods 82
Data Collection 82
Data Analysis 85
Results 86
Variation in Activities 1988-89 86
Comparison with 1973-74 Activity Budgets . 101
Diet Ill
Discussion 124
viii


Table 2.6. Spearman's rank correlation coefficients for mean monthly environmental
variables and mean monthly fruiting scores for Mnazini forest. Variables defined in Table
2.4.
RIPE AND UNRIPE
RIPE ONLY
RIPE AND UNRIPE
RIPE AND UNRIPE
Total Continuous Seasonal
Total Continuous Seasonal
Seasonal
(w/o P. reclinata)
P. reclinata
MINTEMP
0.58*
ns
0.58*
**
0.72
0.60*
0.68*
0.58*
**
0.75
RAIN
ns
ns
ns
ns
ns*
ns
ns
ns
RAINLAG1
ns
ns
ns
ns
0.60
ns
ns
ns
RAINLAG2
ns
ns
ns
ns
ns
ns
ns
ns
RAINLAG3
ns
ns
ns
ns
ns
ns
ns
ns
RAINADV
ns
ns ##
ns
ns
ns *
ns
ns
ns
FLOW
ns
0.78
ns
ns
0.57
ns
ns
ns.
FL0ULAG1
ns
ns
ns
0.59
ns
ns
ns
FL0WLAG2
FL0ULAG3
ns
ns
ns
ns
ns #
-0.59
ns *
-0.63
ns
ns
ns
ns
ns **
-0.71
0.79
FLOWADV
ns
ns
ns
ns
ns
ns
ns
ns
I. P<0-05
p<0.001
*
o


228
computed for 2 time periods, 15 years apart, or
approximately 2 mangabey generations (Table 6.4). Minimum
and maximum population estimates (N) reflect the population
decline over the 15-year period. Values of Nm and Nt for the
2 periods indicate little change in the breeding sex ratio.
Because males had a greater number of breeding females
available in 1973-74, separate estimates of lifetime
reproductive success were calculated for the 2 periods
(Figures 6.1 and 6.2). Male reproductive success is
illustrated for 1973-74 and 1988-89 for three simulations
where a male attains tenure once, twice or three times
during his reproductive lifespan. Both the mean and variance
of male reproductive success increased when a male was given
the opportunity to secure breeding status in a 2nd and 3rd
group. Relative to the mean however, variance decreased with
increasing reproductive success. Variance in progeny number
for males is higher than for females in all situations, as
would be expected for polygamous species (Wade and Arnold
1980), and always exceeds the mean. Variance in female
infant production however, was lower than the mean and
resulted in effective numbers of females (Nc() being slightly
greater than the census numbers (N,) .
Using Eg. 3 for fluctuating population size, I arrive
at a final range of effective population sizes of 87 to 135
mangabeys. This range of estimates does not incorporate the
effects of overlapping generations. Models for overlapping


Figure 5.13. Daily travel paths by North and South Mchelelo
groups during one day of a month of intergroup
fighting. Timing and location of adult male
long-calls are plotted for North (ovals) and
South (rectangles) groups.


218
Table 6.2. Age-sex composition of 7 mangabey groups with
comparative means from 1973-74 (Homewood 1976). AM = adult
male (>4-5 yr); SubM = subadult male (3-5 yr); AF = adult
female (>4 yr); SubF = subadult female (2.5-3 yr); JM =
juvenile male (8 mon-3 yr); JF = juvenile female (8 mon-2.5
yr); IM and IF = infant male and infant female, respectively
(< 7 mon).
Forest
AM
SubM
AF
SubF
JM JF
IM"
IF
Total
Mchelelo
1
0
6
0
5 3
1
1
18
Congolani
C
2
1
6
2
7 6
3
1
28
Kano'
2
0
4
0
3 4
3
1
17
Mnazini
2
0
4
0
3
2
15
Mnazini N
2
2
1
7
0
11
3
24
Guru
2
1
5
0
6
4
18
Baomo N
2
1
8
0
12
2
25
Mean
1.85
0.57
5.85
0.29
8.57
3
. 00
20.7
1973-74
Mean
3.25
3.50
9
. 75
6.00
2
.50
26.5
(n=4 groups)
'main study groups
4 infants not included died before 6 months of age; 3 were
believed to be males, one was not sexed.


128
between 1973-74 and 1988-89 activity budgets may not be due
to sampling bias alone. Intergroup variability was low or
insignificant in both the 1973-74 and 1988-89 studies, and
differences between the 2 studies were highly significant.
Inter-study differences may have been the result of changes
in the forest habitats over the 15-year period or
differences in behavioral scoring by Homewood and myself.
Although the latter possibility was not tested and may play
a role in the observed behavioral changes, inter-study
differences in behaviors that were less likely to be
confused or scored differently by separate observers (e.g.,
resting, grooming, and suckling) were significant and
increases confidence in the comparisons.
The 1988-89 mangabey groups spent significantly more
time eating and searching for food and less time grooming
and inactive than the 1973-74 groups. These results are
similar to those reported for Japanese monkeys (Macaca
fuscata) in habitats of deteriorating quality (Nakagawa
1989) and are consistent with 1988-89 trends between
activity budgets and variation in habitat quality due to
seasonal and forest differences.
Increased search time in 1988-89 was expressed through
increased time spent foraging, but not through increased
movement (with the exception of the S. Mchelelo group that
underwent long movements between 2 forest patches). Based on
optimal foraging theory (Charnov 1976), mangabeys should


Figure 2.1. Map of the study area showing the Mchelelo,
Mnazini, and Nkano forests within the Tana
River National Primate Reserve.


Figure 6.2. Distribution of male lifetime reproductive
success estimated for males with 1, 2 and 3
tenures as a dominant, breeding group male,
a) 1973-74 male lifetime reproductive success
calculated for 1 male:9.75 females; b) 1988-89
male lifetime reproductive success calculated
for 1 male:5.85 females.


112
plant species over 45, 28, and 34 observation days,
respectively (Appendix B). Although the number of plant
species eaten varied monthly (x=16.8 + 3.76 for N. Mchelelo,
x=12.66 + 3.44 for S. Mchelelo, and x=16.25 + 3.19 for
Nkano), the top 5 diet plant species for each month
constituted 72 to 100% of the feeding observations (Table
3.5). Thirteen plant species accounted for 59 of 75 (79%)
potential top rankings in N. Mchelelo, 48 of 60 (80%)
potential top rankings in S. Mchelelo, and 58 of 60 (97%)
potential top rankings in Nkano (number of potential top
rankings = 5 top rank positions number of months in
sample). Phoenix reclinata ranked in the top 5 species every
month for at least 1 of the Mchelelo groups and in 9 of the
12 months sampled for Nkano. Either Ficus svcomorus or Saba
comorensis ranked in the top 5 species of Nkano for 11 of
the 12 months sampled. Thus, although mangabeys exploited a
wide variety of foods, they concentrated feeding activity on
a small number of plant species.
When the use of food items regardless of species was
considered, fruit and seed contributed the most to the 1988-
89 monthly diets (Tables 3.6 and 3.7, Figure 3.8). The
majority of seeds eaten by each of the 1988-89 groups were
unripe while fruits were consumed primarily in the ripe
stage (Tables 3.6 and 3.7, Figure 3.8).


202
The outcome of intergroup fights was not determined by
group size and was not independent of location. Thirty-four
of 40 fights were observed from beginning to end and the
winner of 17 fights was determined. The smaller North group
(n=18 individuals) supplanted South group (n=28 individuals)
in 11 of 17 (65%) fights; South group won the remaining 6
(35%) fights. North group won all fights in the northern
sector of the forest and South group won all fights in the
southern sector of the forest (Figure 5.14). Of the 17
fights in which no clear winner was identified (both groups
withdrew), 14 occurred within a 100 m strip in the middle of
the forest. Of the total recorded fights, 63% (n=25)
occurred within 25 m of an isolated, seasonally fruiting
tree or clusters of fruiting dhoum palms (Hyphaene
compressa).
Monthly intergroup behaviors did not affect the
frequency of long-calls given by either group males or all
calls combined (ANOVA, p>0.05); groups called as often
during avoid months as they did during either fight or merge
months. Group response to long calls, as noted above,
appeared to differ among avoid, merge, and fight months. I
speculated that response differences were due to the
availability and dispersion of fruit resources. I therefore
tested for differences in fruit availability and percent
time spent eating patchy and uniform foods among the 3
intergroup behavior categories.


60
The predominance of unripe fruits may be the result of the
time necessary to ripen a fruit crop and/or asynchronous
ripening within individuals, and fruits being selectively
chosen by animals as soon as they ripen. Nkano and Mnazini
generally had larger biomasses of ripe fruits relative to
Mchelelo, possibly due to the higher density of figs in
Nkano and species such as Mimusops fruticosa and Garcinia
livingstonei in Mnazini, that tend to ripen more
synchronously.
Environmental Correlates with Flowering and Fruiting
Temperature, rainfall and river flow showed strong
seasonality. The hottest months occurred between November
and April and the coolest months occurred between June and
October (Figure 2.16). Rainfall was high during November and
December and again from March-June (Figure 2.17). Rainfall
in Nkano generally was higher than Mchelelo although there
was a strong correlation between monthly totals for the 2
areas (r=0.85, n=ll, p<0.05). River flow was highest April-
June, 1988 and January-April, 1989 (Figure 2.18), and was
correlated with the previous months rainfall in both areas
(r=0.72 and 0.76 for Nkano and Mchelelo, respectively, n=ll,
p<0.05).
Flowering was negatively correlated with rainfall in
Nkano (r=-0.57, n=14, p<0.05) and with river flow in
Mchelelo (r=-0.51, n=16, p<0.05) and Mnazini (r=-0.77, n=10,
p<0.05). When all 3 forests are combined, there were


134
Homewood (1978) described the Tana mangabey as a highly
adaptable, flexible primate that could adjust its behavior
to buffer fluctuations in resource availability in a highly
seasonal habitat. My results corroborate this and further
suggest that the mangabeys' behavioral and dietary
flexibility has enabled it to adapt to large-scale habitat
change. Such adaptability may enhance the mangabeys' chances
for long-term persistence. Both the mangabey and the
sympatric Tana River red colobus monkey (Colobus badius
rufomitratus) experienced population declines between 1976
and 1985. The mangabey, although initially the rarer
species, declined less and is now found in greater numbers
(680-725 individuals, Chapter 6) than the more specialized,
folivorous red colobus (200-300 individuals, Decker 1989).
The mangabeys' greater resilience may have been due to its
more opportunistic diet and greater ability to adjust to
habitat change than the red colobus (Decker and Kinnaird
1990, Marsh 1986).
Why should such an adaptable monkey be so rare and
endangered? The rarity of all 4 subspecies of Cercocebus
galeritus may be attributed, in part, to habitat
requirements. Homewood (1978) speculated that because of
their semi-terrestriality and flexible foraging and social
behaviors, C. galeritus subspecies are well suited to
exploit temporary superabundances of fruit, dense sub
canopies and frequently flooded terrain. She speculated


273
Robinson, J.G. 1986. Seasonal variation in use of time and
space by the Wedge-capped Capuchin monkey, Cebus
olivaceus: Implications for foraging theory.
Smithsonian Contributions to Zoology 431.
Rose, M.D. 1977. Positional behaviour of olive baboons
(Papio anubis) and its relationship to maintenance and
social activities. Primates 18:59-116.
Ryman, N., R. Baccus, C. Reuterwall, and M.H. Smith. 1981.
Effective population size, generation interval, and
potential loss of genetic variability in game species
under different hunting regimes. Oikos 36:257-266.
van Schaik, C.P. 1986. Phenological changes in a Sumatran
rain forest. Journal of Tropical Ecology 2:327-347.
van Schaik, C.P., M.A. van Noordwijk, R.J. de Boer, and I.
den Tondelaar. 1983. The effect of group size on time
budgets and social behaviour in wild long-tailed
macaques (Macaca fascicularis). Behavioral Ecology and
Sociobiology 13:173-181.
SAS. 1985. SAS User's Guide. SAS Institute Incorporated,
Cary, NC.
Shaffer, M.L. 1981. Minimum population sizes for species
conservation. Bioscience 31:131-134.
Shaffer, M.L. and F.B. Samson. 1985. Population size and
extinction: a note on determining critical population
sizes. American Naturalist 125:144-152.
Siegel, S. 1956. Nonparametric Statistics for the
Behavioral Sciences. McGraw-Hill Book Company, New
York, NY.
Silk, J.B. and H.C. Kraemer. 1978. Comparison of mother-
infant proximity among wild and captive chimpanzees.
Pages 115-118 in D.J. Chivers and J. Herbert, editors.
Recent Advances in Primatology, 1. Academic Press, New
York, NY.
Slatkin, M. 1985. Gene flow in natural populations. Ann.
Rev. Ecol. Syst. 16:393-430.
Smythe, N. 1978. The natural history of the Central
American agouti (Dasyprocta punctata). Smithsonian
Contributions to Zoology 257.
Sokal, R.R. and F.J. Rohlf. 1981. Biometry. Freeman
Press, San Francisco, CA.


To Ratilli Kawa
and the women of rural East Africa


Figure 5.1. Cumulative frequency of sightings of one or more North Mchelelo group
mangabeys versus rank order of 0.25 ha quadrats. Quadrats were ranked with
respect to the number of sightings occurring in each. The areas accounting
for 50 and 100% of all sightings are designated.


Figure 2.8. Mean monthly fruit scores from Mchelelo and Mnazini forests for Garcinia
livinqstonei and Sorindeia madaqascariensis. seasonally fruiting species
with 2 fruiting peaks per year.
a) Garcinia livinqstonei in Mchelelo forest; b) Sorindeia madaqascariensis
in Mchelelo forest; c) Garcinia livinqstonei in Mnazini forest; d) Sorindeia
madaqascariensis in Mnazini forest.


Figure 3.9. Percent ripe and unripe fruit and seed items in
the diet.
a) for three 1988-89 mangabey groups; b) for 2
1973-74 mangabey groups.


127
correlation between activity and daily temperature for the
Tana River red colobus (Colobus badius rufomitratus),
seasonal changes in food availability also influenced
activity. Tana mangabeys expended more of their daily
activity budgets foraging during drier seasons when food
resources were less available and more widely dispersed
(Chapter 2). Conversely, when food resources were more
readily available, particularly during the seasonal short-
rains, mangabeys spent more time in social/sexual
activities.
Group differences in activity budgets also may be
related to gross habitat differences. The Nkano group was
less active, particularly during the inter-rains, than the
Mchelelo groups. The Nkano group, although ranging in a less
diverse forest than the Mchelelo groups, had access to a
potentially richer resource base that was more evenly
distributed throughout the year (Chapter 2). The richer,
more reliable resource base may have enabled the Nkano group
to allocate more time to inactivity. Group size also may
have played an important role in the distribution of
activities. The N. Mchelelo and Nkano groups were similar in
size and activity budgets. The larger S. Mchelelo group, in
general, ate less, moved more, and was less active than the
2 smaller groups.
Despite the complexities of inter-study comparisons
(Altmann 1974, Clutton-Brock 1977), the strong contrast.


Table 3.5. Rank order of the five most common plant species eaten and the percentage
the total feeding records for identified plants (N) by group and month. Complete
scientific names of plant species listed in Appendix B.
a) N. Mchelelo
Month
I
II
III
IV
V
Species
X
Species
X
Species
X
Species
X
Species
X
Jan 88
H. comoressa
35
F. sycamorus
16
A. salviifol inn
13
F. bubu
10
Grasses
8
Feb
H. comoressa
34
P. reclinata
32
D. mesoitiformes
8
0. spinosa
7
M. fruticosa
7
Mar
P. reclinata
35
S. madaqascartensis
14
F. sycamorus
11
M. fruticosa
11
Grasses
9
Apr
P. reclinata
59
0. spinosa
10
S. madaqascariensis
6
D. mespiliformes
5
Mushrooms
5
May
P. reclinata
61
0. spinosa
12
S. comorensis
6
H. comoressa
5
A. venosum
4
Jun
P. reclinata
71
S. comorensis
8
0. spinosa
6
F. natalensis
5
D. mesoitiformes
3
Jut
0. spinosa
18
D. mespiliformes
14
Mushrooms
14
P. reclinata
13
T. indica
11
Aug
P. reclinata
42
T. indica
18
H. comoressa
11
D. mespiliformes
6
C. lukei
5
Sep
P. reclinata
43
T. indica
11
D. mespiliformes
11
H. comoressa
8
C. lukei
8
Oct
P. reclinata
45
D. mespiliformes
9
0. spinosa
7
H. comoressa
6
S. comorensis
6
Nov
A. paniculata
53
H. comoressa
19
P. reclinata
6
0. spinosa
4
L. schweinfurthii
4
Dec
0. spinosa
27
L. schweinfurthii
26
F. sycamorus
11
Mushrooms
9
F. natalensis
7
Jan 89
P. reclinata
17
A. salviifoliun
17
0. spinosa
11
H. zevlanica
9
P. msolo
8
Feb
F. natalensis
33
A. robusta
23
H. comoressa
13
A. salviifolinn
7
D. ferrea
6
Mar
P. reclinata
59
H. comoressa
14
H. abyssinica
5
S. comorensis
4
C. rotundifolia
4
Of
N
134
178
207
218
172
217
168
243
196
164
254
229
241
273
219


Table 2.1. Composition of mothly phenology sample by
species for Mchelelo, Nkano and Mnazini forests.
20
Number of Individuals
Taxon Mchelelo Nkano Mnazini
ALANGIACEAE
Alangium salvifolium
ANACARDIACEAE
Sorindeia madaqascariensis
APOCYNACEAE
Saba comorensis
EBENACEAE
Diospyros mespiliformis
FLACOURTIACEAE
Oncoba spinosa
GUTTIFERAE
Garcinia livinqstonei
MIMOSACEAE
Acacia robusta
Albizia gummifera
MORACEAE
Ficus bubu
Ficus bussei
Ficus natalensis
Ficus svcomorus
PALMAE
Hvphaene compressa
Phoenix reclinata
SAPOTACEAE
Mimusops fruticosa
Pachvstela msolo
10 5 -
10-8
7 3 8
13 5 9
10
9-10
10 10
10
13 0
1
6 7
10 8 7
11
10 7 3
8 10
- 10


1400
1200
1000
800
600
400
200
0
Cumulative frequency
10 20 30 40 50 60 70 80
Rank order of quadrats
t->
cn
CTi


86
independent variables (e.g., group, season, time-of-day) on
correlated dependent variables (Sokal and Rohlf 1981). If
MANOVA showed significant effect of independent variables on
the suite of dependent variables, I then used ANOVA to
examine the effects of these independent variables on each
dependent variable. An arcsin sguare-root transformation was
used on proportional data (Sokal and Rohlf 1981). I
calculated significance values from partial sums of squares
(Type III), which adjust each effect for all other effects
in the model and are unrelated to cell frequencies (Freund
and Littell 1981). When analysis of variance indicated
significant differences due to an independent variable, I
used least square means which adjust for unbalanced data, to
determine where differences lay among the independent
variables.
I used the same analysis of variance methods to compare
activity budgets and diet composition between 1988-89 and
1973-74. I used G-tests (Sokal and Rohlf 1981) to test for
differences in the distributions of social activities
between 1988-89 and 1973-74 age-sex classes because repeated
monthly samples were not available for the 1973-74 groups
and only annual means were available.
Results
Variation in Activities 1988/1989
Mangabeys from all groups spent the majority of their
time (58-64%) engaged in feeding behaviors (defined as


209
similar to C. g. galeritus (Horn 1987); the study groups
however, were unhabituated and the data were not collected
systematically. The inconsistency between daily path length
and home range size suggests that the Tana mangabey must
satisfy requirements similar to other closely related
species within a much smaller home range.
Presence of neighboring groups further restricts home
range size, use and movement patterns. South Mchelelo group
strongly influenced North Mchelelo group's ranging patterns.
High rates of backtracking and frequency of long-calls in
Nkano forest also indicate that neighboring groups may have
restricted Nkano group's movements and home range; despite
the larger size of Nkano forest (35 versus 17 ha), the Nkano
group's home range was similar in size to North Mchelelo
group.
Group size also may affect home range and movements.
Waser (1977) showed for Cercocebus albigena that an increase
in group size over a certain threshold significantly
increased daily path length, and potentially annual home
range. South Mchelelo group, the largest of the 3 groups,
moved between 2 forests and occupied the largest home range.
The relative importance of group size and presence of
neighboring groups, however, cannot be determined without
additional data on more groups.
Several primate species exhibit facultative
territoriality, expressing territoriality only under


236
migrants/generation (m=0.58) could potentially gain access
to a group and breed.
The Tana River may form a major impediment to gene
flow. This is indicated by the high FS1 values calculated
when considering combined forest demes on opposite river
banks as 2 subpopulations. Gene flow across the river occurs
during times of major flooding when breaks in river meanders
result in whole forests and their groups shifting from one
side of the river to the other. FIt values were high whether
major floods were assumed to occur on average every 1 in 80
years (Dunne and Leopold 1978) (Fs, = 0.579) or every 1 in 50
years (Kenya Ministry of Water Development 1978) (Fn =
0.492) .
Demographic Extinction Model
Estimates of r calculated from population growth
(r=0.110) and the Cole equation (r=0.108) were similar. I
therefore considered 0.110 an adequate estimate of r for the
extinction model. V was estimated at 0.20058, a value close
to Goodman's (1987b) assumption that V>2r.
I computed the relationship between T and a range of
N's that incorporated maximum population estimates for
mangabeys during 1973-74 and 1988-89 (Figure 6.3). In 15
years (1974-1989), minimum estimated population size
declined from approximately 1,500 to 725 individuals and
mean anticipated population persistence times
correspondingly declined from 425 to less than 200 years. A


25
associations between rainfall, riverflow and temperature,
and flower and fruit scores for trees and lianas for each
forest and all forests combined. I conducted separate
analyses for flowers and fruits for all species combined.
For fruits, I examined seasonal and continuous fruiting
species separately, and ripe versus, unripe fruits. Because
flowering occurred as much as 3 months prior to fruiting,
monthly flowering, rainfall and river flow data were lagged
up to 3 months in each analysis. Associations between
flowering and fruiting and environmental conditions during
the months after flowering or fruiting were examined by
advancing rain and river data by one month. I also conducted
a multiple regression analysis to test whether the
environmental parameters chosen explain a significant amount
of variability in fruiting. A sguare root transformation was
performed on monthly scores because they were discrete data
(Sokal and Rohlf 1981). Monthly rainfall levels were
recorded for Mchelelo and Nkano/Mnazini areas, and Tana
River discharge levels for the Hola Station were obtained
from the Kenya Water Department, Hydrology Section.
Comparisons with 1973-74
I compared monthly fruit scores for 9 species examined
during this study and by Homewood in 1973-74 to determine if
fruiting patterns varied between the 2 study periods. Seven
species from Mnazini and 6 species from Mchelelo forest were
compared. I subsampled from my 1987-89 data to correspond


Mchelelo
Fruit score
6n
Nkano
Fruit score
6-
4 -
3-
NDJFMAMJJASONDJFMAM
8 6 8
7 8 0
Mnazini
Fruit score
6-i
Ullliiiu.
NDJFMAMJJASO
8 8
7 8
Month
M


178
Table 5.1. Mean diversity of quadrat use (H'), mean
number of unique quadrats entered, mean distances travelled,
mean half-hour step distances, mean turning angles, and mean
number of path crossings by month for the 3 study groups.
Month
H'
a)
Quad
Entries
North Mchelelo
Distance 1/2 Hr
Travelled Steps
(m) (m)
Turning
Angles
Path
Cross
Feb 8 8
3.3
35
1237
52
64
7
Mar
3.4
41
1321
55
58
4
Apr
3.7
51
1307
55
71
2
May
3.2
34
1266
53
45
4
Jun
3.6
50
1053
44
52
1
Jul
3.6
44
1282
54
54
3
Aug
3.3
35
939
39
58
2
Sep
3.6
43
1069
46
53
3
Oct
3.4
40
1390
60
55
3
Nov
3.1
27
1129
49
56
4
Dec
3.4
38
1133
50
50
3
Jan 89
3.5
41
1076
47
48
2
Feb
3.1
29
980
44
50
5
Mar
3.4
44
756
34
50
1
b)
South Mchelelo
Feb 88
3.6
43
1368
56
61
7
Mar
3.6
47
1158
46
56
3
Apr
3.1
28
1328
59
55
4
Jun
3.6
49
1042
44
60
2
Jul
3.3
34
1089
47
73
2
Sep
3.4
40
1351
60
55
4
Oct
3.7
45
1589
75
79
3
Nov
3.5
41
1207
51
62
2
Dec
3.0
24
1397
61
45
10
Jan 89
3.6
45
1194
53
52
4
c) Nkano
Mar 88
3.4
34
1112
47
70
2
Apr
3.5
46
1592
67
64
3
Jun
3.4
42
1142
49
65
3
Jul
3.7
51
1268
54
55
2
Aug
2.8
19
955
42
82
2
Sep
3.4
41
1299
56
65
4
Oct
3.6
44
1617
70
70
5
Nov
2.8
23
887
39
72
5
Dec
3.1
31
1060
47
87
6
Jan 89
3.1
31
1040
46
72
4
Feb
3.1
33
740
33
70
3
Mar
3.3
39
860
40
55
0


N. Mcheielo
Frequency
S. Mcheielo
Frequency
eo-i
0 1 2 3 F 6 6 7 8 9 10 11 12 13 14 15 16 17
Nkano
Frequency
70 n
0 1 2 3 4 6 6 7 8 9 10 11 12 13 14 16 16 17
Angle of turn (10 degree Intervals)


205
Total fruit biomass was significantly lower in avoid
months (x=704 kg/ha) than either fight (x=955 kg/ha) or
merge (x=954 kg/ha) months (F=10.09, df=2,63, p<0.0002).
Although total fruit biomass was similar during merge and
fights months, the distribution of foods eaten by mangabeys
differed. Mangabeys spent significantly more time feeding on
uniform food resources during merge months (F=7.03, df=2,59,
p<0.002), and significantly more time feeding on patchy food
resources during fight months (F=3.72, df=2,59, p<0.03;
Figure 5.15).
Discussion
Movements and use of space by Tana mangabeys are
influenced by a complex interaction of the spatial and
temporal distribution of foods, forest patch size, group
size, and intergroup interactions. Results show that fruit
availability and distribution influence mangabey use of
space but have little effect on movement patterns
(backtracking and path length). Because Tana mangabeys
occupy small home ranges and cover much of their range in
one day, they may be able to continuously monitor fruit
production; knowledge of fruit resources would allow them to
move directly to areas of high production and reduce the
need to search. Robinson (1986) suggested that the lack of
consistent correlation between movement patterns and food


Figure 5.11. Daily travel paths by North and South Mchelelo
groups during one day of a month of intergroup
avoidance. Timing and location of adult male
long-calls are plotted for North (ovals) and
South (rectangles) groups.


216
used to develop a range of viable population sizes. I then
evaluate the level of risk facing the mangabey by comparing
calculated population sizes with present day population
estimates and discuss the potential for achieving such
numbers.
Methods
Mangabey Data Set
The number of mangabey groups present in 32 forests
where mangabeys are or were known to exist in the mid-1970s
was determined during forest surveys conducted in April and
May 1989 (Table 6.1). Data for 2 forests surveyed in July
1988 are included with the 1989 data. Comparative data for
the mid-1970s were obtained from surveys conducted by
Homewood (1976) and Marsh (1976, 1978a).
Data on age and sex composition were collected over
20 months for 3 intensively studied groups and 4 additional
groups that were monitored on an opportunistic basis every
few months (Table 6.2). Mangabeys form multi-male, female-
bonded groups (Homewood 1976). Males disperse as subadults
and adult male turnover in groups can be high (Homewood
1976). No data exist as to whether males obtain breeding
status more than once in their reproductive lifetime.
Observations indicate that the dominant male secures the
majority of the copulations (Kinnaird 1990a; but see
Homewood 1976) and I assume throughout the calculations that


224
groups within separate forests and on opposite river banks
should be considered effectively isolated subpopulations. I
chose the one-dimensional stepping stone model instead of a
multi-dimensional one (Kimura and Weiss 1964; Slatkin 1985;
Wright 1943) to examine gene flow along the river banks
because the highly linear arrangement of the forests should
limit most gene-flow to adjacent forests. The island model
was used to examine gene flow across the river because only
2 subpopulations were considered and migration to distant
forests was not possible.
The amount of genetic differentiation between demes is
evaluated with a fixation index, Fst, the component of
genetic variance distributed among demes. Fst can vary from 0
to 1; approaching 1 when demes are sufficiently isolated
from one another to be fixed for alternative alleles.
Although Wright (1978) suggests that FIt values as low as
0.05 (equivalent to 5% of the genetic variation being due to
differences among groups) may indicate differentiation,
other authors give more liberal interpretations of Ftt,
suggesting values ranging from 0.15 to 0.33 would be
necessary to indicate great differentiation (Hartl 1980,
Allendorf 1983). For the linear stepping-stone model, F is
defined as:
F = 1/(1 + 2NcC)
where C = 2(2m m)l/2;


141
I collected data on mangabey diets by scan sampling
(Altmann 1974) during systematic monthly observations
(approximately 1000 hrs.) of 2 groups of mangabeys, 1 each
from the Mchelelo and Nkano forests (see Chapter 3). For
each feeding record, I recorded the item eaten (e.g. leaf,
flower, fruit, seed) and the stage of ripeness if the item
was a fruit or seed. Seeds were recorded as the item eaten
only when they were heard to crack. If a seed was swallowed
with the rest of the fruit, or spit after sucking off the
fleshy mesocarp, I recorded the item eaten as a fruit.
Results
Human use of P. reclinata along the Tana River is
varied, and incorporates all of the 6 broad plant use
categories (Table 4.1). All parts of the palm, including
the roots, trunk, leaves and fruits, are utilized.
The distributions of size classes differ significantly
between Mchelelo and Nkano (G=276.8, df=3, p<0.001) (Figure
4.1). Nkano has a higher than expected number of individuals
in the small size class and a lower than expected number in
the largest size class relative to Mchelelo (pcO.OOl).
There is significant heterogeneity in the distribution
of harvested and unharvested palms by size class (GI=104.4,
df=3, pcO.OOl); all but 1 cell deviates significantly from
expected (p<0.001) with greater than expected numbers of
palms harvested in the 2 largest size classes and fewer


227
using the observed exponential rate of increase. Per-capita
growth rates for the 3 groups were averaged and a variance
calculated. Because the 3 study groups occupied very
different forest habitats, I assumed environmental variation
between them to be independent and that the variance was
characteristic of the population as a whole. I also assumed
that each group sampled sufficient individuals such that the
contribution of independent variation between individuals
would be negligible relative to that attributed to the
environment. I calculated a second estimate of r for
comparison following Cole's (1954) method.
I followed Belovsky's (1987) recommendations for
estimating V by first computing the ratio of variance in r
to mean r in the absence of environmental variation:
(2b/r) 1
where b = birth rate and
r = mean per-capita growth rate.
This ratio is multiplied by the square of the coefficient of
variation of r for the 3 study groups to obtain an estimate
of the ratio between r's variance due to the environment and
mean r. The final estimate of V is obtained by multiplying
by mean r.
Results
Estimates of N.
Parameter values and estimates of effective population
size for Eg. 1 and 2 of Lande and Barrowclough's model are


Mchelelo
Fruit score
6-
4 -
3-
Nkano
Fruit score
Fruit score
6-
4 -
3-
2 -
JjIIi- *
N D J F M A M
8 8
7 8
J A S O N D J
8
9
Month
Mnazini
I 1 1 r
F M A M


139
sweet when ripe. In the Tana River District of Northeastern
Kenya, P. reclinata is abundant, and occurs primarily near
river banks and in narrow belts separating riverine forests
from woodlands and savannas.
Data Collection
Data on palm harvesting by people and fruit and seed
consumption by mangabeys were collected between October 1987
and July 1989 in Mchelelo and Nkano forests. Phoenix
reclinata is the most common plant species in the Mchelelo
forest, occurring at a density of 194 individuals/ha and the
second most common plant species in the Nkano forest,
occurring at a density of 102/individuals ha (Chapter 2).
Mchelelo is not used by local people but Nkano experiences a
high degree of human exploitation.
I gathered information on palm use by people through
personal observation and informal interviews with villagers
living within and near the TRNPR. I divided palm uses into 6
broad categories developed by Prance et al. (1987): 1) food
and beverage, 2) construction material, 3) technology, 4)
remedy, 5) commerce and 6) other.
I collected data on palm harvesting in Nkano forest in
April 1989, after the peak of palm harvesting by local
people. I walked 2 kms of transects at 100 m intervals and
recorded the following data: the number of palms or palm
clusters within 3 m of each side of the transect, the size
category of each palm or cluster, if the palm had been


APPENDIX B
NUMBER OF FEEDING RECORDS AND PERCENT OF TOTAL RECORDS
BY PLANT TAXON AND MANGABEY GROUP
N. Mchelelo S. Mchelelo Nkano
Taxon
No. of
records %
No. of
records %
No. of
records %
ALANGIACEAE
Alanqium salviifolium
ANACARDIACEAE
Lannea schweinfurthii
APOCYNACEAE
Hunteria zevlanica
Rauvolfia rooirtbasiana
Saba comorensis
ARACEAE
Philodendron sp.
BORAGINACEAE
Cordia qoetzei
CAESALPINIACEAE
Cvnometra lukei
COMMELINIACEAE
Commelina sp.
COMPOSITAE
Pluchea dioscoridis
CUCURBITACEAE
Coccinia qrandis
Kedrostis foetidissima
EBENACEAE
Diosovros ferrea
Diospyros mespiliformes
87
3.3
75
4.4
73
2.4
_
_
5 49
1.6
44
2.6
22
0.7
2
0.1
40
1.3
-
-
55
1.8
41
2.4
9
0.3
-
-
22
1.0
-
-
4
0.1
-
-
-
-
1
0.06
1
0.03
17
1.0
12
0.4
11
0.7
15
0.5
_
114
3.6
85
5.0
78 4.1
377 19.7
1 0.05
3 0.2
2 0.1
4 0.2
14 0.7
258


Figure 4.1. The number of harvested and unharvested palms
in Nkano and the number of reproductive and
nonreproductive palms in Mchelelo by size
class.
a) harvested and unharvested palms by size
class in Nkano; b) reproductive and
nonreproductive palms in Mchelelo by size
class.


211
Neighboring groups often overlap, sharing space and
resources but are separate in time (sensu "time-sharing",
Jolly 1972). As fruit availability increases, the
distribution of plant diet species influences the type of
interaction; peaceful, intergroup interactions generally
occur when mangabeys eat uniformly distributed species and
aggressive interactions occur when mangabeys eat species
with patchy distributions.
Similar patterns of intergroup interactions based on
resource availability and/or distribution have been
demonstrated experimentally for other primate species.
Avoidance, mediated through long-range vocalizations, among
groups of grey-cheeked mangabeys (Cercocebus albiqena) has
been demonstrated by Waser (1976). Grey-cheeked mangabeys
occupy large (410 ha) home ranges where costs of defense are
high relative to potential benefits. Robinson (1985) showed
that capuchin monkeys respond to playbacks of spacing
vocalizations more aggressively when feeding on patchy food
resources than when feeding on more uniformly distributed
resources.
Patterns of defense expressed by the Tana mangabey
reflect trade-offs between costs and benefits of defending
resources of differing distributions. Discrete, isolated
resource patches (e.g., fruiting trees) are more easily
monopolized than resources with uniform distributions.
Benefits accrued through defense of ripe fruit trees also


125
widely dispersed, temporally fluctuating food resources. The
inverse association found between eating and other behaviors
in the Tana mangabey is consistent with that shown for a
wide range of primate species feeding primarily on fruits
(Clutton-Brock and Harvey 1977, Struhsaker 1978, 1980).
Because fruits are less evenly distributed and relatively
more unpredictable than foliage (Waser 1976), species
relying primarily on fruits must invest more time searching
for food resources. Primates eating primarily fruits also do
not reguire the extended periods of rest needed for
digestion of foliage. The greater digestion rate in
primarily frugivorous species relative to folivorous species
(Clemons and Phillips 1980, Milton 1981) requires a faster
and/or more consistent input of food to achieve adequate
nutritional intake; this results in a greater amount of the
daily activity budget invested in eating (Clutton-Brock and
Harvey 1977).
Predictable temporal variability in mangabey activity
patterns occurred on a daily and seasonal basis. The
biphasic or triphasic patterns of activities typical of most
primates (Fa 1986, Post 1981) however, were not marked in
the Tana mangabey. Most behaviors, with the exception of
grooming and inactivity, peaked in the cooler, early morning
hours, particularly during the dry season. Eating occupied a
relatively constant proportion of the daylight hours with
the exception of approximately an hour in the early morning


45
Mchelelo
Fruit score
Nkano
Fruit score
8 8 e
7 8 Q
Mnazini
Fruit score
NDJFMAMJJASONDJFMAM
8 8
7 8
Month
8
e


CHAPTER THREE
ACTIVITY PATTERNS AND DIET OF THE TANA RIVER CRESTED
MANGABEY: EFFECTS OF SEASONAL AND LONG-TERM HABITAT CHANGE
Introduction
Comparisons of the proportion of time that animals
spend on different activities are important in assessing
inter- and intraspecific behavioral differences, and may be
used to identify the adaptive nature of variability in
temporal patterning of activities (Fa 1986). Variation in
the ways in which animals partition time among different
activities is seen in a wide range of primate species.
Several generalizations have been made concerning the
relationship between ecological factors and primate activity
budgets (Clutton-Brock 1977, Crook and Gartlan 1966,
Eisenberg et al. 1972, Robinson 1986, Struhsaker and Leland
1979). Diet and habitat structure have been shown to impose
overwhelming constraints on animals use of time (Altmann, S.
1974). Because partitioning of time involves tradeoffs
between metabolic requirements of different activities and
the acquisition of energy to fulfill those requirements,
diet quality and the temporal and spatial distribution of
food strongly influence the amount of time spent in
different activities (Oates 1986).
79


157
I collected ranging data simultaneously on North and
South Mchelelo groups on 25 days with the help of field
assistants. Simultaneous group follows were distributed
throughout the study period and conducted only when the
South group was present in the Mchelelo forest (see Chapter
3). From daily range maps for both groups, I calculated
distance between neighboring groups each half-hour, number
of half-hour periods when groups were less than 100 m apart,
and number of half-hour periods when groups occupied the
same 0.25 ha quadrat. Simultaneous follows allowed me to
measure the effects of neighboring groups on movements and
use of space, and to accurately assess the response of
groups to long-calls of neighboring males.
Data Analysis
To examine variability in mangabey behavior throughout a
group's home range, I tested for differences in the
proportion of time spent in 4 behaviors while located in
high-use quadrats (those accounting for a cumulative 50% of
all sightings) and low-use quadrats. The proportion of time
spent eating, foraging, moving, and engaged in other
activities while the group's center-of-mass was located in a
given quadrat was estimated by calculating the mean percent
time spent in each activity during each behavioral scan (see
Chapter 3). Mean percent diet species and diet item
consumption was calculated similarly.


3


Table 2.3. Estimated canopy surface area (S), canopy volume (V), mean fruit production
(FN) and mean fruit weight (FW) for 16 species. Species are categorized into 4 groups:
those that bear their fruits on the canopy surface (C) or along the branches of the tree
(R), lianas (L) and palms (P).
Taxon
Group
S(m3)
V (m3)
FN(m3)
FW(gr)
Alanaium salvifolium
C
125.9
48
0.86
Sorindeia madaqascariensis
R
-
313.0
142
0.65
Saba comorensis
L
232.0
-
6
119.15
Diospvros mespiliformis
C
1016.3
-
47
1.27
Oncoba spinosa
C
74.3
-
12
29.07
Garcinia livinqstonei
C
567.0
-
47
4.87
Acacia robusta
C
931.2
-
48
0.96
Albizia qummifera
c
731.6
-
26
0.70
Ficus bubu
R
-
412.0
122
7.75
Ficus bussei
C
1965.0
-
55
7.80
Ficus natalensis
C
1004.8
-
176
0.80
Ficus sycomorus
R
1345.0
90
7.86
Hyphaene compressa
P
-
-
980
6.88
Phoenix reclinata
P
-
-
11502
0.81
Mimusops fruticosa
C
695.8
-
67
6.18
Pachvstela msolo
R
627.3
121
0.94
ui
.t*


Table 3.5
Continued
b) S. Mchelelo
Month
I
11
III
IV
V
N
Species
X
Species
X
Species
X
Species
X
Species
X
Jan 88
H. compressa
44
P. reclinata
18
O. spinosa
13
Grasses
10
M. fruticosa
5
62
Feb
M. fruticosa
37
H. compressa
19
P. reclinata
15
F. svcamorus
12
D. mespiliformes
5
155
Mar
P. reclinata
40
F. busei
19
S. madaqascariensis
19
H. compressa
6
M. fruticosa
5
225
Apr
P. reclinata
49
C. aoetzei
13
Grasses
9
F. natalensis
8
F. svcamorus
5
252
Jun
D. mesoiliformes
27
P. reclinata
19
Mushrooms
15
S. comorensis
13
T. indica
5
144
Jul
P. reclinata
50
D. mesoiliformes
14
H. compressa
11
Mushrooms
7
A. salviifoliun
4
175
Sep
P. reclinata
61
T. indica
8
Grasses
8
Mushrooms
7
D. mesoiliformes
4
107
Oct
P. reclinata
60
A. robusta
11
K. foetidissima
9
0. spinosa
4
D. mesoiliformes
4
45
Nov
A. paniculate
39
P. reclinata
27
H. compressa
8
0. spinosa
6
S. comorensis
4
238
Dec
P. reclinaba
31
0. spinosa
26
F. natalensis
24
Mushrooms
11
G. livinqstonei
8
55
Jan 89
A. salviifolium
32
0. spinosa
13
P. reclinata
12
H. compressa
11
F. natalensis
8
181
Apr
P. reclinata
67
Mushrooms
11
K. foetidissima
6
A. venosun
4
Grasses
3
70
c) Nkarto
Mar 88
F. svcamorus
47
P. reclinata
35
S. comorensis
4
P. msolo
4
Grasses
3
158
Apr
P. msolo
43
P. reclinata
25
F. svcamorus
15
S. comorensis
4
Grasses
4
172
Jun
S. comorensis
69
P. reclinata
11
P. msolo
9
Grasses
6
D. mesoiliformes
3
126
Jul
S. comorensis
61
P. reclinata
13
Grasses
11
F. svcamorus
4
D. mesoiliformes
3
151
Aug
P. reclinata
49
P. msolo
19
S. comorensis
12
Grasses
7
unknown vine
7
59
Sep
P. msolo
36
F. svcamorus
30
P. reclinata
30
S. comorensis
5
Mushrooms
1
203
Oct
P. multiflora
32
P. reclinata
27
S. comorensis
19
A. paniculate
11
F. svcamorus
4
170
Nov
A. paniculate
72
F. svcamorus
16
P. multiflora
3
S. comorensis
1
A. salviifolinn
1
245
Dec
F. svcamorus
55
A. paniculate
24
A. salviifolinn
12
Mushrooms
4
P. msolo
3
139
Jan 89
A. salviifolium
40
F. svcamorus
25
S. comorensis
14
P. msolo
7
A. pan i culata
5
123
Feb
P. reclinata
53
S. comorensis
28
P. msolo
8
0. spinosa
4
F. svcamorus
3
182
Mar
P. reclinata
43
S. comorensis
34
F. svcamorus
10
P. msolo
4
R. elliotica
3
190


142
Table 4.1. Human uses of P. reclinata grouped into 6
broad categories. Plant parts utilized and harvesting
methods employed are outlined.
CATEGORY PLANT PART/SPECIFIC USE HARVESTING METHOD
Food and Beverage a) palm heart top palm, cut out palm
infrequently used vegetable source heart, eat raw or boil.
b) fruits
infrequently used diet item
c) vascular bundles
alcholic beverage
Construction a) trunk
poles for building construction
b) leaf rachis
wattle for constructing mud houses
c) leaflets
building ties
d) leaf
door entrances and covers
Technology a) leaf rachis
fish traps
b) leaf
fans for stoking fires
and fanning insects
cut rachillae of unripe fruits,
dip in water and hang until
fruits ripen; eat raw or squeeze
fruits in cloth and use extracted
sweet mesocarp in baking.
top palm, cover and hang gourd
to one side to collect oozing,
exudates, empty fermented
contents from full gourds and cut
a thin layer from top of trunk,
replace cover and gourd and
repeat daily until exudates stop
flowing.
cut tall trunk at root base
(used traditionally for
constructing community or
political buildings)
cut frond at base, remove
leaflets, mount horizontally
between building poles before
applying mud.
strip leaflets from freshly cut
leaf, use to bind poles, tie
thatching to roofs, etc.
cut numerous, long fronds at
base, sun dry, bind together in
bundles, place over entrances.
strip leaflets from freshly cut
leaf, use flexible rachis and
leaflets to construct oblong,
funnel traps.
cut unexpanded 'sword leaf',
plait leaflets into wide, flat
square fan with rachis as handle.
c) leaflets cut unexpanded 'sword leaf', sun
strapping for hanging food and dry, split and plait into wide,
items that attract animals. flat strapping, tie around rings
of bananna leaves and suspend
with strapping from ceiling.
split rachis into numerous
fibers, bundle together, fold in
half and bind under fold with
leaflets.
d) leaf rachis
hand brooms


257
correlated with all other shapes for both the surface area
and volume calculations. I therefore chose to use the
cylinder formula to approximate surface area and volume
(i.e. food-producing area) of trees of differing shapes.
Although this formula will not give the true values for any
given tree due to variation in overall shape and canopy
peculiarities, it should be the most applicable and give the
best approximation. Using a geometric formula such as the
open-bottom cylinder also avoids the problems of linear
indices such as those of Struhsaker and Marsh. These indices
frequently result in ties even when there are large crown
differences. If tree canopies are being ranked, these
indices will not be able to distinguish what may be large
differences in canopy surface area and volume.


77
Pronounced year-to-year variation in the onset and
termination of the dry season could have important
consequences for the flowering and fruiting phenologies of
tree species (e.g. Chivers and Raemaekers 1980). The Tana
riverine forests experience high year-round temperature with
some seasonal variation, twice yearly rains and irregular
flooding, with the severity and timing of rains and floods
varying appreciably between years (Hughes 1985, Homewood
1976, Marsh 1976). The extreme unpredictability and annual
variation in rainfall, temperature and flooding, overlaid
with habitat degradation and fragmentation, make it
difficult to determine causes of significant differences in
fruiting patterns between 1973-74 and 1988-89. Wheelwright
(1986) states that even 7 years of phenological data on
species occurring in a Costa Rican montane forest, a habitat
far less variable than the Tana River, were too few to
determine supra-annual cycles of fruit production. The
critical point however, is that total fruit availability
will decline with decreasing abundances of reproductive
trees. The abundances of several important food species
(e.g. Ficus sycomorus. Ficus natalensis. Acacia robusta,
Albizia glaberrima. and Mimusops fruticosa^ declined between
1973-74 and 1988-89 in at least 1 or more of the study sites
(Decker 1989, Medley 1990), effectively decreasing overall
fruit production. Such long-term changes in the fruit


140
harvested, and the level of harvesting. I used 4 size
categories: 1) small, isolated individuals below 1 m in
height, 2) bushy clusters up to 2 m in height that had not
formed an obvious trunk, 3) palms with short trunks less
than 2 m tall and 4) palms with trunks greater than or equal
to 2 m in height. A palm was considered harvested if any
parts had been cut or removed. Four levels of harvesting
were defined based on the degree of cutting: 1) light (few
leaves removed), 2) medium (up to half of all leaves
removed), 3) heavy (often involving removal of all leaf
fronds) and 4) topped (where the apical meristem of at least
1 trunk was removed).
Similar data, plus additional data on the sex of the
palms, were collected for 1.1 km of transects in Mchelelo
forest in June 1989. Sex was determined for reproductive
plants by the presence of sexually distinct pendulous
rachillae that bear the flowers and fruits; these rachillae
generally wither and remain on the plant for extended
periods after flowering and fruiting. Because the palms in
Mchelelo are not exploited by the local people, Mchelelo was
used as a control for comparison with data from Nkano. I
used G-tests (Sokal and Rohlf 1980) to determine if the
distributions of size classes were different for the 2
forests and if the frequency and level of harvesting among
size classes deviated from uniform distributions.


Figure 2.15. Contribution of ripe and unripe fruits to
monthly fruit biomass index,
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.


I also thank my parents who have endured much, including
my long research stints out of the country. They have
encouraged and supported me in numerous ways throughout all
my endeavors, even when they may not have understood fully
just what drove me.
Finally, I thank the "Makarau Mgogo", those rare and
elusive primates who allowed me, albeit with insouciance, to
investigate their lives. Without their cooperation, this
study would not have been feasible.
Vll


Figure 3.4. Mean percent time spent in 4 major behaviors (eating, foraging, moving and
inactive) by time-of-day and season for the 1988-89 Nkano group.


Figure 5.6. Occupancy of each 0.25 ha quadrat by North
and South Mchelelo groups; percent occupancy is
divided into 1 of 5 intensities of use 173
Figure 5.7. Occupancy of each 0.25 ha quadrat by Nkano
group; percent occupancy is divided into 1 of 5
intensities of use 175
Figure 5.8. Frequency distributions of half-hour
distance steps in intervals of 20 m. Asterisks
indicate values expected if distance steps were
distributed randomly (following a Poisson
distribution). a) Frequency distribution for N.
Mchelelo group; b) Frequency distribution for S.
Mchelelo group; c) Frequency distribution for
Nkano group 182
Figure 5.9. Frequency distributions of half-hour
turning angles in 10 degree intervals. a)
Frequency distribution for N. Mchelelo group; b)
Frequency distribution for S. Mchelelo group; c)
Frequency distribution for Nkano group 185
Figure 5.10. Degree of monthly overlap by North Mchelelo
group (solid line) and South Mchelelo group (broken
line) and the area of overlap in hectares (bars) 189
Figure 5.11. Daily travel paths by North and South
Mchelelo groups during one day of a month of
intergroup avoidance. Timing and location of
adult male long-calls are plotted for North
(ovals) and South (rectangles) groups 196
Figure 5.12. Daily travel paths by North and South
Mchelelo groups during one day of a month of
intergroup merging. Timing and location of adult
male long-calls are plotted for North (ovals) and
South (rectangles) groups 198
Figure 5.13. Daily travel paths by North and South
Mchelelo groups during one day of a month of
intergroup fighting. Timing and location of adult
male long-calls are plotted for North (ovals) and
South (rectangles) groups 201
Figure 5.14. Location and outcome of 34 fights
observed between North and South Mchelelo groups.
N and S indicate that North and South group won
fights, respectively. D indicates that no clear
winner was determined 204
XVI11


92
Table 3.3. ANOVA results for group membership, season, and
time-of-day differences in 7 behaviors. Significance values
are calculated from Type III partial sums of squares.
Source
df
Sum of
Squares
Mean
Squares
F
P
Eat
Group
2
0.218
0.109
3.78
0.023
Season
3
0.096
0.032
1.11
0.345
Time-of-Day
2
0.037
0.018
0.64
0.529
Group Season
6
0.191
0.032
1.10
0.359
Season Time-of-Day
Error
6
441
0.152
12.715
0.025
0.029
0.88
0.512
Forage
Group
2
0.005
0.002
0.12
0.886
Season
3
0.383
0.128
6.29
0.0003
Time-of-Day
2
0.934
0.467
23.03
0.0001
Group Season
6
0.193
0.322
1.58
0.149
Season Time-of-Day
Error
6
441
0.189
8.948
0.031
0.020
1.55
0.160
Move
Group
2
0.139
0.069
2.68
0.069
Season
3
0.094
0.031
1.21
0.307
Time-of-Day
2
0.522
0.260
10.10
0.0001
Group Season
6
0.111
0.019
0.72
0.635
Season Time-of-Day
Error
6
441
0.466
11.382
0.078
0.026
3.01
0.007
Inactive
Group
2
0.316
0.158
6.63
0.002
Season
3
0.069
0.023
0.96
0.413
Time-of-Day
2
0.541
0.270
11.32
0.0001
Group Season
6
0.380
0.063
2.65
0.015
Season Time-of-Day 6
Error 441
0.236
10.536
0.039
0.024
1.64
0.134
Autogroom
Group
2
0.055
0.027
2.73
0.066
Season
3
0.274
0.009
0.91
0.437
Time-of-Day
2
0.053
0.026
2.63
0.073
Group Season
6
0.133
0.022
2.21
0.041
Season Time-of-Day 6
Error 441
0.060
4.434
0.010
0.010
1.00
0.424


265
Fa, J.E. 1986. Use of time and resources by provisioned
troops of monkeys: Social behaviour, time and energy in
the Barbary macaque. Contributions to Primatology 23,
377 pp.
Foster, R.B. 1974. Seasonality of Fruit Production and
Seed Fall in a Tropical Forest Ecosystem in Panama.
Ph.D. dissertation, Duke University, Chapel Hill, NC.
Freund, R.J. and R.C. Littell. 1981. SAS for linear
models. SAS Institute, Raleigh, NC.
Gartlan, J.S. and C.K. Brain. 1968. Ecology and social
variability in Cercooithecus aethiops and C. mitis.
Pages 253-293 in P. Jay, editor. Primates: Studies in
Adaptation and Variability. Holt, Rinehart, and
Winston, New York, NY.
Gartlan, J.S., D. B. McKey, P.G. Waterman, C.N. Mbi, and
T.T. Struhsaker. 1980. A comparative study of the
phytochemistry of two African rain forests. Biochemical
Systematics and Ecology 8:401-422.
Gautier-Hion, A., J.M. Duplantier, L. Emmons, F. Feer, P.
Heckestweiler, A. Moungazi, R. Quris and C. Sourd.
1985. Coadaptation entre rhythmes de fructification et
frugivorie en foret tropicale humide du Gabon: Mythe ou
realite. Rev. Ecol. (Terre Vie) 40:405-434.
Gittins, S.P. and J.J. Raemaekers. 1980. Siamang, lar, and
agile gibbons. Pages 63-106 in D.J. Chivers, editor.
Malayan Forest Primates: Ten Years' Study in Tropical
Rain Forest. Plenum Press, New York, NY.
Goodman, D. 1987a. How do any species persist? Lessons
for conservation biology. Conservation Biology 1:59-62.
Goodman, D. 1987b. The demography of chance extinction.
Pages 11-34 in M.E. Soul, editor. Viable Populations
for Conservation. Cambridge University Press,
Cambridge, UK.
Greig-Smith, P. 1983. Quantitative Plant Ecology.
University of California Press, Berkeley, CA.
Groves, C.P. 1978. Phylogenetic and population systematics
of the mangabeys (Primates: Cercopithecoidea). Primates
19(1):1-34.


199
adult males during merges also was observed; on 4 occasions
adult males were observed being aggressive toward sexually
receptive females, causing the females to move away from the
other group.
Fight months were intermediate to avoid and merge
months in range overlap; South group frequently moved
through the north sector of the forest and each group
occasionally crossed the others' travel path. Vocal
responses to long calls were more frequent than during avoid
months and calls often resulted in direct, long-distance
movements (>150 m) toward the caller (Figure 5.13). When
fighting occurred, members of opposing groups lined-up,
faced one another and lunged with tails arched forward,
heads down, and eye-lids lowered. Occasional, low,
"cackling" vocalizations were given by group females. All
age-sex classes except infants, participated in fights;
adult females participated most in the fighting line-up and
only adult males participated in branch-shaking and active
chases through the canopy. Intragroup herding by males
during fights occurred only once when an adult male forced a
group female of unknown receptivity toward a fight. Physical
contact during fights was rare; adult females and juvenile
males were observed 4 times biting opponents and
standing on their back legs battling each other with their
front legs. Grooming bouts generally occurred between
individuals of the same group after intergroup fights.




91
Table 3.2. MANOVA results for group membership, season, and
time-of-day differences among 1988-89 mangabey activity
budgets.
Source
df
Wilks'
Lambda
F-ratio
P
Group Membership
14
0.89
3.65
0.0001
Season
21
0.87
2.84
0.0001
Time-of-Day
14
0.83
6.23
0.0001
Group Season
42
0.87
1.51
0.019
Group Time-of-Day
28
0.93
1.08
0.343
Season Time-of-Day
42
0.86
1.59
0.001


93
Table 3.3. -- Continued.
Sum of Mean
Source df Squares Squares F P
Allogroom
Group 2
Season 3
Time-of-Day 2
Group Season 6
Season Time-of-Day 6
Error 441
Social/Sexual
Group 2
Season 3
Time-of-Day 2
Group Season 6
Season Time-of-Day 6
Error 441
0.052
0.026
0.98
0.375
0.086
0.029
1.08
0.356
0.183
0.092
3.44
0.033
0.107
0.018
0.67
0.676
0.099
0.027
1.01
0.419
11.730
0.026
0.024
0.012
0.57
0.563
0.659
0.220
10.61
0.0001
0.014
0.007
0.33
0.721
0.101
0.017
0.82
0.555
0.106
0.018
0.85
0.529
9.138
0.021


158
I classified the distribution of selected plant diet
species as patchy or uniform to test the effect of plant
abundance and distribution on mangabey ranging patterns and
intergroup behaviors. I used only those plant diet species
that accounted for more than 2.5% of the estimated annual
diet of both the North and South Mchelelo groups (N=9,
Appendix A). I defined a patchy species as having fewer than
2 individuals/ha or a Morisita's index of dispersion greater
than 2.0. Species that were highly abundant in the forest
(>15 individuals/ha), had Morisita's indices less than 2.0,
or tended to occur in larger, continuous clusters (> 0.25
ha) were defined as uniform. This classification scheme
resulted in one ambiguous species (Oncoba spinosa with 6.49
individuals/ha and I§=2.67) that was dropped from the
analysis.
I used 3 categories to describe monthly intergroup
behaviors according to the extent and outcome of intergroup
interactions: avoid, fight, and merge. Avoid months were
defined as those months in which no intergroup interactions
were recorded and the South group was resident in Mchelelo
forest less than 50% of the days monitored. Fight months
were defined as months in which at least one fight was
recorded between groups. Merge months were defined as those
months in which no fights occurred when both groups
simultaneously occupied the same 0.25 ha quadrat or
residency of South Mchelelo group was greater than 50%.


182
N. Mchelelo
Frequency
S. Mchelelo
Frequency
Nkano
Frequency
1/2 hour distance steps (20 m intervals)


Figure 2.12. Mean monthly fruit scores for Ficus svcomorus.
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.


270
Medley, K.E., M.F. Kinnaird and B.S. Decker. 1989. A
survey of the riverine forests in the Wema/Hewani
vicinity, with reference to development and the
preservation of endemic primates and human resources.
Utafiti: Occasional Papers of the National Museums of
Kenya 2(1):1-6.
Medway, L. 1972. Phenology of a tropical rain forest in
Malaya. Biological Journal of the Linnaean Society
4:117-146.
Mehlman, P.T. 1989. Comparative density, demography, and
ranging behavior of Barbary Macaques (Macaca svlvanus)
in marginal and Prime conifer habitats. International
Journal of Primatology 10(4):269-291.
Milton, K. 1981. Food choice and digestive strategies of
two sympatric primate species. American Naturalist
117:496-505.
Mitani, J.C. and P.S. Rodman. 1979. Territoriality: the
relation of ranging pattern and home range size to
defendability, with an analysis of territoriality among
primate species. Behavioral Ecology and Sociobiology
5:241-251.
Mitani, M. 1989. Cercocebus torcruatus: adaptive feeding
and ranging behaviors related to seasonal fluctuations
of food resources in the tropical rain forest of south
western Cameroon. Primates 30(3):307-323.
Mittermeier, R.A. 1981. A global overview of primate
conservation. Pages 325-340 in J.G. Else and P.C. Lee.
Primate Ecology and Conservation, Selected Proceedings
of the Tenth Congress of the International
Primatological Society, Nairobi, Kenya; Vol. 2.
Cambridge University Press, Cambridge, U.K.
Mittermeier, R.A. 1990. Hunting and its effect on wild
primate populations in Suriname. Pages 128-143 .in J.G.
Robinson and K.H. Redford, editors. Neotropical
Wildlife Use and Conservation. University of Chicago
Press, Chicago, IL.
Monasterio, M. and G. Sarmiento. 1976. Phenological
strategies of plant species in the tropical savanna
and the semi-deciduous forest of the Venezuelan
Llanos. Journal of Biogeography 3:325-356.


246
Demographic Extinction Model
The results of my data applied to Goodman's (1987b)
model suggest that Tana mangabeys have a high probability of
extinction within the next 50-100 years given their present
population status. The model required very high population
numbers for long-term persistence; a population of 8000
mangabeys was necessary to ensure a 95% probability of
survival over the next 100 years. The Tana mangabey
population, however, could not presently attain, and very
likely never has attained such high numbers, given its
limited range. Goodman (1987a) found that his model often
predicts impossibly large numbers for long-term persistence.
He suggests that rare species may persist in spite of
predictions because they may not be governed by dynamics
with strong environmental variance like those developed in
his demographic model. The Tana mangabey is a highly
adaptable, generalist species within it's habitat (Homewood
1976). It may be capable for example, of major dietary
shifts in response to habitat change (Chapter 3) and
therefore may be less sensitive to environmental variance.
Population sizes predicted by Goodman's model were not
unrealistically high for the mangabey when lower
probabilities of long-term persistence were considered. If
an 80% probability of survival over the next 100 years is an
acceptable management objective, then a mean persistence
time of around 400 years with a corresponding population of


19
patchy distributions, respectively. Deviation of the index
from unity is tested with the F statistic.
Although Morisita's Index is relatively independent of
quadrat size, Poole (1974) suggests that quadrats be large
enough to contain at least one individual of one of the more
common species. A quadrat size of 0.25 ha meets this
requirement: the first and third most abundant species in
Mchelelo forest, Phoenix reclinata and Hyphaene compressa.
occur at least once in over 90% and 70% of the quadrats,
respectively; at least one individual of one of the 3 most
abundant species occurs in over 90% of the quadrats. The 3
most common species in Nkano forest, Alanqium salviifolium.
Phoenix reelinata and Pachystela msolo. all occur in over
90% of the quadrats and all quadrats contain at least 1
individual of 1 of these species.
Temporal Distribution. Temporal fluctuations in the
availability of flowers and fruits were tabulated for 240
individuals in 16 species within the 3 forest areas: 126
individuals in 15 species for Mchelelo forest, 49
individuals in 9 species for Nkano forest, and 65
individuals in 8 species for Mnazini South forest (Table
2.1). I chose the species based on their presence in each
study site and their importance as mangabey foods (Homewood
1976). The more common species were represented by at least
10 individuals; species with fewer than 10 individuals in
the study areas were represented by the total number of




Figure 1.1. Adult male Tana River crested mangabey,
Cercocebus galeritus qaleritus. seated in
Hypheane compressa palm.


252
The best remaining habitat outside the reserve
(Wema/Hewani) is being seriously affected by the development
of the Tana Delta Rice Irrigation Scheme (Medley et al.
1989) ; forests have been eliminated or fragmented,
selectively harvested, burned, and exposed to entirely
different flooding regimes depending on where they occur in
relation to the irrigation dikes. Kinnaird (1990b)
speculated that mangabey groups living in these forests were
critical for improving mangabey population viability by
increasing overall population size and supplying additional
pools of potentially successful immigrants. The degradation
of Wema/Hewani forest area underscores the importance of the
remaining forests of the TRNPR.
Clearly, the situation is critical. Immediate action
towards habitat preservation and enhancement of existing
forest area will be necessary for the conservation of the
Tana River crested mangabey and its riverine forest habitat.
Kinnaird et al. (1990) have developed management
recommendations for the TRNPR that are hoped to form the
basis of a government action plan. The recommendations
center on preservation of existing habitat through
prevention of further forest clearing and taking of forest
products. Resettlement assistance for agriculturalists
living or farming within the reserve and establishment of
alternate sources of tree products such as subsidized
aluminum canoes and woodlots for propagating building poles


276
Waser, P. and K. Homewood. 1979. Cost-benefit approaches
to territoriality: a test with forest primates.
Behavioral Ecology and Sociobiology 6:115-119.
Waser, P. and R.H. Wiley. 1980. Mechanisms and evolution
of spacing in animals. Pages 159-221 in P. Marler and
J.G. Vandenbergh, editors. Handbook of Behavioral
Neurobiology Vol. 3. Plenum Press, New York, NY.
Waterman, P. 1984. Food acquisition and processing as a
function of plant chemistry. Pages 177-211 in D.J.
Chivers, B.A. Wood and A. Bilsborough, editors. Food
Acquisition and Processing in Primates. Plenum Press,
New York, NY.
Wheelwright, N.T. 1986. A seven-year study of individual
variation in fruit production in tropical bird-dispersed
tree species in the family Lauraceae. Pages 19-35 in A.
Estrada and T.H. Fleming, editors. Frugivores and seed
dispersal. Dr. W. Junk, Publishers, Dordrecht, The
Netherlands.
Wood, J.W. 1987. The genetic demography of the Gainj of
Papua New Guinea. 2. Determinants of effective
population size. American Naturalist 129:165-187.
Wrangham, R.W. 1980. An ecological model of female-bonded
primate groups. Behaviour 75:262-300.
Wright, S. 1931. Evolution in Mendelian populations.
Genetics 16:97-159.
Wright, S. 1943. Isolation by distance. Genetics 28:114-
38.
Wright, S. 1978. Evolution and the Genetics of
Populations, Vol 4. Variability within and among
Natural Populations. University of Chicago Press,
Chicago, IL.
Yoshiba, K. 1968. Local and intertroop variability in
ecology and social behavior of common Indian langurs.
Pages 217-242 in P.C. Jay, editor. Primates: studies in
adaptability and variability. Holt, Rinehart and
Winston, New York, NY.


Figure 5.14.
Location and outcome of 34 fights observed
between North and South Mchelelo groups. N
and S indicate that North and South group won
fights, respectively. D indicates that no
clear winner was determined.


183
previous half hour) more than backtrack (>90) (Figure 5.9).
Turning angles, however, were not always good indices of
backtracking. A series of small turning angles, all in the
same direction, often indicated forward movement but
resulted in backtracking. The number of times a group
crossed its own path therefore, may be a better index of
backtracking. Number of daily path crossings varied from ID-
13 (x=3.4) for the North Mchelelo group, 0-10 (x=3.6) for
the South Mchelelo group, and 0-9 (x=3.4) for the Nkano
group. Although number of daily path crossings was
significantly related to mean turning angle, turning angle
was a weak predictor of path crossings, accounting for only
17% of the variance.
Despite differences in group size and habitat, there
were few significant effects of group membership on daily
distance travelled and measures of travel pattern. Group
membership had a significant effect only on turning angle
(F=8.51, df=2,ll, p<0.002); least square means comparisons
showed the Nkano group had higher average turning angles
than either of the Mchelelo groups.
Seasonal variation. There was considerable monthly
variation in the average daily distance travelled, mean rate
of group movement, mean turning angles, and mean path
crossing for all 3 groups (Table 5.1). I examined
associations between mean daily path length and travel


130
Studies on captive and provisioned versus free-ranging
Barbary macaques (Macaca svlvanus) have shown that food
quality strongly influences activity budgets; macaques
provisioned with high quality foods spend significantly less
time eating and more time inactive than their free-ranging
counterparts (Fa 1986). The switch by 1988-89 groups to
primarily unripe seed and fruit may have resulted in a lower
quality diet. Ripe fruits, because of their concentration of
nutrients and readily metabolized simple sugars, offer an
outstanding source of available energy (Waterman 1984).
Unripe seed and fruits are less desirable energetically and
high levels of secondary metabolites present in unripe seeds
and fruits further reduce diet quality. Secondary
metabolites such as tannins, that inhibit digestion or
alkaloids that interfere with specific physiological
processes, have been shown to be higher in unripe verses
ripe fruits (Gartlan et al. 1980). The fact that unripe
fleshy fruits are more likely to contain inhibitory
secondary metabolites than other fruit types (McKey et al.
1981) may explain why mangabeys rarely consumed unripe
fleshy fruits (e.g., Ficus spp.) and fed primarily on non-
fleshy fruits or seeds in the unripe stage (Table 3.7).
Diet composition may result from a critical balance
between nutrient content and tannin/alkaloid content of food
items, such that food items become unattractive when a
certain cost/benefit ratio is exceeded (Gartlan et al. 1980,


13
variable in less seasonal forests (Hilty 1980, Heideman
1989) The phenological patterns of seasonal, ground-water
dependent forests are not well documented and may be
influenced more strongly by variation in river water level
than other climatic factors.
The Tana River forests occur in a semi-arid environment
with low annual rainfall (about 400 mm/year) and are
dependent on ground-water from the river for their existence
(Homewood 1976, Marsh 1976, Hughes, 1985). The occurrence of
ground-water forest communities along the Tana appears to be
related to the forest height above flood levels and duration
of flooding, soil texture and therefore ability to hold
water, soil chemistry, age of substrate and the history of
human activities within the forest (Hughes 1985).
Little information exists on the influence of forest
fragmentation, isolation or habitat change on tree
phenologies. Increased leaf fall has been noted in Brazilian
forests after fragmentation (Lovejoy et al. 1986) but even
qualitative changes in plant reproduction have not been
investigated. Varying degrees of habitat degradation,
fragmentation and isolation within the Tana River forests
over the last 15 years provide a basis for examining the
effects of habitat change on plant reproduction.
In this chapter I describe the phenological and spatial
patterns of fruit species eaten by Tana mangabeys in 3
forests. I derive 2 indices of monthly fruit production in


CHAPTER TWO
ABUNDANCE AND DISTRIBUTION OF FRUIT RESOURCES
IN THREE TANA RIVERINE FORESTS
Introduction
Temporal patterns of resource abundance and scarcity,
coupled with the spatial distribution of fruit trees, are
considered crucial to molding the ranging and foraging
behaviors of frugivorous birds and primates (Chivers and
Raemaekers 1980, Leighton and Leighton 1983, Terborgh 1983,
Levey 1988). In most tropical trees, flowering and fruiting
is episodic and seasonal peaks in fruit abundance have been
recorded for many forests of the Neotropics (Foster 1974,
Terborgh 1983, Levy 1988) and Old World tropics (Koelmeyer
1960, Medway 1972, Struhsaker 1975, Leighton and Leighton
1983, Gautier-Hion et al. 1985, van Schaik 1986). The timing
of flowering and fruiting in tropical trees has been
ascribed to climatic, edaphic and biotic factors (Rathche
and Lacey 1985). In most tropical forests, variation in
rainfall appears to be the most significant climatic factor
influencing the phenologies of flowering and fruiting
(Foster 1974, Hilty 1980, Borchert 1983), although the
relationship between rainfall and plant reproduction is
12


->
Intergroup merge site
begin ni ng of dag travel
end of dag travel
direction of travel North group
direction of travel South group


256
1/2 prolate spheroid (for formulas see CRC Standard
Mathematical Tables 1970). I also tested 2 canopy indices
developed by Struhsaker (1976) and Marsh (1978a) that make
few assumptions about tree shape. The Struhsaker and Marsh
indices are calculated as the sum and the product,
respectively, of tree depth and tree width.
The descriptive statistics and simple correlations for
the area and volume calculations are as follows:
Surface Area Volume N
Formula
Mean (S.D.)
Minimum
Maximum
Mean
(S.D.)
Minimum
Maximum
Cylinder
38.7 (18.4)
9.4
84.8
27.2
(19.8)
3.1
84.8
27
Cone
32.7 (16.0)
7.6
68.3
9.1
(6.6)
1.0
28.3
27
1/2 Sphere
27.2 (14.9)
6.3
56.6
21.0
(16.4)
2.1
56.6
27
Sphere Zone
28.3 (12.9)
6.3
56.6
39.8
(28.0)
4.2
113.1
27
1/2 Oblate Spheriod
24.7 (10.9)
11.1
44.5
20.0
(14.7)
2.1
56.6
11
1/2 Prolate Spheriod
27.3 (11.7)
10.7
44.6
13.2
(13.9)
2.1
56.6
11
Struhsaker
4.0 (1.0)
2.0
6.0
4.0
(1.0)
2.0
6.0
27
Marsh
4.0 (2.1)
1.0
9.0
4.0
(2.1)
1.0
9.0
27
Surface Area Correlations:
1/2 1/2 1/2 Cylinder
Cone Sphere Sphere Zone Oblate Prolate Struhsaker
Cylinder
Cone
0.91
1/2 Sphere
0.80
0.98
Sphere Zone
0.92
0.74
0.58
1/2 Oblate
0.99
0.96
0.93
0.92
1/2 Prolate
0.99
0.99
0.97
0.99
Struhsaker
0.92
0.73
0.57
0.97
0.98 0.95
Marsh
0.94
0.72
0.55
1.00
0.92 0.99 0.97
Volume Correlations:
1/2
1/2 1/2
Cylinder
Cone
Sphere
Sphere Zone
Oblate Prolate Struhsaker
Cylinder
Cone
1.00
1/2 Sphere
0.75
0.75
Sphere Zone
0.94
0.94
0.52
1/2 Oblate
1.00
1.00
0.97
0.95
1/2 Prolate
0.95
0.95
0.61
0.98
Struhsaker
0.90
0.90
0.55
0.96
0.89 0.89
Marsh
0.95
0.95
0.53
0.98
0.98 0.98 0.97
All geometric shapes are highly correlated, however,
the open-bottom cylinder is the most highly and consistently


135
further that they may be less competitive in higher stature
or more open forests than other primates (e.g., guenons and
baboons) that are better suited for a more arboreal or
terrestrial lifestyle and rely to a greater extent on more
readily available and/or lower-energy foods. That yellow
baboons and Sykes monkeys (Cercopithecus mitis) have not
declined along the Tana River supports the idea that C.
oaleritus is more affected by competition.
The Tana mangabeys' endangered status is primarily the
result of the highly restricted distribution of Tana forest
habitat and the rapid destruction of this habitat. Forest
patch size and degree of isolation may influence the
presence or persistence of mangabey groups (Chapter 5).
Small patches may not contain enough individuals of rare
diet species (e.g., Ficus spp.) or extensive enough stands
of stable food items (e.g., Phoenix reclinata) to support
groups of mangabeys. Isolation of patches also may limit
opportunities for colonization or recolonization of formerly
occupied forests; mangabeys readily cross open woodland, but
they have never been observed crossing extensive grasslands
or non-forested distances greater than 0.5 km.


264
Davis, T.A. 1988. Uses of semi-wild palms in Indonesia and
elsewhere in South and Southeast Asia. Advances in
Economic Botany 6:98-118.
Decker, B.S. 1989. Effects of Habitat Disturbance on the
Behavioral Ecology and Demographics of the Tana River
Red Colobus (Colobus badius rufomitratus). Ph.D.
dissertation, Emory University, Atlanta, GA.
Decker, B.S. and M.F. Kinnaird. 1990. Tana River red
colobus and crested mangabey: results of recent
censuses. Submitted to the Journal of American
Primatology.
DeMason, D.A. and K.N. Chandra Sekhar. 1988. The breeding
system in the date palm (Phoenix dactvlifera L.) and its
recognition by early cultivators. Advances in Economic
Botany 6:20-35.
Dobroruka, L.J. and J. Badalec. 1966. Zur Artbildung der
Mangaben Gattung Cercocebus (Cercopithecidae, Primates).
Rev. Zool. Bot. Afr. 73:345-350.
Dobson, A.P. and A.M. Lyles. 1989. The population dynamics
and conservation of primate populations. Conservation
Biology 3(4):362-380.
Dunbar, R. and P. Dunbar. 1974. Ecological relations and
niche separation between sympatric terrestrial primates
in Ethiopia. Folia Primatologica 21:36-61.
Dunne, T. and L. B. Leopold. 1978. Water in Environmental
Planning. Freeman Press, Chicago, IL.
Eisenberg, J.F. 1980. The density and biomass of tropical
mammals. Pages 35-56 in M.E. Soul and B.A. Wilcox,
editors. Conservation Biology: An Evolutionary-
Ecological Perspective. Sinauer Associates, Inc.,
Sunderland, MA.
Eisenberg, J.F., N.A. Muckenhirn and R. Rudran. 1972. The
relation between ecology and social structure in
primates. Science 176:863-874.
Ewens, W.J., P.J. Brockwell, J.M. Gani and S.I. Resnick.
1987. Minimum viable population size in the presence of
catastrophes. Pages 59-68 in M.E. Soul, editor. Viable
Populations for Conservation. Cambridge University
Press, Cambridge, UK.


10
provided measures of variance in mangabey behaviors due to
seasonal habitat change alone.
I begin in Chapter 2 by describing the abundance and
distribution of important mangabey fruit resources and
examining associations between the temporal availability of
these resources and various environmental parameters. I also
examine changes over the 15-year period in the temporal
pattern of fruit availability. In Chapter 3, I investigate
the effect of changes in forest structure and the food
resource base on activity budgets. I also examine the
temporal distribution (diurnal and seasonal) of activity
patterns and the influence of diet on time budgets. In
Chapter 4, I present an example of the impact of human
exploitation of forest products on mangabeys. Specifically,
I present data on competing uses of an important mangabey
diet species, Phoenix reclinata, by mangabeys and humans. I
quantify mangabey ranging patterns and defense of space in
Chapter 5 and incorporate data from Chapters 2 and 3 to test
the influence of seasonal changes in food availability and
distribution on variation in defense of space by mangabeys.
In Chapter 6, I present 2 approaches, demographic and
genetic, for estimating population sizes necessary for the
long-term persistence of the Tana mangabey. In this chapter,
I evaluate the mangabey's risk of extinction by comparing
derived population sizes with present population estimates.
I summarize in Chapter 7 the major results of the body of


225
m = the fraction of members exchanged with adjacent forests
each generation; and
m,, = the fraction of members a forest exchanges with the
population as a whole (long distance migration).
When the number of demes, or forests, is finite, the actual
expected differentiation is approximated by the product of
the above estimate of Fst and a correction factor for finite
forest number. For a linear case, the correction factor is:
[1 1/n (2/n2) (N-i) r(i)]
where n is the number of demes and
r(i) = exp (2maj/m)1/2i
Wright's (1943) island model is computed as:
F = l/(4Nem + 1)
Estimating model parameters. Little data are available
on the dispersal patterns of male mangabeys and indirect
estimates of numbers of migrants (M) must be used. One
estimate of M is the potential number of males produced by a
deme in one generation that could migrate to a neighboring
deme and breed. Given a 1:1 sex ratio at birth, this is
calculated as:
M = [ (t) (Af) (Rf)/2] (SJ
where t = generation length in years
A, = average number of reproductive
females/group
R, = average female reproduction/year


39
occurred from June to October in Mchelelo and Nkano and from
July through November in Mnazini.
Several species showed seasonal patterns of fruiting
(Figures 2.7-2.10). The majority of seasonal fruiters bore
fruits only once in a year; however, individuals of 2
species (Sorindeia madagascariensis and Garcinia
livingstonei) produced fruits twice within a year (Figure
2.8). Of the seasonal fruiters, only Phoenix reclinata and
Diospyros mespiliformes produced fruit during the cool, dry
months between June and September (Figures 2.9 and 2.10).
Three species appeared to fruit continuously (Figures 2.11
and 2.12) due to 1) continuous production of fruit by all
individuals (Oncoba spinosa and Ficus svcomorus); 2)
retention of fruits over long periods of time (Hyphaene
compressa); and 3) unsynchronized bouts of fruiting (Ficus
natalensis).
Species sampled in more than 1 of the forests tended to
show similar patterns of fruiting among forests. One
exception was the liana, Saba comorensis. Saba comorensis
exhibited highly seasonal fruiting in Mnazini, seasonal but
extended fruiting in Nkano, and continuous fruiting in
Mchelelo (Figure 2.13).
Estimates of fruit biomass. Estimated maximum fruit
numbers, fruit weight and canopy size varied widely between
species and shape categories (Table 2.3), and strongly
influenced monthly biomass indices of fruit (Figure 2.14).


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
BEHAVIORAL AND DEMOGRAPHIC RESPONSES TO HABITAT CHANGE
BY THE TANA RIVER CRESTED MANGABEY
(CERCOCEBUS GALERITUS GALERITUS)
By
Margaret F. Kinnaird
December 1990
Chairperson: Dr. John F. Eisenberg
Major Department: Forest Resources and Conservation
(Wildlife and Range Sciences)
The Tana River crested mangabey (Cercocebus qaleritus
qaleritus) is an endangered primate found only in small
forests of the Tana River in Kenya. The mangabey is afforded
some protection in the Tana River National Primate Reserve.
Since the establishment of the reserve in 1976 the mangabey
population has decreased by 45%. The reduction in habitat
size and quality is believed to have been a major factor
contributing to the population decline. I compare data on
mangabey behavior and demography with a study conducted in
1973-74 to investigate the effects of habitat loss and
reductions in important diet species on mangabey behavior
and risk of extinction.
xx


Figure 6.1. Distribution of estimated female lifetime
reproductive success. Infant values are
rounded to the nearest whole number.


6
occurrence of favorable conditions for seed germination and
seedling establishment. Data from Medley (1990) indicate
that upstream dams may be having an effect; the Tana River
has shortened its course in recent years as a result of less
meandering and sediments loads delivered during floods have
declined. Agricultural practices also influence forest
regeneration. Sites of active forest regeneration (i.e.
point-bars) are selected for farming because much of the
crop production by the local people is dependent on biannual
flooding and the nutrient rich soils near the river.
The Tana River National Primate Reserve (TRNPR; Figure
1.2) was gazetted in 1976 by the government of Kenya to
protect the best remaining stands of riverine forest and
populations of the Tana River crested mangabey and red
colobus monkey (Colobus badius rufomitratus). At the time of
establishment, the reserve occupied approximately 171 km2 of
forest, dry woodland and savanna habitat on the east and
west banks of the river (Marsh 1976). The forest habitat
covered an estimated 17.5 km2, broken into approximately 40
patches. In the 15 years since the establishment of the
reserve, there has been a decline in forest area and
increased fragmentation of the remaining forest stands
(Kinnaird et al. 1990, Medley 1990). Primate populations
also have declined by an estimated 83% for the red
colobus (Marsh 1986, Decker 1989) and 25-45% for the
mangabey (Chapter 6, Marsh 1986). The mangabey population


Figure 5.15. Percent time North and South Mchelelo
groups spent feeding on foods with different
distributions as a function of intergroup
behavior. a) Percent time feeding on foods with
uniform distributions; b) Percent time feeding on
foods with patchy distributions 207
Figure 6.1. Distribution of estimated female lifetime
reproductive success. Infant values are rounded
to the nearest whole number 231
Figure 6.2. Distribution of male lifetime reproductive
success estimated for males with 1, 2 and 3 tenures as
a dominant, breeding group male. a) 1973-74 male
lifetime reproductive success calculated for 1 male:
9.75 females; b) 1988-89 male lifetime reproductive
success calculated for 1 male:5.85 females .... 233
Figure 6.3. Persistence time in years as a function of
population size (N). Dashed lines connect 1973-74
and 1988-89 population sizes to estimated mean
persistence times 238
Figure 6.4. Probability of extinction p(E) within 100,
50, 25, and 10 years (denoted by labeled curves)
for populations with mean persistence times (T) of
200 to 2000 years 241
Figure 6.5. Effective population size (NJ as a
function of number of breeders (Nb) and variance
of male lifetime reproductive success (Vm) .... 244
xix


Short Rains
Long Rains
Morning
Mid-day
Afternoon
Inter-rains
Dry Season


Figure 5.5. Frequency distributions of intensity of quadrat
use. Asterisks indicate the values expected if
quadrat occupancy were random (following a
Poisson distribution).
a) Frequency distribution for N. Mchelelo
group; b) Frequency distribution for S.
Mchelelo group; c) Frequency distribution for
Nkano group.


268
Kenya Ministry of Water Development. 1978. Extreme flow
hydrology of the Tana River Basin. Government of Kenya,
Nairobi.
Kimura, M. and G.H. Weiss. 1964. The stepping stone model
of population structure and the decrease of genetic
correlation with distance. Genetics 49:561-76.
Kinnaird, M.F. 1990a. Pregnancy, gestation and parturition
in free-ranging Tana River Crested Mangabeys (Cercocebus
qaleritus qaleritus). American Journal of Primatology.
In press.
Kinnaird, M.F. 1990b. Estimates of effective population
size for a forest primate, the Tana River crested
mangabey (Cercocebus qaleritus qaleritus). Primate
Conservation, in press.
Kinnaird, M.F., K.E. Medley, B.S. Decker and W.O. Ochiago.
1990. Management Issues and Recommendations for the
Tana River National Primate Reserve, Kenya. Report
submitted to Wildlife Conservation International, New
York Zoological Society and The Kenya Wildlife Service,
Nairobi, Kenya.
Koelmeyer, K.O. 1960. The periodicity of leaf change and
flowering in the principal forest communities of Ceylon
II. The Ceylon Forester 4(4):308-364.
Lande, R. 1988. Genetics and demography in biological
conservation. Science 241:1455-1460.
Lande, R. and G.F. Barrowclough. 1987. Effective
population size, genetic variation, and their use in
population management. Pages 87-124 in M.E. Soul,
editor. Viable Populations for Conservation. Cambridge
University Press, Cambridge, UK.
Lee, P.C. 1984. Ecological constraints on the social
development of vervet monkeys. Behaviour 91:245-262.
Lee, P.C., J. Thornback and E.L. Bennett. 1988. Threatened
Primates of Africa. The IUCN Red Data Book. IUCN
Gland, Switzerland and Cambridge, U.K.
Leighton, M. and D.R. Leighton. 1983. Vertebrate responses
to fruiting seasonality within a Bornean rain forest.
Pages 181-196 in S.L. Sutton, T.C. Whitmore and A.C.
Chadwick, editors. Tropical Rain Forest: Ecology and
Management. Blackwell Scientific Publications, Oxford,
UK.


process of habituating elusive mangabeys would have been a
much harder and more arduous task.
Kim Medley, Barbara Decker, Odhiambo Ochiago, Paula
Kahumbu, Francis Muli, and Alex Njue shared the camp with me
as fellow students and bush companions for various periods
of time. Their company was greatly appreciated, especially
that of Kim Medley, who shared the first year and many first
experiences with me.
My experience living on the Tana River was made rich
and rewarding by a host of kind and generous people. At the
top of the list comes the Kawa family, who allowed me to
construct a small, auxiliary campsite on their shamba. I
could never repay them for their overwhelming hospitality.
I thank Father Spiro, Sisters Romona and Evelyn, and
the late Sister Agnes of the Wema Catholic Mission for
opening their home to me on numerous occasions when I was
surveying the forests of the lower Tana. Their strength will
always be remembered as an inspiration to me.
I single out Tim O'Brien for very special and sincere
thanks. This product would not have been the same without
his input. He has helped in literally all stages--from
assisting in the field to advising on statistical analysis,
writing numerous computer programs, and reading every
chapter. He suffered through all phases of the project and,
in spite of it all, stuck beside me.
vi


215
population sizes (Goodman 1987a, b, Lande 1988, Shaffer and
Samson 1985). Population extinction is by definition a
demographic process (Lande 1988) ; the amount of genetic
variation within a population becomes irrelevant if a
population goes extinct for reasons such as habitat
destruction. Demographic characteristics are subject to
natural selection and are therefore affected by the genetic
composition of a population as well as other factors such as
habitat quality and environmental variation (Shaffer 1981,
Soul 1980, Stearns 1976, Goodman 1987b).
Rigid adherence to the results of either approach is
potentially dangerous given the confusion over which models
produce the best estimates (Harris and Allendorf 1989) and
the uncertainties that arise when using limited data bases
characteristic of long-lived species studied for very short
periods (Dobson and Lyles 1989). Lande (1988) calls for "a
realistic integration of demography and population genetics"
for examining population viability, although he states that
the task of integration has proven formidable and has
largely eluded ecologists. Part of the problem is the lack
of common variables in genetic and demographic models. A
first approach is to perform both analyses and compare the
results.
In this chapter, I conduct demographic and genetic
analyses separately for the Tana River crested mangabey.
Estimates generated from the two approaches are compared and


BIOGRAPHICAL SKETCH
Margaret Kinnaird was born on 26 November 1956 in
Ashland, Kentucky. She spent the first two years of her
undergraduate career at Stetson University in DeLand,
Florida, and graduated with a Bachelor of Science in zoology
from the University Florida in 1978. Margaret worked as a
naturalist guide and conducted research in the Galapagos
Islands, Ecuador, from 1978 to 1980. She completed her
thesis on the cooperative breeding of the Galapagos
mockingbird and graduated from the University of Michigan
with a Master of Science in ecology and evolutionary biology
in 1981. Margaret then worked with the Cooperative Fish and
Wildlife Research Unit at the University of Florida,
researching boat-related mortality of the Florida manatee.
In 1984, Margaret accepted a position with Cornell
University as assistant field director for a project in East
Africa on a cooperatively breeding bird, the White-fronted
bee-eater. In 1985 she returned to the University of Florida
and entered the Ph.D. program in the Department of Wildlife
and Range Sciences.
277


220
Table 6.3. Reproductive parameters
mangabeys.
for male and
female
Parameter
Male
Female
Mean age of first breeding (y)
7
6.5
Maximum recorded lifespan (y)
19
19
Maximum estimated number
of reproductive years
12
12.5
Gestation length (d)

171
Number of offspring
per litter

1
Interbirth interval (y)

2
Birth season
Aug-Apr
*Sources: Harvey and Clutton-Brock (1987) for Cj. qaleritus
spp.. Homewood (1976), Kinnaird (1990a), and this study.


Ill
Although 1988-89 infants were carried by their mothers less
in the first months of infancy, they continued to cling and
suckle in their sixth and seventh months, indicating that
1988-89 infants were weaned later than 1973-74 infants
(Figure 3.7).
In summary, over a 15 year period, mangabeys appear to
have made dramatic shifts in the allocation of time spent in
various activities. The 1973-74 groups were, in general,
less involved in feeding behaviors (eating and foraging)
than the 1988-89 groups, and concomitantly spent more time
in resting behaviors (grooming and inactive). Juveniles and
infants showed the greatest differences in interactions with
other group members. The 1988-89 juveniles and infants
appeared to be less involved in the overall social network
of their groups; juveniles reduced the amount of time
grooming or increased the amount of time within the peer
group, and infants associated more and longer with their
mothers.
Diet
Diet composition. Of the combined eating observations
for the 3 1988-89 groups, 6664 (73%) were plants, 887 (10%)
were animals (primarily invertebrates); 1591 (17%) were
unidentified items and 3 (<1%) were regurgitated matter. I
observed mangabeys eating more than 58 different plant
species during 107 observation days. The N. Mchelelo, S.
Mchelelo and Nkano groups consumed 44, 36, and 31 of these


Ficus natalensis
34
Ficus natalensis
11-20
1
6-10
20+


Table 3.6. Percentage of diet items eaten by month and group. Unidentified items are
not included.
a) N. Mchetelo
Ripe Unripe Ripe Unripe Sprouts Wood
Month
Fruit
Fruit
Seeds
Seeds
Flowers
Leaves
Shoots
Bark
Gun
Animal
N
Jan 88
19.9
17.7
12.8
32.6
3.5
0.0
11.3
0.0
0.0
10.6
141
Feb
13.9
1.0
7.5
59.7
5.5
2.5
2.5
0.0
0.0
7.5
201
Mar
26.9
0.0
1.4
34.4
0.5
0.5
9.9
0.9
0.0
14.6
212
Apr
13.9
6.8
0.7
50.7
0.4
0.7
1.4
1.1
0.0
24.3
280
May
24.6
27.2
2.1
22.1
4.6
1.5
1.5
6.6
0.5
9.2
195
Jun
15.4
5.8
1.2
59.8
0.0
0.4
1.5
1.2
0.0
14.7
259
Jul
21.3
14.9
18.4
9.2
0.0
2.3
14.4
4.6
0.0
14.9
174
Aug
20.6
27.2
1.2
40.5
1.6
1.2
1.2
1.2
2.3 .
3.1
257
Sep
30.6
14.8
6.2
31.6
2.9
2.9
2.9
0.5
0.5
7.2
209
Oct
22.2
6.2
19.6
24.7
7.2
1.0
1.0
4.1
2.1
11.8
194
Nov
10.9
0.4
53.8
21.5
1.1
0.0
0.4
1.8
2.5
7.6
275
Dec
62.8
10.6
9.7
5.3
0.0
1.3
1.3
0.4
0.0
8.4
226
Jan 89
43.7
8.3
2.0
9.9
15.1
2.8
5.2
2.0
0.8
10.3
252
Feb
38.1
5.8
2.7
27.1
4.6
2.1
0.6
1.8
2.4
14.6
328
Mar
6.1
3.4
10.9
51.9
2.7
3.4
1.5
6.1
0.0
14.0
264
b) S.
Mchelelo
Jan 88
21.5
7.7
12.3
41.5
0.0
4.6
10.8
0.0
0.0
1.5
65
Feb
39.5
3.7
3.7
23.7
6.3
0.5
3.7
1.0
0.0
17.9
190
Mar
11.7
19.9
1.1
47.5
0.0
1.9
3.0
2.6
0.0
12.1
265
Apr
7.0
10.2
2.9
57.0
2.2
0.4
6.3
5.1
0.0
8.8
272
Jun
14.7
9.0
7.1
34.6
1.3
2.6
4.5
3.8
0.0
22.4
156
Jul
20.2
4.0
5.8
54.9
0.6
1.2
6.4
2.9
0.0
4.0
173
Sep
44.7
9.7
0.0
28.2
0.9
0.0
7.8
6.8
0.0
1.9
103
Oct
11.7
9.8
29.4
21.6
0.0
5.9
1.9
0.0
7.8
11.8
51
Nov
3.9
3.9
68.4
9.1
0.9
0.0
1.7
0.9
0.0
10.4
231
Dec
49.2
1.5
22.4
4.5
0.0
0.0
0.0
1.5
0.0
20.9
67
Jan 89
50.3
4.6
2.6
10.8
12.3
0.5
9.7
3.6
0.0
5.6
195
115


Figure 2.5.
Mean flower scores averaged over all
month.
a) Mchelelo forest; b) Nkano forest;
forest.
species by
c) Mnazini


251
of suitable forest patches. Demographic and genetic models
of population persistence indicate that the present
population size may not be adeguate to avert loss of genetic
variability or probable extinction. Low dispersal rates
between isolated forest patches or groups on opposite banks
of the river also may pose additional problems for the long
term persistence of the population.
If the Tana mangabey is to persist over the next 80-100
years, an increase in suitable habitat will be necessary.
Presently, however, forest area is decreasing, and the
remaining forest is becoming more fragmented (personal
observation). These processes are attributed, in part, to
changes in the rivers course and patterns of flooding.
Selective felling of trees and forest clearing for
agriculture, however, are more immediate and perhaps more
pervasive causes of habitat degradation and loss.
The extraction of canopy trees for canoes and beehives,
the cutting of trees for building poles, or the intensive
use of palm fronds for roofing and weaving material result
in the loss of trees important in the mangabey's diet and,
in the case of palms, reduced fruit production. Low and
unpredictable crop yields and human population growth
encourage forest removal to maintain adequate agricultural
production. Loss of forest area, for whatever cause, is not
balanced by regeneration of new forest (Hughes 1985) ;
forests are being lost faster than they are regenerating.


Mchelelo
Temperature ( C)
10
0 "1 I 1 I 1 ¡ i I I I I I I I 1 T 1 1
SONDJFMAMJ JASONDJFMA
8 8 8
7 8 9
Nkano
Temperature ( C)
40 i
1
0 ~ Mean maximum
Mean minimum
0 r T T 1
SONDJFMAMJJASONDJFMA
8 8
7 8
Month
8
9


Figure 2.7. Mean monthly fruit scores from Mchelelo
and Mnazini forests for Mimusops fruticosa and
Acacia robusta. seasonally fruiting species with 1
peak in fruiting per year. a) Mimusops fruiticosa
in Mchelelo forest; b) Acacia robusta in Mchelelo
forest; c) Mimusops fruticosa in Mnazini forest;
d) Acacia robusta in Mnazini forest 41
Figure 2.8. Mean monthly fruit scores from Mchelelo
and Mnazini forests for Garcinia livingstonei and
Sorindeia madaqascariensis. seasonally fruiting
species with 2 fruiting peaks per year, a) Garcinia
livingstonei in Mchelelo forest; b) Sorindeia
madaqascariensis in Mchelelo forest; c) Garcinia
livingstonei in Mnazini forest; d) Sorindeia
madaqascariensis in Mnazini forest 43
Figure 2.9. Mean monthly fruit scores for Phoenix
reclinata. a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest 45
Figure 2.10. Mean monthly fruit scores for Diospyros
mespiliformes. a) Mchelelo forest; b) Nkano forest;
c) Mnazini forest 47
Figure 2.11. Mean monthly fruit scores for 3 species
showing patterns of continuous fruit production,
a) continuous fruiting due to asynchrony among
individuals of the species (Ficus natalensis); b)
continuous fruiting due to retention of fruits
over time (Hyphaene compressa); c) continuous
fruiting due to rapid turnover of fruits (Oncoba
spinosa) 49
Figure 2.12. Mean monthly fruit scores for Ficus
sycomorus. a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest 51
Figure 2.13. Seasonality of mean monthly fruit scores
for Saba comorensis in the 3 study forests, a)
seasonal fruiting in Mnazini forest; b) seasonal
but extended fruiting in Nkano forest; c)
continuous fruiting in Mchelelo forest 53
Figure 2.14. Contribution to monthly fruit biomass
indices by different species categories, a)
Mchelelo forest; b) Nkano forest; c) Mnazini
forest 56
xv


269
Levey, D.J. 1988. Spatial and temporal variation in Costa
Rican fruit and fruit-eating bird abundance. Ecological
Monographs 58:251-269.
Lovejoy, T.E., R.O. Bierregaard Jr., A.B. Rylands, J.R.
Malcolm, D.E. Quintela, L.H. Harper, K.S. Brown Jr.,
A.H Powell, G.V.N. Powell, H.O.R. Schubart and M.B.
Hays. 1986. Edge and other effects of isolation on
Amazon forest fragments. Pages 257-285 in M.E. Soule,
editor. Conservation Biology. Sinauer Associates,
Inc., Sunderland, MA.
Marsh, C.W. 1976. A Management Plan for the Tana River
Game Reserve. Report to the Kenya Department of
Wildlife Conservation and Management, Nairobi.
Marsh, C.W. 1978a. Ecology and Social Organization of the
Tana River Red Colobus (Colobus badius rufomitratus).
PhD dissertation, University of Bristol, Bristol, UK.
Marsh, C.W. 1978b. Comparative activity budgets of red
colobus. Pages 249-251 in D.J. Chivers and J. Herbert,
editors. Recent Advances in Primatology, Vol. 1,
Behaviour. Academic Press, London, UK.
Marsh, C.W. 1981a. Time budget of the Tana River Red
Colobus. Folia Primatologica 35:30-50.
Marsh, C.W. 1981b. Ranging behaviour and its relation to
diet selection in Tana River red colobus (Colobus badius
rufomitratus). Journal of Zoology 195:473-492.
Marsh, C.W. 1986. A resurvey of Tana River primates and
their habitat. Primate Conservation 7:72-81.
McCurrach, J.C. 1960. Palms of the World. Harper and
Brothers, New York, NY.
McKey, D.B. 1979. The distribution of secondary compounds
within plants. Pages 98-101 in G.A. Rosenthal and D.H.
Janzen, editors. Herbivores, Their Interaction with
Secondary Plant Metabolites. Academic Press, London, UK.
McKey, D.B., J.S. Gartlan, P.G. Waterman and G.M. Choo.
1981. Food selection by black colobus monkeys (Colobus
satanas) in relation to plant chemistry. Biological
Journal of the Linnean Society 16:115-146.
Medley, K.E. 1990. Forest Ecology and Conservation in the
Tana River National Primate Reserve, Kenya. Unpublished
Ph.D. dissertation, Michigan State University.


149
Discussion
Phoenix reclinata is a very useful and important plant
species for the people of the lower Tana River. Although
direct economic benefits are limited to the sale of mats and
baskets, the indirect benefits of free construction material
and dietary supplements are numerous.
Although many of the human uses involve only the
removal of leaves, several of the harvesting technigues are
destructive. Cutting palm trunks for building, palm heart
extraction, beer production, or for access to high growing
leaves, damages the apical meristem and prevents further
growth. Excessive leaf removal also may have a significant
impact on reproduction and growth. Leaves are removed
primarily from the 2 largest size classes. Data from
Mchelelo indicate that fruit production by P. reclinata is
rare until after a small trunk has begun to form.
Preferential and excessive harvesting of the reproductive
size classes may prevent or limit fruit production; no
individuals were observed fruiting within 6 months after
complete or near complete removal of leaves. This is also
true for other palm species. Excessive harvesting of Sabal
causiarum Cook leaves for hats in Puerto Rico also
influences flowering and fruiting (Read 1988), and Woka
palms (Livistona rotundifolia Lamark) of Indonesia
frequently die as the result of excessive pruning (Davis
1988) .


57
Palms, particularly Phoenix reclinata. often accounted for a
large proportion of the monthly biomass indices in Mchelelo
because of their potential for large fruit crops and their
high density in the forest. Although ramiflorous species
were not abundant in Mchelelo, their capacity for large
fruit crops results in a relatively large contribution to
the monthly biomass indices. In Nkano, ramiflorous species
such as Pachvstela msolo and Ficus svcomorus were abundant
and they dominated the monthly biomass index.
Although fewer species were sampled in Nkano, the
forest produced a much greater biomass of fruits than either
Mchelelo or Mnazini for all months of the year with the
exception of August. The high fruit biomass production by
Nkano was primarily the result of high densities of Ficus
svcomorus. a species that produces large fruit crops
throughout the year, with a decline in production during
June and July (Figure 2.12). Mchelelo had a more consistent
production of fruits across the months due to the
availability of palm fruits between May and October, a
period of relative scarcity in the other 2 forests. Mnazini,
with a low abundance of figs and palms, was more seasonal in
fruit production than Mchelelo or Nkano and experienced a 4
month period (May-Aug) of fruit scarcity.
Unripe fruits contributed the most to the biomass index
over all months in all study forests; ripe fruits were
entirely unavailable during several months (Figure 2.15).


LIST OF FIGURES
Figure 1.1. Adult male Tana River crested mangabey,
Cercocebus galeritus qaleritus. seated in a
Hvpheane compressa palm 3
Figure 1.2. Map of Kenya and location of the Tana
River National Primate Reserve (TRNPR; map
modified from Hughes 1985) 8
Figure 2.1. Map of the study area showing the
Mchelelo, Mnazini, and Nkano forests within the
Tana River National Primate Reserve 16
Figure 2.2. Number of individuals/0.25 ha of 4 common
species within Mchelelo forest showing clumped
spatial distributions a) Hyphaene compressa;
b)Alangium salviifolium; c)Phoenix reclinata; d)
Oncoba spinosa 30
Figure 2.3. Number of individuals/0.25 ha of 4 common
species within Nkano forest showing clumped
spatial distributions, a) Ficus svcomorus; b)
Alangium salviifolium; c) Phoenix reclinata; d)
Pachvstela msolo 32
Figure 2.4. Examples of randomly distributed Ficus
spp. a) Ficus natalensis in Mchelelo forest; b)
Ficus svcomorus in Mchelelo forest; c) Ficus
natalensis in Nkano forest 34
Figure 2.5. Mean flower scores averaged over all
species by month, a) Mchelelo forest; b) Nkano
forest; c) Mnazini forest 36
Figure 2.6. Mean fruiting scores averaged over all
species by month, a) Mchelelo forest; b) Nkano
forest; c) Mnazini forest 38
xiv


18
reproductive size for each species were enumerated. Foot
trails cut along compass bearings at 50m intervals were used
to delineate quadrats of 0.25 ha. I enumerated 23 plant
species within 65 quadrats (16.25 ha) for Mchelelo and 14
species within 54 quadrats (13.5 ha) for Nkano. Nine species
present in Mchelelo were not found in Nkano. Average
density was calculated by summing the number of individuals
for a particular species in all quadrats and dividing by the
total number of hectares sampled.
Spatial Distribution. I used Morisita's (1959) index of
spatial pattern (Ig) to measure the distribution of each
species in the two forests. I chose Morisita's index over
the more commonly used variance to mean ratio because many
of the species I considered had calculated means less than
1.0 individual/ha. Greig-Smith (1983) has shown the variance
to mean ratio to behave erratically when the mean is less
than 1.0. Morisita's index provides a measure of the extent
to which individuals of a particular species are nonrandomly
distributed among identically-sized quadrats. The index is
calculated as
I§ = [^ni(ni l)/n (n-1) ] N
where n^ is the number of individuals in the itl_1 quadrat, n
is the total number of individuals in all quadrats, and N is
the number of quadrats. Morisita's I§ ranges from 1/N to N;
values approaching 1.0 indicate a random distribution and
those less than or greater than 1.0 indicate regular and


Figure 5.9. Frequency distributions of half-hour turning
angles in 10 degree intervals,
a) Frequency distribution for N. Mchelelo
group; b) Frequency distribution for S.
Mchelelo group; c) Frequency distribution for
Nkano group.


36
Mchelelo
Flower score
6-
4 -
3-
Nkano
Flower score
8 6 8
7 8 9
Mnazini
Flower score
8 8 8
7 8 9
Month


168


132
the probability of finding another individual not only in
fruit, but bearing ripe fruits. More synchronously ripening
species (e.g. Garcinia livingstonei and Phoenix reelinata)
that offered rich sources of ripe fruits often were
unavailable to mangabeys because of competition with yellow
baboons. Baboons, in groups of more than 80 individuals,
were always successful in supplanting mangabeys from
contested food patches. Mangabeys therefore may have been
unable to wait for preferred ripe fruit items and switched
to more abundant and less contested diet items.
Diet quality and consequent shifts in activity budgets
may influence infant dependency and schedules of
development. Silk and Kraemer (1978) showed that chimpanzee
mother/infant associations were influenced by environmental
conditions. Infants gained independence more quickly among
captive, provisioned chimps than their wild counterparts.
Mangabey infants in 1988-89 associated with their mothers
for longer and were weaned at a later age than 1973-74
infants. If the opportunity for an infant to move away from
its mother and interact with other group members is greater
when a mother is inactive, then the more mobile 1988-89
mothers may have limited those opportunities. Struhsaker
(1980) found that red colobus (Colobus badius) infants
played more than infant redtail monkeys (Cercooithecus
ascanius), and speculated that the more stationary red


14
each forest and examine the associations between flowering
and fruiting phenology, river flow, temperature and
rainfall. I also make comparisons with similar data
collected during 1973-74 (Homewood 1976) to determine if
fruiting patterns differ between the 2 time periods.
Methods
Study Sites
Three forests were chosen for intensive study:
Mchelelo, Nkano and Mnazini (Figure 2.1). Baseline data were
available for these forests from Homewood's (1976) 1973-74
study. The forests vary in size, vegetation, degree of
flooding, human disturbance, and mangabey density. Mchelelo
is a small forest (17 ha) with a diverse plant species
composition. It is on a raised river levee and experiences
minimal flooding. Mchelelo presently is occupied by 1-2
groups of mangabeys, depending on the season and is not
disturbed by local people. Plant species composition has
changed since the mid-1970s (Marsh 1986, Medley 1990) and
several important mangabey food species have declined in
abundance.
Nkano is a larger forest (38 ha) with a much less
diverse plant species composition than Mchelelo (Homewood
1976, Medley 1990). Nkano presently is occupied by 2
mangabey groups and experiences a high degree of flooding
and human exploitation. Although forest area in Nkano has
decreased by an estimated 24% due to a fire in 1981 (Marsh


217
Table 6.1. Minimum and maximum numbers of mangabey groups
censused in forests on the east and west banks of the lower
Tana River in 1988-89 and 1973-74. Forests not censused or
indicated with (-); those no longer standing indicated with
(x). Data from 1973-74 are taken from Homewood (1976) and
Marsh (1976, 1978).
East Bank
West Bank
Forest Name
Year
Forest Name Year
1988
1973
1988
1973
-89
-74
-89
-74
Kanj onga
2
2
Guru
2
3-4
Wenj e
2
2
Mchelelo
1
2
Maroni
1
1
Sifa
0
1
Guru
1
1
Congolani
0
1
Sifa
2-3
2-3
Congolani C.
2
2
Baomo
1
1
Maridadi
0
1
Marembo
1
1
Baomo N.
1
1-2
Mnazini
2
2-3
Baomo S.
3-4
5-6
Bubesa
0
1
Kitere
1
1
Lemu
-
1
Mnazini/Kano
2
2
Wema 1
1
1
Mnazini
1
2
Home
1
2-3
Woodlands
1
1
Lango la
Simba -
1
Kinyadu
0
1-2
Bubesa
0
1
Subtotal
14-15
18-19
Matalani
X
1
Nkame
1
0
Marembo
1
1-2
Maziwa
1-2
1-2
Hewani 1
1
1-2
Hewani 2
0
1
Subtotal
18-20
29-3'
Totals for both banks 1988-89: 32-35 groups
1973-74: 47-57 groups


75
natalensis), however, can affect the temporal variation in
fruit availability by dampening the severity of periods of
scarcity. A critical number of individuals, however, is
required to ensure that fruits are available each month. As
these species decline in abundance, the probability that the
species population has an individual bearing fruits during
bottlenecks decreases. This is apparent in the monthly
phenological scores for F. svcomorus individuals from
Mnazini (n=7) and Mchelelo (n=10) forests (Fig. 12). Figs
were unavailable during 3 months in Mnazini but at least 1
fig was fruiting in all months sampled in Mchelelo.
Species such as Phoenix reclinata and Diospyros
mespiliformes that fruit at times when other species are not
in fruit also can form important and critical food resources
for mangabeys during months of otherwise low fruit
production. The large fruit crop of P. reclinata. coupled
with its extremely high density in the Mchelelo forest,
resulted in abundant fruit during months of relative food
scarcity in Nkano and Mnazini, forests with lower densities
of reproductive palms (Fig. 13). The abundance of species
such as P. reclinata and Ficus spp. may be a key factor in
explaining the presence or absence of mangabeys in the
various forest patches along the Tana River.
Many tropical forest species show adaptations toward
fruit-drop and/or seed germination near the beginning of
heavy rains (Rathche and Lacey 1985). Flowering occurs


210
particular environmental or social conditions. Butynski
(1990), Gartlan and Brain (1968), and Yoshiba (1968)
described facultative territoriality for several primate
species in response to differing population densities.
Others (Hamilton et al. 1976, Kavanagh 1981, Chapman and
Fedigan 1984) have documented variable territoriality
between species populations in response to differing food
resources. Competition for limited resources is essential
for the evolution of territoriality, but whether
territoriality will evolve is ultimately dependent on the
'economic defendability' of these resources (Brown 1964,
Horn 1968). Mitani and Rodman (1979) developed an "index of
defendability" (D) based on the frequency with which a group
encounters the boundaries of its range; they consider the
potential for frequent contacts with range boundaries (range
defense) high when D > 1.0. Indices of defendability for
North Mchelelo (D=4.5), South Mchelelo (D=1.5) and Nkano
(D=4.5) groups indicate that their home ranges are
defendable. Although the conditions for territoriality by
mangabeys are present, defense should be expressed only when
resources are limited.
My results and those of Homewood's (1976) suggest that
seasonal variation in food availability and distribution
result in facultative defense of space by Tana mangabeys.
When fruit resources are low, or limited, intergroup
interactions are rare and groups actively avoid one another.


85
Data Analysis
I estimated the amount of time spent in different
activities by calculating the proportions of each activity
recorded for all individuals during each day of a month, and
taking the mean of the daily proportions. Data also were
summarized by time-of-day by calculating the proportions of
each activity recorded during each hour of the day in a
given month and taking the mean of the hourly estimates.
Mean hourly estimates were grouped into 3 time periods for
analysis: morning (0700-1100 hrs.), mid-day (1100-1400 hrs.)
and afternoon (1500-1800 hrs.). I further grouped data into
4 seasons: short rains (Nov-Jan), inter-rains (Feb), long
rains (Mar-May), and dry season (Jun-Oct). Seasonal activity
budgets were calculated by taking the mean of the
appropriate monthly estimates. Finally, annual activity
budgets were calculated by taking the mean of the all
monthly estimates. This procedure corrects for the potential
bias due to an unequal distribution of observation time
among different months or hours of the day (Altmann and
Altmann 1970).
I used analysis of variance methods (ANOVA, MANOVA) to
simultaneously evaluate the effects of suites of independent
variables on monthly activity budgets and diet composition
(Neter et al. 1985). The use of MANOVA reduced the
likelihood of Type I error resulting from multiple tests,
and allowed simultaneous measurement of the effects of


262
Belovsky, G.E. 1987. Extinction models and mammalian
persistence. Pages 35-58 in M.E. Soul, editor. Viable
Populations for Conservation. Cambridge University
Press, Cambridge, UK.
Bennett, E.L. 1986. Environmental correlates of ranging
behaviour in the banded langur, Presbvtis melalophos.
Folia Primatologica 47:26-38.
Berenstain, L. 1986. Responses of long-tailed macaques to
drought and fire in Eastern Borneo: A preliminary
report. Biotropica 18(3):257-262.
Bernstein, I.S. 1976. Activity patterns in a Sooty
mangabey group. Folia Primatologica 26:185-206.
Borchert, R. 1983. Phenology and control of flowering in
tropical trees. Biotropica 15(2):81-89.
Brown, J.L. 1964. The evolution of diversity in avian
territorial systems. Wilson Bulletin 76:160-169.
Butynski, T.M. 1990. Comparative ecology of blue monkeys
(Cercopithecus mrtis) in high- and low-density
subpopulations. Ecological Monographs 60(1)1-26.
Chapman, C. and L.M. Fedigan. 1984. Territoriality in the
St. Kits Vervet, Cercopithecus aethiops. Journal of
Human Evolution 13:677-686.
Charnov, E.L. 1976. Optimal foraging: the marginal value
theorem. Theoretical Population Biology 9:129-136.
Chepko-Sade, B.D., W.M. Shields, J. Berger, Z.T. Halpin,
W.T. Jones, L.L. Rogers, J.P. Rood and A.T. Smith.
1987. The effects of dispersal and social structure on
effective population size. Pages 287-322 in B.D.
Chepko-Sade and Z.T. Halpin, editors. Mammalian
Dispersal Patterns: The Effects of Social Structure on
Population Genetics. University of Chicago Press,
Chicago, IL.
Chivers, D.J. 1974. The siamang in Malaya: a field study
of a primate in a tropical rainforest. Contributions to
Primatology 4:1-335. S. Karger, Basel.
Chivers, D.J. and J.J. Raemaekers. 1980. Long-term changes
in behavior. Pages 209-260 in D.J. Chivers, editor.
Malayan Forest Primates: Ten Years' Study in Tropical
Rain Forest. Plenum Press, New York, NY.


82
I ask whether short-term associations between activity
budgets and food resources are consistent with those
observed after long-term habitat change.
Methods
Data Collection
I studied 2 groups of mangabeys in the Mchelelo forest
and 1 group in the Nkano forest. Group size averaged 18
(range 13-19) and 28 (range 25-28) for the North and South
Mchelelo groups, respectively, and 16 (range 12-17) for the
Nkano group. Activity and diet data for 2 mangabey groups
ranging in the Mchelelo and Nkano forests during 1973-74
were taken from Homewood (1976). Group size averaged 36 and
17 for these groups, respectively. Group age-sex composition
did not differ significantly between the 2 studies (G=8.4,
df=5, p>0.05). Although the 1973-74 Nkano group spent the
majority of the observation months (66%) in an adjacent
forest (Mnazini) 0.5 km south of Nkano, monthly activity
budgets were not significantly different when the group
ranged entirely in Nkano or Mnazini (G=3.1, df=7, p>0.05); I
therefore used all monthly data for this group in the
following analyses.
I collected data over 15 months on the Mchelelo groups
(January 1988-March 1989) and 13 months on the Nkano group
(March 1988-March 1989). Data were collected using methods
similar to the previous study by Homewood (1976). I followed
groups for 3 consecutive days each month from 0700 to 1830


Mchelelo
Fruit score
3-1
2
Fruit score
Nkano
I r
A M
Mnazini
Fruit score
3-|
llllllll.l.a
NDJFMAMJJASON
I I 1 1 1 r
D J F M A M
8
7
8
8
Month
8
9


Homewood, K.M and W.A. Rodgers. 1981. A previously
undescribed Mangabey from southern Tanzania.
International Journal of Primatology 2:47-55.
267
Horn, A. D. 1987. The socioecology of the black mangabey
(Cercocebus aterrimus) near Lake Tumba, Zaire. American
Journal of Primatology 12:165-180.
Horn, H. 1968. The adaptive significance of colonial
nesting in the brewers blackbird (Euphagus
cvanocephalus). Ecology 49:682-694.
Hubbel, S.P. and R.B. Foster. 1986. Commonness and rarity
in a Neotropical forest: Implications for tropical tree
conservation. Pages 205-232 in M.E. Soule, editor.
Conservation Biology. Sinauer Associates, Inc.,
Sunderland, MA.
Hughes, F.M.R. 1985. The Tana River Floodplain Forest,
Kenya: Ecology and the Impact of Development.
Unpublished Ph.D. dissertation, University of Cambridge,
Cambridge, UK.
Hughes, F.M.R. 1987. Conflicting uses for forest resources
in the Lower Tana River basin of Kenya. Pages 211-227
in D. Anderson and R. Grove, editors. Conservation in
Africa. Cambridge University Press, Cambridge, UK.
Hughes, F.M.R. 1990. The influence of flooding regimes on
forest distribution and composition in the Tana River
floodplain, Kenya. Journal of Applied Ecology 27:475-
491.
Isbell, L.A. 1983. Daily ranging behavior of red colobus
(Colobus badius tephrosceles) in Kibale Forest, Uganda.
Folia Primatologica 41:34-48.
International Union for the Conservation of Nature and
Natural Resources (IUCN). 1976. The Tana Mangabey.
Red Data Book, Vol. 1: Mammalia. IUCN, Gland,
Switzerland.
Iwamoto, T. 1975. Food resource and the feeding activity.
Proceedings of the Fifth International Congress of
Primatology 475-481.
Jolly, A. 1972. Troop continuity and troop spacing in
Propithecus verreauxi and Lemur catta at Berenty
(Madagascar). Folia Primatologica 17:335-362.
Kavanagh, M. 1981. Variable territoriality among Tantalus
monkeys in Cameroon. Folia Primatologica 36:76-98.


Table 2.4. Spearman's rank correlation coefficients for mean monthly environmental
variables and mean monthly fruiting scores for Mchelelo forest. MINTEMP=minimum monthly
temperature; RAIN=total monthly rainfall (mm); RAINLAG1, RAINLAG2 AND RAINLAG3=monthly
rainfall lagged by 1,2, and 3 months, respectively; RAINADV=previous months' rainfall;
FLOW=mean monthly river flow (mcm); FLOWLAG1, FLOWLAG2, FLOWLAG3=mean monthly river flow
lagged by 1,2 and 3 months, respectively; FLOWADV^previous months' mean river flow.
RIPE AND UNRIPE
RIPE ONLY
RIPE AND UNRIPE
RIPE AND UNRIPE
Total Continuous Seasonal
Total Continuous Seasonal
Seasonal
(w/o P. reelinafa)
P. reclinata
MINTEMP
0.54*
ns
ns
ns
ns
ns
ns
ns
RAIN
ns
ns
ns
ns
ns
ns
ns
ns
RAINLAG1
ns
ns
ns
ns
ns
ns
ns
ns
RAINLAG2
ns
ns
ns
ns
ns
ns
ns
ns
RAINLAG3
ns
ns*
ns
ns
ns
ns
ns
ns
RAI NADV
ns
.74
ns
ns
ns
ns
ns
ns
FLOW
ns
ns
ns
ns
ns *
ns
ns
ns
FL0ULAG1
ns
ns
ns
ns
0.64
ns
ns
0.78
FL0WLAG2
ns
ns
ns
ns
ns
ns
ns
0.83
**
0.79
FL0WLAG3
ns
ns
ns
0.48
ns
ns
-0.59
FLOWADV
ns
ns
ns
ns
ns
ns
ns
ns
*. P^.05
p<0.001
cr>
oo


28
madagascariensis. Diospyros inespiliformis. Oncoba spinosa
and Garcinia livingstonei) occurred at much higher densities
in Mchelelo than Nkano (Table 2.2).
The majority of species in both forests were patchily
or randomly distributed in space and no species showed a
uniform spatial pattern (Table 2.1, Figures 2.2-2.4). The
tendency towards patchy distributions of species was
consistent on a larger spatial scale: quadrats with high
numbers of individuals of a species tended to be clumped
throughout the forests (Figures 2.2 and 2.3). Five species
showed different patterns of spatial distribution between
the 2 forests. Diospyros mespiliformis. Garcinia
livingstonei. and Blighia uniiugata occurred in clumps in
Mchelelo but were randomly distributed in Nkano. Pachvstela
msolo and Ficus svcomorus. on the other hand, occurred in
clumps in Nkano but were randomly distributed in Mchelelo.
These apparently random distributions, however, may be an
artifact of small sample size.
Temporal Distribution of Resources
Patterns of flowering and fruiting. Flower and fruit
scores averaged over all species by month suggest
seasonality of flowering and fruiting in the 3 forests
(Figures 2.5 and 2.6). Increased flowering occurred from
August to December with minor increases in February and
March. Fruiting was high from January to May of 1988, and
November to March of 1988-89; low levels of fruiting


223
Those males who had a cumulative tenure of less than 9 years
after 2 runs were given a final opportunity to secure a 3rd
group for a maximum of 6 years and a total tenure of no more
than 12 years. Tenure time was multiplied by the mean number
of reproductive females available to a male and the mean
number of infants produced annually by a female. Infant
production then was devalued by the probability of survival
of male and female infants. The mean and variance of infant
production by males was calculated for three scenarios:
males attain tenure only once in their lifetime; males
attain breeding status in a 2nd group after losing tenure in
the first; males secure multiple groups (maximum of 3)
throughout their reproductive lifetime.
Population subdivision
The effective size of a population will be influenced
greatly by the degree of population subdivision. If gene
flow between subpopulations, or demes, is so low that they
become fixed for alternative alleles, then the effective
size of each deme, rather than that of the population as a
whole, may need to be considered. Because the Tana mangabey
population is distributed among forest patches separated by
a major river, limited and non-random gene flow among
forests and particularly across the river could create
isolated subpopulations. I used a one-dimensional stepping-
stone model developed by Kimura and Weiss (1964) and
Wright's (1943) island model to determine whether or not


Figure 5.4. Taut-string line (shaded with diagonal lines)
enclosing all ad libitum sightings of South
Mchelelo group when the group ranged outside
the Mchelelo forest.


90
Table 3.1. Mean percent time spent in each activity by
month for mangabeys in the 1988-89 N. Mchelelo, S. Mchelelo,
and Nkano study groups. Standard errors are in parentheses.
See text for definition of activities.
a) N. Mchelelo
Mon/Y N(d) Eat Forage Move Inactive Autogroom Allogroom Soc/Sex
Jan
88
3
50
.9
(4
.1)
13.
.9
(2.7)
20,
.7
(2.
.2)
4.
,5
(1
.0)
0
.5
(0.3)
2,
.8
(0.6)
6.
.6
(1
.5)
Feb
3
52
.7
(4
.7)
18.
.1
(2.4)
13
.9
(2
.9)
4.
,0
(0
.4)
1
.3
(0.7)
4.
.2
(1.4)
5.
.1
(1
.8)
Mar
3
43
.5
(2
.5)
19,
.1
(2.3)
19
.3
(3,
.0)
4.
.6
(1
.2)
0
.8
(0.5)
6,
.2
(2.0)
5.
.4
(1
.9)
Apr
3
54
.6
(1
.9)
12,
.5
(1.5)
16
.5
(0.
.7)
8.
,4
(0
.9)
0
.9
(0.6)
2.
.4
(0.9)
4.
.4
(1
.7)
May
3
42
.4
(3.
.9)
21.
.8
(0.8)
12,
.8
(1,
.7)
8.
,7
(3
.3)
0
.7
(0.2)
5.
.5
(0.9)
7.
.5
(0,
.7)
Jun
3
52
.3
(1,
.2)
14,
.8
(0.6)
16
.6
(2.
.2)
5.
,5
(1
.4)
0
.7
(0.3)
4,
.3
(1.5)
4.
.9
(1
.1)
Jul
3
48
.7
(1
.7)
18,
.2
(1.5)
13
.1
(1,
.7)
8.
,7
(0
.5)
1
.6
(0.9)
3,
.0
(0.6)
6.
.6
(2
.0)
Aug
3
46,
.6
(3
.2)
17,
.0
(2.5)
13,
.2
(1,
.1)
9.
.8
(2
.5)
1
.4
(0.5)
4,
.2
(1.0)
7.
.1
(1
.7)
Sep
3
44
.1
(2.
.1)
21,
.6
(1.6)
11.
.9
(1,
.4)
8.
.2
(1,
.3)
0
.7
(0.4)
7.
.1
(1.1)
6.
.0
(1
.5)
Oct
3
42,
.7
(1
.7)
19.
.6
(3.1)
15,
.2
(2
.0)
7.
,2
(1,
.9)
1,
.2
(0.7)
4.
.3
(1.8)
9.
.7
(1
.1)
Nov
3
49,
.8
(3
.0)
8.
.9
(1.4)
10.
.2
(2,
.2)
8.
,1
(0.
.4)
1,
.7
(0.7)
5.
.2
(0.9)
15.
.7
(1
.5)
Dec
3
48,
.2
(2
.1)
11.
.1
(0.4)
14.
.2
(2.
.1)
5.
,2
(1,
.1)
3.
.3
(0.9)
4.
.1
(1.2)
13.
,7
(0,
.6)
Jan
89
3
46,
.9
(3.
.2)
15,
.1
(1.9)
10,
.6
(1.
.2)
8.
8
(0,
.9)
1,
.3
(0.3)
3,
.8
(1.9)
12.
.7
(0
.5)
Feb
3
53,
.0
(1
.6)
13.
.3
(2.4)
12,
.9
(0.
.4)
4.
8
(1.
.3)
1.
.3
(0.3)
4.
.8
(2.1)
9.
.4
(0.
.6)
Mar
3
44,
.6
(3,
.0)
16.
.0
(1.7)
9.
.8
(1,
.4)
9.
6
(2.
.4)
1.
.9
(0.5)
4.
.2
(1.3)
12.
.9
(1.
.7)
b)
S.
, Mchelelo
Jan
88
2
35.
.6
(5.
.2)
19.
.4
(1.6)
30.
.0
(2.
.9)
7.
8
(2.
.9)
0.
.9
(0.3)
1.
.9
(1.9)
4.
,4
(0.
.6)
Feb
3
45,
.4
(4.
.6)
23.
.3
(2.3)
19.
.7
(6
.5)
2.
2
(1.
.1)
0.
.3
(0.2)
1.
.8
(0.9)
6.
9
(2.
.7)
Mar
3
45.
.4
(3
.9)
18.
.7
(0.4)
18.
.7
(0,
.6)
7.
5
(2.
.0)
0.
.3
(0.2)
5.
,9
(1.2)
3.
,4
(1,
.7)
Apr
3
49.
.5
(1.
.7)
12.
,2
(1.2)
20.
.4
(2.
.3)
5.
2
(1.
.8)
1.
.8
(0.6)
3.
,0
(1.0)
7.
3
(1.
.3)
Jun
3
44.
.5
(4.
.6)
15.
.0
(1.9)
18.
.6
(4.
,2)
7.
3
(1.
.4)
1.
.4
(0.2)
4.
.0
(1.4)
8.
5
(1.
.3)
Jul
3
51.
.1
(6.
.4)
16.
.4
(2.6)
19.
.4
(2.
,7)
2.
4
(1,
.5)
1.
.2
(0.2)
3.
9
(0.8)
4.
8
(1.
.4)
Sep
3
34.
.6
(1.
.4)
12.
,6
(1.9)
26.
.5
(2.
,3)
8.
3
(0.
.7)
0.
.6
(0.3)
3.
8
(0.9)
11.
6
(0.
4)
Oct
1
30.
.1
24.
,5
-
23.
,1
7.
4
2.
.6
-
3.
,1
-
9.
2
Nov
3
48.
.3
(1.
.2)
12.
,4
(1.0)
14.
,0
(0.
,9)
5.
0
(0.
.3)
1.
.9
(0.1)
4.
.2
(0.4)
13.
9
(0.
.6)
Dec
1
43.
,8
12.
,4
-
15.
,1
7.
3
3.
,7
-
5.
,0
-
14.
6
Jan
89
3
41.
.6
(1.
.7)
14.
,6
(1.9)
17.
,7
(5.
,1)
4.
5
(2.
.5)
2.
.4
(0.7)
7.
1
(1.9)
11.
4
(0.
.8)
c) Nkano
Mar 88 3
Apr 3
Jun 3
Jul 3
Aug 1
Sep 3
Oct 3
Nov 3
Dec 3
Jan 3
Feb 3
Mar 89 3
46.5 (1.6)
42.3 (3.6)
45.6 (2.4)
46.4 (0.5)
50.31
52.4 (2.0)
45.7 (2.2)
58.7 (2.4)
40.4 (0.4)
35.7 (1.4)
45.4 (0.9)
48.3(5.8)
17.0 (3.1)
13.6 (1.1)
18.0 (1.1)
16.1(1.9)
15.9 -
14.6 (1.6)
13.8(0.7)
9.0 (1.6)
9.8(1.0)
17.4 (1.2)
15.5 (1.4)
17.4(5.7)
16.5(3.9)
22.4 (2.0)
20.1 (2.5)
19.7 (1.0)
11.5
17.3 (1.3)
19.3 (3.3)
10.8 (1.1)
13.1 (2.9)
16.4 (0.4)
14.4 (0.8)
10.1 (0.8)
6.4 (0.3)
9.4 (0.4)
6.1 (0.7)
8.0 (0.7)
7.0
4.2 (0.8)
7.3 (0.3)
7.2(1.8)
10.4(0.7)
13.3 (1.8)
10.4 (1.3)
6.8(0.9)
1.7 (0.7)
1.4 (0.2)
2.0 (0.4)
0.7 (0.5)
1.3 -
1.3(0.1)
1.2(0.4)
1.2(0.3)
2.5 (1.9)
1.2 (0.9)
2.8 (1.2)
2.2 (0.1)
6.7(1.0)
4.4 (1.6)
2.0 (0.5)
4.1 (1.5)
3.2 -
4.5 (0.4)
6.3 (0.5)
2.4 (0.7)
6.7 (0.6)
4.8 (0.6)
3.1(1.0)
5.5 (2.5)
5.1 (1.6)
6.4 (1.4)
6.1 (0.4)
4.7(1.2)
10.0 -
5.5 (0.5)
6.0 (0.5)
10.5(0.9)
16.4(3.0)
11.2(1.5)
8.1 (0.4)
9.2 (1.4)


177
of Saba comorensis in high- versus low-use quadrats (t=2.6,
df=64, p=0.02), and the Nkano group ate significantly higher
proportions of Pachvstela msolo (t=3.0, df=76, p=0.004) and
Aporrhiza paniculata in high- versus low-use quadrats
(t=3.5, df=7 6, p=0.002) .
In general, there appear to be no particular attributes
that characterize high- and low-use quadrats. Use of space
by mangabeys may be influenced more by the seasonal
availability of food resources.
Seasonal variation. The diversity of quadrats used by
mangabey groups (Shannon-Wiener Index of diversity, H')
varied between months; in some months mangabeys spent a
disproportionate amount of time in few quadrats (Table 5.1).
If mangabeys track seasonally available fruit resources,
then their use of space at any given time should correlate
with the spatial distribution of fruits being eaten at that
time (Robinson 1986). I examined associations between
frequency of quadrat use by each mangabey group and the
frequency of occurrence in a given quadrat of the top 5 diet
species of each month (Table 3.5). I used only those months
for which data on frequencies of all trees in the top 5 diet
species were available, and I excluded Phoenix reelinata
from the North and South Mchelelo group analyses because of
the overwhelming abundance of the palm throughout the
forest. There were significant correlations for 8 of 11
months, 3 of 9 months, and 7 of 12 months (for which data


129
spend more time exploiting a fruiting tree and less time
traveling between trees as the probability of encountering a
fruiting tree or a tree in the preferred phenological state
declines. The cost of leaving a fruiting tree and moving to
another increases as trees become rarer and farther apart.
Mangabeys made direct inter-patch movements only when
several large trees of 1 or 2 widely dispersed species
(e.g., Ficus spp.) were fruiting synchronously; otherwise
they made slow, non-direct movements. Slow inter-patch
movement, during which mangabeys continuously foraged for
seeds and invertebrates on the forest floor, should have
minimized the cost of travel and maximized food intake.
Once food was encountered, the 1988-89 mangabeys spent
more time eating than 1973-74 mangabeys. Increased time
spent eating in 1988-89 may have resulted from greater
handling times of common diet items or an actual increase in
intake to compensate for lower diet quality. Although
feeding rates are not available for the 2 studies, the
overwhelming dependence on Phoenix reclinata fruits in 1988-
89 would have lowered, not increased, handling times.
Phoenix reclinata fruits are small and grow in easily
harvested clusters. Handling time is brief even when
mangabeys are selectively feeding on unripe seeds, and the
small fruit size allows mangabeys to fill their cheek
pouches and continue foraging for other food items while
processing palm fruits.


260
APPENDIX B.
Continued
N. Mchelelo
S. Mchelelo
Nkano
Taxon
No. of
No. of
No. of
records %
records %
records %
SAPINDACEAE
Aporrhiza paniculata 140
Blighia uniiuqata
Chvthranthus obliquinervis
Maiidea zanquebarica
Tamarindus indica 97
SAPOTACEAE
Mimusops fruitcosa 41
Pachvstela msolo 19
SIMAROUBACEAE
Harrisonia abvssinica 16
STERCULIACEAE
Sterculia rhvnchocarpa
RUBIACEAE
Rubiaceae sp. 1
Pavetta sphaerobotrvs 8
Polvsphaeria multiflora 21
URTICACEAE
stinging nettle 22
VITACEAE
Cissus rotundifolia 23
Cvphostemma iiquu 2
VIOLACEAE
Rinorea elliptica
CODED TAXA
vine 1 10
vine 2 1
vine 3
Mwimbo vine 5
Code MK1
Code ALE 1
mushrooms (all spp.) 104
Total 3070
4.5
94
5.5
233
12.2
-
-
-
1
0.05
-
-
-
2
0.1
-
1
0.06
-
-
3.1
22
1.3
-

1.3
74
4.3
-
0.6
2
0.1
219 11.4
0.5
5
0.6
0.03
10
0.6
_
0.3
6
0.4
1
0.05
0.7
10
0.6
63
3.3
0.7
8
0.5
-
-
0.7
16
0.9
6
0.3
0.07
-



9
0.5
0.3
4
0.2
0.03
-
-
-
-
-
-
-
8
0.4
0.03
-
-
-
-
-
-
-
3
0.2
0.03
-
-

-
3.3
72
CM
13
0.7
1678 1916


Mchelelo
Biomass Index of fruits (kg)/ha
6000-
4600-
4000-
3600-
3000-
2600-
2000-
1600-
1000
600
0
NDJFMAMJJASONDJFMAM
8 8 8
7 8 9
Nkano
Biomass index of fruits (kg)/he
6000
4600
4000
3600
3000
2500
2000
1600
1000
600
0
NDJFMAMJJA
8 8
7 8
Mnazinl
Biomass Index of fruits (kg)/he
6000-)
4600
4000
3600
3000
2600
2000
1600
1000
600
0
NDJFMAMJJASONDJFMAM
8 8 8
7 8 9
Month
liana
-- - -l palm
ElSS surface
ramlflorou8


Figure 5.7. Occupancy of each 0.25 ha quadrat by Nkano
group; percent occupancy is divided into 1 of 5
intensities of use.


Table 5.5. Intergroup behavior categories and measures of proximity and fighting by
month. F, M, and A designate fight, merge, and avoid months, respectively. N includes
all 12-hr observation days and days for which only presence or absence of South group was
determined.
Month
Intergroup
Behavior
Category
%
S. Group
Residency
% Scans
Groups
Merged
% Scans
Groups
clOOm Apart
% Encounters
Resulting
in Fights
N
(days)
Jan 88
F
62
10
40
13
Feb
F
50
7
15
40
7
Mar
F
94
40
53
56
9
Apr
F
35
0
1
100
13
Jun
M
55
0
14
0
9
Jul
M
43
4
28
0
7
Sep
M
54
7
13
0
11
Oct
M
50
6
21
0
7
Nov
F
93
7
18
22
7
Dec
A
18
0
11
0
10
Jan 89
F
52
5
4
100
13
Feb
A
0
0
0
0
8
Mar
A
19
0
0
0
13
Apr
F
14
4
4
100
9
194


150
Preferential harvesting also may affect palm population
structure. Differences in the distributions of size classes
between Nkano and Mchelelo may be explained, in part, by the
excessive harvesting of the larger size classes in Nkano.
The lack of tall, trunked palms in Nkano is due to the
extraction of this size class from the forest and possibly
to growth suppression in other size classes due to frequent
leaf pruning. The abundance of small, young plants in Nkano
indicates that conditions for seedling establishment may be
better in Nkano than Mchelelo and/or that undisturbed palms
in Mchelelo tend to reproduce vegetatively. The more open
understory and higher disturbance regime due to flooding in
Nkano may enhance seedling establishment (Uhl and Dransfield
1987). There is no evidence, however, to suggest that the
abundant young plants in Nkano will successfully reach the
larger size classes; the majority of young plants occur in
the center of the forest, an area where few large palms have
established.
Throughout most of the year P. reclinata fruits and
seeds are important food items for Tana mangabeys. Because
P. reclinata is seasonal, producing fruit out of synchrony
with other tree species (Chapter 2), mangabeys rely heavily
on its fruits and seeds when few other fruits are available.
The lower consumption of P. reclinata in Nkano relative to
Mchelelo may be due to the lower abundance of reproductive
individuals in Nkano, a higher abundance of alternative


CHAPTER FOUR
COMPETING USES OF A FOREST PALM, Phoenix reclinata N.J.
JACQUIN, BY HUMANS AND TANA RIVER CRESTED MANGABEYS
Introduction
Non-destructive, sustained-yield harvesting of plant
and animal resources by rural people could play an important
role in tropical forest conservation (e.g., Myers 1984,
Peters et al. 1989, Vasquez and Gentry 1989). Regulated
harvesting may have minimal impacts on tropical forests
while generating substantial economic returns to rural
people, thereby providing alternatives to more destructive
types of land use. Peters et al. (1989), for example, show
that market-oriented extraction of fruits by villagers from
some Amazonian forests provides a promising, economic
alternative to deforestation. Other field workers however,
have illustrated the potential negative aspects of such an
economy. Redford and Robinson (1987) and others (Smythe
1978, Mittermeier in press) have shown that harvesting of
vertebrate species by many people living in Central and
South American forests is frequently non-sustainable.
Vasquez and Gentry (1989) also question the sustainability
of fruit harvesting practices by rural people of Amazonian
Peru because of excessive and destructive exploitation. The
136


Mchelelo
Fruit score
8 0 8
7 8 0
Nkano
Fruit score
6-
4 -
NDJFMAMJ
6 8
7 8
JASONDJFMAM
8
9
Mnazini
Fruit score
5n
III.!
N D J F M A M
8 8
7 8
JASON
Month
1 1 1 1 1 r
D J F M A M
8
9


Figure 1.2.
Map of Kenya and location of the
National Primate Reserve (TRNPR;
from Hughes 1985).
Tana River
map modified


Figure 5.3. Cumulative frequency of sightings of one or more Nkano group mangabeys
versus rank order of 0.25 ha quadrats. Quadrats were ranked with respect to
the number of sightings occurring in each. The areas accounting for 50 and
100% of all sightings are designated.




Figure 2.11. Mean monthly fruit scores for 3 species
showing patterns of continuous fruit
production.
a) continuous fruiting due to asynchrony among
individuals of the species (Ficus natalensis);
b) continuous fruiting due to retention of
fruits over time (Hyphaene compressa); c)
continuous fruiting due to rapid turnover of
fruits (Oncoba spinosa).


219
the dominant male fathers all infants born in his group
during his tenure. Basic reproductive parameters for male
and female mangabeys are summarized in Table 6.3.
Population Genetics Model
Effective population size
The first step in developing criteria for minimum population
size is determining how much inbreeding, or loss of genetic
variability, is acceptable. The quantity of genetic
variation in a population is a function of effective
population size (NJ the size of the ideal population that
would undergo the same amount of random genetic drift as the
actual, or census, population (Wright 1931). To estimate Ne
for the mangabey, I used models developed by Lande and
Barrowclough (1987). I chose these models because they
incorporate the mean and variance of male and female
reproductive success, terms that have been shown to be
exceedingly important in estimating Nc (Chepko-Sade et al.
1987, Wood 1987, Harris and Allendorf 1989) and avoided the
greater number of assumptions necessary for other models
(e.g., Hill 1972 and Ryman et al. 1981). Lande and
Barrowclough's approach simultaneously incorporates the
effects of variance in progeny number (Eq. 1), unequal sex
ratio (Eq. 2) and fluctuating population size (Eq. 3):
Equation 1. Ncm = (NmKm l)/[Km + K^KJ 1] and
Ncf = (NrK( 1) / [ K( + KVKf) 1]


207
% time eati
60
50 -
40 -
30 -
20 -
10 -
ng
I
0
i
Avoid
Uniform food resources
I
Fight
1
Merge
Patchy food resources
o
Avoid
1
Fight
Intergroup behavior
1
Merge


BEHAVIORAL AND DEMOGRAPHIC RESPONSES TO HABITAT CHANGE
BY THE TANA RIVER CRESTED MANGABEY
(Cercocebus galeritus qaleritus)
By
MARGARET F. KINNAIRD
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1990


101
readily available (Chapter 2). Habitat and group size also
may influence activity budgets. The Nkano group, which
inhabits a larger forest with a richer resource base
(Chapter 2), allocated more time to inactivity than either
of the Mchelelo groups. The larger S. Mchelelo group ate
less, was less active and moved more than either of the
smaller groups.
Comparisons with 1973-74 Activity Budgets
Overall activity budget. MANOVA showed a significant
effect of group (Wilks' Lambda=0.047, F=8.14, df=28,171,
pcO.OOOl) on the overall activity budgets from two 1973-74
and three 1988-89 mangabey groups. One-way ANOVA showed
significant group differences for all behaviors but
social/sexual (F=10.4, 4.16, 7.69, 16.14, 23.72 and 4.59 for
eating, foraging, moving, inactive, autogrooming, and
allogrooming, respectively, df=4,53, p<0.005). Least square
means comparisons showed that most of the significant
differences (p<0.05) existed between the 1973-74 and 1988-89
groups, with fewer differences among groups of the same
study. The 1973-74 groups ate less, were less active, and
spent more time grooming than the 1988-89 groups (Figure
3.1). The differences between foraging and moving for the 5
groups were more complicated. The 1973-74 Mchelelo group
moved more than all other groups with the exception of the
1988-89 S. Mchelelo group and foraged less than all other
groups.


193
Variability in intergroup interactions. Fourteen months
of observations from Mchelelo forest were classified by
intergroup interactions; May and August 1989 were excluded
because of severe flooding and small sample size,
respectively. Seven of 14 months were classified as fight
months, 4 as merge months, and 3 as avoid months (Table
5.5) .
Avoid months were characterized by limited, non
overlapping ranging patterns; the North group remained
mostly in the north sector of the forest and the South group
remained in the south sector (Figure 5.11). Although long
calls were frequent, immediate vocal response (within 2 min.
of the initial call) by neighboring group males was rare and
groups tended to move away from the direction of the call.
Merge months were characterized by more overlap in
ranging patterns; both groups frequently crossed the other's
travel path and often traveled in parallel progressions
(Figure 5.12). Immediate vocal responses to long calls were
more frequent than during avoid months and calls often
resulted in approaches by the neighboring group. When groups
merged, they formed close, non-aggressive associations
moving together for periods of up to several hours. During
these associations, each group remained largely discrete but
occasionally individuals of the 2 groups intermingled,
foraging side-by-side, grooming, and sexually presenting to
members of the neighboring group. Intragroup herding by


110
Table 3.4. ANOVA results for study (Homewood's 1973-74
study and the present 1988-89 study) and age differences in
3 measures of mother/infant association: a) amount of time
infant is on its' mother, b) amount of time infant has
mother as nearest neighbor and c) amount of time infant is
separate from mother. Data were nested by infants within
studies.
Source
df
Sum of
Squares
Mean
Squares
F
P
Association A
Study1
1
0.009
0.009
0.07
0.793
Infant (Study)
10
1.227
0.123
2.78
0.007
Age
6
8.212
1.369
30.99
0.0001
Study Age
6
2.172
0.362
8.20
0.0001
Error
59
2.606
0.044
Association B
Study1
1
1.043
1.043
16.77
0.002
Infant (Study)
10
0.622
1.043
33.25
0.0001
Age
6
1.277
0.213
6.78
0.0001
Study Age
6
0.058
0.009
0.31
0.929
Error
59
1.852
0.031
Association C
Study1
1
0.759
0.759
7.97
0.018
Infant (Study)
10
0.952
0.952
2.44
0.017
Age
6
11.206
1.868
47.88
0.0001
Study Age
6
1.519
0.253
6.49
0.0001
Error
59
2.301
0.390
^-statistic calculated using infant (study) as the mean-
square error term.


Figure 6.4. Probability of extinction p(E) within 100, 50, 25 and 10 years (denoted by
labeled curves) for populations with mean persistence times (T) of 200 to
2000 years.


109
Association A
Association B
Association C


275
Uhl, N.W. and J. Dransfield. 1987. Genera Palmarum: A
Classification of Palms Based on the Work of Harold E.
Moore, Jr. Allen Press, Lawrence, KA.
Vasquez, R. and A.H. Gentry. 1989. Use and misuse of
forest harvested fruits in the Iquitos area.
Conservation Biology: 3 (4):350-349.
Wade, M.J. 1979. Sexual selection and variance in
reproductive success. American Naturalist 114:742-747.
Wade, M.J. and S.J. Arnold. 1980. The intensity of sexual
selection in relation to male sexual behaviour, female
choice, and sperm precedence. Animal Behaviour 28:446-
461.
Walters, J.F. 1986. Transition to adulthood. Pages 358
369 in B.B. Smuts, D.L. Cheney, R.M. Seyfarth, R.W.
Wrangham and T.T. Struhsaker, editors. Primate
Societies. University of Chicago Press, Chicago, IL.
Waser, P. 1975. Monthly variations in the feeding and
activity patterns of the mangabey, Cercocebus albiaena.
East African Wildlife Journal 13:249-265.
Waser, P. 1976. Cercocebus albiqena: Site attachment,
avoidance, and intergroup spacing. American Naturalist
110:911-935.
Waser, P. 1977. Feeding, ranging and group size in the
mangabey Cercocebus albiqena. Pages 183-222 in T.
Clutton-Brock, editor. Primate Ecology. Academic
Press, New York, NY.
Waser, P. 1982. The evolution of male loud calls among
mangabeys and baboons. Pages 117-143 in C.T. Snowdon,
C.H. Brown and M.R. Petersen, editors. Primate
Communication. Cambridge University Press, Cambridge,
U.K.
Waser, P. 1984. Ecological differences and behavioral
contrasts between two mangabey species. Pages 198-216
in P.S. Rodman and J.G.H. Cant, editors. Adaptations
for Foraging in Non-human Primates. Contributions to an
Organismal Biology of Prosimians, Monkeys and Apes.
Columbia University Press, New York, NY.
Waser, P. and O. Floody. 1974. Ranging patterns of the
mangabey, Cercocebus albiqena. in the Kibale Forest,
Uganda. Z. Tierpsychol. 35:85-101.


Figure 2.13. Seasonality of mean monthly fruit scores for
Saba comorensis in the 3 study forests,
a) seasonal fruiting in Mnazini forest; b)
seasonal but extended fruiting in Nkano forest
c) continuous fruiting in Mchelelo forest.


272
Peters, C.M., M.J. Balick, F. Kahn, A.B. Anderson. 1989.
Oligarchic forests of economic plants in Amazonia:
utilization and conservation of an important tropical
resource. Conservation Biology: 3(4):341-349.
Pielou, E.C. 1969. An Introduction to Mathematical
Ecology. John Wiley and Sons, New York, NY.
Post, D.G. 1981. Activity patterns of yellow baboons
(Papio cvnocephalus) in the Amboseli National Park,
Kenya. Animal Behaviour 29:357-374.
Poole, R. W. 1974. An Introduction to Quantitative
Ecology. McGraw-Hill, Inc., New York, NY.
Prance, G.T., W. Balee, B.M. Boom and R.L. Garneiro. 1989.
Quantitative ethnobotany and the case for conservation
in Amazonia. Conservation Biology 1(4):296-310.
Quris, R. 1975. Ecologie et organisation sociale de
Cercocebus qaleritus dans le nordest du Gabon. Terre et
Vie 29:337-398.
Rasmussen, D.R. 1979. Correlates of patterns of range use
of a troop of yellow baboons (Papio cvnocephalus) I:
sleeping sites, impregnable females, births, and male
emigrations and immigrations. Animal Behaviour 27:1098-
1112.
Rasmussen, D.R. 1983. Correlates of patterns of range use
of a troop of yellow baboons (Papio cvnocephalus) II:
Spatial structure, cover density, food gathering and
individual behaviour patterns. Animal Behaviour 31:834-
856.
Rathche, B. and E.P. Lacey. 1985. Phenological patterns of
terrestrial plants. Annual Review of Ecology and
Systematics 16:179-214.
Read, R.W. 1988. Utilization of indigenous palms in the
Caribbean (In relation to their abundance). Advances in
Economic Botany 6:137-143.
Redford, K.H. and J.G. Robinson. 1987. The game of choice:
patterns of indian and colonist hunting in the
Neotropics. American Anthropologist 89(3):650-666.
Robinson, J.G. 1985. Expected benefits determine area
defense: experiments with Capuchin monkeys. National
Geographic Research Report 21:421-424.


CHAPTER SEVEN
CONCLUSION
This study has demonstrated the flexibility of Tana
mangabey activity patterns and feeding, ranging, and
intergroup behaviors in response to seasonal and large-scale
habitat changes. Mangabeys adjust activity patterns not only
to buffer fluctuations in resource availability in a highly
seasonal habitat, but to compensate for permanent loss of
habitat and critical food resources. As resources decline in
availability, mangabeys spend more time feeding and less
time inactive or engaged in social grooming. Mangabeys are
opportunistic feeders; they have a relatively diverse diet
but concentrate on what is most available at any given time
and are capable of responding to habitat alterations through
major dietary shifts. A decline in numbers of important diet
trees over the last 15 years appears to have caused
mangabeys to shift from primarily ripe, fleshy fruits to
unripe seeds and insects.
The Tana mangabey's flexibility of ranging and
intergroup behaviors are additional features of adaptation
to a highly seasonal and restricted habitat. My results
suggest that seasonal variation in food availability and
249


187
home range by North Mchelelo group was less extensive
(approximately 25%) because the North group never left
Mchelelo forest. The amount of overlap between the 2 groups
varied considerably by month, with the South Mchelelo group
generally spending a larger proportion of time in the
overlap zone than the North Mchelelo group (Figure 5.10).
The Nkano group also overlapped with an unhabituated
mangabey group that ranged primarily in the northern section
of the Nkano forest and possibly a third group that
occasionally visited from the nearby Mnazini forest.
I tested for differences in quadrat use, travel patterns
and distances traveled by the North Mchelelo group when
South Mchelelo group was present and absent from Mchelelo
forest to determine if the presence of a neighboring group
influenced North group's movements and/or use of space.
After controlling for the effects of month, the presence of
South group appeared not to influence distance traveled, but
did influence travel patterns and use of space (ANOVA, Table
5.3). Least square means showed that the amount of
backtracking, as measured by mean turning angle (p=0.007)
and number of path crossings (p=0.04), was significantly
higher when South group was present. North group also
restricted its use of space when South group was present;
the diversity of quadrats used and the number of unique
quadrats entered by North group were significantly lower
when South group was present.


94
occupied a greater proportion of the activity budget during
the drier seasons than the rainy seasons (Figures 3.2-3.4),
and the groups were more social during the short rains than
all other seasons of the year. Interactions between forest
and season were significant for inactivity; the Nkano group
spent a greater amount of time inactive during the inter
rains than either N. or S. Mchelelo.
Time-of-day was significant for foraging, moving,
inactivity and allogrooming (ANOVA, Table 3.3). On average,
foraging occurred primarily during the morning (0700-1000)
and mid-day hours (1100-1400), and groups became
increasingly less active by the afternoon hours (1500-1800)
when allogrooming increased (Figures 3.2-3.4). There also
were significant interactions between season and time-of-day
on moving. Groups moved more in the morning during the drier
seasons, but showed a late-afternoon peak in movement during
the rainy seasons (Figures 3.2-3.4).
In summary, there appear to be predictable hourly and
seasonal patterns for certain activities. Although the 3
groups allocate differing amounts of time to various
activities, they engage in similar activities during
particular hours of the day. Mangabeys expend more of their
daily activity budget in food gathering behaviors (e.g.
foraging and moving) during drier seasons when food
resources are less available, and engage more in social
behaviors during the short rains when food resources are


Figure 2.2.
Number of individuals/0.25 ha quadrat of 4
common species within Mchelelo forest showing
clumped spatial distributions,
a) Hvohaene compressa; b) Alanoium
salviifolium; c) Phoenix reclinata; d) Oncoba
spmosa.


186
patterns and monthly measures of food availability (Chapter
2) to determine if mangabeys adjusted movements to take
advantage of resource availability. If movements depend on
encounter rates with food (see Robinson 1986), then
mangabeys should travel further and show less tendency to
backtrack during months when resources are scarce. The
results were equivocal; mean path length was significantly
correlated with ripe fruit availability for the South
Mchelelo group only (r=-0.66, n=10, p<0.04).
Alternatively, monthly movements may be influenced by
the spatial distribution and renewability of food items used
each month. Because foods of plant origin, especially ripe
fruit, often are less homogeneously distributed and/or have
slower renewal rates than invertebrates (Robinson 1986,
Waser and Wiley 1980), mangabeys should backtrack less, have
longer daily path lengths and longer half-hour step
distances when feeding on ripe fruit trees. This association
was not apparent for any of the 3 groups; there were no
significant correlations between mean monthly path length or
tendency to backtrack and the proportion of ripe fruits or
invertebrates in the diet (Spearman's Rank correlations,
p>0.05).
Intergroup Behavior
Group range overlap. The home range of South Mchelelo
group overlapped almost completely with that of North
Mchelelo (Figure 5.6). Overlap of South Mchelelo group's


1973-74
(1 Male:9.75 Females)
233
250
No. of males
200 -
150 -
100 -
50 -
Ii 1 Tenure
2 Tenure
3 Tenure
No. of males
250 -i
200 -
1988-89
(1 Male:5.85 Females)
150
100 -
6 10 13
Infants per male


196


Figure 2.9. Mean monthly fruit scores for Phoenix
reclinata.
a) Mchelelo forest; b) Nkano forest; c)
Mnazini forest.


REFERENCES
Aldrich-Blake, P. T. Bunn, R. Dunbar and P. Headley. 1971.
Observations on baboons, Papio anubis. in an arid region
in Ethiopia. Folia Primatologica 15:1-35.
Allendorf, F.W. 1983. Isolation, gene flow, and genetic
differentiation among populations. Pages 51-65 in C.M.
Schonewald-Cox, S.M. Chambers, B. MacBryde and L.
Thomas, editors. Genetics and Conservation: A
Reference for Managing Wild Animal and Plant
Populations. Benjamin/Cummings Publishing Company
Inc., Menlo Park, CA.
Altmann, J. 1974. Observational study of behavior:
Sampling methods. Behaviour 49:227-267.
Altmann, S. 1974. Baboons, space, time, and energy.
American Zoologist 14:221-248.
Altmann, S. and J. Altmann. 1970. Baboon Ecology.
University of Chicago Press, Chicago, IL.
Anderson, A.B. 1988. Use and management of native forests
dominated by Acai Palm (Euterpe olercea Mart.) in the
Amazon Estuary. Advances in Economic Botany 6:144-154.
Andrews, P., C.P. Groves and J.F.M. Horne. 1975. Ecology
of the lower Tana River floodplain. Journal of East
African Natural History Society Natural History Museum
151:1-31.
Ayres, J.M.C. 1986. Uakaris and Amazonian Flooded Forest.
Unpubl. Ph.D. dissertation, Cambridge University,
Cambridge, UK.
Baldwin, J.D. and J.I. 1973. The role of play in social
organization: Comparative observations on squirrel
monkeys (Saimiri). Primates 14(4)369-381.
Bates, D.M. 1988. Utilization pools: a framework for
comparing and evaluating the economic importance of
palms. Advances in Economic Botany 6:56-64.
261


263
Clemons, E. and B. Phillips. 1980. Organic acid production
and digesta movement in gastrointestinal tracts of the
baboon and sykes monkey. Comparative Biochemical
Physiology 66:529-532.
Clutton-Brock, T.H.C. 1974. Activity patterns of red
colobus (Colobus badius tephroseles). Folia
Primatologica 21:161-187.
Clutton-Brock, T.H.C. 1977. Some aspects of intraspecific
variation in feeding and ranging behaviour in primates.
Pages 539-556 in T.H.C. Clutton-Brock, editor. Primate
Ecology: Studies of Feeding and Ranging Behaviour in
Lemurs, Monkeys and Apes. Academic Press, New York, NY.
Clutton-Brock, T.H.C. and P. Harvey. 1977. Species
differences in feeding and ranging behaviour in
primates. Pages 557-584 in T.H.C. Clutton-Brock,
editor. Primate Ecology: Studies of Feeding and Ranging
Behaviour in Lemurs, Monkeys and Apes. Academic Press,
New York, NY.
Cole, L.C. 1954. The population consequences of life
history phenomena. Quarterly Review of Biology 29:103-
137.
Croat, T.B. 1975. Phenological behavior of habit and
habitat classes on Barro Colorado Island (Panama Canal
Zone). Biotropica 7(4):270-277.
Crook, J.H. and P. Aldrich-Blake. 1968. Ecological and
behavioural contrasts between sympatric ground dwelling
primates in Ethiopia. Folia Primatologica 8:192-227.
Crook, J.H. and J.S. Gartlan. 1966. Evolution of primate
societies. Nature 210:1200-1203.
Dandelot, P. 1974. Part 3: Order Primates. Pages xx-xx in
J. Meester and H.W. Setzer, editors. The Mammals of
Africa: An Identification Manual. Smithsonian
Institution Press, Washington, D.C.
Davidge, C. 1978. Activity patterns of chacma baboons
(Papio ursinus) at Cape Point. Zoolgica Africana
13:329-350.
Davies, A.G., E.L. Bennett and P.G. Waterman. 1988. Food
selection by two South-east Asian colobine monkeys
(Presbvtis rubicunda and Presbvtis melalophos) in
relation to plant chemistry. Biological Journal of the
Linnaean Society 34:33-56.


67
significant negative correlations between flowering and both
rainfall and river flow (r=0.-41 and -0.35 for rainfall and
river flow, respectively, n=41, p<0.05). Correlations
between flowering and fruiting were significant when flower
scores were lagged by 3 months in Mchelelo, Nkano and for
all 3 forests combined (r=0.76, p<0.05 for Mchelelo, r=0.55,
p<0.05 for Nkano and r=.45, n=41, p<0.05 for all forests
combined).
There were no consistent associations between monthly
fruit scores and either rainfall or river flow for the 3
forests (Tables 2.4, 2.5 and 2.6). Minimum monthly
temperature was positively correlated with monthly fruit
scores for all species combined in the 3 separate forests.
There were also positive correlations between minimum
monthly temperature and scores from seasonal fruiting
species in Nkano and Mnazini. When the palm, reclinata.
was removed from the analysis, there were significant
negative correlations between fruiting of seasonal species
and river flow lagged by 3 months. Phoenix reclinata showed
significant positive correlations with river flow lagged
only by 1 or 2 months.
When data for all forests were combined, many of the
results were consistent with the analysis performed on the
forests separately (Table 2.7). There were significant
positive correlations between fruit scores for continuous
species and rainfall. Minimum monthly temperature showed


Figure 3.5. Mean percent time spent in 7 social behaviors
by 1973-74 and 1988-89 adult males and females,
a) adult males; b) adult females.


Figure 3.2. Mean percent time spent in 4 major behaviors (eating, foraging, moving and
inactive) by time-of-day and season for the 1988-89 N. Mchelelo group.


23
has multiple crowns, the above value was multiplied by the
number of crowns per individual.
Fruit counts for all species, excluding palms, were
multiplied by an estimate of relative food-producing area to
calculate the total number of fruits potentially produced by
an average individual of a particular species. Fruit
producing area was calculated using surface area and volume
formulas for an open-bottomed cylinder, a shape that best
approximates a variety of differing tree shapes (Appendix
A). Surface areas were calculated for the category of
species bearing their fruits primarily on the surface of the
canopy and volumes were calculated for ramiflorous species.
Branches of ramiflorous species were subjectively estimated
to occupy little over 1/4 of the entire volume of the
canopy; volume estimates, therefore, were divided by 4. The
food-producing area of the liana species, which grows over
the top of tree canopies, was calculated by estimating the
amount of surface area of the tree canopy covered by the
liana. I did not correct for canopy size in palms because a
more direct count of fruit numbers was possible.
Samples of 10 or more fruits were collected and weighed
from a maximum of 10 individuals per species. Only those
fruit parts consumed by mangabeys were weighed and parts
were weighed only during the stage of ripeness when
mangabeys were feeding on them. Weights were averaged for
individuals and individual averages were pooled to calculate


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
F. Eisenberg, Chai]
Katharine Ordway PlroT^or
of Ecosystem Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree g^y^^ctpr of Plj'osophy.
G. Robinson
/ Associate Professor of Forest
j Resources and Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Michael W. Collopy/7^
Professor of Fore&t^Resources
and Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.
Th ostras
Assistant Professor of Forest
Resources and Conservation
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree o_£ Doctor of Philosophy.
y\ i
rancis E.Putz)
ssociate Professor of Botany


239
population of 725 individuals however, will not necessarily
survive over the next 200 years. Expected persistence time
is distributed as a negative exponential; for a population
of a given size with a large expected persistence time, a
majority of populations of that size will persist for less
time and a few will persist for more time. To achieve a
given persistence time with a high degree of certainty (e.g.
95%), the average persistence time will need to be much
greater.
I generated distributions of persistence times to
determine probabilities of extinction within a given time
period for mean persistence times of 200, 400, 1000 and 2000
years (Figure 6.4). For each mean persistence time, 10
samples of 1000 values were randomly generated using a
negative exponential distribution. For each sample, the
proportion of persistence times less than 100, 50, 25 and 10
years respectively, were computed. These proportions were
averaged over the 10 samples and the averages were used as
the probabilities of going extinct within a given time
period. The results show that for a mean persistence time of
200, the probability of extinction over the next 100 years
is approximately 40%. Shorter extinction intervals have
higher probabilities of persistence but it is not until the
very short-term extinction interval of 10 years that there
is a 95% probability of the population persisting. To assure
a 95% probability of population persistence over the next


While in Kenya, my life was made enormously easier by
the NMK and IPR. Dr. Jim Else dealt with every millimeter of
red tape in Nairobi while I was isolated in the bush. Dr.
Mohammed Isahakia, Dr. Steven Njuguna, Richard Leakey,
Margaret Omoto, and Mary Sefu also played critical roles in
securing the necessary visas, permits and affilliations.
Mary Matibo, Dorothy McCullough and Debbie Snelson of the
African Wildlife Foundation kindly dealt with licenses and
insurance for my jeep and saw that I was legal on the road.
My stays in Nairobi were memorable and warm due to the
hospitality and generosity of Jim, Margaret and Jessie Else.
In Malindi, Elsa Friman provided me with a luxurious home
base. She, Romolo, and their employee Sammi, renewed me in
ways they will never know. I also thank Ann and Ian
Robertson for the many wonderful afternoons and discussions
on their Malindi porch.
My research was made much easier by the comfortable
accomodations of the Tana River Research Camp. Fred Decker
designed, built and kept the camp in working order for most
of my stay; the fact that all ran smoothly in his absence is
testimony to his skill. I also thank Alex Njui for managing
the camp after Fred's departure. To our camp staff Mustafa
Lordis, Bwana Mzee, Jumaa Galana, and Abio I give thanks for
their help and friendship. I also owe very special thanks to
my field assistants Galole Galana and Bakari Kawa. They
taught me much about jungle life and without their help, the
v


254
coupled with incorporation of surrounding villages into a
buffer zone could improve regional conservation and benefit
villagers from activities directly and indirectly related to
the park (e.g. tourism). Recent, growing concern by the
Kenyan government and its Wildlife Service for the TRNPR,
and their possible implementation of the above management
recommendations gives cause for optimism for the future of
the reserve and its endangered primates.


Figure 2.3.
Number of individuals/0.25 ha quadrat of 4
common species within Nkano forest showing
clumped spatial distributions,
a) Ficus svcomorus; b) Alangium salviifolium
c) Phoenix reclinata; d) Pachvstela msolo.


Mchelelo
6000 n
4600
4000
3600
3000
2600
2000
1600
1000
600
Biomass Index of fruits (kg)/he
Nkano
Mnazini
Month


CHAPTER SIX
POPULATION VIABILITY OF THE TANA RIVER CRESTED MANGABEY
Introduction
A major goal in the study of endangered species is to
determine the minimum population sizes that can ensure the
long-term persistence of the species (Soul 1980). Although
problems exist in the estimation of viable population sizes,
there is universal agreement about the importance of such
estimates in making informed management decisions (Harris
and Allendorf 1989, Soul 1980).
Genetics and demography offer two very different
approaches to the problem of estimating viable populations
(Ewens et al. 1987). The genetic approach is concerned with
the maintenance of genetic variation and the rate at which
this variation is lost through inbreeding or random genetic
drift. The demographic approach is concerned with the
probability of complete extinction of a population through
random demographic events. Although the effects of
inbreeding depression and the maintenance of genetic
variability have received far more attention than demography
in estimating viable population size, researchers have begun
to argue that demography may be of more immediate importance
than population genetics in determining minimum viable
214


Figure 5.13. Daily travel paths by North and South Mchelelo
groups during one day of a month of intergroup
fighting. Timing and location of adult male
long-calls are plotted for North (ovals) and
South (rectangles) groups.


26
with Homewood's 12 month data from Mnazini forest and 7
month data from Mchelelo forest. Kolmogrov-Smirnov two-
sample tests (Hollander and Wolfe 1973) were performed by-
species to test for differences in the distributions of
fruiting between the 2 study periods. Wilcoxon signed-ranks
tests also were performed by species to test whether fruit
scores were consistently higher during 1 study period than
the other.
Results
Abundance and Spatial Distribution of Resources
Density values ranged widely for enumerated species
both within and between forests (Table 2.2). Over 50% of the
species enumerated in Mchelelo (n=13 of 25) and Nkano (n=8
of 15) were rare, occurring at densities of less than 1
individual per ha. The palm, Phoenix reclinata. was one of
the most common species, occurring at the highest and second
highest densities (193.6 and 101.6 individuals/ha) of all
species in Mchelelo and Nkano, respectively. Alangium
salviifolium. an understory species, also occurred at very
high densities (33.6-112.3 individuals/ha) in both forests.
Five species present in both forests (Pachvstela msolo.
Ficus svcomorus. Alangium salviifolium. Aporrhiza paniculata
and Saba comorensis) occurred at much higher densities in
Nkano than Mchelelo; 4 additional species (Sorindeia


146
harvested in the smallest size class. Although there is no
significant heterogeneity in the overall distribution of
damage level by size class (p>0.05), there is significant
within-cell heterogeneity due to a high frequency of topping
in the largest size class (GI=17.7, df=3, p<0.005).
The majority of reproductive palms (84%) belonged to
the 2 trunk-bearing size classes (Figure 4.1). Eighty
percent (n=20) of the reproductive individuals counted in
the smaller, trunkless size class were male. The proportion
of male (60%) and female (40%) reproductive individuals was
more equitable in the larger size class.
Phoenix reclinata accounts for up to 62% and 42% of the
monthly diet of mangabeys in Mchelelo and Nkano,
respectively. Phoenix reclinata accounts for 26% of the
annual diet in Mchelelo and is eaten in some form during
every month of the year. In Nkano, P. reclinata accounts
for 18% of the annual diet and is eaten in some form during
all but 1 month of the year. Ripe and unripe seeds account
for the majority of the items taken in Mchelelo (84%) and
Nkano (96%) and consumption of ripe fruits accounts for less
than 3% in either forest (Figure 4.2), indicating mangabeys
are major seed predators of P. reclinata. The 'other'
category accounts for less than 10% of the total items
consumed in either forest and includes thorns, open and
unopened flowers, the pith of the leaf rachis, and
seedlings.


156
group (Chapter 3). I followed groups from 0700-1830 hrs each
day; every half hour, following a behavioral scan (Chapter
3), I plotted the location of all individuals that could be
found in a 5-minute period. I noted all quadrats that were
occupied by at least 1 individual and determined the quadrat
location of the group's center-of-mass (see Altmann and
Altmann 1970, Waser and Floody 1974). I assumed the distance
moved by the group in each half-hour period equalled the
straight line distance between consecutive locations of the
center-of-mass. The sum of all half-hour "step distances"
was used as an estimate of total daily distance moved. I
also calculated the turning angle between each pair of 1/2
hr steps (not calculated or included in summary statistics
if group did not move) and counted the number of times a
group crossed its own path.
I recorded the time and location of all adult male
mangabey long-calls (defined as loud-call or long-range
vocalization by Homewood 1976 and Waser 1982) on daily range
maps. The group which initiated a long call was determined
when possible and response to the long calls by the study
group was recorded.
Time, duration, and outcome of all inter-group
interactions were recorded ad libitum. I also noted age,
sex, and behavior of interactants and plotted the location
of all interactions on daily range maps.


24
a species mean. The density of potentially reproductive
individuals in each species for Mchelelo and Nkano study
sites were taken from the species enumerations described
above. Species densities for Mnazini were taken from Medley
(1990).
A final monthly biomass (kg/ha) index of fruits of
species sampled was calculated as:
FB = [£'{(Xsi*Dsi.SAsi*FPsi)/Isi)] +
[£{(Xri*Dri*Vri*FPri)/Iri)] +
[< {£ ( (X11*Dli*SAli*FPli)/Ilin where
X = monthly fruit score for species i / 5 (to scale the
scores to maximum production);
I = number of individuals scored for species i;
D = density of species i per hectare;
SA = surface area of species i;
V = branch volume of species i;
FP = total fruit production measured as mean fruit count *
mean fruit weight of species i;
and subscripts s, r, p and 1 denote surface, ramiflorous,
palm and liana categories, respectively.
Environmental Correlates with Flowering and Fruiting
I examined associations between 3 environmental
parameters and the temporal patterns of flowering and
fruiting in the 3 study sites. A Spearman's rank correlation
analysis (Siegel 1956) was conducted to examine possible


Short Rains
Long Rains
Morning
Mid-day
Afternoon
Inter-rains
Dry Season
100


235
Table 6.5. Parameter values and estimates of the genetic
fixation index, Fsl, for forests on the east and west banks
and for west bank and east bank subpopulations. East vs west
bank figures are calculated for group migration from one
bank to the other for two estimates of high flood frequency.
Number of
Mean Ne/
forests
forest
m
m
Fst
10 (East Bank)
3.19
. 522
.00004
. 036
13 (West Bank)
3.02
.516
.00004
. 048
2 (East and West
Bank)
80 year floods
35.9
. 005'
-
. 579
50 year floods
35.9
. 007*
-
. 495
'migration rates adjusted to account for an average of 6.85
breeding individuals (males and females) transferring
instead of males alone.


138
centuries and the leaves are widely used in Africa as a
source of material for weaving (DeMason and Sekhar 1988, Uhl
and Dransfield 1987). In this paper I outline the use of
Phoenix reclinata Jacq. by the people of the Tana River in
northeastern Kenya and by an endangered forest primate, the
Tana River crested mangabey (Cercocebus qaleritus
galeritus). The Tana mangabey, endemic to a 60 km stretch of
the lower Tana River, is a fruit and seed specialist that
feeds heavily on P. reclinata (Homewood 1976). I examine the
impact on the palm population of harvesting by people and
palm fruit consumption by mangabeys. Specifically, I address
questions concerning the nature of the harvesting, the
preferential harvesting of different palm size classes, the
role of mangabeys in seed dispersal and the potential
effects of palm harvesting on fruit availability to the
endangered primates.
Methods
Study Species
Phoenix reclinata is a dioecious, cluster palm with
slender (<16 cm in diameter), often inclined trunks reaching
over 10 m in height (McCurrach 1960). Leaves grow up to 2.5
m in length, are pinnate with proximal leaflets modified
into spines, and show induplicate venation. Fruits are small
(1.3-1.7 cm long by 0.9-1.3 cm wide) fleshy drupes, and vary
in color from pale yellow to orange or dull red depending on
ripeness. The mesocarp is 1-2 mm thick and is moist and


241


192
Table 5.4. Number of adult male long-calls by study group
and month. North refers to long-calls given by the North
Mchelelo group males. South M refers to long-calls given by
the South Mchelelo group males while present in Mchelelo
forest and South C refers to long-calls given by the South
Mchelelo group males while in the neighboring Congolani
forest. Alien refers to long-calls given by males of a
second, unhabituated group in Nkano forest, and Unk
represents long-calls of unidentified males in the Nkano
forest.
a) North and South Mchelelo
Month
North
South
M South C
Total
N (days)
Feb 88
17
9
1
28
5
Mar
19
19
0
42
6
Apr
18
28
13
61
6
May
16
18
0
34
3
Jun
14
20
8
43
5
Jul
9
37
14
62
6
Aug
13
4
0
17
3
Sep
15
37
1
54
5
Oct
22
14
0
38
4
Nov
34
31
1
69
6
Dec
17
9
0
26
4
Jan 89
24
25
0
49
5
Feb
8
2
0
10
3
Mar
9
5
0
14
3
b)
Nkano
Month
Nkano
Alien
Unk
Total
N (days)
Mar 88
6
0
49
55
3
Apr
36
25
3
64
3
Jun
43
27
3
73
3
Jul
24
15
7
46
3
Aug
5
9
7
21
1
Sep
20
28
20
68
3
Oct
20
22
1
43
3
Nov
37
85
15
137
3
Dec
32
28
5
65
3
Jan 89
37
30
7
74
3
Feb
30
36
5
71
3
Mar
19
16
5
40
3


Nkano
Frequency
400 n
306
Small (<1 m) Trunkless (>1 m)
Harvested
Nonhar vested
Mchelelo
Frequency
Small (<1 m) Trunkless (>1 m) Trunk (<2 m) Trunk (>2 m)
Size class


271
Moore, H.E. Jr. 1973. Palms in the tropical forest
ecosystems of Africa and South America. Pages 63-88 in
B.J. Meggers, E.S. Ayensu and W.D. Duckworth, editors.
Tropical Forest Ecosystems in Africa and South America:
A Comparative Review. Smithsonian Institution Press,
Washington, D.C.
Morisita, M. 1959. Measuring the dispersion of individuals
and analysis of the distributional patterns. Mem.
Fac. Sci. Kjyushu Univ. Ser. E (Biol) 2:215-235.
Myers, N. 1984. The Primary Source: Tropical Forests and
Our Future. Norton, New York, NY.
Nagel, U. 1973. A comparison of anubis baboons, hamadryas
baboons, and their hybrids at a species border in
Ethiopia. Folia Primatologica 19:104-165.
Nakagawa, N. 1989. Feeding strategies of Japanese monkeys
against deterioration of habitat guality. Primates
30(1):1-16.
National Research Council. 1981. Techniques for the Study
of Primate Population Ecology. National Academy Press,
Washington, D.C.
Neter, J., W. Wasserman and M.H. Kutner. 1985. Applied
Linear Statistical Models. Richard D. Irwin, Inc.,
Homewood, IL.
Oates, J. 1977. The guereza and its food. Pages 276-323
in T. H. Clutton-Brock, editor. Primate Ecology:
Studies of Feeding and Ranging Behaviour in Lemurs,
Monkeys, and Apes. Academic Press, London, UK.
Oates, J.F. 1985. Action Plan for African Primate
Conservation: 1986-90. IUCN/SSC Primate Specialist
Group.
Oates, J.F. 1986. Food distribution and foraging behavior.
Pages 197-209 in B.B. Smuts, D.L. Cheney, R.M. Seyfarth,
R.W. Wrangham and T.T. Struhsaker, editors. Primate
Societies. University of Chicago Press, Chicago, IL.
O'Brien, T.G. 1990. The Determinants and Consequences of
Social Structure in a Neotropical Primate, Cebus
olivaceus. Unpublished Ph.D. dissertation. University
of Florida, Gainesville, FL.


2000
1500
1000
500
0
Cumulative frequency
80
ON
to


Figure 5.10. Degree of monthly overlap by North Mchelelo
group (solid line) and South Mchelelo group
(broken line) and the area of overlap in
hectares (bars).


169
I tested the distribution of sightings throughout each
group's range against a Poisson distribution to determine if
their use of space was random. I found significant
deviations in the distribution of sightings from the
expected Poisson distributions for all 3 groups (N.
Mchelelo: X2=29.8, df=6, P<0.0001; S. Mchelelo: X2=31.4,
df=6, pcO.OOOl; Nkano: X2=55.6, df=7, p<0.0001). As
indicated above, some guadrats were used much more than
expected, some less than expected (Figure 5.5). This uneven
pattern of use is evident also from the percent of total
sightings recorded within each quadrat for the 3 study
groups (Figures 5.6 and 5.7). Few quadrats were used
intensely (accounting for >4% of the total observations),
many were used less frequently.
Attributes of high-use quadrats. I tested for
differences in mangabey behaviors in high-use quadrats
(quadrats accounting for a cumulative 50% of total sightings
and individually > 2% of total sightings) and low-use
quadrats to determine if mangabeys frequented certain
quadrats to feed (eat and forage) or simply moved through
them to reach other locations. There were no significant
differences in the amount of time spent eating, foraging or
moving in high- versus low-use quadrats for any of the study
groups (One-way Student's t-tests, p>0.05). The variance in
behaviors, however, was significantly less in high- versus
low-use quadrats for each study group (F-ratio for equal


Figure 6.3. Persistence time in years as a function of
population size (N). Dashed lines connect 1973-
74 and 1988-89 population sizes to estimated
mean persistence times.


Figure 2.18. Mean monthly river flow for the Tana River
Hola Station.


124
Mchelelo (H'= 1.89, t=-0.24, df=18, p>0.8) or the S.
Mchelelo H'=1.95, t=0.16, df=ll, p>0.8) group. Mean monthly
diet diversity, however, decreased significantly between
1973-74 and 1988-89 for the Nkano group (H'=1.40, 1988-89
and H'=2.12, 1973-74, t=6.04, df=21, p<0.0001).
In summary, differences in activity budgets and diet
among the 1988-89 groups were similar to differences in
behavior and diet observed between the 1973-74 and 1988-89
study periods. There was an association between increased
feeding time and decreased fruit availability both within
and between studies. As ripe fruit and seed in the diet
decreased within and between years, the amount of time spent
in feeding behaviors also increased. Increased time spent in
feeding behaviors, concomitantly decreased the amount of
time available for grooming and inactivity both among the
1988-89 study groups and between the 1988-89 and 1973-74
studies. The only inconsistent association was between diet
diversity and fruit availability; 1988-89 diet diversity did
not increase with decreased fruit availability, and mean
monthly diet diversity for Nkano was lower in 1988-89 than
1973-74.
Discussion
Many studies have shown that feeding behaviors take
precedence over social or other activities, particularly for
species such as mangabeys (Waser 1975), baboons (Altmann
1974, Post 1981), and macaques (Fa 1986) that concentrate on


160
calls given as a function of intergroup behavior (avoid,
fight, and merge) were made when ANOVA indicated significant
differences. All analyses were performed on the SAS
statistical package (SAS 1985).
Results
Use of Space
Mangabeys from the North Mchelelo, South Mchelelo and
Nkano groups were sighted in a total of 75, 72, and 80
quadrats, respectively (Figures 5.1-5.3). This represents
home range areas of 19 and 20 ha for the North Mchelelo and
Nkano groups, and a range area of 18 ha for the South
Mchelelo group only when in the Mchelelo forest. A "taut-
string line" (Waser and Floody 1974), enclosing all ad
libitum South Mchelelo group sightings outside the Mchelelo
forest, increased the home range estimate for this group to
70 ha (Figure 5.4).
Use of space by each group was not distributed evenly.
The number of observations spent in a quadrat plotted
against the rank of that quadrat showed that North Mchelelo
group mangabeys spent over 50% of their time in 14 quadrats,
which corresponds to 3.5 ha, or 19% of their total home
range (Figure 5.1). The South Mchelelo and Nkano groups
spent over 50% of their time in 16 quadrats. This
corresponds to 4 ha, or 22% of the South group's total range
in the Mchelelo forest (Figure 5.2), and 4 ha, or 20% of the
total home range of the Nkano group (Figure 5.3).


248
large that it often will be impossible to contain the
required number of individuals in reserves and sanctuaries
(Soul 1987). This is particularly true for the Tana
mangabey with a distribution along approximately 60 km of
floodplain forest, less than 1/3 of which is protected
within the Tana River National Primate Reserve. It would be
an unrealistic management goal to strive for population
numbers as great as those required by the models given the
arguments made above and the mangabeys1 severe habitat
constraints. The projections allow us, however, to compare
ideal population sizes to real populations and to use the
results as a guide to determine future actions necessary for
conservation.
In the case of the Tana mangabey, the results of the
models do not give an optimistic picture; Tana mangabeys may
not survive without drastic management measures. Because the
mangabey is a generalist species that easily exploits new
habitats (Homewood 1976) it should respond well to habitat
management. The fact that the framework for such management
exists within the Tana River National Primate Reserve gives
cause for hope.


1988/89
60 -
K3

N. Mchelelo
S. Mchelelo
Nkano
1973/74
80-i
Fruit Seed Flower Gum Bark Shoot Leaves Animal


Figure 2.7. Mean monthly fruit scores from Mchelelo and Mnazini forests for Mimusops
fruticosa and Acacia robusta. seasonally fruiting species with 1 peak in
fruiting per year.
a) Mimusops fruiticosa in Mchelelo forest; b) Acacia robusta in Mchelelo
forest; c) Mimusops fruticosa in Mnazini forest; d) Acacia robusta in Mnazini
forest.


221
where N = number of breeding males or females (denoted
by subscript) in the population;
Nc = effective number of males or females (denoted by
subscript) in the population;
K = mean number of progeny produced by an individual
male or female during its lifetime; and
= variance of K for each sex.
Equation 2 Nc = 4[(1/Ncm) + (1/NJ]1
Equation 3 Nc = 1/2 [1 (II [1 l/2Nt (i) ] J17']'1
where Nc(i) = the effective population numbers in different
generations; and
t = generation length in years;
Estimating model parameters. Total population size (N)
was estimated by multiplying the number of groups within the
mangabeys total geographic range (Table 6.1) by the average
group size (Table 6.2). Nf and Nm were estimated by
multiplying the mean number of breeding females or males per
group (Table 6.2) by the total number of groups. Separate
calculations were made using minimum and maximum estimates
of group numbers and numbers of male and female breeders for
both 1973-74 and 1988-89.
Female lifetime reproductive success was simulated
for 1000 adult females. Because female mangabeys are capable
of producing a maximum of 6.25 infants in their lifetime
(Table 6.3), a potential 7 infants were assigned to each of
the 1000 females. Assuming a 1:1 sex ratio at birth, sex was


159
I used both parametric and nonparametric statistical
methods for data analyses. Spearman's Rank correlations were
used to examine associations between daily and monthly
travel patterns and the proportion of time spent in various
activities (e.g., eating, foraging and moving), the type of
dietary items consumed (e.g., fruit versus invertebrate),
and measures of fruit availability (Chapter 2). I also used
Spearman's Rank correlations to examine associations between
patterns of travel and the frequency of long-calls.
Student's t-tests were used to test for differences in the
means of tree density, tree species diversity, and percent
time spent in various behaviors between high-use and low-use
quadrats.
I used analysis-of-variance techniques (ANOVA) to
evaluate the effects of independent variables (month,
season, group membership, neighboring group presence, and
intergroup behaviors) on mangabey use of space, patterns of
travel, and frequency of long-calls. Arcsin square-root and
square-root transformations were used on proportional and
count data, respectively, to meet the assumptions of equal
variances (Sokal and Rohlf 1981). I tested for significance
with F-tests using partial sums of squares (Type III), which
adjust each effect for all other effects in the model and
are unrelated to cell frequencies (Freund and Littell 1981).
Multiple means comparisons of the distribution of foods
eaten (uniform versus patchy) and the frequency of long-


89
1988/89
1973/74
Eat Forage Move Inact Autogrm Allogrm Soc/Sex
Behavior



PAGE 1

%(+$9,25$/ $1' '(02*5$3+,& 5(63216(6 72 +$%,7$7 &+$1*( %< 7+( 7$1$ 5,9(5 &5(67(' 0$1*$%(< &HUFRFHEXV JDOHULWXV TDOHULWXVf %\ 0$5*$5(7 ) .,11$,5' $ ',66(57$7,21 35(6(17(' 72 7+( *5$'8$7( 6&+22/ 2) 7+( 81,9(56,7< 2) )/25,'$ ,1 3$57,$/ )8/),//0(17 2) 7+( 5(48,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

7R 5DWLOOL .DZD DQG WKH ZRPHQ RI UXUDO (DVW $IULFD

PAGE 3

$&.12:/('*0(176 0\ UHVHDUFK ZDV IXQGHG E\ :LOGOLIH &RQVHUYDWLRQ ,QWHUQDWLRQDO 1HZ
PAGE 4

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

PAGE 5

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nV GHSDUWXUH 7R RXU FDPS VWDII 0XVWDID /RUGLV %ZDQD 0]HH -XPDD *DODQD DQG $ELR JLYH WKDQNV IRU WKHLU KHOS DQG IULHQGVKLS DOVR RZH YHU\ VSHFLDO WKDQNV WR P\ ILHOG DVVLVWDQWV *DOROH *DODQD DQG %DNDUL .DZD 7KH\ WDXJKW PH PXFK DERXW MXQJOH OLIH DQG ZLWKRXW WKHLU KHOS WKH Y

PAGE 6

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n%ULHQ IRU YHU\ VSHFLDO DQG VLQFHUH WKDQNV 7KLV SURGXFW ZRXOG QRW KDYH EHHQ WKH VDPH ZLWKRXW KLV LQSXW +H KDV KHOSHG LQ OLWHUDOO\ DOO VWDJHVIURP DVVLVWLQJ LQ WKH ILHOG WR DGYLVLQJ RQ VWDWLVWLFDO DQDO\VLV ZULWLQJ QXPHURXV FRPSXWHU SURJUDPV DQG UHDGLQJ HYHU\ FKDSWHU +H VXIIHUHG WKURXJK DOO SKDVHV RI WKH SURMHFW DQG LQ VSLWH RI LW DOO VWXFN EHVLGH PH YL

PAGE 7

, DOVR WKDQN P\ SDUHQWV ZKR KDYH HQGXUHG PXFK LQFOXGLQJ P\ ORQJ UHVHDUFK VWLQWV RXW RI WKH FRXQWU\ 7KH\ KDYH HQFRXUDJHG DQG VXSSRUWHG PH LQ QXPHURXV ZD\V WKURXJKRXW DOO P\ HQGHDYRUV HYHQ ZKHQ WKH\ PD\ QRW KDYH XQGHUVWRRG IXOO\ MXVW ZKDW GURYH PH )LQDOO\ WKDQN WKH 0DNDUDX 0JRJR WKRVH UDUH DQG HOXVLYH SULPDWHV ZKR DOORZHG PH DOEHLW ZLWK LQVRXFLDQFH WR LQYHVWLJDWH WKHLU OLYHV :LWKRXW WKHLU FRRSHUDWLRQ WKLV VWXG\ ZRXOG QRW KDYH EHHQ IHDVLEOH 9OO

PAGE 8

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

PAGE 9

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

PAGE 10

/,67 2) 7$%/(6 7DEOH &RPSRVLWLRQ RI PRQWKO\ SKHQRORJ\ VDPSOH E\ VSHFLHV IRU 0FKHOHOR 1NDQR DQG 0QD]LQL IRUHVWV 7DEOH 'HQVLW\ RI UHSURGXFWLYH WUHHV DQG OLDQDV HQXPHUDWHG LQ KD RI 0FKHOHOR IRUHVW DQG KD LQ 1NDQR IRUHVW DQG WKH GHJUHH RI FOXPSLQJ RI HDFK VSHFLHV XVLQJ 0RULVLWDnV LQGH[ ,if ,I LV VLJQLILFDQWO\ JUHDWHU RU OHVV WKDQ XQLW\ XVLQJ WKH )WHVW WKH VSHFLHV LV FRQVLGHUHG FOXPSHG RU HYHQO\ GLVWULEXWHG LQ VSDFH UHVSHFWLYHO\ ,I LV QRW VLJQLILFDQWO\ GLIIHUHQW IURP XQLW\ WKDQ WKH VSHFLHV LV FRQVLGHUHG UDQGRPO\ GLVWULEXWHG 7DEOH (VWLPDWHG FDQRS\ VXUIDFH DUHD 6f FDQRS\ YROXPH 9f PHDQ IUXLW SURGXFWLRQ )1f DQG PHDQ IUXLW ZHLJKW ):f IRU VSHFLHV 6SHFLHV DUH FDWHJRUL]HG LQWR JURXSV WKRVH WKDW EHDU WKHLU IUXLWV RQ WKH FDQRS\ VXUIDFH &f RU DORQJ WKH EUDQFKHV RI WKH WUHH 5f OLDQDV /f DQG SDOPV 3f 7DEOH 6SHDUPDQnV UDQN FRUUHODWLRQ FRHIILFLHQWV IRU PHDQ PRQWKO\ HQYLURQPHQWDO YDULDEOHV DQG PHDQ PRQWKO\ IUXLWLQJ VFRUHV IRU 0FKHOHOR IRUHVW 0,17(03 PLQLPXP PRQWKO\ WHPSHUDWXUH 5$,1 WRWDO PRQWKO\ UDLQIDOO PPf 5$,1/$* 5$,1/$* $1' 5$,1/$* PRQWKO\ UDLQIDOO ODJJHG E\ DQG PRQWKV UHVSHFWLYHO\ 5$,1$'9 SUHYLRXV PRQWKVn UDLQIDOO )/: PHDQ PRQWKO\ ULYHU IORZ PFPf )/2:/$* )/2:/$* )/2:/$* PHDQ PRQWKO\ ULYHU IORZ ODJJHG E\ DQG PRQWKV UHVSHFWLYHO\ )/2:$'9 SUHYLRXV PRQWKVn PHDQ ULYHU IORZ 7DEOH 6SHDUPDQnV UDQN FRUUHODWLRQ FRHIILFLHQWV IRU PHDQ PRQWKO\ HQYLURQPHQWDO YDULDEOHV DQG PHDQ PRQWKO\ IUXLWLQJ VFRUHV IRU 1NDQR IRUHVW 9DULDEOHV GHILQHG LQ 7DEOH 7DEOH 6SHDUPDQnV UDQN FRUUHODWLRQ FRHIILFLHQWV IRU PHDQ PRQWKO\ HQYLURQPHQWDO YDULDEOHV DQG PHDQ PRQWKO\ IUXLWLQJ VFRUHV IRU 0QD]LQL IRUHVW 9DULDEOHV GHILQHG LQ 7DEOH [

PAGE 11

7DEOH 6SHDUPDQn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nV VWXG\ DQG WKH SUHVHQW VWXG\f DQG DJH GLIIHUHQFHV LQ PHDVXUHV RI PRWKHULQIDQW DVVRFLDWLRQ Df DPRXQW RI WLPH LQIDQW LV RQ LWVn PRWKHU Ef DPRXQW RI WLPH LQIDQW KDV PRWKHU DV QHDUHVW QHLJKERU DQG Ff DPRXQW RI WLPH LQIDQW LV VHSDUDWH IURP PRWKHU 'DWD ZHUH QHVWHG E\ LQIDQWV ZLWKLQ VWXGLHV 7DEOH 5DQN RUGHU RI WKH ILYH PRVW FRPPRQ SODQW VSHFLHV HDWHQ DQG WKH SHUFHQWDJH RI WKH WRWDO IHHGLQJ UHFRUGV IRU LGHQWLILHG SODQWV 1f E\ JURXS DQG PRQWK &RPSOHWH VFLHQWLILF QDPHV RI SODQW VSHFLHV OLVWHG LQ $SSHQGL[ % 7DEOH 3HUFHQWDJH RI GLHW LWHPV HDWHQ E\ PRQWK DQG JURXS 8QLGHQWLILHG LWHPV DUH QRW LQFOXGHG 7DEOH )UHTXHQF\ RI SODQW SDUWV HDWHQ IRU FRPPRQ GLHW VSHFLHV RI WKH VWXG\ JURXSV 7DEOH +XPDQ XVHV RI 3 UHHOLQDWD JURXSHG LQWR EURDG FDWHJRULHV 3ODQW SDUWV XWLOL]HG DQG KDUYHVWLQJ PHWKRGV HPSOR\HG DUH RXWOLQHG 7DEOH 0HDQ GLYHUVLW\ RI TXDGUDW XVH +nf PHDQ QXPEHU RI XQLTXH TXDGUDWV HQWHUHG PHDQ GLVWDQFHV WUDYHOOHG PHDQ KDOIKRXU VWHS GLVWDQFHV PHDQ WXUQLQJ DQJOHV DQG PHDQ QXPEHU RI SDWK FURVVLQJV E\ PRQWK IRU WKH VWXG\ JURXSV [L

PAGE 12

7DEOH 6SHDUPDQnV 5DQN FRUUHODWLRQ FRHIILFLHQWV UAf EHWZHHQ IUHTXHQF\ RI TXDGUDW XVH E\ PDQJDEH\V DQG IUHTXHQF\ RI RFFXUUHQFH RI WRS SODQW GLHW VSHFLHV HDFK PRQWK IRU WKH 1RUWK 0FKHOHOR 6RXWK 0FKHOHOR DQG 1NDQR JURXSV f LQGLFDWHV WKDW UDQJLQJ GDWD RU SODQW IUHTXHQF\ GDWD ZHUH QRW DYDLODEOH IRU DOO WRS GLHW VSHFLHV WKDW PRQWK 7DEOH $129$ UHVXOWV IRU HIIHFW RI 6RXWK 0FKHOHOR JURXS SUHVHQFH DQG PRQWK RQ GLVWDQFH WUDYHOHG DQG PHDVXUHV RI WUDYHO SDWWHUQ DQG XVH RI VSDFH E\ 1RUWK 0FKHOHOR JURXS )VWDWLVWLFV DUH FDOFXODWHG IURP 7\SH ,,, SDUWLDO VXPV RI VTXDUHV 7DEOH 1XPEHU RI DGXOW PDOH ORQJFDOOV E\ VWXG\ JURXS DQG PRQWK 1RUWK UHIHUV WR ORQJFDOOV JLYHQ E\ WKH 1RUWK 0FKHOHOR JURXS PDOHV 6RXWK 0 UHIHUV WR ORQJFDOOV JLYHQ E\ WKH 6RXWK 0FKHOHOR JURXS PDOHV ZKLOH SUHVHQW LQ 0FKHOHOR IRUHVW DQG 6RXWK & UHIHUV WR ORQJFDOOV JLYHQ E\ WKH 6RXWK 0FKHOHOR JURXS PDOHV ZKLOH LQ WKH QHLJKERULQJ &RQJRODQL IRUHVW $OLHQ UHIHUV WR ORQJFDOOV JLYHQ E\ PDOHV RI D VHFRQG XQKDELWXDWHG JURXS LQ 1NDQR IRUHVW DQG 8QN UHSUHVHQWV ORQJFDOOV RI XQLGHQWLILHG PDOHV LQ WKH 1NDQR IRUHVW 7DEOH ,QWHUJURXS EHKDYLRU FDWHJRULHV DQG PHDVXUHV RI SUR[LPLW\ DQG ILJKWLQJ E\ PRQWK ) 0 DQG $ GHVLJQDWH ILJKWV PHUJH DQG DYRLG PRQWKV UHVSHFWLYHO\ 1 LQFOXGHV DOO KU REVHUYDWLRQ GD\V DQG GD\V IRU ZKLFK RQO\ SUHVHQFH RU DEVHQFH RI 6RXWK JURXS ZDV GHWHUPLQHG 7DEOH 0LQLPXP DQG PD[LPXP QXPEHUV RI PDQJDEH\ JURXSV FHQVXVHG LQ IRUHVWV RQ WKH HDVW DQG ZHVW EDQNV RI WKH ORZHU 7DQD 5LYHU LQ DQG )RUHVWV QRW FHQVXVHG LQGLFDWHG ZLWK f WKRVH QR ORQJHU VWDQGLQJ LQGLFDWHG ZLWK [f 'DWD IURP WDNHQ IURP +RPHZRRG f DQG 0DUVK f 7DEOH $JHVH[ FRPSRVLWLRQ RI PDQJDEH\ JURXSV ZLWK FRPSDUDWLYH PHDQV IURP +RPHZRRG f $0 DGXOW PDOH \Uf 6XE0 VXEDGXOW PDOH \Uf $) DGXOW IHPDOH \Uf 6XE) VXEDGXOW IHPDOH \Uf -0 MXYHQLOH PDOH PRQ \Uf -) MXYHQLOH IHPDOH PRQ \Uf ,0 DQG ,) LQIDQW PDOH DQG LQIDQW IHPDOH UHVSHFWLYHO\ PRQf 7DEOH 5HSURGXFWLYH SDUDPHWHUV IRU PDOH DQG IHPDOH PDQJDEH\V ;OO

PAGE 13

7DEOH 0LQLPXP DQG PD[LPXP SDUDPHWHU YDOXHV DQG HVWLPDWHV RI HIIHFWLYH SRSXODWLRQ VL]H 1 FHQVXV QXPEHU RI PDOHV RU IHPDOHV HVWLPDWH RI OLIHWLPH UHSURGXFWLYH VXFFHVV RI PDOHV RU IHPDOHV YDULDQFH LQ OLIHWLPH UHSURGXFWLYH VXFFHVV DQG 1f HIIHFWLYH SRSXODWLRQ VL]H 6XEVFULSWV GHQRWH SDUDPHWHUV IRU PDOHV YV IHPDOHV 7DEOH 3DUDPHWHU YDOXHV DQG HVWLPDWHV RI WKH JHQHWLF IL[DWLRQ LQGH[ )f IRU IRUHVWV RQ HDVW DQG ZHVW EDQNV DQG IRU ZHVW EDQN DQG HDVW EDQN VXESRSXODWLRQV (DVW YV ZHVW EDQN ILJXUHV DUH FDOFXODWHG IRU JURXS PLJUDWLRQ IURP RQH EDQN WR WKH RWKHU IRU WZR HVWLPDWHV RI KLJK IORRG IUHTXHQF\ [LLL

PAGE 14

/,67 2) ),*85(6 )LJXUH $GXOW PDOH 7DQD 5LYHU FUHVWHG PDQJDEH\ &HUFRFHEXV JDOHULWXV TDOHULWXV VHDWHG LQ D +YSKHDQH FRPSUHVVD SDOP )LJXUH 0DS RI .HQ\D DQG ORFDWLRQ RI WKH 7DQD 5LYHU 1DWLRQDO 3ULPDWH 5HVHUYH 75135 PDS PRGLILHG IURP +XJKHV f )LJXUH 0DS RI WKH VWXG\ DUHD VKRZLQJ WKH 0FKHOHOR 0QD]LQL DQG 1NDQR IRUHVWV ZLWKLQ WKH 7DQD 5LYHU 1DWLRQDO 3ULPDWH 5HVHUYH )LJXUH 1XPEHU RI LQGLYLGXDOV KD RI FRPPRQ VSHFLHV ZLWKLQ 0FKHOHOR IRUHVW VKRZLQJ FOXPSHG VSDWLDO GLVWULEXWLRQV Df +\SKDHQH FRPSUHVVD Ef$ODQJLXP VDOYLLIROLXP Ff3KRHQL[ UHFOLQDWD Gf 2QFRED VSLQRVD )LJXUH 1XPEHU RI LQGLYLGXDOV KD RI FRPPRQ VSHFLHV ZLWKLQ 1NDQR IRUHVW VKRZLQJ FOXPSHG VSDWLDO GLVWULEXWLRQV Df )LFXV VYFRPRUXV Ef $ODQJLXP VDOYLLIROLXP Ff 3KRHQL[ UHFOLQDWD Gf 3DFKYVWHOD PVROR )LJXUH ([DPSOHV RI UDQGRPO\ GLVWULEXWHG )LFXV VSS Df )LFXV QDWDOHQVLV LQ 0FKHOHOR IRUHVW Ef )LFXV VYFRPRUXV LQ 0FKHOHOR IRUHVW Ff )LFXV QDWDOHQVLV LQ 1NDQR IRUHVW )LJXUH 0HDQ IORZHU VFRUHV DYHUDJHG RYHU DOO VSHFLHV E\ PRQWK Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW )LJXUH 0HDQ IUXLWLQJ VFRUHV DYHUDJHG RYHU DOO VSHFLHV E\ PRQWK Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW [LY

PAGE 15

)LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IURP 0FKHOHOR DQG 0QD]LQL IRUHVWV IRU 0LPXVRSV IUXWLFRVD DQG $FDFLD UREXVWD VHDVRQDOO\ IUXLWLQJ VSHFLHV ZLWK SHDN LQ IUXLWLQJ SHU \HDU Df 0LPXVRSV IUXLWLFRVD LQ 0FKHOHOR IRUHVW Ef $FDFLD UREXVWD LQ 0FKHOHOR IRUHVW Ff 0LPXVRSV IUXWLFRVD LQ 0QD]LQL IRUHVW Gf $FDFLD UREXVWD LQ 0QD]LQL IRUHVW )LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IURP 0FKHOHOR DQG 0QD]LQL IRUHVWV IRU *DUFLQLD OLYLQJVWRQHL DQG 6RULQGHLD PDGDTDVFDULHQVLV VHDVRQDOO\ IUXLWLQJ VSHFLHV ZLWK IUXLWLQJ SHDNV SHU \HDU Df *DUFLQLD OLYLQJVWRQHL LQ 0FKHOHOR IRUHVW Ef 6RULQGHLD PDGDTDVFDULHQVLV LQ 0FKHOHOR IRUHVW Ff *DUFLQLD OLYLQJVWRQHL LQ 0QD]LQL IRUHVW Gf 6RULQGHLD PDGDTDVFDULHQVLV LQ 0QD]LQL IRUHVW )LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IRU 3KRHQL[ UHFOLQDWD Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW )LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IRU 'LRVS\URV PHVSLOLIRUPHV Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW )LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IRU VSHFLHV VKRZLQJ SDWWHUQV RI FRQWLQXRXV IUXLW SURGXFWLRQ Df FRQWLQXRXV IUXLWLQJ GXH WR DV\QFKURQ\ DPRQJ LQGLYLGXDOV RI WKH VSHFLHV )LFXV QDWDOHQVLVf Ef FRQWLQXRXV IUXLWLQJ GXH WR UHWHQWLRQ RI IUXLWV RYHU WLPH +\SKDHQH FRPSUHVVDf Ff FRQWLQXRXV IUXLWLQJ GXH WR UDSLG WXUQRYHU RI IUXLWV 2QFRED VSLQRVDf )LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IRU )LFXV V\FRPRUXV Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW )LJXUH 6HDVRQDOLW\ RI PHDQ PRQWKO\ IUXLW VFRUHV IRU 6DED FRPRUHQVLV LQ WKH VWXG\ IRUHVWV Df VHDVRQDO IUXLWLQJ LQ 0QD]LQL IRUHVW Ef VHDVRQDO EXW H[WHQGHG IUXLWLQJ LQ 1NDQR IRUHVW Ff FRQWLQXRXV IUXLWLQJ LQ 0FKHOHOR IRUHVW )LJXUH &RQWULEXWLRQ WR PRQWKO\ IUXLW ELRPDVV LQGLFHV E\ GLIIHUHQW VSHFLHV FDWHJRULHV Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW [Y

PAGE 16

)LJXUH &RQWULEXWLRQ RI ULSH DQG XQULSH IUXLWV WR PRQWKO\ IUXLW ELRPDVV LQGH[ Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW )LJXUH 0RQWKO\ PHDQ PLQLPXP DQG PHDQ PD[LPXP DLU WHPSHUDWXUH LQ GHJUHHV FHQWLJUDGH Df 0FKHOHOR IRUHVW Ef 1NDQR0QD]LQL IRUHVW DUHDV )LJXUH 7RWDO PRQWKO\ UDLQIDOO LQ PLOOLPHWHUV Df 0FKHOHOR IRUHVW Ef 1NDQR0QD]LQL IRUHVW DUHD )LJXUH 0HDQ PRQWKO\ ULYHU IORZ IRU WKH 7DQD 5LYHU +ROD 6WDWLRQ )LJXUH 0HDQ DQQXDO SHUFHQW WLPH VSHQW DPRQJ EHKDYLRUV E\ WKUHH PDQJDEH\ JURXSV DQG WZR PDQJDEH\ JURXSV Df Ef )LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ PDMRU EHKDYLRUV HDWLQJ IRUDJLQJ PRYLQJ DQG LQDFWLYHf E\ WLPHRIGD\ DQG VHDVRQ IRU WKH 1 0FKHOHOR JURXS )LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ PDMRU EHKDYLRUV HDWLQJ IRUDJLQJ PRYLQJ DQG LQDFWLYHf E\ WLPHRIGD\ DQG VHDVRQ IRU WKH 6 0FKHOHOR JURXS )LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ PDMRU EHKDYLRUV HDWLQJ IRUDJLQJ PRYLQJ DQG LQDFWLYHf E\ WLPHRIGD\ DQG VHDVRQ IRU WKH 1NDQR JURXS )LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ VRFLDO EHKDYLRUV E\ DQG DGXOW PDOHV DQG IHPDOHV Df DGXOW PDOHV Ef DGXOW IHPDOHV )LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ VRFLDO EHKDYLRUV E\ WKUHH DJHVH[ FODVVHV LQ DQG Df MXYHQLOH PDOHV Ef MXYHQLOH IHPDOHV Ff PDOH DQG IHPDOH LQIDQWV FRPELQHG )LJXUH 3HUFHQW WLPH DQG LQIDQWV VSHQW LQ OHYHOV RI DVVRFLDWLRQ ZLWK WKHLU PRWKHUV $VWHULVNV GHQRWH PRQWKV ZKHQ OHYHOV RI DVVRFLDWLRQ ZHUH VLJQLILFDQWO\ GLIIHUHQW EHWZHHQ WKH VWXG\ SHULRGV Df $VVRFLDWLRQ $ WKH DPRXQW RI WLPH DQ LQIDQW ZDV RQ LWV PRWKHU Ef $VVRFLDWLRQ % WKH DPRXQW RI WLPH DQ LQIDQW KDG LWV PRWKHU DV QHDUHVW QHLJKERU Ff $VVRFLDWLRQ & WKH DPRXQW RI WLPH LQIDQWV ZHUH VHSDUDWH IURP WKHLU PRWKHUV [YL

PAGE 17

)LJXUH 3HUFHQW FRPSRVLWLRQ RI GLHW E\ LWHP IRU DQG Df Ef )LJXUH 3HUFHQW ULSH DQG XQULSH IUXLW DQG VHHG LWHPV LQWKH GLHW Df IRU WKUHH PDQJDEH\ JURXSV Ef IRU PDQJDEH\ JURXSV )LJXUH 7KH QXPEHU RI KDUYHVWHG DQG XQKDUYHVWHG SDOPV LQ 1NDQR DQG WKH QXPEHU RI UHSURGXFWLYH DQG QRQUHSURGXFWLYH SDOPV LQ 0FKHOHOR E\ VL]H FODVV Df KDUYHVWHG DQG XQKDUYHVWHG SDOPV E\ VL]H FODVV LQ 1NDQR Ef UHSURGXFWLYH DQG QRQUHSURGXFWLYH SDOPV LQ 0FKHOHOR E\ VL]H FODVV )LJXUH 7KH QXPEHU RI IHHGLQJ REVHUYDWLRQV RQ 3 UHFOLQDWD E\ PRQWK DQG LWHP IRU PDQJDEH\V LQ WKH 0FKHOHOR DQG 1NDQR IRUHVWV )LJXUH &XPXODWLYH IUHTXHQF\ RI VLJKWLQJV RI RQH RU PRUH 1RUWK 0FKHOHOR JURXS PDQJDEH\V YHUVXV UDQN RUGHU RI K TXDGUDWV 4XDGUDWV ZHUH UDQNHG ZLWK UHVSHFW WR WKH QXPEHU RI VLJKWLQJV RFFXUULQJ LQ HDFK 7KH DUHDV DFFRXQWLQJ IRU DQG b RI DOO VLJKWLQJV DUH GHVLJQDWHG )LJXUH &XPXODWLYH IUHTXHQF\ RI VLJKWLQJV RI RQH RU PRUH 6RXWK 0FKHOHOR JURXS PDQJDEH\V YHUVXV UDQN RUGHU RI K TXDGUDWV 4XDGUDWV ZHUH UDQNHG ZLWK UHVSHFW WR WKH QXPEHU RI VLJKWLQJV RFFXUULQJ LQ HDFK 7KH DUHDV DFFRXQWLQJ IRU DQG b RI DOO VLJKWLQJV DUH GHVLJQDWHG )LJXUH &XPXODWLYH IUHTXHQF\ RI VLJKWLQJV RI RQH RU PRUH 1NDQR JURXS PDQJDEH\V YHUVXV UDQN RUGHU RI KD TXDGUDWV 4XDGUDWV ZHUH UDQNHG ZLWK UHVSHFW WR WKH QXPEHU RI VLJKWLQJV RFFXUULQJ LQ HDFK 7KH DUHDV DFFRXQWLQJ IRU DQG b RI DOO VLJKWLQJV DUH GHVLJQDWHG )LJXUH 7DXWVWULQJ OLQH VKDGHG ZLWK GLDJRQDO OLQHVf HQFORVLQJ DOO DG OLELWXP VLJKWLQJV RI 6RXWK 0FKHOHOR JURXS ZKHQ WKH JURXS UDQJHG RXWVLGH WKH 0FKHOHOR IRUHVW )LJXUH )UHTXHQF\ GLVWULEXWLRQV RI LQWHQVLW\ RI TXDGUDW XVH $VWHULVNV LQGLFDWH WKH YDOXHV H[SHFWHG LI TXDGUDW RFFXSDQF\ ZHUH UDQGRP IROORZLQJ D 3RLVVRQ GLVWULEXWLRQfDf )UHTXHQF\ GLVWULEXWLRQ IRU 1 0FKHOHOR JURXS Ef )UHTXHQF\ GLVWULEXWLRQ IRU 6 0FKHOHOR JURXS Ff )UHTXHQF\ GLVWULEXWLRQ IRU 1NDQR JURXS [YLL

PAGE 18

)LJXUH 2FFXSDQF\ RI HDFK KD TXDGUDW E\ 1RUWK DQG 6RXWK 0FKHOHOR JURXSV SHUFHQW RFFXSDQF\ LV GLYLGHG LQWR RI LQWHQVLWLHV RI XVH )LJXUH 2FFXSDQF\ RI HDFK KD TXDGUDW E\ 1NDQR JURXS SHUFHQW RFFXSDQF\ LV GLYLGHG LQWR RI LQWHQVLWLHV RI XVH )LJXUH )UHTXHQF\ GLVWULEXWLRQV RI KDOIKRXU GLVWDQFH VWHSV LQ LQWHUYDOV RI P $VWHULVNV LQGLFDWH YDOXHV H[SHFWHG LI GLVWDQFH VWHSV ZHUH GLVWULEXWHG UDQGRPO\ IROORZLQJ D 3RLVVRQ GLVWULEXWLRQf Df )UHTXHQF\ GLVWULEXWLRQ IRU 1 0FKHOHOR JURXS Ef )UHTXHQF\ GLVWULEXWLRQ IRU 6 0FKHOHOR JURXS Ff )UHTXHQF\ GLVWULEXWLRQ IRU 1NDQR JURXS )LJXUH )UHTXHQF\ GLVWULEXWLRQV RI KDOIKRXU WXUQLQJ DQJOHV LQ GHJUHH LQWHUYDOV Df )UHTXHQF\ GLVWULEXWLRQ IRU 1 0FKHOHOR JURXS Ef )UHTXHQF\ GLVWULEXWLRQ IRU 6 0FKHOHOR JURXS Ff )UHTXHQF\ GLVWULEXWLRQ IRU 1NDQR JURXS )LJXUH 'HJUHH RI PRQWKO\ RYHUODS E\ 1RUWK 0FKHOHOR JURXS VROLG OLQHf DQG 6RXWK 0FKHOHOR JURXS EURNHQ OLQHf DQG WKH DUHD RI RYHUODS LQ KHFWDUHV EDUVf )LJXUH 'DLO\ WUDYHO SDWKV E\ 1RUWK DQG 6RXWK 0FKHOHOR JURXSV GXULQJ RQH GD\ RI D PRQWK RI LQWHUJURXS DYRLGDQFH 7LPLQJ DQG ORFDWLRQ RI DGXOW PDOH ORQJFDOOV DUH SORWWHG IRU 1RUWK RYDOVf DQG 6RXWK UHFWDQJOHVf JURXSV )LJXUH 'DLO\ WUDYHO SDWKV E\ 1RUWK DQG 6RXWK 0FKHOHOR JURXSV GXULQJ RQH GD\ RI D PRQWK RI LQWHUJURXS PHUJLQJ 7LPLQJ DQG ORFDWLRQ RI DGXOW PDOH ORQJFDOOV DUH SORWWHG IRU 1RUWK RYDOVf DQG 6RXWK UHFWDQJOHVf JURXSV )LJXUH 'DLO\ WUDYHO SDWKV E\ 1RUWK DQG 6RXWK 0FKHOHOR JURXSV GXULQJ RQH GD\ RI D PRQWK RI LQWHUJURXS ILJKWLQJ 7LPLQJ DQG ORFDWLRQ RI DGXOW PDOH ORQJFDOOV DUH SORWWHG IRU 1RUWK RYDOVf DQG 6RXWK UHFWDQJOHVf JURXSV )LJXUH /RFDWLRQ DQG RXWFRPH RI ILJKWV REVHUYHG EHWZHHQ 1RUWK DQG 6RXWK 0FKHOHOR JURXSV 1 DQG 6 LQGLFDWH WKDW 1RUWK DQG 6RXWK JURXS ZRQ ILJKWV UHVSHFWLYHO\ LQGLFDWHV WKDW QR FOHDU ZLQQHU ZDV GHWHUPLQHG ;9,

PAGE 19

)LJXUH 3HUFHQW WLPH 1RUWK DQG 6RXWK 0FKHOHOR JURXSV VSHQW IHHGLQJ RQ IRRGV ZLWK GLIIHUHQW GLVWULEXWLRQV DV D IXQFWLRQ RI LQWHUJURXS EHKDYLRU Df 3HUFHQW WLPH IHHGLQJ RQ IRRGV ZLWK XQLIRUP GLVWULEXWLRQV Ef 3HUFHQW WLPH IHHGLQJ RQ IRRGV ZLWK SDWFK\ GLVWULEXWLRQV )LJXUH 'LVWULEXWLRQ RI HVWLPDWHG IHPDOH OLIHWLPH UHSURGXFWLYH VXFFHVV ,QIDQW YDOXHV DUH URXQGHG WR WKH QHDUHVW ZKROH QXPEHU )LJXUH 'LVWULEXWLRQ RI PDOH OLIHWLPH UHSURGXFWLYH VXFFHVV HVWLPDWHG IRU PDOHV ZLWK DQG WHQXUHV DV D GRPLQDQW EUHHGLQJ JURXS PDOH Df PDOH OLIHWLPH UHSURGXFWLYH VXFFHVV FDOFXODWHG IRU PDOH IHPDOHV Ef PDOH OLIHWLPH UHSURGXFWLYH VXFFHVV FDOFXODWHG IRU PDOH IHPDOHV )LJXUH 3HUVLVWHQFH WLPH LQ \HDUV DV D IXQFWLRQ RI SRSXODWLRQ VL]H 1f 'DVKHG OLQHV FRQQHFW DQG SRSXODWLRQ VL]HV WR HVWLPDWHG PHDQ SHUVLVWHQFH WLPHV )LJXUH 3UREDELOLW\ RI H[WLQFWLRQ S(f ZLWKLQ DQG \HDUV GHQRWHG E\ ODEHOHG FXUYHVf IRU SRSXODWLRQV ZLWK PHDQ SHUVLVWHQFH WLPHV 7f RI WR \HDUV )LJXUH (IIHFWLYH SRSXODWLRQ VL]H 1DV D IXQFWLRQ RI QXPEHU RI EUHHGHUV 1Ef DQG YDULDQFH RI PDOH OLIHWLPH UHSURGXFWLYH VXFFHVV 9Pf [L[

PAGE 20

$EVWUDFW RI 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH 6FKRRO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WKH 'HJUHH RI 'RFWRU RI 3KLORVRSK\ %(+$9,25$/ $1' '(02*5$3+,& 5(63216(6 72 +$%,7$7 &+$1*( %< 7+( 7$1$ 5,9(5 &5(67(' 0$1*$%(< &(5&2&(%86 *$/(5,786 *$/(5,786f %\ 0DUJDUHW ) .LQQDLUG 'HFHPEHU &KDLUSHUVRQ 'U -RKQ ) (LVHQEHUJ 0DMRU 'HSDUWPHQW )RUHVW 5HVRXUFHV DQG &RQVHUYDWLRQ :LOGOLIH DQG 5DQJH 6FLHQFHVf 7KH 7DQD 5LYHU FUHVWHG PDQJDEH\ &HUFRFHEXV TDOHULWXV TDOHULWXVf LV DQ HQGDQJHUHG SULPDWH IRXQG RQO\ LQ VPDOO IRUHVWV RI WKH 7DQD 5LYHU LQ .HQ\D 7KH PDQJDEH\ LV DIIRUGHG VRPH SURWHFWLRQ LQ WKH 7DQD 5LYHU 1DWLRQDO 3ULPDWH 5HVHUYH 6LQFH WKH HVWDEOLVKPHQW RI WKH UHVHUYH LQ WKH PDQJDEH\ SRSXODWLRQ KDV GHFUHDVHG E\ b 7KH UHGXFWLRQ LQ KDELWDW VL]H DQG TXDOLW\ LV EHOLHYHG WR KDYH EHHQ D PDMRU IDFWRU FRQWULEXWLQJ WR WKH SRSXODWLRQ GHFOLQH FRPSDUH GDWD RQ PDQJDEH\ EHKDYLRU DQG GHPRJUDSK\ ZLWK D VWXG\ FRQGXFWHG LQ WR LQYHVWLJDWH WKH HIIHFWV RI KDELWDW ORVV DQG UHGXFWLRQV LQ LPSRUWDQW GLHW VSHFLHV RQ PDQJDEH\ EHKDYLRU DQG ULVN RI H[WLQFWLRQ [[

PAGE 21

7KH 7DQD PDQJDEH\n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n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

PAGE 22

PDQJDEH\ LV D JHQHUDOLVW VSHFLHV WKDW HDVLO\ H[SORLWV QHZ KDELWDW LW VKRXOG UHVSRQG ZHOO WR KDELWDW PDQDJHPHQW ;;,,

PAGE 23

&+$37(5 21( ,1752'8&7,21 7KH 7DQD 5LYHU FUHVWHG PDQJDEH\ RU 7DQD PDQJDEH\ &HUFRFHEXV TDOHULWXV TDOHULWXV 3HWHUV f LV D UDUH DQG KLJKO\ HQGDQJHUHG FHUFRSLWKHFLQH PRQNH\ ,8&1 /HH HW DO f IRXQG RQO\ LQ VPDOO IRUHVW SDWFKHV RI WKH 7DQD IORRGSODLQ LQ QRUWKHDVWHUQ .HQ\D +RPHZRRG )LJXUH f 0LWWHUPHLHU f UDQNHG & J TDOHULWXV DPRQJ WKH PRVW HQGDQJHUHG SULPDWHV LQ WKH ZRUOG DQG WKH ,8&166& 3ULPDWH 6SHFLDOLVW *URXS OLVWHG LW LQ WKH KLJKHVW SULRULW\ FDWHJRU\ IRU FRQVHUYDWLRQ 2DWHV f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f 7KH 7DQD PDQJDEH\ LV WKRXJKW WR EH D UHOLFW SRSXODWLRQ RI D RQFH ZLGHO\ GLVWULEXWHG VSHFLHV +RPHZRRG f

PAGE 24

)LJXUH $GXOW PDOH 7DQD 5LYHU FUHVWHG PDQJDEH\ &HUFRFHEXV JDOHULWXV TDOHULWXV VHDWHG LQ +\SKHDQH FRPSUHVVD SDOP

PAGE 26

RWKHU VXEVSHFLHV & J FKUYVRTDVWHU DQG & J DJLOLV LQKDELW WKH &RQJROHVH IRUHVW EORFN IURP &DPHURRQ WKURXJK *DERQ &RQJR %UD]]DYLOOHf DQG =DLUH 'DQGHORW /HH HW DO f $ IRXUWK VXEVSHFLHV & J VDQLHL RFFXUV LQ WKH 8]XQJZD 0RXQWDLQV RI 7DQ]DQLD +RPHZRRG DQG 5RGJHUV /HH HW DO f $OO VXEVSHFLHV RI & TDOHULWXV DUH VHPLn WHUUHVWULDO DQG EURDGO\ VLPLODU LQ HFRORJ\ DQG VRFLDO EHKDYLRU +RPHZRRG 4XULV f EXW YDU\ LQ PRUSKRORJ\ 'REURUXND DQG %DGDOHF f DQG *URYHV f KDYH FODVVLILHG WKH 7DQD PDQJDEH\ DV D VHSDUDWH VSHFLHV EDVHG RQ GLIIHUHQFHV LQ VNXOO PRUSKRORJ\ ,UUHVSHFWLYH RI WKH WD[RQRP\ XVHG WKH LPSRUWDQW SRLQW LV WKDW WKH 7DQD PDQJDEH\ LV D JHRJUDSKLFDOO\ DQG SHUKDSV JHQHWLFDOO\ GLVWLQFW SRSXODWLRQ DQG VKRXOG EH FRQVLGHUHG VHSDUDWHO\ IURP LWV FRQVSHFLILFV LQ WKH FRQVHUYDWLRQ DUHQD 7KH 7DQD PDQJDEH\ RFFXUV RQO\ ZLWKLQ D NP VWUHWFK RI WKH ORZHU 7DQD 5LYHU $QGUHZV HW DO +RPHZRRG f ZKHUH JURXQGZDWHU VHHSDJH IURP WKH ULYHU LV VXIILFLHQW WR VXSSRUW SDWFKHV RI ULYHULQH IRUHVW LQ D UHJLRQ RI RWKHUZLVH VHPLDULG EXVKODQG :LWKLQ WKH PDQJDEH\Vn UDQJH WKHUH DUH DQ HVWLPDWHG IRUHVW SDWFKHV DYHUDJLQJ KD LQ VL]H UDQJH KD 0DUVK D f VHSDUDWHG E\ VWUHWFKHV RI GHFLGXRXV ZRRGODQG EXVK DQG JUDVV )RUHVW SDWFKHV RU IUDJPHQWV UHVXOW LQ SDUW IURP FKDQJHV LQ WKH ULYHUnV FRXUVH DQG SDWWHUQV RI IORRGLQJ &KDQJHV LQ WKH ULYHUnV FRXUVH GXULQJ KLJK IORRGV OHDYHV VRPH IRUHVWV IDU IURP WKH OLIH

PAGE 27

VXSSRUWLQJ JURXQGZDWHU VXSSO\ RWKHUV DUH FXW DQG HURGHG E\ WKH VFRXULQJ DFWLRQ RI WKH ULYHU ,Q UHFHQW \HDUV QDWXUDO IRUHVW IUDJPHQWDWLRQ KDV EHHQ H[DFHUEDWHG E\ LQFUHDVLQJ KXPDQ H[SORLWDWLRQ DQG ODQG XVH +RPHZRRG +XJKHV .LQQDLUG HW DO 0DUVK f $Q LQFUHDVLQJ ORFDO SRSXODWLRQ HJ WKH QXPEHU RI IDPLOLHV LQ RQH YLOODJH 1NDQR LQFUHDVHG IURP WR EHWZHHQ DQG SHUVRQDO REVHUYDWLRQ 77 6WUXKVDNHUf GHSHQGHQW RQ VODVKDQGEXUQ DJULFXOWXUH KDV UHVXOWHG LQ LQFUHDVHV LQ IRUHVW FOHDULQJ DQG WUHH IHOOLQJ IRU KRXVHV DQG GXJRXW FDQRHV .LQQDLUG HW DO f 0HGOH\ f VKRZHG D b UHGXFWLRQ LQ IRUHVW DUHD DQG IUDJPHQWDWLRQ RI IRUHVW DUHDV LQWR SDWFKHV EHWZHHQ DQG 7KLV ORVV ZDV DWWULEXWHG LQ SDUW WR FKDQJHV LQ WKH ULYHUV FRXUVH 0HGOH\ f DQG WR WKH GHVWUXFWLRQ RI IRUHVW IRU DJULFXOWXUH 'HFNHU f )RUHVWV DOVR KDYH EHHQ FOHDUHG IRU ODUJHVFDOH LUULJDWLRQ VFKHPHV 0HGOH\ HW DO f DQG SRVVLEO\ KDYH H[SHULHQFHG D UHGXFWLRQ LQ UHJHQHUDWLRQ GXH WR DOWHUDWLRQV LQ WKH ULYHUV IORZ E\ XSVWUHDP GDPV +XJKHV /HH HW DO 0DUVK f $OWKRXJK FKDQJHV LQ WKH IORRGLQJ UHJLPH GXH WR XSVWUHDP GDPV KDYH QRW EHHQ YHULILHG +XJKHV f KDV VSHFXODWHG WKDW D UHGXFWLRQ LQ SHDN IORRGV GXH WR GDPPLQJ ZLOO UHGXFH WKH VHGLPHQWQXWULHQW ORDG DQG WKH IUHTXHQF\ RI PHDQGHU IRUPDWLRQ DQG FXWRII WKHUHE\ UHGXFLQJ WKH G\QDPLF QDWXUH RI WKH IORRGSODLQ DQG WKH

PAGE 28

RFFXUUHQFH RI IDYRUDEOH FRQGLWLRQV IRU VHHG JHUPLQDWLRQ DQG VHHGOLQJ HVWDEOLVKPHQW 'DWD IURP 0HGOH\ f LQGLFDWH WKDW XSVWUHDP GDPV PD\ EH KDYLQJ DQ HIIHFW WKH 7DQD 5LYHU KDV VKRUWHQHG LWV FRXUVH LQ UHFHQW \HDUV DV D UHVXOW RI OHVV PHDQGHULQJ DQG VHGLPHQWV ORDGV GHOLYHUHG GXULQJ IORRGV KDYH GHFOLQHG $JULFXOWXUDO SUDFWLFHV DOVR LQIOXHQFH IRUHVW UHJHQHUDWLRQ 6LWHV RI DFWLYH IRUHVW UHJHQHUDWLRQ LH SRLQWEDUVf DUH VHOHFWHG IRU IDUPLQJ EHFDXVH PXFK RI WKH FURS SURGXFWLRQ E\ WKH ORFDO SHRSOH LV GHSHQGHQW RQ ELDQQXDO IORRGLQJ DQG WKH QXWULHQW ULFK VRLOV QHDU WKH ULYHU 7KH 7DQD 5LYHU 1DWLRQDO 3ULPDWH 5HVHUYH 75135 )LJXUH f ZDV JD]HWWHG LQ E\ WKH JRYHUQPHQW RI .HQ\D WR SURWHFW WKH EHVW UHPDLQLQJ VWDQGV RI ULYHULQH IRUHVW DQG SRSXODWLRQV RI WKH 7DQD 5LYHU FUHVWHG PDQJDEH\ DQG UHG FROREXV PRQNH\ &ROREXV EDGLXV UXIRPLWUDWXVf $W WKH WLPH RI HVWDEOLVKPHQW WKH UHVHUYH RFFXSLHG DSSUR[LPDWHO\ NP RI IRUHVW GU\ ZRRGODQG DQG VDYDQQD KDELWDW RQ WKH HDVW DQG ZHVW EDQNV RI WKH ULYHU 0DUVK f 7KH IRUHVW KDELWDW FRYHUHG DQ HVWLPDWHG NP EURNHQ LQWR DSSUR[LPDWHO\ SDWFKHV ,Q WKH \HDUV VLQFH WKH HVWDEOLVKPHQW RI WKH UHVHUYH WKHUH KDV EHHQ D GHFOLQH LQ IRUHVW DUHD DQG LQFUHDVHG IUDJPHQWDWLRQ RI WKH UHPDLQLQJ IRUHVW VWDQGV .LQQDLUG HW DO 0HGOH\ f 3ULPDWH SRSXODWLRQV DOVR KDYH GHFOLQHG E\ DQ HVWLPDWHG b IRU WKH UHG FROREXV 0DUVK 'HFNHU f DQG b IRU WKH PDQJDEH\ &KDSWHU 0DUVK f 7KH PDQJDEH\ SRSXODWLRQ

PAGE 29

)LJXUH 0DS RI .HQ\D DQG ORFDWLRQ RI WKH 1DWLRQDO 3ULPDWH 5HVHUYH 75135 IURP +XJKHV f 7DQD 5LYHU PDS PRGLILHG

PAGE 31

GHFOLQHG WKURXJK D ORVV LQ WKH WRWDO QXPEHU RI JURXSV EXW WKH UHPDLQLQJ JURXSV VKRZHG QR VLJQLILFDQW FKDQJHV LQ VL]H 'HFNHU DQG .LQQDLUG f .LQQDLUG HW DO f VSHFXODWHG WKDW WKH H[WHQVLYH ORVV RI IRUHVW LQ WKH ODWH V OHG WR D FRPSUHVVLRQ RI WKH SULPDWH SRSXODWLRQV DQG XQXVXDOO\ KLJK SRSXODWLRQ GHQVLWLHV ,Q WKH VXEVHTXHQW \HDUV WKHVH SRSXODWLRQV ZHUH EHOLHYHG WR KDYH GHFOLQHG DQG VWDELOL]HG DW OHYHOV LQ HTXLOLEULXP ZLWK WKH QHZ FDUU\LQJ FDSDFLW\ RI WKH UHGXFHG IRUHVW DUHD /RVV RI IRUHVW DUHD OLNHO\ ZDV H[DFHUEDWHG E\ UHGXFWLRQ LQ KDELWDW TXDOLW\ GXH WR IRUHVW VHQHVFHQFH +XJKHV 0DUVK f DQG SRVVLEO\ E\ VHYHUH GURXJKW FRQGLWLRQV LQ WKH HDUO\ V 'HFNHU DQG .LQQDLUG 0DUVK f 7KH SUHVHQW VWXG\ H[DPLQHV WKH HIIHFWV RI KDELWDW FKDQJH RQ PDQJDEH\ EHKDYLRU DQG GHPRJUDSK\ 7KH XQGHUO\LQJ WKHPH RI WKH GLVVHUWDWLRQ LV FRPSDUDWLYH XVLQJ EDVHOLQH GDWD RQ EHKDYLRU DQG GHPRJUDSK\ FROOHFWHG GXULQJ D VWXG\ E\ +RPHZRRG f WKH RQO\ RWKHU VWXG\ WR GDWH RI WKH 7DQD PDQJDEH\ &RPSDULVRQ RI P\ GDWD ZLWK WKDW RI +RPHZRRGnV f DOORZHG PH WR DGGUHVV TXHVWLRQV FRQFHUQLQJ WKH ORVV RI PDQJDEH\ JURXSV DQG HIIHFWV RI KDELWDW FKDQJH HJ ORVV RI IRUHVW DUHD DQG UHGXFWLRQV LQ LPSRUWDQW PDQJDEH\ GLHW VSHFLHVf RQ PDQJDEH\ DFWLYLW\ EXGJHWV GLHW UDQJLQJ EHKDYLRU DQG ULVN RI H[WLQFWLRQ 'HWDLOHG DQDO\VLV RI HFRORJLFDO DQG EHKDYLRUDO GDWD FROOHFWHG GXULQJ DOVR

PAGE 32

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f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nV ULVN RI H[WLQFWLRQ E\ FRPSDULQJ GHULYHG SRSXODWLRQ VL]HV ZLWK SUHVHQW SRSXODWLRQ HVWLPDWHV VXPPDUL]H LQ &KDSWHU WKH PDMRU UHVXOWV RI WKH ERG\ RI

PAGE 33

WKH GLVVHUWDWLRQ DQG EURDGO\ GLVFXVV PDQDJHPHQW UHFRPPHQGDWLRQV

PAGE 34

&+$37(5 7:2 $%81'$1&( $1' ',675,%87,21 2) )58,7 5(6285&(6 ,1 7+5(( 7$1$ 5,9(5,1( )25(676 ,QWURGXFWLRQ 7HPSRUDO SDWWHUQV RI UHVRXUFH DEXQGDQFH DQG VFDUFLW\ FRXSOHG ZLWK WKH VSDWLDO GLVWULEXWLRQ RI IUXLW WUHHV DUH FRQVLGHUHG FUXFLDO WR PROGLQJ WKH UDQJLQJ DQG IRUDJLQJ EHKDYLRUV RI IUXJLYRURXV ELUGV DQG SULPDWHV &KLYHUV DQG 5DHPDHNHUV /HLJKWRQ DQG /HLJKWRQ 7HUERUJK /HYH\ f ,Q PRVW WURSLFDO WUHHV IORZHULQJ DQG IUXLWLQJ LV HSLVRGLF DQG VHDVRQDO SHDNV LQ IUXLW DEXQGDQFH KDYH EHHQ UHFRUGHG IRU PDQ\ IRUHVWV RI WKH 1HRWURSLFV )RVWHU 7HUERUJK /HY\ f DQG 2OG :RUOG WURSLFV .RHOPH\HU 0HGZD\ 6WUXKVDNHU /HLJKWRQ DQG /HLJKWRQ *DXWLHU+LRQ HW DO YDQ 6FKDLN f 7KH WLPLQJ RI IORZHULQJ DQG IUXLWLQJ LQ WURSLFDO WUHHV KDV EHHQ DVFULEHG WR FOLPDWLF HGDSKLF DQG ELRWLF IDFWRUV 5DWKFKH DQG /DFH\ f ,Q PRVW WURSLFDO IRUHVWV YDULDWLRQ LQ UDLQIDOO DSSHDUV WR EH WKH PRVW VLJQLILFDQW FOLPDWLF IDFWRU LQIOXHQFLQJ WKH SKHQRORJLHV RI IORZHULQJ DQG IUXLWLQJ )RVWHU +LOW\ %RUFKHUW f DOWKRXJK WKH UHODWLRQVKLS EHWZHHQ UDLQIDOO DQG SODQW UHSURGXFWLRQ LV

PAGE 35

YDULDEOH LQ OHVV VHDVRQDO IRUHVWV +LOW\ +HLGHPDQ f 7KH SKHQRORJLFDO SDWWHUQV RI VHDVRQDO JURXQGZDWHU GHSHQGHQW IRUHVWV DUH QRW ZHOO GRFXPHQWHG DQG PD\ EH LQIOXHQFHG PRUH VWURQJO\ E\ YDULDWLRQ LQ ULYHU ZDWHU OHYHO WKDQ RWKHU FOLPDWLF IDFWRUV 7KH 7DQD 5LYHU IRUHVWV RFFXU LQ D VHPLDULG HQYLURQPHQW ZLWK ORZ DQQXDO UDLQIDOO DERXW PP\HDUf DQG DUH GHSHQGHQW RQ JURXQGZDWHU IURP WKH ULYHU IRU WKHLU H[LVWHQFH +RPHZRRG 0DUVK +XJKHV f 7KH RFFXUUHQFH RI JURXQGZDWHU IRUHVW FRPPXQLWLHV DORQJ WKH 7DQD DSSHDUV WR EH UHODWHG WR WKH IRUHVW KHLJKW DERYH IORRG OHYHOV DQG GXUDWLRQ RI IORRGLQJ VRLO WH[WXUH DQG WKHUHIRUH DELOLW\ WR KROG ZDWHU VRLO FKHPLVWU\ DJH RI VXEVWUDWH DQG WKH KLVWRU\ RI KXPDQ DFWLYLWLHV ZLWKLQ WKH IRUHVW +XJKHV f /LWWOH LQIRUPDWLRQ H[LVWV RQ WKH LQIOXHQFH RI IRUHVW IUDJPHQWDWLRQ LVRODWLRQ RU KDELWDW FKDQJH RQ WUHH SKHQRORJLHV ,QFUHDVHG OHDI IDOO KDV EHHQ QRWHG LQ %UD]LOLDQ IRUHVWV DIWHU IUDJPHQWDWLRQ /RYHMR\ HW DO f EXW HYHQ TXDOLWDWLYH FKDQJHV LQ SODQW UHSURGXFWLRQ KDYH QRW EHHQ LQYHVWLJDWHG 9DU\LQJ GHJUHHV RI KDELWDW GHJUDGDWLRQ IUDJPHQWDWLRQ DQG LVRODWLRQ ZLWKLQ WKH 7DQD 5LYHU IRUHVWV RYHU WKH ODVW \HDUV SURYLGH D EDVLV IRU H[DPLQLQJ WKH HIIHFWV RI KDELWDW FKDQJH RQ SODQW UHSURGXFWLRQ ,Q WKLV FKDSWHU GHVFULEH WKH SKHQRORJLFDO DQG VSDWLDO SDWWHUQV RI IUXLW VSHFLHV HDWHQ E\ 7DQD PDQJDEH\V LQ IRUHVWV GHULYH LQGLFHV RI PRQWKO\ IUXLW SURGXFWLRQ LQ

PAGE 36

HDFK IRUHVW DQG H[DPLQH WKH DVVRFLDWLRQV EHWZHHQ IORZHULQJ DQG IUXLWLQJ SKHQRORJ\ ULYHU IORZ WHPSHUDWXUH DQG UDLQIDOO DOVR PDNH FRPSDULVRQV ZLWK VLPLODU GDWD FROOHFWHG GXULQJ +RPHZRRG f WR GHWHUPLQH LI IUXLWLQJ SDWWHUQV GLIIHU EHWZHHQ WKH WLPH SHULRGV 0HWKRGV 6WXG\ 6LWHV 7KUHH IRUHVWV ZHUH FKRVHQ IRU LQWHQVLYH VWXG\ 0FKHOHOR 1NDQR DQG 0QD]LQL )LJXUH f %DVHOLQH GDWD ZHUH DYDLODEOH IRU WKHVH IRUHVWV IURP +RPHZRRGnV f VWXG\ 7KH IRUHVWV YDU\ LQ VL]H YHJHWDWLRQ GHJUHH RI IORRGLQJ KXPDQ GLVWXUEDQFH DQG PDQJDEH\ GHQVLW\ 0FKHOHOR LV D VPDOO IRUHVW KDf ZLWK D GLYHUVH SODQW VSHFLHV FRPSRVLWLRQ ,W LV RQ D UDLVHG ULYHU OHYHH DQG H[SHULHQFHV PLQLPDO IORRGLQJ 0FKHOHOR SUHVHQWO\ LV RFFXSLHG E\ JURXSV RI PDQJDEH\V GHSHQGLQJ RQ WKH VHDVRQ DQG LV QRW GLVWXUEHG E\ ORFDO SHRSOH 3ODQW VSHFLHV FRPSRVLWLRQ KDV FKDQJHG VLQFH WKH PLGV 0DUVK 0HGOH\ f DQG VHYHUDO LPSRUWDQW PDQJDEH\ IRRG VSHFLHV KDYH GHFOLQHG LQ DEXQGDQFH 1NDQR LV D ODUJHU IRUHVW KDf ZLWK D PXFK OHVV GLYHUVH SODQW VSHFLHV FRPSRVLWLRQ WKDQ 0FKHOHOR +RPHZRRG 0HGOH\ f 1NDQR SUHVHQWO\ LV RFFXSLHG E\ PDQJDEH\ JURXSV DQG H[SHULHQFHV D KLJK GHJUHH RI IORRGLQJ DQG KXPDQ H[SORLWDWLRQ $OWKRXJK IRUHVW DUHD LQ 1NDQR KDV GHFUHDVHG E\ DQ HVWLPDWHG b GXH WR D ILUH LQ 0DUVK

PAGE 37

)LJXUH 0DS RI WKH VWXG\ DUHD VKRZLQJ WKH 0FKHOHOR 0QD]LQL DQG 1NDQR IRUHVWV ZLWKLQ WKH 7DQD 5LYHU 1DWLRQDO 3ULPDWH 5HVHUYH

PAGE 38

N

PAGE 39

'HFNHU f SODQW VSHFLHV FRPSRVLWLRQ KDV QRW FKDQJHG VLJQLILFDQWO\ VLQFH WKH PLGV 0HGOH\ f 7KH KD 0QD]LQL IRUHVW LV OHVV WKDQ NP VRXWK RI WKH 1NDQR IRUHVW EHLQJ VHSDUDWHG IURP LW E\ DQ ROG ULYHU FRXUVH DQG DQ DUHD RI JUDVVODQG 0QD]LQL LV D ORZ GLYHUVLW\ IRUHVW +RPHZRRG 0HGOH\ f WKDW H[SHULHQFHV KLJK OHYHOV RI IORRGLQJ DQG KXPDQ H[SORLWDWLRQ 7KH IRUHVW KDV GHFUHDVHG LQ VL]H VLQFH WKH PLGV E\ DSSUR[LPDWHO\ b GXH WR IORRGV DQG ILUHV LQ DQG UHVSHFWLYHO\ 7KH QXPEHU RI PDQJDEH\ JURXSV LQ 0QD]LQL KDV GHFOLQHG IURP WR VLQFH WKH PLGV 7KH FOLPDWH LV VWURQJO\ VHDVRQDO 5DLQIDOO DYHUDJHV PP SHU \HDU PRVW RI ZKLFK IDOOV GXULQJ WKH ORQJ UDLQV 0DUFK0D\f DQG WKH VKRUW UDLQV 1RYHPEHU-DQXDU\f ZLWK LQWHUYHQLQJ GU\ SHULRGV JHQHUDOO\ RFFXUULQJ GXULQJ )HEUXDU\ LQWHUUDLQVf DQG -XQH2FWREHU GU\ VHDVRQf +XJKHV f )ORRGV DUH KLJKO\ YDULDEOH LQ LQWHQVLW\ DQG WLPLQJ EXW LQ JHQHUDO ZDWHU OHYHOV SHDN GXULQJ WKH ORQJ DQG VKRUW UDLQV +XJKHV f $EXQGDQFH DQG 'LVWULEXWLRQ RI 5HVRXUFHV 'HQVLW\ HQXPHUDWHG VHOHFWHG WUHH VSHFLHV LQ RI WKH VWXG\ VLWHV ZKHUH PDQJDEH\ JURXSV ZHUH LQWHQVLYHO\ VWXGLHG 0FKHOHOR DQG 1NDQRf WR FDOFXODWH WKH DEXQGDQFH DQG GHVFULEH WKH GLVSHUVLRQ SDWWHUQV RI PDQJDEH\ IRRGV 6SHFLHV ZHUH VHOHFWHG EDVHG RQ WKHLU LPSRUWDQFH LQ WKH GLHW GXULQJ +RPHZRRG f RU WKH SUHVHQW VWXG\ 2QO\ WKRVH WUHHV RI

PAGE 40

UHSURGXFWLYH VL]H IRU HDFK VSHFLHV ZHUH HQXPHUDWHG )RRW WUDLOV FXW DORQJ FRPSDVV EHDULQJV DW P LQWHUYDOV ZHUH XVHG WR GHOLQHDWH TXDGUDWV RI KD HQXPHUDWHG SODQW VSHFLHV ZLWKLQ TXDGUDWV KDf IRU 0FKHOHOR DQG VSHFLHV ZLWKLQ TXDGUDWV KDf IRU 1NDQR 1LQH VSHFLHV SUHVHQW LQ 0FKHOHOR ZHUH QRW IRXQG LQ 1NDQR $YHUDJH GHQVLW\ ZDV FDOFXODWHG E\ VXPPLQJ WKH QXPEHU RI LQGLYLGXDOV IRU D SDUWLFXODU VSHFLHV LQ DOO TXDGUDWV DQG GLYLGLQJ E\ WKH WRWDO QXPEHU RI KHFWDUHV VDPSOHG 6SDWLDO 'LVWULEXWLRQ XVHG 0RULVLWDnV f LQGH[ RI VSDWLDO SDWWHUQ ,Jf WR PHDVXUH WKH GLVWULEXWLRQ RI HDFK VSHFLHV LQ WKH WZR IRUHVWV FKRVH 0RULVLWDnV LQGH[ RYHU WKH PRUH FRPPRQO\ XVHG YDULDQFH WR PHDQ UDWLR EHFDXVH PDQ\ RI WKH VSHFLHV FRQVLGHUHG KDG FDOFXODWHG PHDQV OHVV WKDQ LQGLYLGXDOKD *UHLJ6PLWK f KDV VKRZQ WKH YDULDQFH WR PHDQ UDWLR WR EHKDYH HUUDWLFDOO\ ZKHQ WKH PHDQ LV OHVV WKDQ 0RULVLWDnV LQGH[ SURYLGHV D PHDVXUH RI WKH H[WHQW WR ZKLFK LQGLYLGXDOV RI D SDUWLFXODU VSHFLHV DUH QRQUDQGRPO\ GLVWULEXWHG DPRQJ LGHQWLFDOO\VL]HG TXDGUDWV 7KH LQGH[ LV FDOFXODWHG DV ,i >AQLQL OfQ Qf @ 1 ZKHUH QA LV WKH QXPEHU RI LQGLYLGXDOV LQ WKH LWOB TXDGUDW Q LV WKH WRWDO QXPEHU RI LQGLYLGXDOV LQ DOO TXDGUDWV DQG 1 LV WKH QXPEHU RI TXDGUDWV 0RULVLWDnV ,i UDQJHV IURP 1 WR 1 YDOXHV DSSURDFKLQJ LQGLFDWH D UDQGRP GLVWULEXWLRQ DQG WKRVH OHVV WKDQ RU JUHDWHU WKDQ LQGLFDWH UHJXODU DQG

PAGE 41

SDWFK\ GLVWULEXWLRQV UHVSHFWLYHO\ 'HYLDWLRQ RI WKH LQGH[ IURP XQLW\ LV WHVWHG ZLWK WKH ) VWDWLVWLF $OWKRXJK 0RULVLWDnV ,QGH[ LV UHODWLYHO\ LQGHSHQGHQW RI TXDGUDW VL]H 3RROH f VXJJHVWV WKDW TXDGUDWV EH ODUJH HQRXJK WR FRQWDLQ DW OHDVW RQH LQGLYLGXDO RI RQH RI WKH PRUH FRPPRQ VSHFLHV $ TXDGUDW VL]H RI KD PHHWV WKLV UHTXLUHPHQW WKH ILUVW DQG WKLUG PRVW DEXQGDQW VSHFLHV LQ 0FKHOHOR IRUHVW 3KRHQL[ UHFOLQDWD DQG +\SKDHQH FRPSUHVVD RFFXU DW OHDVW RQFH LQ RYHU b DQG b RI WKH TXDGUDWV UHVSHFWLYHO\ DW OHDVW RQH LQGLYLGXDO RI RQH RI WKH PRVW DEXQGDQW VSHFLHV RFFXUV LQ RYHU b RI WKH TXDGUDWV 7KH PRVW FRPPRQ VSHFLHV LQ 1NDQR IRUHVW $ODQTLXP VDOYLLIROLXP 3KRHQL[ UHHOLQDWD DQG 3DFK\VWHOD PVROR DOO RFFXU LQ RYHU b RI WKH TXDGUDWV DQG DOO TXDGUDWV FRQWDLQ DW OHDVW LQGLYLGXDO RI RI WKHVH VSHFLHV 7HPSRUDO 'LVWULEXWLRQ 7HPSRUDO IOXFWXDWLRQV LQ WKH DYDLODELOLW\ RI IORZHUV DQG IUXLWV ZHUH WDEXODWHG IRU LQGLYLGXDOV LQ VSHFLHV ZLWKLQ WKH IRUHVW DUHDV LQGLYLGXDOV LQ VSHFLHV IRU 0FKHOHOR IRUHVW LQGLYLGXDOV LQ VSHFLHV IRU 1NDQR IRUHVW DQG LQGLYLGXDOV LQ VSHFLHV IRU 0QD]LQL 6RXWK IRUHVW 7DEOH f FKRVH WKH VSHFLHV EDVHG RQ WKHLU SUHVHQFH LQ HDFK VWXG\ VLWH DQG WKHLU LPSRUWDQFH DV PDQJDEH\ IRRGV +RPHZRRG f 7KH PRUH FRPPRQ VSHFLHV ZHUH UHSUHVHQWHG E\ DW OHDVW LQGLYLGXDOV VSHFLHV ZLWK IHZHU WKDQ LQGLYLGXDOV LQ WKH VWXG\ DUHDV ZHUH UHSUHVHQWHG E\ WKH WRWDO QXPEHU RI

PAGE 42

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

PAGE 43

LQGLYLGXDOV SUHVHQW FKRVH RQO\ WKRVH LQGLYLGXDOV EHOLHYHG WR EH RI D VSHFLHVn UHSURGXFWLYH VL]H DQG LQGLYLGXDOV ZKRVH FDQRSLHV ZHUH UHODWLYHO\ HDV\ WR YLHZ 7KH SKHQRORJLFDO VWDWH RI HDFK LQGLYLGXDO ZDV DVVHVVHG PRQWKO\ IRU 0FKHOHOR Q PRQWKVf 1NDQR Q PRQWKVf DQG 0QD]LQL Q PRQWKVf EHJLQQLQJ 2FWREHU XVHG D VFDOH GHYHORSHG DQG XVHG E\ +RPHZRRG f LQ WR YLVXDOO\ DVVHVV WKH SUHVHQFH RI IORZHUV DQG ULSH DQG XQULSH IUXLWV RQ HDFK LQGLYLGXDO 7KH VFDOH LV H[SUHVVHG DV D SURSRUWLRQ RI WKH WRWDO FDQRS\ FRYHUHG E\ IORZHUV RU IUXLWV ZKHUH QR IORZHUV RU IUXLWV b FRYHUDJH b FRYHUDJH b FRYHUDJH b FRYHUDJH DQG b FRYHUDJH )ORZHUV DQG IUXLWV ZHUH H[DPLQHG VHSDUDWHO\ HDFK UHFHLYLQJ D SRWHQWLDO PD[LPXP VFRUH RI 7ZR LQGLFHV RI IORZHU DQG IUXLW DYDLODELOLW\ ZHUH FDOFXODWHG IRU HDFK VWXG\ IRUHVW XVLQJ WKH IUXLW VFDOH 7KH ILUVW LQGH[ LV FRPSDUDEOH WR WKDW XVHG E\ +RPHZRRG f DQG SURYLGHV D UHODWLYH PHDVXUH RI YDULDWLRQ LQ IORZHULQJ DQG IUXLWLQJ 7KLV LQGH[ LV FDOFXODWHG E\ SRROLQJ PRQWKO\ IORZHU RU IUXLW VFRUHV IRU LQGLYLGXDOV RI HDFK VSHFLHV DQG DYHUDJLQJ WR JLYH D PHDQ VSHFLHV VFRUH IRU HDFK PRQWK $Q LQGH[ RI WRWDO PRQWKO\ IORZHULQJ RU IUXLWLQJ LV FDOFXODWHG E\ VXPPLQJ PRQWKO\ PHDQV IRU DOO VSHFLHV DQG GLYLGLQJ E\ WKH QXPEHU RI VSHFLHV LQ WKH VDPSOH 7KH VHFRQG LQGH[ XVHG RQO\ IRU IUXLWV FRQVLGHUV VSHFLHV GLIIHUHQFHV LQ DEXQGDQFH DQG SURGXFWLYLW\ DQG

PAGE 44

SURYLGHV DQ LQGLFDWLRQ RI PRQWKO\ IUXLW ELRPDVV SURGXFWLRQ 7KH LQGH[ LQFRUSRUDWHV HVWLPDWHG QXPEHUV RI IUXLWV SURGXFHG E\ D VSHFLHV GXULQJ PD[LPXP IUXLW SURGXFWLRQ WKH HVWLPDWHG ZHLJKW RI IUXLWV RU IUXLW SDUWV HDWHQ E\ PDQJDEH\V DQG WKH DEXQGDQFH RI UHSURGXFWLYH LQGLYLGXDOV )UXLW ELRPDVV LQGLFHV ZHUH FDOFXODWHG IRU FDWHJRULHV RI VSHFLHV GHSHQGLQJ RQ KRZ IUXLWV ZHUH GLVWULEXWHG WKURXJKRXW WKH FDQRS\ 6SHFLHV EHDULQJ IUXLWV SULPDULO\ RQ WKH RXWHU VXUIDFH RI WKH FDQRS\ Q OOf ZHUH FDWHJRUL]HG DV VXUIDFH IUXLWHUV 6SHFLHV EHDULQJ IUXLWV DORQJ WKHLU EUDQFKHV HLWKHU LQ FOXVWHUV RU HYHQO\ ZHUH FODVVLILHG DV UDPLIORURXV IUXLWHUV Q f 3DOPV Q f DQG OLDQDV Q Of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

PAGE 45

KDV PXOWLSOH FURZQV WKH DERYH YDOXH ZDV PXOWLSOLHG E\ WKH QXPEHU RI FURZQV SHU LQGLYLGXDO )UXLW FRXQWV IRU DOO VSHFLHV H[FOXGLQJ SDOPV ZHUH PXOWLSOLHG E\ DQ HVWLPDWH RI UHODWLYH IRRGSURGXFLQJ DUHD WR FDOFXODWH WKH WRWDO QXPEHU RI IUXLWV SRWHQWLDOO\ SURGXFHG E\ DQ DYHUDJH LQGLYLGXDO RI D SDUWLFXODU VSHFLHV )UXLW SURGXFLQJ DUHD ZDV FDOFXODWHG XVLQJ VXUIDFH DUHD DQG YROXPH IRUPXODV IRU DQ RSHQERWWRPHG F\OLQGHU D VKDSH WKDW EHVW DSSUR[LPDWHV D YDULHW\ RI GLIIHULQJ WUHH VKDSHV $SSHQGL[ $f 6XUIDFH DUHDV ZHUH FDOFXODWHG IRU WKH FDWHJRU\ RI VSHFLHV EHDULQJ WKHLU IUXLWV SULPDULO\ RQ WKH VXUIDFH RI WKH FDQRS\ DQG YROXPHV ZHUH FDOFXODWHG IRU UDPLIORURXV VSHFLHV %UDQFKHV RI UDPLIORURXV VSHFLHV ZHUH VXEMHFWLYHO\ HVWLPDWHG WR RFFXS\ OLWWOH RYHU RI WKH HQWLUH YROXPH RI WKH FDQRS\ YROXPH HVWLPDWHV WKHUHIRUH ZHUH GLYLGHG E\ 7KH IRRGSURGXFLQJ DUHD RI WKH OLDQD VSHFLHV ZKLFK JURZV RYHU WKH WRS RI WUHH FDQRSLHV ZDV FDOFXODWHG E\ HVWLPDWLQJ WKH DPRXQW RI VXUIDFH DUHD RI WKH WUHH FDQRS\ FRYHUHG E\ WKH OLDQD GLG QRW FRUUHFW IRU FDQRS\ VL]H LQ SDOPV EHFDXVH D PRUH GLUHFW FRXQW RI IUXLW QXPEHUV ZDV SRVVLEOH 6DPSOHV RI RU PRUH IUXLWV ZHUH FROOHFWHG DQG ZHLJKHG IURP D PD[LPXP RI LQGLYLGXDOV SHU VSHFLHV 2QO\ WKRVH IUXLW SDUWV FRQVXPHG E\ PDQJDEH\V ZHUH ZHLJKHG DQG SDUWV ZHUH ZHLJKHG RQO\ GXULQJ WKH VWDJH RI ULSHQHVV ZKHQ PDQJDEH\V ZHUH IHHGLQJ RQ WKHP :HLJKWV ZHUH DYHUDJHG IRU LQGLYLGXDOV DQG LQGLYLGXDO DYHUDJHV ZHUH SRROHG WR FDOFXODWH

PAGE 46

D VSHFLHV PHDQ 7KH GHQVLW\ RI SRWHQWLDOO\ UHSURGXFWLYH LQGLYLGXDOV LQ HDFK VSHFLHV IRU 0FKHOHOR DQG 1NDQR VWXG\ VLWHV ZHUH WDNHQ IURP WKH VSHFLHV HQXPHUDWLRQV GHVFULEHG DERYH 6SHFLHV GHQVLWLHV IRU 0QD]LQL ZHUH WDNHQ IURP 0HGOH\ f $ ILQDO PRQWKO\ ELRPDVV NJKDf LQGH[ RI IUXLWV RI VSHFLHV VDPSOHG ZDV FDOFXODWHG DV )% >en^;VLr'VL6$VLr)3VLf,VLf@ >e^;ULr'ULr9ULr)3ULf,ULf@ >‘;SLr'SLr)3SLf,SO`@ ^e ;r'OLr6$OLr)3OLf,OLQ ZKHUH ; PRQWKO\ IUXLW VFRUH IRU VSHFLHV L WR VFDOH WKH VFRUHV WR PD[LPXP SURGXFWLRQf QXPEHU RI LQGLYLGXDOV VFRUHG IRU VSHFLHV L GHQVLW\ RI VSHFLHV L SHU KHFWDUH 6$ VXUIDFH DUHD RI VSHFLHV L 9 EUDQFK YROXPH RI VSHFLHV L )3 WRWDO IUXLW SURGXFWLRQ PHDVXUHG DV PHDQ IUXLW FRXQW r PHDQ IUXLW ZHLJKW RI VSHFLHV L DQG VXEVFULSWV V U S DQG GHQRWH VXUIDFH UDPLIORURXV SDOP DQG OLDQD FDWHJRULHV UHVSHFWLYHO\ (QYLURQPHQWDO &RUUHODWHV ZLWK )ORZHULQJ DQG )UXLWLQJ H[DPLQHG DVVRFLDWLRQV EHWZHHQ HQYLURQPHQWDO SDUDPHWHUV DQG WKH WHPSRUDO SDWWHUQV RI IORZHULQJ DQG IUXLWLQJ LQ WKH VWXG\ VLWHV $ 6SHDUPDQnV UDQN FRUUHODWLRQ DQDO\VLV 6LHJHO f ZDV FRQGXFWHG WR H[DPLQH SRVVLEOH

PAGE 47

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f 0RQWKO\ UDLQIDOO OHYHOV ZHUH UHFRUGHG IRU 0FKHOHOR DQG 1NDQR0QD]LQL DUHDV DQG 7DQD 5LYHU GLVFKDUJH OHYHOV IRU WKH +ROD 6WDWLRQ ZHUH REWDLQHG IURP WKH .HQ\D :DWHU 'HSDUWPHQW +\GURORJ\ 6HFWLRQ &RPSDULVRQV ZLWK FRPSDUHG PRQWKO\ IUXLW VFRUHV IRU VSHFLHV H[DPLQHG GXULQJ WKLV VWXG\ DQG E\ +RPHZRRG LQ WR GHWHUPLQH LI IUXLWLQJ SDWWHUQV YDULHG EHWZHHQ WKH VWXG\ SHULRGV 6HYHQ VSHFLHV IURP 0QD]LQL DQG VSHFLHV IURP 0FKHOHOR IRUHVW ZHUH FRPSDUHG VXEVDPSOHG IURP P\ GDWD WR FRUUHVSRQG

PAGE 48

ZLWK +RPHZRRGnV PRQWK GDWD IURP 0QD]LQL IRUHVW DQG PRQWK GDWD IURP 0FKHOHOR IRUHVW .ROPRJURY6PLUQRY WZR VDPSOH WHVWV +ROODQGHU DQG :ROIH f ZHUH SHUIRUPHG E\ VSHFLHV WR WHVW IRU GLIIHUHQFHV LQ WKH GLVWULEXWLRQV RI IUXLWLQJ EHWZHHQ WKH VWXG\ SHULRGV :LOFR[RQ VLJQHGUDQNV WHVWV DOVR ZHUH SHUIRUPHG E\ VSHFLHV WR WHVW ZKHWKHU IUXLW VFRUHV ZHUH FRQVLVWHQWO\ KLJKHU GXULQJ VWXG\ SHULRG WKDQ WKH RWKHU 5HVXOWV $EXQGDQFH DQG 6SDWLDO 'LVWULEXWLRQ RI 5HVRXUFHV 'HQVLW\ YDOXHV UDQJHG ZLGHO\ IRU HQXPHUDWHG VSHFLHV ERWK ZLWKLQ DQG EHWZHHQ IRUHVWV 7DEOH f 2YHU b RI WKH VSHFLHV HQXPHUDWHG LQ 0FKHOHOR Q RI f DQG 1NDQR Q RI f ZHUH UDUH RFFXUULQJ DW GHQVLWLHV RI OHVV WKDQ LQGLYLGXDO SHU KD 7KH SDOP 3KRHQL[ UHFOLQDWD ZDV RQH RI WKH PRVW FRPPRQ VSHFLHV RFFXUULQJ DW WKH KLJKHVW DQG VHFRQG KLJKHVW GHQVLWLHV DQG LQGLYLGXDOVKDf RI DOO VSHFLHV LQ 0FKHOHOR DQG 1NDQR UHVSHFWLYHO\ $ODQJLXP VDOYLLIROLXP DQ XQGHUVWRU\ VSHFLHV DOVR RFFXUUHG DW YHU\ KLJK GHQVLWLHV LQGLYLGXDOVKDf LQ ERWK IRUHVWV )LYH VSHFLHV SUHVHQW LQ ERWK IRUHVWV 3DFKYVWHOD PVROR )LFXV VYFRPRUXV $ODQJLXP VDOYLLIROLXP $SRUUKL]D SDQLFXODWD DQG 6DED FRPRUHQVLVf RFFXUUHG DW PXFK KLJKHU GHQVLWLHV LQ 1NDQR WKDQ 0FKHOHOR DGGLWLRQDO VSHFLHV 6RULQGHLD

PAGE 49

7DEOH 'HQVLW\ RI UHSURGXFWLYH WUHHV DQG OLDQDV HQXPHUDWHG LQ KD RI 0FKHOHOR IRUHVW DQG KD LQ 1NDQR IRUHVW DQG WKH GHJUHH RI FOXPSLQJ RI HDFK VSHFLHV XVLQJ 0RULVLWDnV LQGH[ ,if ,I ,i LV VLJQLILFDQWO\ JUHDWHU RU OHVV WKDQ XQLW\ XVLQJ WKH )WHVW WKH VSHFLHV LV FRQVLGHUHG FOXPSHG RU HYHQO\ GLVWULEXWHG LQ VSDFH UHVSHFWLYHO\ ,I ,i LV QRW VLJQLILFDQWO\ GLIIHUHQW IURP XQLW\ WKDQ WKH VSHFLHV LV FRQVLGHUHG UDQGRPO\ GLVWULEXWHG 0FKHOHOR 1NDQR 7D[RQ &RGH 'HQVLW\ ni )WHVW Sf 'HQVLW\ 7HVW Sf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

PAGE 50

PDGDJDVFDULHQVLV 'LRVS\URV LQHVSLOLIRUPLV 2QFRED VSLQRVD DQG *DUFLQLD OLYLQJVWRQHLf RFFXUUHG DW PXFK KLJKHU GHQVLWLHV LQ 0FKHOHOR WKDQ 1NDQR 7DEOH f 7KH PDMRULW\ RI VSHFLHV LQ ERWK IRUHVWV ZHUH SDWFKLO\ RU UDQGRPO\ GLVWULEXWHG LQ VSDFH DQG QR VSHFLHV VKRZHG D XQLIRUP VSDWLDO SDWWHUQ 7DEOH )LJXUHV f 7KH WHQGHQF\ WRZDUGV SDWFK\ GLVWULEXWLRQV RI VSHFLHV ZDV FRQVLVWHQW RQ D ODUJHU VSDWLDO VFDOH TXDGUDWV ZLWK KLJK QXPEHUV RI LQGLYLGXDOV RI D VSHFLHV WHQGHG WR EH FOXPSHG WKURXJKRXW WKH IRUHVWV )LJXUHV DQG f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f ,QFUHDVHG IORZHULQJ RFFXUUHG IURP $XJXVW WR 'HFHPEHU ZLWK PLQRU LQFUHDVHV LQ )HEUXDU\ DQG 0DUFK )UXLWLQJ ZDV KLJK IURP -DQXDU\ WR 0D\ RI DQG 1RYHPEHU WR 0DUFK RI ORZ OHYHOV RI IUXLWLQJ

PAGE 51

)LJXUH 1XPEHU RI LQGLYLGXDOV KD TXDGUDW RI FRPPRQ VSHFLHV ZLWKLQ 0FKHOHOR IRUHVW VKRZLQJ FOXPSHG VSDWLDO GLVWULEXWLRQV Df +YRKDHQH FRPSUHVVD Ef $ODQRLXP VDOYLLIROLXP Ff 3KRHQL[ UHFOLQDWD Gf 2QFRED VSPRVD

PAGE 53

)LJXUH 1XPEHU RI LQGLYLGXDOV KD TXDGUDW RI FRPPRQ VSHFLHV ZLWKLQ 1NDQR IRUHVW VKRZLQJ FOXPSHG VSDWLDO GLVWULEXWLRQV Df )LFXV VYFRPRUXV Ef $ODQJLXP VDOYLLIROLXP Ff 3KRHQL[ UHFOLQDWD Gf 3DFKYVWHOD PVROR

PAGE 55

)LJXUH ([DPSOHV RI UDQGRPO\ GLVWULEXWHG )LFXV VSS Df )LFXV QDWDOHQVLV LQ 0FKHOHOR IRUHVW Ef )LFXV VYFRPRUXV LQ 0FKHOHOR IRUHVW" Ff )LFXV QDWDOHQVLV LQ 1NDQR IRUHVW

PAGE 56

)LFXV QDWDOHQVLV )LFXV QDWDOHQVLV

PAGE 57

)LJXUH 0HDQ IORZHU VFRUHV DYHUDJHG RYHU DOO PRQWK Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW IRUHVW VSHFLHV E\ Ff 0QD]LQL

PAGE 58

0FKHOHOR )ORZHU VFRUH 1NDQR )ORZHU VFRUH 0QD]LQL )ORZHU VFRUH 0RQWK

PAGE 59

)LJXUH 0HDQ IUXLWLQJ VFRUHV DYHUDJHG RYHU DOO VSHFLHV E\ PRQWK Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW

PAGE 60

0FKHOHOR )UXLW VFRUH )UXLW VFRUH 1NDQR f, U $ 0 0QD]LQL )UXLW VFRUH ‘OOOOOOOOOD 1'-)0$0--$621 , U ) 0 $ 0 0RQWK

PAGE 61

RFFXUUHG IURP -XQH WR 2FWREHU LQ 0FKHOHOR DQG 1NDQR DQG IURP -XO\ WKURXJK 1RYHPEHU LQ 0QD]LQL 6HYHUDO VSHFLHV VKRZHG VHDVRQDO SDWWHUQV RI IUXLWLQJ )LJXUHV f 7KH PDMRULW\ RI VHDVRQDO IUXLWHUV ERUH IUXLWV RQO\ RQFH LQ D \HDU KRZHYHU LQGLYLGXDOV RI VSHFLHV 6RULQGHLD PDGDJDVFDULHQVLV DQG *DUFLQLD OLYLQJVWRQHLf SURGXFHG IUXLWV WZLFH ZLWKLQ D \HDU )LJXUH f 2I WKH VHDVRQDO IUXLWHUV RQO\ 3KRHQL[ UHFOLQDWD DQG 'LRVS\URV PHVSLOLIRUPHV SURGXFHG IUXLW GXULQJ WKH FRRO GU\ PRQWKV EHWZHHQ -XQH DQG 6HSWHPEHU )LJXUHV DQG f 7KUHH VSHFLHV DSSHDUHG WR IUXLW FRQWLQXRXVO\ )LJXUHV DQG f GXH WR f FRQWLQXRXV SURGXFWLRQ RI IUXLW E\ DOO LQGLYLGXDOV 2QFRED VSLQRVD DQG )LFXV VYFRPRUXVf f UHWHQWLRQ RI IUXLWV RYHU ORQJ SHULRGV RI WLPH +\SKDHQH FRPSUHVVDf DQG f XQV\QFKURQL]HG ERXWV RI IUXLWLQJ )LFXV QDWDOHQVLVf 6SHFLHV VDPSOHG LQ PRUH WKDQ RI WKH IRUHVWV WHQGHG WR VKRZ VLPLODU SDWWHUQV RI IUXLWLQJ DPRQJ IRUHVWV 2QH H[FHSWLRQ ZDV WKH OLDQD 6DED FRPRUHQVLV 6DED FRPRUHQVLV H[KLELWHG KLJKO\ VHDVRQDO IUXLWLQJ LQ 0QD]LQL VHDVRQDO EXW H[WHQGHG IUXLWLQJ LQ 1NDQR DQG FRQWLQXRXV IUXLWLQJ LQ 0FKHOHOR )LJXUH f (VWLPDWHV RI IUXLW ELRPDVV (VWLPDWHG PD[LPXP IUXLW QXPEHUV IUXLW ZHLJKW DQG FDQRS\ VL]H YDULHG ZLGHO\ EHWZHHQ VSHFLHV DQG VKDSH FDWHJRULHV 7DEOH f DQG VWURQJO\ LQIOXHQFHG PRQWKO\ ELRPDVV LQGLFHV RI IUXLW )LJXUH f

PAGE 62

)LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IURP 0FKHOHOR DQG 0QD]LQL IRUHVWV IRU 0LPXVRSV IUXWLFRVD DQG $FDFLD UREXVWD VHDVRQDOO\ IUXLWLQJ VSHFLHV ZLWK SHDN LQ IUXLWLQJ SHU \HDU Df 0LPXVRSV IUXLWLFRVD LQ 0FKHOHOR IRUHVW Ef $FDFLD UREXVWD LQ 0FKHOHOR IRUHVW Ff 0LPXVRSV IUXWLFRVD LQ 0QD]LQL IRUHVW Gf $FDFLD UREXVWD LQ 0QD]LQL IRUHVW

PAGE 63

0FKHOHOR )UXLW VFRUH 0QD]OQL )UXLW VFRUH 0RQWK 0FKHOHOR Q )UXLW VFRUH ,,, L Ua ‘OLOLO 1'-)0$0--$621'-)0$0 0QD]OQL )UXLW VFRUH 0RQWK

PAGE 64

)LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IURP 0FKHOHOR DQG 0QD]LQL IRUHVWV IRU *DUFLQLD OLYLQTVWRQHL DQG 6RULQGHLD PDGDTDVFDULHQVLV VHDVRQDOO\ IUXLWLQJ VSHFLHV ZLWK IUXLWLQJ SHDNV SHU \HDU Df *DUFLQLD OLYLQTVWRQHL LQ 0FKHOHOR IRUHVW Ef 6RULQGHLD PDGDTDVFDULHQVLV LQ 0FKHOHOR IRUHVW Ff *DUFLQLD OLYLQTVWRQHL LQ 0QD]LQL IRUHVW Gf 6RULQGHLD PDGDTDVFDULHQVLV LQ 0QD]LQL IRUHVW

PAGE 65

0FKHOHOR )UXLW VFRUH J 0QD]LQL )UXLW VFRUH 0RQWK 0FKHOHOR )UXLW VFRUH 1'-)0$0--$621'-)0$0 0QD]LQL )UXLW VFRUH 0RQWK A X!

PAGE 66

)LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IRU 3KRHQL[ UHFOLQDWD Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW

PAGE 67

0FKHOHOR )UXLW VFRUH 1NDQR )UXLW VFRUH H 4 0QD]LQL )UXLW VFRUH 1'-)0$0--$621'-)0$0 0RQWK H

PAGE 68

)LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IRU 'LRVS\URV PHVSLOLIRUPHV Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW

PAGE 69

0FKHOHOR )UXLW VFRUH Q 1NDQR )UXLW VFRUH 1'-)0$0--$621'-)0$0 0QD]LQL )UXLW VFRUH L 8OOOLLLX 1'-)0$0--$62 0RQWK 0

PAGE 70

)LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IRU VSHFLHV VKRZLQJ SDWWHUQV RI FRQWLQXRXV IUXLW SURGXFWLRQ Df FRQWLQXRXV IUXLWLQJ GXH WR DV\QFKURQ\ DPRQJ LQGLYLGXDOV RI WKH VSHFLHV )LFXV QDWDOHQVLVf Ef FRQWLQXRXV IUXLWLQJ GXH WR UHWHQWLRQ RI IUXLWV RYHU WLPH +\SKDHQH FRPSUHVVDf Ff FRQWLQXRXV IUXLWLQJ GXH WR UDSLG WXUQRYHU RI IUXLWV 2QFRED VSLQRVDf

PAGE 71

)LFXV QDWDOHQcV )UXLW VFRUH 1'-)0$0-$621'-)0$0 +\SKDHQH FRPSUHVVD )UXLW VFRUH 1'-)0$0--$621'-)0$0 2QFRED VSLQRVD )UXLW VFRUH L 8OOLOLLLLLLOLOLOLL 1'-)0$0--$621'-)0$0 0RQWK

PAGE 72

)LJXUH 0HDQ PRQWKO\ IUXLW VFRUHV IRU )LFXV VYFRPRUXV Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW

PAGE 73

0FKHOHOR )UXLW VFRUH 1NDQR )UXLW VFRUH )UXLW VFRUH -M,,L r 1 ) 0 $ 0 $ 6 2 1 0RQWK 0QD]LQL f, U ) 0 $ 0

PAGE 74

)LJXUH 6HDVRQDOLW\ RI PHDQ PRQWKO\ IUXLW VFRUHV IRU 6DED FRPRUHQVLV LQ WKH VWXG\ IRUHVWV Df VHDVRQDO IUXLWLQJ LQ 0QD]LQL IRUHVW Ef VHDVRQDO EXW H[WHQGHG IUXLWLQJ LQ 1NDQR IRUHVW Ff FRQWLQXRXV IUXLWLQJ LQ 0FKHOHOR IRUHVW

PAGE 75

0FKHOHOR )UXLW VFRUH 1NDQR )UXLW VFRUH 1'-)0$0-$621'-)0$0 0QD]LQL )UXLW VFRUH Q ,,, 1 ) 0 $ 0 -$621 0RQWK U ) 0 $ 0

PAGE 76

7DEOH (VWLPDWHG FDQRS\ VXUIDFH DUHD 6f FDQRS\ YROXPH 9f PHDQ IUXLW SURGXFWLRQ )1f DQG PHDQ IUXLW ZHLJKW ):f IRU VSHFLHV 6SHFLHV DUH FDWHJRUL]HG LQWR JURXSV WKRVH WKDW EHDU WKHLU IUXLWV RQ WKH FDQRS\ VXUIDFH &f RU DORQJ WKH EUDQFKHV RI WKH WUHH 5f OLDQDV /f DQG SDOPV 3f 7D[RQ *URXS 6Pf 9 Pf )1Pf ):JUf $ODQDLXP VDOYLIROLXP & 6RULQGHLD PDGDTDVFDULHQVLV 5 6DED FRPRUHQVLV / 'LRVSYURV PHVSLOLIRUPLV & 2QFRED VSLQRVD & *DUFLQLD OLYLQTVWRQHL & $FDFLD UREXVWD & $OEL]LD TXPPLIHUD F )LFXV EXEX 5 )LFXV EXVVHL & )LFXV QDWDOHQVLV & )LFXV V\FRPRUXV 5 +\SKDHQH FRPSUHVVD 3 3KRHQL[ UHFOLQDWD 3 0LPXVRSV IUXWLFRVD & 3DFKYVWHOD PVROR 5 XL Wr

PAGE 77

)LJXUH &RQWULEXWLRQ WR PRQWKO\ IUXLW ELRPDVV LQGLFHV E\ GLIIHUHQW VSHFLHV FDWHJRULHV Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW

PAGE 78

0FKHOHOR %LRPDVV ,QGH[ RI IUXLWV NJfKD 1'-)0$0--$621'-)0$0 1NDQR %LRPDVV LQGH[ RI IUXLWV NJfKH 1'-)0$0--$ 0QD]LQO %LRPDVV ,QGH[ RI IUXLWV NJfKH f 1'-)0$0--$621'-)0$0 0RQWK OLDQD  O SDOP (O66 VXUIDFH UDPOIORURX

PAGE 79

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f 0FKHOHOR KDG D PRUH FRQVLVWHQW SURGXFWLRQ RI IUXLWV DFURVV WKH PRQWKV GXH WR WKH DYDLODELOLW\ RI SDOP IUXLWV EHWZHHQ 0D\ DQG 2FWREHU D SHULRG RI UHODWLYH VFDUFLW\ LQ WKH RWKHU IRUHVWV 0QD]LQL ZLWK D ORZ DEXQGDQFH RI ILJV DQG SDOPV ZDV PRUH VHDVRQDO LQ IUXLW SURGXFWLRQ WKDQ 0FKHOHOR RU 1NDQR DQG H[SHULHQFHG D PRQWK SHULRG 0D\$XJf RI IUXLW VFDUFLW\ 8QULSH IUXLWV FRQWULEXWHG WKH PRVW WR WKH ELRPDVV LQGH[ RYHU DOO PRQWKV LQ DOO VWXG\ IRUHVWV ULSH IUXLWV ZHUH HQWLUHO\ XQDYDLODEOH GXULQJ VHYHUDO PRQWKV )LJXUH f

PAGE 80

)LJXUH &RQWULEXWLRQ RI ULSH DQG XQULSH IUXLWV WR PRQWKO\ IUXLW ELRPDVV LQGH[ Df 0FKHOHOR IRUHVW Ef 1NDQR IRUHVW Ff 0QD]LQL IRUHVW

PAGE 81

0FKHOHOR Q %LRPDVV ,QGH[ RI IUXLWV NJfKH 1NDQR 0QD]LQL 0RQWK

PAGE 82

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f 5DLQIDOO ZDV KLJK GXULQJ 1RYHPEHU DQG 'HFHPEHU DQG DJDLQ IURP 0DUFK-XQH )LJXUH f 5DLQIDOO LQ 1NDQR JHQHUDOO\ ZDV KLJKHU WKDQ 0FKHOHOR DOWKRXJK WKHUH ZDV D VWURQJ FRUUHODWLRQ EHWZHHQ PRQWKO\ WRWDOV IRU WKH DUHDV U Q OO Sf 5LYHU IORZ ZDV KLJKHVW $SULO -XQH DQG -DQXDU\$SULO )LJXUH f DQG ZDV FRUUHODWHG ZLWK WKH SUHYLRXV PRQWKV UDLQIDOO LQ ERWK DUHDV U DQG IRU 1NDQR DQG 0FKHOHOR UHVSHFWLYHO\ Q OO Sf )ORZHULQJ ZDV QHJDWLYHO\ FRUUHODWHG ZLWK UDLQIDOO LQ 1NDQR U Q Sf DQG ZLWK ULYHU IORZ LQ 0FKHOHOR U Q Sf DQG 0QD]LQL U Q Sf :KHQ DOO IRUHVWV DUH FRPELQHG WKHUH ZHUH

PAGE 83

)LJXUH 0RQWKO\ PHDQ PLQLPXP DQG PHDQ PD[LPXP DLU WHPSHUDWXUH LQ GHJUHHV FHQWLJUDGH Df 0FKHOHOR IRUHVW Ef 1NDQR0QD]LQL IRUHVW DUHDV

PAGE 84

0FKHOHOR 7HPSHUDWXUH &f , c L , , , 7 621'-)0$0-$621'-)0$ 1NDQR 7HPSHUDWXUH &f L a f§f§ 0HDQ PD[LPXP f§ 0HDQ PLQLPXP U n 7 7 f 621'-)0$0--$621'-)0$ 0RQWK

PAGE 85

)LJXUH 7RWDO PRQWKO\ UDLQIDOO LQ PLOOLPHWHUV Df 0FKHOHOR IRUHVW Ef 1NDQR0QD]LQL IRUHVW DUHD

PAGE 86

0FKHOHOR 5DLQIDOO PPf /, OO/O $621'-)0$0--$621'-)0$0 1NDQR 5DLQIDOO PPf M 21'-)0$0-$621'-)0 0RQWK

PAGE 87

)LJXUH 0HDQ PRQWKO\ ULYHU IORZ IRU WKH 7DQD 5LYHU +ROD 6WDWLRQ

PAGE 88

0LOOLRQ &XELF 0HWHUV L 0RQWK

PAGE 89

VLJQLILFDQW QHJDWLYH FRUUHODWLRQV EHWZHHQ IORZHULQJ DQG ERWK UDLQIDOO DQG ULYHU IORZ U DQG IRU UDLQIDOO DQG ULYHU IORZ UHVSHFWLYHO\ Q Sf &RUUHODWLRQV EHWZHHQ IORZHULQJ DQG IUXLWLQJ ZHUH VLJQLILFDQW ZKHQ IORZHU VFRUHV ZHUH ODJJHG E\ PRQWKV LQ 0FKHOHOR 1NDQR DQG IRU DOO IRUHVWV FRPELQHG U S IRU 0FKHOHOR U S IRU 1NDQR DQG U Q S IRU DOO IRUHVWV FRPELQHGf 7KHUH ZHUH QR FRQVLVWHQW DVVRFLDWLRQV EHWZHHQ PRQWKO\ IUXLW VFRUHV DQG HLWKHU UDLQIDOO RU ULYHU IORZ IRU WKH IRUHVWV 7DEOHV DQG f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f 7KHUH ZHUH VLJQLILFDQW SRVLWLYH FRUUHODWLRQV EHWZHHQ IUXLW VFRUHV IRU FRQWLQXRXV VSHFLHV DQG UDLQIDOO 0LQLPXP PRQWKO\ WHPSHUDWXUH VKRZHG

PAGE 90

7DEOH 6SHDUPDQnV UDQN FRUUHODWLRQ FRHIILFLHQWV IRU PHDQ PRQWKO\ HQYLURQPHQWDO YDULDEOHV DQG PHDQ PRQWKO\ IUXLWLQJ VFRUHV IRU 0FKHOHOR IRUHVW 0,17(03 PLQLPXP PRQWKO\ WHPSHUDWXUH 5$,1 WRWDO PRQWKO\ UDLQIDOO PPf 5$,1/$* 5$,1/$* $1' 5$,1/$* PRQWKO\ UDLQIDOO ODJJHG E\ DQG PRQWKV UHVSHFWLYHO\ 5$,1$'9 SUHYLRXV PRQWKVn UDLQIDOO )/2: PHDQ PRQWKO\ ULYHU IORZ PFPf )/2:/$* )/2:/$* )/2:/$* PHDQ PRQWKO\ ULYHU IORZ ODJJHG E\ DQG PRQWKV UHVSHFWLYHO\ )/2:$'9ASUHYLRXV PRQWKVn PHDQ ULYHU IORZ 5,3( $1' 815,3( 5,3( 21/< 5,3( $1' 815,3( 5,3( $1' 815,3( 7RWDO &RQWLQXRXV 6HDVRQDO 7RWDO &RQWLQXRXV 6HDVRQDO 6HDVRQDO ZR 3 UHHOLQDIDf 3 UHFOLQDWD 0,17(03 r QV QV QV QV QV QV QV 5$,1 QV QV QV QV QV QV QV QV 5$,1/$* QV QV QV QV QV QV QV QV 5$,1/$* QV QV QV QV QV QV QV QV 5$,1/$* QV QVr QV QV QV QV QV QV 5$, 1$'9 QV QV QV QV QV QV QV )/2: QV QV QV QV QV r QV QV QV )/8/$* QV QV QV QV QV QV )/:/$* QV QV QV QV QV QV QV rr )/:/$* QV QV QV QV QV )/2:$'9 QV QV QV QV QV QV QV QV r 3A S FU! RR

PAGE 91

7DEOH 6SHDUPDQnV UDQN FRUUHODWLRQ FRHIILFLHQWV IRU PHDQ PRQWKO\ HQYLURQPHQWDO YDULDEOHV DQG PHDQ PRQWKO\ IUXLWLQJ VFRUHV IRU 1NDQR IRUHVW 9DULDEOHV GHILQHG LQ 7DEOH 5,3( $1' 815,3( 5,3( 21/< 5,3( $1' 815,3( 5,3( $1' 815,3( 7RWDO &RUQ LQXRXV 6HDVRQDO 7RWDO &RQWLQXRXV 6HDVRQDO 6HDVRQDO ZR 3 UHHOLQDWDf 3 UHHOLQDWD r r r rr 0,17(03 QV QV QV 5$,1 QV QV QV QV QV QV QV QV 5$,1/$* QV QV QV QV QV QV QV 5$,1/$* QV QV QV QV QV QV QV QV 5$,1/$* QV QV QV QV QV QV QV QV 5$,1$'9 QV QV QV QV QV QV QV QV )/2: QV QV QV QV QV QV QV QV r )/:/$* QV QV QV QV QV QV QV )/:/$* QV QV QV QV QV QV QV r rr )/:/$* QV QV QV QV QV QV )/2:$'9 QV QV QV QV QV QV QV QV r fr S S &? !e!

PAGE 92

7DEOH 6SHDUPDQnV UDQN FRUUHODWLRQ FRHIILFLHQWV IRU PHDQ PRQWKO\ HQYLURQPHQWDO YDULDEOHV DQG PHDQ PRQWKO\ IUXLWLQJ VFRUHV IRU 0QD]LQL IRUHVW 9DULDEOHV GHILQHG LQ 7DEOH 5,3( $1' 815,3( 5,3( 21/< 5,3( $1' 815,3( 5,3( $1' 815,3( 7RWDO &RQWLQXRXV 6HDVRQDO 7RWDO &RQWLQXRXV 6HDVRQDO 6HDVRQDO ZR 3 UHFOLQDWDf 3 UHFOLQDWD 0,17(03 r QV r rr r r r rr 5$,1 QV QV QV QV QVr QV QV QV 5$,1/$* QV QV QV QV QV QV QV 5$,1/$* QV QV QV QV QV QV QV QV 5$,1/$* QV QV QV QV QV QV QV QV 5$,1$'9 QV QV QV QV QV r QV QV QV )/2: QV QV QV QV QV QV )/8/$* QV QV QV QV QV QV )/:/$* )/8/$* QV QV QV QV QV QV r QV QV QV QV QV rr )/2:$'9 QV QV QV QV QV QV QV QV 3 S r R

PAGE 93

7DEOH 6SHDUPDQnV UDQN FRUUHODWLRQ FRHIILFLHQWV IRU PHDQ PRQWKO\ HQYLURQPHQWDO YDULDEOHV DQG PHDQ PRQWKO\ IUXLWLQJ VFRUHV IRU DOO IRUHVWV FRPELQHG 9DULDEOHV GHILQHG LQ 7DEOH 5,3( $1' 815,3( 5,3( 21/< 5,3( $1' 815,3( 5,3( $1' 815,3( 7RWDO &RQWLQXRXV 6HDVRQDO 7RWDO &RQWLQXRXV 6HDVRQDO 6HDVRQDO XR 3 UHFOLQDWDf 3 UHFOLQDWD r 0,17(03 QV QV QV QV QV QV 5$,1 QV QV QV QV QV f QV QV QV 5$,1/$* QV QV QV QV QV QV QV 5$,1/$* QV QV QV QV QV QV QV QV f 5$,1/$* QV QV QV QV QV QV QV 5$,1$'9 QV QV QV QV QV QV QV QV )/2: QV QV QV QV QV QV QV QV r )/:/$* QV QV QV QV QV QV QV )/:/$* QV QV QV QV QV QV )/:/$* QV QV QV QV QV )/2:$'9 QV QV QV QV QV QV QV QV r 3 S M +

PAGE 94

VLJQLILFDQW SRVLWLYH FRUUHODWLRQV DQG ULYHU IORZ ODJJHG E\ DQG PRQWKV VKRZHG VLJQLILFDQW QHJDWLYH FRUUHODWLRQV ZLWK VHDVRQDO IUXLWLQJ VSHFLHV H[FOXGLQJ 3 UHFOLQDWD &RPELQHG 3 UHFOLQDWD VFRUHV IRU WKH VLWHV ZHUH VLJQLILFDQWO\ FRUUHODWHG ZLWK UDLQIDOO ODJJHG E\ PRQWKV DQG ULYHU IORZ ODJJHG E\ DW OHDVW PRQWK $ PXOWLSOH UHJUHVVLRQ PRGHO LQFRUSRUDWLQJ PLQLPXP WHPSHUDWXUHV DQG ULYHU IORZ ODJJHG E\ PRQWKV WKH YDULDEOHV PRVW KLJKO\ FRUUHODWHG ZLWK IUXLW VFRUHV H[SODLQHG b RI WKH YDULDQFH LQ PRQWKO\ IUXLW VFRUHV IRU VHDVRQDO VSHFLHV H[FOXGLQJ UHFOLQDWD ) S f 7HPSHUDWXUH H[SODLQHG b DQG ULYHU IORZ b RI WKH WRWDO YDULDQFH &RPSDULVRQV ZLWK 'LVWULEXWLRQV RI IUXLW VFRUHV GLIIHUHG VLJQLILFDQWO\ EHWZHHQ DQG IRU RI WKH VSHFLHV H[DPLQHG $FDFLD UREXVWD 3DFKYVWHOD PVROR DQG )LFXV V\FRPRUXV RI 0QD]LQL DQG 'LRVS\URV PHVSLOLIRUPHV DQG )LFXV QDWDOHQVLV RI 0FKHOHOR KDG VLJQLILFDQWO\ GLIIHUHQW GLVWULEXWLRQV RI IUXLW VFRUHV EHWZHHQ DQG .ROPRJURY6PLUQRY -n DQG IRU $ UREXVWD 3 PVROR ) V\FRPRUXV PHVSLOLIRUPHV DQG ) QDWDOHQVLV UHVSHFWLYHO\ Sf $FDFLD UREXVWD KDG DQ H[WHQGHG IUXLWLQJ VHDVRQ LQ UHODWLYH WR DQG 3DFKYVWHOD PVROR KDG DQ H[WHQGHG IUXLWLQJ VHDVRQ LQ UHODWLYH WR 'LRVS\URV PHVSLOLIRUPHV )LFXV V\FRPRUXV DQG )BcB QDWDOHQVLV KDG PRUH

PAGE 95

XQLIRUP GLVWULEXWLRQV RI IUXLWLQJ LQ WKDQ LQ 7KHUH ZHUH VLJQLILFDQW GLIIHUHQFHV LQ PRQWKO\ IUXLW VFRUHV EHWZHHQ DQG IRU RI WKH VSHFLHV LQ 0QD]LQL DQG IRU RI WKH VSHFLHV LQ 0FKHOHOR )LFXV VYFRPRUXV DQG $FDFLD UREXVWD LQ 0QD]LQL UHFHLYHG VLJQLILFDQWO\ KLJKHU PHDQ PRQWKO\ IUXLWLQJ VFRUHV LQ WKDQ 7 DQG IRU )?B VYFRPRUXV DQG $ UREXVWD UHVSHFWLYHO\ Sf ZKHUHDV LQGLYLGXDOV RI 0LPXVRSV IUXWLFRVD UHFHLYHG VLJQLILFDQWO\ KLJKHU IUXLWLQJ VFRUHV LQ WKDQ 7 Sf 7KH VDPH VSHFLHV LQ 0FKHOHOR VKRZHG QR VLJQLILFDQW GLIIHUHQFHV LQ IUXLWLQJ VFRUHV 'LRVS\URV PHVSLOLIRUPHV DQG WKH SDOP +\SKDHQH FRPSUHVVD KDG VLJQLILFDQWO\ KLJKHU PHDQ IUXLWLQJ VFRUHV LQ WKDQ LQ 0FKHOHOR 7 DQG IRU 'AB PHVSLOLIRUPHV DQG FRPSUHVVD UHVSHFWLYHO\ Sf ,QGLYLGXDOV RI 'AB PHVSLO LI RUPHV LQ 0QD]LQL VKRZHG QR VLJQLILFDQW GLIIHUHQFHV LQ PRQWKO\ IUXLWLQJ VFRUHV EHWZHHQ WKH WLPH SHULRGV 'LVFXVVLRQ 0DQJDEH\ IRRGV ZHUH KLJKO\ YDULDEOH LQ WKHLU WHPSRUDO DEXQGDQFH DQG VSDWLDO GLVWULEXWLRQ 7KH PDMRULW\ RI IRRG VSHFLHV H[DPLQHG ZHUH UDUH DQG VSDWLDOO\ FOXPSHG 5DULW\ DQG VSDWLDO DJJUHJDWLRQ DUH FKDUDFWHULVWLF RI PDQ\ WURSLFDO WUHH VSHFLHV +XEEHO DQG )RVWHU 5RELQVRQ 6WUXKVDNHU f DQG PD\ EH WKH UHVXOW RI FKDQFH RU SDUWLFXODU

PAGE 96

PLFURKDELWDW UHTXLUHPHQWV +XEEHO DQG )RVWHU f +XJKHV f VKRZHG WKDW VHYHUDO VSHFLHV RFFXUULQJ LQ WKH 7DQD 5LYHU IRUHVWV KDYH YHU\ SDUWLFXODU VRLO DQG ZDWHU UHTXLUHPHQWV IRU JHUPLQDWLRQ DQG JURZWK %HFDXVH WKH IUHTXHQF\ DQG GHSWK RI IORRGLQJ VRLO W\SH DQG VRLO PRLVWXUH KROGLQJ FDSDFLW\ YDU\ ZLGHO\ EHWZHHQ IRUHVWV DQG RIWHQ ZLWKLQ D VLQJOH IRUHVW VWDQG GLIIHUHQW YHJHWDWLYH DVVHPEODJHV RFFXU DPRQJ IRUHVWV DQG FHUWDLQ SODQW VSHFLHV PD\ RFFXU RQO\ LQ SDUWLFXODU DUHDV ZLWKLQ D IRUHVW +XJKHV 0DUVK D +RPHZRRG f 7KH VSDWLDO GLVWULEXWLRQ RI PDQJDEH\ IRRG VSHFLHV VKRXOG DIIHFW UDQJLQJ EHKDYLRU WKURXJKRXW WKH \HDU EXW EHFDXVH RI WKH VPDOO VL]H RI WKH IRUHVWV DQG WKH PDQJDEH\Vn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f DQG UHODWLYH VFDUFLW\ -XQH 2FWREHUf 7KH DEXQGDQFH RI FRQWLQXRXV DV\QFKURQRXVO\ IUXLWLQJ VSHFLHV HJ )LFXV V\FRPRUXV DQG )LFXV

PAGE 97

QDWDOHQVLVf KRZHYHU FDQ DIIHFW WKH WHPSRUDO YDULDWLRQ LQ IUXLW DYDLODELOLW\ E\ GDPSHQLQJ WKH VHYHULW\ RI SHULRGV RI VFDUFLW\ $ FULWLFDO QXPEHU RI LQGLYLGXDOV KRZHYHU LV UHTXLUHG WR HQVXUH WKDW IUXLWV DUH DYDLODEOH HDFK PRQWK $V WKHVH VSHFLHV GHFOLQH LQ DEXQGDQFH WKH SUREDELOLW\ WKDW WKH VSHFLHV SRSXODWLRQ KDV DQ LQGLYLGXDO EHDULQJ IUXLWV GXULQJ ERWWOHQHFNV GHFUHDVHV 7KLV LV DSSDUHQW LQ WKH PRQWKO\ SKHQRORJLFDO VFRUHV IRU ) VYFRPRUXV LQGLYLGXDOV IURP 0QD]LQL Q f DQG 0FKHOHOR Q f IRUHVWV )LJ f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f 7KH DEXQGDQFH RI VSHFLHV VXFK DV 3 UHFOLQDWD DQG )LFXV VSS PD\ EH D NH\ IDFWRU LQ H[SODLQLQJ WKH SUHVHQFH RU DEVHQFH RI PDQJDEH\V LQ WKH YDULRXV IRUHVW SDWFKHV DORQJ WKH 7DQD 5LYHU 0DQ\ WURSLFDO IRUHVW VSHFLHV VKRZ DGDSWDWLRQV WRZDUG IUXLWGURS DQGRU VHHG JHUPLQDWLRQ QHDU WKH EHJLQQLQJ RI KHDY\ UDLQV 5DWKFKH DQG /DFH\ f )ORZHULQJ RFFXUV

PAGE 98

SULPDULO\ GXULQJ SHULRGV RI GURXJKW DQG ORZ VRLO PRLVWXUH $\UHV 0RQDVWHULR DQG 6DUPLHQWR &URDW )RVWHU .RHOPH\HU f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bf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f

PAGE 99

3URQRXQFHG \HDUWR\HDU YDULDWLRQ LQ WKH RQVHW DQG WHUPLQDWLRQ RI WKH GU\ VHDVRQ FRXOG KDYH LPSRUWDQW FRQVHTXHQFHV IRU WKH IORZHULQJ DQG IUXLWLQJ SKHQRORJLHV RI WUHH VSHFLHV HJ &KLYHUV DQG 5DHPDHNHUV f 7KH 7DQD ULYHULQH IRUHVWV H[SHULHQFH KLJK \HDUURXQG WHPSHUDWXUH ZLWK VRPH VHDVRQDO YDULDWLRQ WZLFH \HDUO\ UDLQV DQG LUUHJXODU IORRGLQJ ZLWK WKH VHYHULW\ DQG WLPLQJ RI UDLQV DQG IORRGV YDU\LQJ DSSUHFLDEO\ EHWZHHQ \HDUV +XJKHV +RPHZRRG 0DUVK f 7KH H[WUHPH XQSUHGLFWDELOLW\ DQG DQQXDO YDULDWLRQ LQ UDLQIDOO WHPSHUDWXUH DQG IORRGLQJ RYHUODLG ZLWK KDELWDW GHJUDGDWLRQ DQG IUDJPHQWDWLRQ PDNH LW GLIILFXOW WR GHWHUPLQH FDXVHV RI VLJQLILFDQW GLIIHUHQFHV LQ IUXLWLQJ SDWWHUQV EHWZHHQ DQG :KHHOZULJKW f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f HIIHFWLYHO\ GHFUHDVLQJ RYHUDOO IUXLW SURGXFWLRQ 6XFK ORQJWHUP FKDQJHV LQ WKH IUXLW

PAGE 100

UHVRXUFH EDVH PD\ EH VKDSLQJ WKH SUHVHQW GD\ IHHGLQJ HFRORJ\ RI WKH 7DQD PDQJDEH\

PAGE 101

&+$37(5 7+5(( $&7,9,7< 3$77(516 $1' ',(7 2) 7+( 7$1$ 5,9(5 &5(67(' 0$1*$%(< ())(&76 2) 6($621$/ $1' /21*7(50 +$%,7$7 &+$1*( ,QWURGXFWLRQ &RPSDULVRQV RI WKH SURSRUWLRQ RI WLPH WKDW DQLPDOV VSHQG RQ GLIIHUHQW DFWLYLWLHV DUH LPSRUWDQW LQ DVVHVVLQJ LQWHU DQG LQWUDVSHFLILF EHKDYLRUDO GLIIHUHQFHV DQG PD\ EH XVHG WR LGHQWLI\ WKH DGDSWLYH QDWXUH RI YDULDELOLW\ LQ WHPSRUDO SDWWHUQLQJ RI DFWLYLWLHV )D f 9DULDWLRQ LQ WKH ZD\V LQ ZKLFK DQLPDOV SDUWLWLRQ WLPH DPRQJ GLIIHUHQW DFWLYLWLHV LV VHHQ LQ D ZLGH UDQJH RI SULPDWH VSHFLHV 6HYHUDO JHQHUDOL]DWLRQV KDYH EHHQ PDGH FRQFHUQLQJ WKH UHODWLRQVKLS EHWZHHQ HFRORJLFDO IDFWRUV DQG SULPDWH DFWLYLW\ EXGJHWV &OXWWRQ%URFN &URRN DQG *DUWODQ (LVHQEHUJ HW DO 5RELQVRQ 6WUXKVDNHU DQG /HODQG f 'LHW DQG KDELWDW VWUXFWXUH KDYH EHHQ VKRZQ WR LPSRVH RYHUZKHOPLQJ FRQVWUDLQWV RQ DQLPDOV XVH RI WLPH $OWPDQQ 6 f %HFDXVH SDUWLWLRQLQJ RI WLPH LQYROYHV WUDGHRIIV EHWZHHQ PHWDEROLF UHTXLUHPHQWV RI GLIIHUHQW DFWLYLWLHV DQG WKH DFTXLVLWLRQ RI HQHUJ\ WR IXOILOO WKRVH UHTXLUHPHQWV GLHW TXDOLW\ DQG WKH WHPSRUDO DQG VSDWLDO GLVWULEXWLRQ RI IRRG VWURQJO\ LQIOXHQFH WKH DPRXQW RI WLPH VSHQW LQ GLIIHUHQW DFWLYLWLHV 2DWHV f

PAGE 102

0RQWKO\ YDULDWLRQ LQ DFWLYLW\ EXGJHWV SDUWLFXODUO\ LQ WKH WLPH VSHQW IHHGLQJ KDV EHHQ GHPRQVWUDWHG IRU VHYHUDO SULPDWH VSHFLHV DQG VKRZQ WR FRYDU\ ZLWK PHDVXUHV RI IRRG DYDLODELOLW\ HJ +DUULVRQ +RPHZRRG 2DWHV 5RELQVRQ 7HUERUJK :DVHU f ,Q JHQHUDO VSHFLHV UHO\LQJ KHDYLO\ RQ IROLDJH WHQG WR GHFUHDVH WKHLU OHYHO RI DFWLYLW\ GXULQJ PRQWKV ZKHQ KLJKTXDOLW\ IRRG LV VFDUFH ZKLOH VSHFLHV WKDW UHO\ KHDYLO\ RQ IUXLWV RU LQVHFWV LQFUHDVH WKH DPRXQW RI WLPH VSHQW ORRNLQJ IRU DQG SURFHVVLQJ IRRG GXULQJ WLPHV RI VFDUFLW\ 2DWHV f 7KH VSDWLDO GLVWULEXWLRQ RI IRRG DOVR LQIOXHQFHV WKH DPRXQW RI WLPH LQGLYLGXDOV VSHQG PRYLQJ LQGLYLGXDOV RI VSHFLHV WKDW IHHG RQ KLJK GHQVLW\ HYHQO\ GLVSHUVHG IRRG VSHQG OHVV WLPH PRYLQJ WKDQ WKRVH RI VSHFLHV WKDW IHHG RQ ZLGHO\ GLVSHUVHG IRRG 6WUXKVDNHU 6WUXKVDNHU DQG /HODQG f 2SWLPDO IRUDJLQJ WKHRU\ SUHGLFWV WKDW DQ DQLPDO ZLOO DFW WR PD[LPL]H HQHUJ\ LQWDNH SHU XQLW WLPH WR PHHW GHPDQGV IRU PDLQWHQDQFH PRYHPHQW DQG UHSURGXFWLRQ 6WHSKHQV DQG .UHEV f $FWLYLW\ EXGJHWV WKHUHIRUH VKRXOG UHIOHFW WKHVH DFWLRQV $V WKH GLVWULEXWLRQ DQG DYDLODELOLW\ RI IRRG FKDQJHV WLPH VSHQW LQ HQHUJ\DFFUXLQJ EHKDYLRUV HDWLQJf YHUVXV PRUH FRVWO\ EHKDYLRUV RI VHDUFKLQJ IRU DQG PRYLQJ EHWZHHQ SDWFKHV RI IRRG VKRXOG FKDQJH 7KH SURSRUWLRQ RI WLPH VSHQW LQ YDULRXV DFWLYLWLHV IRUPV D PXWXDOO\ GHSHQGHQW VHW LH WLPH VSHQW LQ RQH DFWLYLW\ DIIHFWV WKH WLPH DYDLODEOH IRU RWKHU DFWLYLWLHV $OWPDQQ DQG $OWPDQQ f

PAGE 103

7KHUHIRUH ZKHQ DQLPDOV PXVW JR WKURXJK D ORQJ VHDUFK RU KDQGOLQJ WLPH SHU XQLW HQHUJ\ PDMRU VKLIWV LQ DFWLYLW\ EXGJHWV ZLOO RFFXU $OWKRXJK VHDVRQDO DQG GDLO\ YDULDWLRQ LQ DFWLYLW\ SDWWHUQV DUH ZHOO GRFXPHQWHG IRU VHYHUDO SULPDWH VSHFLHV YHU\ IHZ VWXGLHV KDYH LQYHVWLJDWHG ORQJWHUP YDULDWLRQ LQ DFWLYLW\ EXGJHWV 2DWHV f )XUWKHUPRUH QRQH KDV DGGUHVVHG WKH TXHVWLRQ RI ZKHWKHU ODUJHVFDOH KDELWDW FKDQJH HJ IRUHVW ORVV DQG IUDJPHQWDWLRQf DQG VXEVHTXHQW FKDQJH LQ WKH IRRG UHVRXUFH EDVH HJ ORZHU UHVRXUFH DYDLODELOLW\f FDXVHV SULPDWHV WR UHDOORFDWH WLPH VSHQW LQ GLIIHUHQW DFWLYLWLHV 7KH 7DQD 5LYHU FUHVWHG PDQJDEH\ &HUFRFHEXV JDOHULWXVf LV DQ LGHDO VSHFLHV IRU DGGUHVVLQJ WKLV TXHVWLRQ 7DQD PDQJDEH\V KDYH H[SHULHQFHG VHYHUH KDELWDW DOWHUDWLRQ RYHU WKH ODVW \HDUV GXH WR IRUHVW VHQHVFHQFH IRUHVW FXWWLQJ IRU DJULFXOWXUH DQG VHOHFWLYH IHOOLQJ DQG SUXQLQJ RI VSHFLHV IRU EXLOGLQJ PDWHULDOV 0DUVK f 7KHVH KDELWDW FKDQJHV ZKLFK LQYROYHG WKH ORVV RI LPSRUWDQW PDQJDEH\ GLHW VSHFLHV KDYH UHVXOWHG LQ DQ RYHUDOO UHGXFWLRQ LQ WKH IRRG UHVRXUFH EDVH ,Q WKLV SDSHU H[DPLQH WKH HIIHFWV RI WKHVH \HDU FKDQJHV LQ WKH IRUHVW VWUXFWXUH DQG IRRG UHVRXUFH EDVH RQ ORQJWHUP YDULDWLRQ LQ DFWLYLW\ EXGJHWV DOVR H[DPLQH WKH WHPSRUDO GLVWULEXWLRQ GLXUQDO DQG VHDVRQDOf RI DFWLYLW\ SDWWHUQV ZLWKLQ DQG DPRQJ PDQJDEH\ JURXSV REVHUYHG LQ DQG WKH LQIOXHQFH RI GLHW RQ WLPH EXGJHWV )LQDOO\

PAGE 104

, DVN ZKHWKHU VKRUWWHUP DVVRFLDWLRQV EHWZHHQ DFWLYLW\ EXGJHWV DQG IRRG UHVRXUFHV DUH FRQVLVWHQW ZLWK WKRVH REVHUYHG DIWHU ORQJWHUP KDELWDW FKDQJH 0HWKRGV 'DWD &ROOHFWLRQ VWXGLHG JURXSV RI PDQJDEH\V LQ WKH 0FKHOHOR IRUHVW DQG JURXS LQ WKH 1NDQR IRUHVW *URXS VL]H DYHUDJHG UDQJH f DQG UDQJH f IRU WKH 1RUWK DQG 6RXWK 0FKHOHOR JURXSV UHVSHFWLYHO\ DQG UDQJH f IRU WKH 1NDQR JURXS $FWLYLW\ DQG GLHW GDWD IRU PDQJDEH\ JURXSV UDQJLQJ LQ WKH 0FKHOHOR DQG 1NDQR IRUHVWV GXULQJ ZHUH WDNHQ IURP +RPHZRRG f *URXS VL]H DYHUDJHG DQG IRU WKHVH JURXSV UHVSHFWLYHO\ *URXS DJHVH[ FRPSRVLWLRQ GLG QRW GLIIHU VLJQLILFDQWO\ EHWZHHQ WKH VWXGLHV GI S!f $OWKRXJK WKH 1NDQR JURXS VSHQW WKH PDMRULW\ RI WKH REVHUYDWLRQ PRQWKV bf LQ DQ DGMDFHQW IRUHVW 0QD]LQLf NP VRXWK RI 1NDQR PRQWKO\ DFWLYLW\ EXGJHWV ZHUH QRW VLJQLILFDQWO\ GLIIHUHQW ZKHQ WKH JURXS UDQJHG HQWLUHO\ LQ 1NDQR RU 0QD]LQL GI S!f WKHUHIRUH XVHG DOO PRQWKO\ GDWD IRU WKLV JURXS LQ WKH IROORZLQJ DQDO\VHV FROOHFWHG GDWD RYHU PRQWKV RQ WKH 0FKHOHOR JURXSV -DQXDU\ 0DUFK f DQG PRQWKV RQ WKH 1NDQR JURXS 0DUFK 0DUFK f 'DWD ZHUH FROOHFWHG XVLQJ PHWKRGV VLPLODU WR WKH SUHYLRXV VWXG\ E\ +RPHZRRG f IROORZHG JURXSV IRU FRQVHFXWLYH GD\V HDFK PRQWK IURP WR

PAGE 105

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f DQG UHFRUGHG WKH IROORZLQJ GDWD IRU HDFK LQGLYLGXDO ORFDWHG GXULQJ D VDPSOH f DJH f VH[ f LGHQWLW\ f KHLJKW DERYH WKH JURXQG f ILUVW DFWLYLW\ VXVWDLQHG IRU VHFRQGV f VSHFLHV DQG LWHP RI IRRG LI HDWLQJ ULSH RU XQULSH IUXLW ULSH RU XQULSH VHHG IORZHU EXG RU IORZHU VKRRWV VWHPV DQG VSURXWV OHDYHV JXP EDUN RU GHDG ZRRG DQG DQLPDO SUH\f DQG f DJH VH[ DQG LGHQWLW\ RI WKH QHDUHVW QHLJKERU ZLWKLQ D P UDGLXV JURXSHG DFWLYLWLHV LQWR WKH VDPH FDWHJRULHV XVHG E\ +RPHZRRG f f (DWLQJ 7KLV DFWLYLW\ LQFOXGHG DQLPDOV LQJHVWLQJ RU FKHZLQJ IRRG LWHPV $QLPDOV VFRUHG DV HDWLQJ ZKLOH PRYLQJ ZHUH FODVVLILHG DV HDWLQJ

PAGE 106

f )RUDJLQJ 7KLV FDWHJRU\ LQFOXGHG ERWK PDQLSXODWRU\ DQGRU VHDUFKLQJ DFWLYLW\ H[FOXGLQJ LQJHVWLRQ 0DQLSXODWRU\ IRUDJLQJ ZDV DQ\ PRYHPHQW ZKLFK LQYROYHG PDQLSXODWLQJ D SRWHQWLDO IRRG REMHFW 6HDUFKLQJ ZDV GHILQHG DV FORVHO\ LQVSHFWLQJ RU VLIWLQJ WKURXJK OHDI OLWWHU RU IROLDJH f 0RYLQJ 0RYHPHQW LQFOXGHG DOO ORFRPRWLRQ H[FHSW VKRUW PRYHPHQWV GXULQJ IRUDJLQJ DQG ORFRPRWLRQ GXULQJ DFWLYLWLHV VXFK DV SOD\LQJ f ,QDFWLYH 7KLV FDWHJRU\ LQFOXGHG DOO VWDWLRQDU\ QRQVRFLDO DFWLYLW\ $QLPDOV VOHHSLQJ VLWWLQJ O\LQJ RU VWDQGLQJ ZHUH FRQVLGHUHG LQDFWLYH f $XWRJURRPLQJ $Q\ DQLPDO VFUDWFKLQJ RU RWKHUZLVH VHOIn FOHDQLQJ ZDV DXWRJURRPLQJ f $OORJURRPLQJ $OORJURRPLQJ LQFOXGHG DQLPDOV JURRPLQJ RU EHLQJ JURRPHG E\ DQRWKHU DQLPDO f 6RFLDO6H[XDO 7KLV ZDV D FRPSRVLWH FDWHJRU\ LQFOXGLQJ Df DSSURDFK LQGLYLGXDOV JLYLQJ RU UHFHLYLQJ IULHQGO\ RU VH[XDO SUHVHQWDWLRQV RU HPEUDFHV Ef DJJUHVVLRQ LQGLYLGXDOV VXSSODQWLQJ DQRWKHU IURP DQ RFFXSLHG SRVLWLRQ RU OXQJLQJ DW FKDVLQJ RU ELWLQJ DQRWKHU DQG LQGLYLGXDOV EHKDYLQJ VXEPLVVLYHO\ WR DQRWKHU E\ SRVWXULQJ RU UHWUHDWLQJ IURP D SRWHQWLDO FRQIOLFW Ff SOD\ LQGLYLGXDOV HQJDJHG LQ PRFNFKDVLQJ RU PRFNILJKWLQJ Gf VH[XDO EHKDYLRUV PRXQWLQJ RU FRSXODWLQJ Hf PDWHUQDO EHKDYLRUV QXUVLQJ SLFNLQJ XS LQIDQW HWF DQG If YRFDOL]LQJ JLYLQJ DODUP FDOOV DQG DGXOW PDOH ORQJFDOOV

PAGE 107

'DWD $QDO\VLV HVWLPDWHG WKH DPRXQW RI WLPH VSHQW LQ GLIIHUHQW DFWLYLWLHV E\ FDOFXODWLQJ WKH SURSRUWLRQV RI HDFK DFWLYLW\ UHFRUGHG IRU DOO LQGLYLGXDOV GXULQJ HDFK GD\ RI D PRQWK DQG WDNLQJ WKH PHDQ RI WKH GDLO\ SURSRUWLRQV 'DWD DOVR ZHUH VXPPDUL]HG E\ WLPHRIGD\ E\ FDOFXODWLQJ WKH SURSRUWLRQV RI HDFK DFWLYLW\ UHFRUGHG GXULQJ HDFK KRXU RI WKH GD\ LQ D JLYHQ PRQWK DQG WDNLQJ WKH PHDQ RI WKH KRXUO\ HVWLPDWHV 0HDQ KRXUO\ HVWLPDWHV ZHUH JURXSHG LQWR WLPH SHULRGV IRU DQDO\VLV PRUQLQJ KUVf PLGGD\ KUVf DQG DIWHUQRRQ KUVf IXUWKHU JURXSHG GDWD LQWR VHDVRQV VKRUW UDLQV 1RY-DQf LQWHUUDLQV )HEf ORQJ UDLQV 0DU0D\f DQG GU\ VHDVRQ -XQ2FWf 6HDVRQDO DFWLYLW\ EXGJHWV ZHUH FDOFXODWHG E\ WDNLQJ WKH PHDQ RI WKH DSSURSULDWH PRQWKO\ HVWLPDWHV )LQDOO\ DQQXDO DFWLYLW\ EXGJHWV ZHUH FDOFXODWHG E\ WDNLQJ WKH PHDQ RI WKH DOO PRQWKO\ HVWLPDWHV 7KLV SURFHGXUH FRUUHFWV IRU WKH SRWHQWLDO ELDV GXH WR DQ XQHTXDO GLVWULEXWLRQ RI REVHUYDWLRQ WLPH DPRQJ GLIIHUHQW PRQWKV RU KRXUV RI WKH GD\ $OWPDQQ DQG $OWPDQQ f XVHG DQDO\VLV RI YDULDQFH PHWKRGV $129$ 0$129$f WR VLPXOWDQHRXVO\ HYDOXDWH WKH HIIHFWV RI VXLWHV RI LQGHSHQGHQW YDULDEOHV RQ PRQWKO\ DFWLYLW\ EXGJHWV DQG GLHW FRPSRVLWLRQ 1HWHU HW DO f 7KH XVH RI 0$129$ UHGXFHG WKH OLNHOLKRRG RI 7\SH HUURU UHVXOWLQJ IURP PXOWLSOH WHVWV DQG DOORZHG VLPXOWDQHRXV PHDVXUHPHQW RI WKH HIIHFWV RI

PAGE 108

LQGHSHQGHQW YDULDEOHV HJ JURXS VHDVRQ WLPHRIGD\f RQ FRUUHODWHG GHSHQGHQW YDULDEOHV 6RNDO DQG 5RKOI f ,I 0$129$ VKRZHG VLJQLILFDQW HIIHFW RI LQGHSHQGHQW YDULDEOHV RQ WKH VXLWH RI GHSHQGHQW YDULDEOHV WKHQ XVHG $129$ WR H[DPLQH WKH HIIHFWV RI WKHVH LQGHSHQGHQW YDULDEOHV RQ HDFK GHSHQGHQW YDULDEOH $Q DUFVLQ VJXDUHURRW WUDQVIRUPDWLRQ ZDV XVHG RQ SURSRUWLRQDO GDWD 6RNDO DQG 5RKOI f FDOFXODWHG VLJQLILFDQFH YDOXHV IURP SDUWLDO VXPV RI VTXDUHV 7\SH ,,,f ZKLFK DGMXVW HDFK HIIHFW IRU DOO RWKHU HIIHFWV LQ WKH PRGHO DQG DUH XQUHODWHG WR FHOO IUHTXHQFLHV )UHXQG DQG /LWWHOO f :KHQ DQDO\VLV RI YDULDQFH LQGLFDWHG VLJQLILFDQW GLIIHUHQFHV GXH WR DQ LQGHSHQGHQW YDULDEOH XVHG OHDVW VTXDUH PHDQV ZKLFK DGMXVW IRU XQEDODQFHG GDWD WR GHWHUPLQH ZKHUH GLIIHUHQFHV OD\ DPRQJ WKH LQGHSHQGHQW YDULDEOHV XVHG WKH VDPH DQDO\VLV RI YDULDQFH PHWKRGV WR FRPSDUH DFWLYLW\ EXGJHWV DQG GLHW FRPSRVLWLRQ EHWZHHQ DQG XVHG *WHVWV 6RNDO DQG 5RKOI f WR WHVW IRU GLIIHUHQFHV LQ WKH GLVWULEXWLRQV RI VRFLDO DFWLYLWLHV EHWZHHQ DQG DJHVH[ FODVVHV EHFDXVH UHSHDWHG PRQWKO\ VDPSOHV ZHUH QRW DYDLODEOH IRU WKH JURXSV DQG RQO\ DQQXDO PHDQV ZHUH DYDLODEOH 5HVXOWV 9DULDWLRQ LQ $FWLYLWLHV 0DQJDEH\V IURP DOO JURXSV VSHQW WKH PDMRULW\ RI WKHLU WLPH bf HQJDJHG LQ IHHGLQJ EHKDYLRUV GHILQHG DV

PAGE 109

HDWLQJ DQG IRUDJLQJ FDWHJRULHV FRPELQHGf VRFLDO EHKDYLRUV E\ FRPSDULVRQ VRFLDOVH[XDO DQG JURRPLQJf RFFXSLHG YHU\ OLWWOH WLPH bf )LJXUH 7DEOH f 7KHUH ZHUH VLJQLILFDQW QHJDWLYH FRUUHODWLRQV EHWZHHQ SURSRUWLRQ RI WLPH VSHQW HDWLQJ DQG Df LQDFWLYLW\ U Q Sf Ef PRYLQJ U Q SF2222Of DQG Ff IRUDJLQJ U Q Sf VXJJHVWLQJ WKDW WLPH QRW GHYRWHG WR HDWLQJ ZDV VSHQW UHVWLQJ RU WUDYHOLQJ EHWZHHQ DQG VHDUFKLQJ IRU IRRG VRXUFHV EXW QRW QHFHVVDULO\ VRFLDOL]LQJ 7KH SURSRUWLRQ RI WLPH VSHQW LQ HDFK DFWLYLW\ IRU DOO JURXS PHPEHUV H[FOXGLQJ LQIDQWVf YDULHG DPRQJ PRQWKV DQG JURXSV GXULQJ WKH ORZ VWDQGDUG HUURUV KRZHYHU LQGLFDWH OLWWOH ZLWKLQ PRQWK YDULDWLRQ 7DEOH f 0$129$ VKRZHG KLJKO\ VLJQLILFDQW HIIHFWV RI DOO LQGHSHQGHQW YDULDEOHV DQG WKHLU LQWHUDFWLRQV JURXS PHPEHUVKLS VHDVRQ DQG WLPHRIGD\f RQ WKH DFWLYLW\ EXGJHW ZLWK WKH H[FHSWLRQ RI WKH JURXS E\ WLPHRIGD\ LQWHUDFWLRQ 7DEOH f $129$ VKRZHG JURXS PHPEHUVKLS WR KDYH D VLJQLILFDQW HIIHFW RQ HDWLQJ DQG LQDFWLYLW\ 7DEOH f 6LJQLILFDQW OHDVW VTXDUHG PHDQV FRPSDULVRQV Sf VKRZHG WKDW WKH 1 0FKHOHOR JURXS VSHQW D KLJKHU SURSRUWLRQ RI WLPH HDWLQJ WKDQ HLWKHU 6 0FKHOHOR RU 1NDQR DQG 1NDQR ZDV OHVV DFWLYH WKDQ HLWKHU RI WKH 0FKHOHOR JURXSV )LJXUH f 6HDVRQDO GLIIHUHQFHV ZHUH VLJQLILFDQW IRU IRUDJLQJ DQG VRFLDOVH[XDO DFWLYLWLHV $129$ 7DEOH f )RUDJLQJ

PAGE 110

)LJXUH 0HDQ DQQXDO SHUFHQW WLPH VSHQW DPRQJ EHKDYLRUV E\ WKUHH PDQJDEH\ JURXSV DQG WZR PDQJDEH\ JURXSV Df Ef

PAGE 111

(DW )RUDJH 0RYH ,QDFW $XWRJUP $OORJUP 6RF6H[ %HKDYLRU

PAGE 112

7DEOH 0HDQ SHUFHQW WLPH VSHQW LQ HDFK DFWLYLW\ E\ PRQWK IRU PDQJDEH\V LQ WKH 1 0FKHOHOR 6 0FKHOHOR DQG 1NDQR VWXG\ JURXSV 6WDQGDUG HUURUV DUH LQ SDUHQWKHVHV 6HH WH[W IRU GHILQLWLRQ RI DFWLYLWLHV Df 1 0FKHOHOR 0RQ< 1Gf (DW )RUDJH 0RYH ,QDFWLYH $XWRJURRP $OORJURRP 6RF6H[ -DQ f f f f f f f )HE f f f f f f f 0DU f f f f f f f $SU f f f f f f f 0D\ f f f f f f f -XQ f f f f f f f -XO f f f f f f f $XJ f f f f f f f 6HS f f f f f f f 2FW f f f f f f f 1RY f f f f f f f 'HF f f f f f f f -DQ f f f f f f f )HE f f f f f f f 0DU f f f f f f f Ef 6 0FKHOHOR -DQ f f f f f f f )HE f f f f f f f 0DU f f f f f f f $SU f f f f f f f -XQ f f f f f f f -XO f f f f f f f 6HS f f f f f f f 2FW 1RY f f f f f f f 'HF -DQ f f f f f f f Ff 1NDQR 0DU $SU -XQ -XO $XJ 6HS 2FW 1RY 'HF -DQ )HE 0DU f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f f

PAGE 113

7DEOH 0$129$ UHVXOWV IRU JURXS PHPEHUVKLS VHDVRQ DQG WLPHRIGD\ GLIIHUHQFHV DPRQJ PDQJDEH\ DFWLYLW\ EXGJHWV 6RXUFH GI :LONVn /DPEGD )UDWLR 3 *URXS 0HPEHUVKLS 6HDVRQ 7LPHRI'D\ *URXS r 6HDVRQ *URXS r 7LPHRI'D\ 6HDVRQ r 7LPHRI'D\

PAGE 114

7DEOH $129$ UHVXOWV IRU JURXS PHPEHUVKLS VHDVRQ DQG WLPHRIGD\ GLIIHUHQFHV LQ EHKDYLRUV 6LJQLILFDQFH YDOXHV DUH FDOFXODWHG IURP 7\SH ,,, SDUWLDO VXPV RI VTXDUHV 6RXUFH GI 6XP RI 6TXDUHV 0HDQ 6TXDUHV ) 3 (DW *URXS 6HDVRQ 7LPHRI'D\ *URXS r 6HDVRQ 6HDVRQ r 7LPHRI'D\ (UURU )RUDJH *URXS 6HDVRQ 7LPHRI'D\ *URXS r 6HDVRQ 6HDVRQ r 7LPHRI'D\ (UURU 0RYH *URXS 6HDVRQ 7LPHRI'D\ *URXS r 6HDVRQ 6HDVRQ r 7LPHRI'D\ (UURU ,QDFWLYH *URXS 6HDVRQ 7LPHRI'D\ *URXS r 6HDVRQ 6HDVRQ r 7LPHRI'D\ (UURU $XWRJURRP *URXS 6HDVRQ 7LPHRI'D\ *URXS r 6HDVRQ 6HDVRQ r 7LPHRI'D\ (UURU

PAGE 115

7DEOH &RQWLQXHG 6XP RI 0HDQ 6RXUFH GI 6TXDUHV 6TXDUHV ) 3 $OORJURRP *URXS 6HDVRQ 7LPHRI'D\ *URXS r 6HDVRQ 6HDVRQ r 7LPHRI'D\ (UURU 6RFLDO6H[XDO *URXS 6HDVRQ 7LPHRI'D\ *URXS r 6HDVRQ 6HDVRQ r 7LPHRI'D\ (UURU

PAGE 116

RFFXSLHG D JUHDWHU SURSRUWLRQ RI WKH DFWLYLW\ EXGJHW GXULQJ WKH GULHU VHDVRQV WKDQ WKH UDLQ\ VHDVRQV )LJXUHV f DQG WKH JURXSV ZHUH PRUH VRFLDO GXULQJ WKH VKRUW UDLQV WKDQ DOO RWKHU VHDVRQV RI WKH \HDU ,QWHUDFWLRQV EHWZHHQ IRUHVW DQG VHDVRQ ZHUH VLJQLILFDQW IRU LQDFWLYLW\ WKH 1NDQR JURXS VSHQW D JUHDWHU DPRXQW RI WLPH LQDFWLYH GXULQJ WKH LQWHUn UDLQV WKDQ HLWKHU 1 RU 6 0FKHOHOR 7LPHRIGD\ ZDV VLJQLILFDQW IRU IRUDJLQJ PRYLQJ LQDFWLYLW\ DQG DOORJURRPLQJ $129$ 7DEOH f 2Q DYHUDJH IRUDJLQJ RFFXUUHG SULPDULO\ GXULQJ WKH PRUQLQJ f DQG PLGGD\ KRXUV f DQG JURXSV EHFDPH LQFUHDVLQJO\ OHVV DFWLYH E\ WKH DIWHUQRRQ KRXUV f ZKHQ DOORJURRPLQJ LQFUHDVHG )LJXUHV f 7KHUH DOVR ZHUH VLJQLILFDQW LQWHUDFWLRQV EHWZHHQ VHDVRQ DQG WLPHRIGD\ RQ PRYLQJ *URXSV PRYHG PRUH LQ WKH PRUQLQJ GXULQJ WKH GULHU VHDVRQV EXW VKRZHG D ODWHDIWHUQRRQ SHDN LQ PRYHPHQW GXULQJ WKH UDLQ\ VHDVRQV )LJXUHV f ,Q VXPPDU\ WKHUH DSSHDU WR EH SUHGLFWDEOH KRXUO\ DQG VHDVRQDO SDWWHUQV IRU FHUWDLQ DFWLYLWLHV $OWKRXJK WKH JURXSV DOORFDWH GLIIHULQJ DPRXQWV RI WLPH WR YDULRXV DFWLYLWLHV WKH\ HQJDJH LQ VLPLODU DFWLYLWLHV GXULQJ SDUWLFXODU KRXUV RI WKH GD\ 0DQJDEH\V H[SHQG PRUH RI WKHLU GDLO\ DFWLYLW\ EXGJHW LQ IRRG JDWKHULQJ EHKDYLRUV HJ IRUDJLQJ DQG PRYLQJf GXULQJ GULHU VHDVRQV ZKHQ IRRG UHVRXUFHV DUH OHVV DYDLODEOH DQG HQJDJH PRUH LQ VRFLDO EHKDYLRUV GXULQJ WKH VKRUW UDLQV ZKHQ IRRG UHVRXUFHV DUH

PAGE 117

)LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ PDMRU EHKDYLRUV HDWLQJ IRUDJLQJ PRYLQJ DQG LQDFWLYHf E\ WLPHRIGD\ DQG VHDVRQ IRU WKH 1 0FKHOHOR JURXS

PAGE 118

6KRUW 5DLQV /RQJ 5DLQV 0RUQLQJ 0LGGD\ $IWHUQRRQ ,QWHUUDLQV 'U\ 6HDVRQ 0RUQLQJ 0LGGD\ $IWHUQRRQ

PAGE 119

)LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ PDMRU EHKDYLRUV HDWLQJ IRUDJLQJ PRYLQJ DQG LQDFWLYHf E\ WLPHRIGD\ DQG VHDVRQ IRU WKH 6 0FKHOHOR JURXS

PAGE 120

6KRUW 5DLQV /RQJ 5DLQV 0RUQLQJ 0LGGD\ $IWHUQRRQ ,QWHUUDLQV 'U\ 6HDVRQ

PAGE 121

)LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ PDMRU EHKDYLRUV HDWLQJ IRUDJLQJ PRYLQJ DQG LQDFWLYHf E\ WLPHRIGD\ DQG VHDVRQ IRU WKH 1NDQR JURXS

PAGE 122

6KRUW 5DLQV /RQJ 5DLQV 0RUQLQJ 0LGGD\ $IWHUQRRQ ,QWHUUDLQV 'U\ 6HDVRQ

PAGE 123

UHDGLO\ DYDLODEOH &KDSWHU f +DELWDW DQG JURXS VL]H DOVR PD\ LQIOXHQFH DFWLYLW\ EXGJHWV 7KH 1NDQR JURXS ZKLFK LQKDELWV D ODUJHU IRUHVW ZLWK D ULFKHU UHVRXUFH EDVH &KDSWHU f DOORFDWHG PRUH WLPH WR LQDFWLYLW\ WKDQ HLWKHU RI WKH 0FKHOHOR JURXSV 7KH ODUJHU 6 0FKHOHOR JURXS DWH OHVV ZDV OHVV DFWLYH DQG PRYHG PRUH WKDQ HLWKHU RI WKH VPDOOHU JURXSV &RPSDULVRQV ZLWK $FWLYLW\ %XGJHWV 2YHUDOO DFWLYLW\ EXGJHW 0$129$ VKRZHG D VLJQLILFDQW HIIHFW RI JURXS :LONVn /DPEGD ) GI SF2222Of RQ WKH RYHUDOO DFWLYLW\ EXGJHWV IURP WZR DQG WKUHH PDQJDEH\ JURXSV 2QHZD\ $129$ VKRZHG VLJQLILFDQW JURXS GLIIHUHQFHV IRU DOO EHKDYLRUV EXW VRFLDOVH[XDO ) DQG IRU HDWLQJ IRUDJLQJ PRYLQJ LQDFWLYH DXWRJURRPLQJ DQG DOORJURRPLQJ UHVSHFWLYHO\ GI Sf /HDVW VTXDUH PHDQV FRPSDULVRQV VKRZHG WKDW PRVW RI WKH VLJQLILFDQW GLIIHUHQFHV Sf H[LVWHG EHWZHHQ WKH DQG JURXSV ZLWK IHZHU GLIIHUHQFHV DPRQJ JURXSV RI WKH VDPH VWXG\ 7KH JURXSV DWH OHVV ZHUH OHVV DFWLYH DQG VSHQW PRUH WLPH JURRPLQJ WKDQ WKH JURXSV )LJXUH f 7KH GLIIHUHQFHV EHWZHHQ IRUDJLQJ DQG PRYLQJ IRU WKH JURXSV ZHUH PRUH FRPSOLFDWHG 7KH 0FKHOHOR JURXS PRYHG PRUH WKDQ DOO RWKHU JURXSV ZLWK WKH H[FHSWLRQ RI WKH 6 0FKHOHOR JURXS DQG IRUDJHG OHVV WKDQ DOO RWKHU JURXSV

PAGE 124

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f FRPELQHG GDWD IRU P\ JURXSV DQG FDOFXODWHG DQQXDO PHDQV IRU VRFLDO EHKDYLRUV IRU HDFK DJHVH[ FODVV 7KHUH ZHUH QR VLJQLILFDQW GLIIHUHQFHV LQ WKH GLVWULEXWLRQ RI VRFLDO EHKDYLRUV EHWZHHQ DQG IRU HLWKHU DGXOW PDOHV GI S!f RU IHPDOHV GI S! )LJXUH f 'LIIHUHQFHV H[LVW KRZHYHU EHWZHHQ WKH VWXGLHV IRU MXYHQLOH PDOHV GI Sf ZKR VSHQW D JUHDWHU WKDQ H[SHFWHG DPRXQW RI WLPH SOD\LQJ LQ WKDQ GI O Sf DQG MXYHQLOH IHPDOHV GI Sf ZKR HQJDJHG LQ JURRPLQJ PRUH LQ WKDQ )LJXUH f 7KHUH DOVR ZDV VLJQLILFDQW KHWHURJHQHLW\ LQ WKH GLVWULEXWLRQV RI LQIDQW VRFLDO EHKDYLRUV EHWZHHQ DQG GI Sf ,QIDQWV RI DSSHDUHG WR EH PRUH

PAGE 125

)LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ VRFLDO EHKDYLRUV E\ DQG DGXOW PDOHV DQG IHPDOHV Df DGXOW PDOHV Ef DGXOW IHPDOHV

PAGE 126

$GXOW 0DOH $GXOW )HPDOH $JJUHVV $JJUnG $SSURDFK $SSUnG *URRP *URRPHG 3OD\

PAGE 127

)LJXUH 0HDQ SHUFHQW WLPH VSHQW LQ VRFLDO EHKDYLRUV E\ WKUHH DJHVH[ FODVVHV LQ DQG Df MXYHQLOH PDOHV Ef MXYHQLOH IHPDOHV Ff PDOH DQG IHPDOH LQIDQWV FRPELQHG

PAGE 128

-XYHQLOH 0DOH -XYHQLOH )HPDOH L $RRUHVV $IOJUnG $SSURDFK $SSUR *URRP *URRPHG 3OD\


222
assigned randomly to the infants. Sex-specific probabilities
of survival were assigned at random among the 7 infants with
values for the 7th infant devalued by 0.75 to account for
the probability of a 7th infant not being born. If the
generated survival value of any infant was lower than the
actual survival rate calculated for each sex from the known
group structures (using the ratio of adults to subadults,
juveniles and infants, Table 6.2), the infant was considered
to have survived. The number of surviving infants was summed
for each of the 1000 females; and the mean and variance of
female lifetime reproductive success calculated.
A male mangabey's lifetime reproductive success (K,,,) is
influenced by his ability to secure and retain tenure within
a group, the number and reproductive potential of available
breeding females, and the survival rate of the offspring
(Wade 1979). The probability of a male retaining tenure for
at least one year is estimated as:
p(t) = 1 (# group takeovers/# of breeding male years).
Assuming this probability is constant and independent
between years, then the probability of retaining tenure for
n years is p(t)n. Based on the calculated distribution of
p(t)n, I randomly distributed tenure time among 1000 males.
Males who retained tenure for less than 9 years after the
first run were given the opportunity to secure another group
and breed for up to 6 years (approximately 1 generation),
such that cumulative tenure was never greater than 12 years.


155
Specifically, I address the question of how the availability
and distribution of food resources influence ranging
patterns and spatial defense, and how intergroup
interactions, in turn, influence ranging patterns.
Methods
Data Collection
I studied 2 groups of mangabeys in Mchelelo forest and
1 group in Nkano forest. Group sizes averaged 18 (range 14-
18) and 28 (range 24-29) for the North and South Mchelelo
groups, respectively, and 16 (range 14-17) for the Nkano
group. Age-sex composition was similar for the 3 groups (see
Chapter 6).
A grid of trails was cut at approximately 50 and 100 m
intervals in both forests. Plastic flagging or marked
aluminum strips were placed every 50 m along trails or along
compass bearings where no trails were cut. The study areas
were mapped and overlaid with a grid representing 0.25 ha
(50m x 50m) quadrats. I chose a relatively small quadrat
size (see Waser and Floody 1974) because of the small size
of the study forests (17 and 38 ha for Mchelelo and Nkano,
respectively). A 0.25 ha quadrat allowed me to compare the
group's use of space to the distributions of selected tree
species, which were analyzed at this scale.
I collected data on mangabey movements and use of space
for 3 days each month from January 1988-March 1989 for each
Mchelelo group, and March 1988-March 1989 for the Nkano


143
Table 4.1. Continued.
CATEGORY
PLANT PART/SPECIFIC USE
HARVESTING METHOD
Remedy
a) leaflets
protecting healthy or
curing diseased mango trees
cut leaflets, tie in a ring
around lower trunk of mango tree,
leave throughout flowering/
fruiting season.
Commerce
a) leaf
sleeping and prayer mats
baskets
cut unexpanded 'sword leaf', sun
dry, split and plait strips for
mats or weave into baskets.
b) rachis
floor mats
cut leaf, remove leaflets, split
rachis linearly into thin strips,
loosely bind strips into large
mats.
Other
a) leaf
ornamentation
cut leaf at base, bind, fold or
weave as desired.
b) roots
insect larvae for fishing
dig through soil and leaf
litter at root base to extract
large wood grubs.
c) leaf
fold in a variety of ways to make
small dolls.


Figure 3.1. Mean annual percent time spent among 7
behaviors by three 1988-89 mangabey groups and
two 1973-74 mangabey groups,
a) 1988-89; b) 1973-74.


213
that strong responses are predicted only when food
availability is high. Waser and Homewood conducted their
experiments during the dry season when measures of food
availability were "approximately 1/2 that of the wet
season", and a low response, therefore, would be expected.
Additional data by Homewood (1976) demonstrated that Tana
mangabeys occasionally engaged in aggressive interactions
and that peaceful, intergroup interactions occurred
primarily during the wet season when fruit resources were
high.
The Tana mangabey has been described as an adaptable,
generalist species that adjusts to a wide range of
environmental conditions (Homewood 1976). Behavioral and
dietary flexibility have been shown to buffer seasonal
fluctuations in resource availability (Homewood 1978,
Chapter 3) and aid in adapting to large-scale habitat change
(Chapter 3). The Tana mangabey's flexibility of ranging and
intergroup behaviors are additional features of adaptation
to a highly seasonal, restricted habitat where food
resources fluctuate widely and awareness of neighboring
groups is high.


Figure 5.2. Cumulative frequency of sightings of one or more South Mchelelo group
mangabeys versus rank order of 0.25 h quadrats. Quadrats were ranked
with respect to the number of sightings occurring in each. The areas
accounting for 50 and 100% of all sightings are designated.


64
Mchelelo
Rainfall (mm)
300 -
250 -
200 -
150 -
L.I-. ll.L.-.l.-
ASONDJFMAMJJASONDJFMAM
8 8 8
7 8 9
Nkano
Rainfall (mm)
300
250 -j
ONDJFMAMJ JASONDJFM
8 8
7 8
Month
8
9


76
primarily during periods of drought and low soil moisture
(Ayres 1986, Monasterio and Sarmiento 1976, Croat 1975,
Foster 1974, Koelmeyer 1960). This phenological pattern does
not appear to prevail within the Tana River study sites.
Although flowering occurs during times of low river water
and rainfall, when soil moisture is low, there are few
consistent associations between fruiting and environmental
conditions. Fruit production by species that produce fruit
continuously throughout the year, particularly Ficus spp.,
appear to be associated with the onset of the rains but the
more seasonal species show no associations with rain. River
flow and temperature also explained very little of the
variance (25.6%) in a multiple regression model. It is
likely that fruiting patterns of the Tana forests are
influenced by a complex interaction of the variables
examined here and perhaps more importantly, the position of
the forest relative to the water table, the occurrence of
floods, the depth and length of inundation during flooding
and soil type. Fruiting time also may be more strongly
correlated with elapsed time from germination than with
environmental variables. Environmental cues seldom stimulate
the onset of fruit ripening; onset is determined primarily
by internal factors that control the rate of fruit
development (e.g., production of sugars; Rathche and Lacey
1985).


171
N. Mchelelo
Frequency
S. Mchelelo
Frequency
Nkano
Frequency
Number of times each quadrat used


CHAPTER FOUR COMPETING USES OF A FOREST PALM, PHOENIX
RECLINATA N. J. JACQUIN, BY HUMANS AND TANA RIVER
CRESTED MANGABEYS 136
Introduction 136
Methods 138
Study Species 138
Data Collection 139
Results 141
Discussion 149
CHAPTER FIVE TANA RIVER CRESTED MANGABEY HOME RANGE
AND SPATIAL DEFENSE: EFFECTS OF FRUIT ABUNDANCE
AND DISTRIBUTION 153
Introduction 153
Methods 155
Data Collection 155
Data Analysis 157
Results 160
Use of Space 160
Movements 179
Intergroup Behavior 186
Discussion 205
CHAPTER SIX POPULATION VIABILITY OF THE TANA RIVER
CRESTED MANGABEY 214
Introduction 214
Methods 216
Mangabey Data Set 216
Population Genetics Model 219
Demographic Extinction Model 226
Results 227
Estimates of Ne 227
Population Subdivision 224
Demographic Extinction Model 236
Discussion 242
Effective Population Size 242
Demographic Extinction Model 246
CHAPTER SEVEN CONCLUSION 249
APPENDIX A ESTIMATING TREE CANOPY SURFACE AREA AND
VOLUME 255
APPENDIX B NUMBER OF FEEDING RECORDS AND PERCENT OF TOTAL
RECORDS BY PLANT TAXON AND MANGABEY GROUP 258
REFERENCES 261
BIOGRAPHICAL SKETCH 277
lx


Mchelelo
Fruit score
6 8 8
7 8 g
Mnazini
Fruit score
8 8 8
7 8 9
Month
Mchelelo
Fruit score
NDJFMAMJJASONDJFMAM
8 8 8
7 8 9
Mnazini
Fruit score
8 8 8
7 8 9
Month ^
u>


mangabey is a generalist species that easily exploits new
habitat, it should respond well to habitat management.
XXII


87
eating and foraging categories combined); social behaviors
by comparison (social/sexual and grooming) occupied very
little time (13-14%) (Figure 3.1, Table 3.1). There were
significant negative correlations between proportion of time
spent eating and a) inactivity (r=-0.41, n=38, p<0.009), b)
moving (r=-0.29, n=38, pcO.OOOl), and c) foraging (r=-0.43,
n=38, p<0.007), suggesting that time not devoted to eating
was spent resting or traveling between and searching for
food sources, but not necessarily socializing.
The proportion of time spent in each activity for all
group members (excluding infants) varied among months and
groups during 1988-89; the low standard errors, however,
indicate little within month variation (Table 3.1). MANOVA
showed highly significant effects of all independent
variables and their interactions (group membership, season,
and time-of-day) on the activity budget with the exception
of the group by time-of-day interaction (Table 3.2). ANOVA
showed group membership to have a significant effect on
eating and inactivity (Table 3.3). Significant least squared
means comparisons (p<0.05) showed that the N. Mchelelo group
spent a higher proportion of time eating than either S.
Mchelelo or Nkano, and Nkano was less active than either of
the Mchelelo groups (Figure 3.1).
Seasonal differences were significant for foraging and
social/sexual activities (ANOVA, Table 3.3). Foraging


Mchelelo
Frequency
160-1
Nkano
Frequency
Month


Table 3.7. Frequency of plant parts eaten for common diet species of the 3 study groups
a) N. Mchelelo
Species
Ripe
Unripe
Ripe
Unripe
Sprouts
Wood
Fruits
Fruits
Seeds
Seeds
Flowers
Leaves
Shoots
Bark
Gun
Total
Phoenix reclinaba
29
50
83
761
53
1
12
23
0
1012
Hyphaene compressa
49
9
24
230
4
0
0
0
0
322
Oncoba spinosa
239
10
0
0
0
0
0
1
0
251
Aporrhiza paniculate
0
0
134
0
0
0
5
1
0
140
Ficus natalensis
124
4
0
0
0
0
0
0
0
128
Diosovros mespiliformes
59
0
10
43
0
0
1
0
0
113
Alanaiun salviifoliun
73
0
5
2
1
0
21
2
0
104
Acacia robusta
0
0
5
65
0
2
0
0
29
101
Tamarindus indica
14
82
0
0
0
0
0
0
0
96
Ficus sycamorus
37
53
0
0
0
0
0
0
0
90
Lannea schweinfurthii
54
0
8
11
0
0
0
0
0
73
Saba comorensis
19
18
0
7
4
4
0
2
0
54
Sorindeia madaaascariensis
10
30
3
0
0
5
0
0
0
49
b)
S. Mchelelo
Phoenix reclnate
47
3
106
390
30
2
6
18
0
602
Hvphaene compressa
32
0
9
96
0
0
0
0
0
137
Aporrhiza paniculate
0
0
93
0
0
0
1
0
0
94
Diosovros mespiliformes
18
2
5
58
0
0
0
2
0
85
Alanaiun salviifoliun
57
0
5
1
0
0
11
1
0
75
Mimusops fru teosa
72
0
0
1
0
0
0
0
0
73
Oncoba spinosa
66
3
0
0
0
0
0
0
0
69
Ficus bussei
11
27
0
5
0
0
0
0
0
43
Cordi a aoetzei
0
0
0
33
0
0
0
0
0
33
Tamarindus indica
0
18
0
3
1
0
0
0
0
22
Acacia robusta
0
1
1
0
0
0
0
0
4
6
c) Nkano
Phoenix reclinata
2
1
40
389
8
0
4
3
447
Saba comorensis
169
0
2
165
5
11
3
22
377
Ficus sycamorus
247
94
0
0
0
1
0
2
342
Aporrhiza paniculate
0
4
175
14
0
0
39
1
233
Pachvstela msolo
88
6
2
7
96
1
4
15
219
Alanaiun salviifolturn
63
0
3
1
1
0
8
2
78
Polvsphaeria multiflora
61
2
0
0
0
0
0
0
63


245
parameter estimates of breeder population size and male and
female lifetime reproductive success, are far below Lande
and Barrowclough's (1987) suggested effective population
size of >500 for maintenance of quantitative variance under
stabilizing or neutral selection. That groups on opposite
banks of the river are effectively isolated does not improve
the situation and suggests that the effective population
size of concern may be that for the separate river bank
subpopulations. Effective population sizes for the east and
west bank demes average around 36 individuals; such small
subpopulations may not protect against the loss of the
majority of genetic variation (Soul 1980).
Harris and Allendorf (1989) state that "for management
purposes, it is probably unnecessary to strive for great
precision in Ne estimates. Given the likely uncertainties in
data necessary for its calculation by any method,
excessively rigid dependence on even the best of estimations
is unwarranted". My data result from a relatively short-term
study of a long-lived species, therefore much of the
demographic data are limited in quality and consequently
there are uncertainties in the parameter estimates. The Nc' s
reported here should nevertheless reflect the best estimates
of effective population size of the Tana mangabey and
provide good first approximations of the potential magnitude
of the problem and the range of genetic loss likely to take
place.


Table 2.5. Spearman's rank correlation coefficients for mean monthly environmental
variables and mean monthly fruiting scores for Nkano forest. Variables defined in Table
2.4.
RIPE AND UNRIPE
RIPE ONLY
RIPE AND UNRIPE
RIPE AND UNRIPE
Total
Corn inuous
Seasonal
Total
Continuous
Seasonal
Seasonal
(w/o P. reelinata)
P. reelinata
*
*
*
**
MINTEMP
0.72
0.63
0.61
ns
ns
ns
0.74
0.75
RAIN
ns
ns
ns
ns
ns #
ns
ns
ns
RAINLAG1
ns
ns
ns
ns
0.60
ns
ns
ns
RAINLAG2
ns
ns
ns
ns
ns
ns
ns
ns
RAINLAG3
ns
ns
ns
ns
ns
ns
ns
ns
RAINADV
ns
ns
ns
ns
ns
ns
ns
ns
FLOW
ns
ns
ns
ns
ns
ns
ns
ns <*
FL0WLAG1
ns
ns
ns
ns
ns
ns
ns
0.78
FL0WLAG2
ns
ns
ns
ns
ns
ns
ns *
0.83**
FL0WLAG3
ns
ns
ns
ns
ns
ns
-0.59
0.79
FLOWADV
ns
ns
ns
ns
ns
ns
ns
ns
*
*
p<0.05
p<0.001
C\
>£>


123
1988/89
80 i
N. Mchelelo S. Mchelelo Nkano
Unripe Seed HZ Unripe Fruit Mi Ripe Seed
1973/74
N. Mchelelo
HH Ripe Fruit


266
Hall, K.R.L. 1962. Numerical data, maintenance activities
and locomotion in the wild chacma baboon (Papio
ursinus). Proceedings of the Zoological Society of
London 139:181-220.
Hamilton, W.J. Ill, R.E. Buskirk, and W.H. Buskirk. 1976.
Defence of space and resources by chacma (Papio ursinus)
baboon troops in an African desert and swamp. Ecology
57:1263-1272.
Harding, R.S.O. 1976. Ranging patterns of a troop of
baboons (Papio anubus) in Kenya. Folia Primatologica
25:143-185.
Harris, R.B. and F.W. Allendorf. 1989. Genetically
effective population size of large mammals: an
assessment of estimators. Conservation Biology 3:181-
191.
Harrison, M. 1985. Time budget of the green monkey,
Cercopithecus sabaeus: Some optimal strategies.
International Journal of Primatology 6(4):351-376.
Hartl, D.L. 1980. Principles of Population Genetics.
Sinauer Associates, Inc. Sunderland, MA.
Heideman, P.D. 1989. Temporal and spatial variation in the
phenology of flowering and fruiting in a tropical
rainforest. Journal of Ecology 77(4):1059-1079.
Hill, W.C.O. 1974. Primates: Comparative Anatomy and
Taxonomy, VII. Cynopithecinae. Edinburgh University
Press, Edinburgh, UK.
Hill, W.G. 1972. Effective size of populations with
overlapping generations. Theoretical Population Biology
3:278-289.
Hilty, S.L. 1980. Flowering and fruiting periodicity in a
premontane forest in Pacific Colombia. Biotropica
12(4):292-306.
Homewood, K.M. 1975. Can the Tana mangabey survive? Oryx
13:53-59.
Homewood, K.M. 1976. Ecology and Behavior of the Tana
Mangabey (Cercocebus galeritus galeritus). Unpublished
Ph.D. dissertation, University of London, UK.
Homewood, K.M. 1978. Feeding strategy of the Tana mangabey
(Cercocebus galeritus galeritus)(Mammalia: Primates).
Journal of Zoology, London 186:375-391.


131
Davies et al. 1988). Seeds may have a lower cost/benefit
ratio because they are rich protein sources and, relative to
other plant items, have fewer digestion-inhibitors (McKey et
al. 1981, Waterman 1984). Although toxic fatty acids and
alkaloids are present in unripe seeds, most tannins are
found in easily discarded seed coats (McKey 1979, Waterman
1984). Perhaps, mangabeys as monogastric primates, are not
as adept at dealing with secondary compounds as the
folivorous Colobine monkeys. A shift to unripe seeds and
other starch- or protein-rich diet items such as
invertebrates (even with the added cost of increased
foraging time) may constitute the best alternative strategy
when ripe fruits are unavailable. Berenstain (1986) showed a
similar diet shift from primarily ripe fruits and seeds to
unripe seeds and insects by long-tailed macaques (Macaca
fascicularis) in Borneo after a severe fire when ripe fruits
were unavailable.
The lower availability of ripe fruits to mangabeys in
1988-89 may have been due both to reduced numbers of
fruiting trees by habitat loss or increased competition for
ripe fruits with yellow baboons. Fruit abortion and
frugivory generally results in unripe fruits being
consistently more abundant than ripe fruits. Many species
fruit asynchronously (e.g. Ficus spp.), both within and
between individuals; a reduction in the total number of
individuals in a given population, therefore, will reduce