Citation
Seed dispersal and post-dispersal seed fate of four tree species in a neotropical cloud forest

Material Information

Title:
Seed dispersal and post-dispersal seed fate of four tree species in a neotropical cloud forest
Creator:
Wenny, Daniel G
Publication Date:
Language:
English
Physical Description:
ix, 179 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Birds ( jstor )
Ecology ( jstor )
Fruits ( jstor )
Germination ( jstor )
Seed dispersal ( jstor )
Seed predation ( jstor )
Seedlings ( jstor )
Seeds ( jstor )
Species ( jstor )
Trees ( jstor )
Cloud forests -- Costa Rica ( lcsh )
Dissertations, Academic -- Zoology -- UF ( lcsh )
Trees -- Seeds -- Dispersal -- Costa Rica ( lcsh )
Zoology thesis, Ph. D ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1998.
Bibliography:
Includes bibliographical references (leaves 152-177).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Daniel G. Wenny.

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:
029248131 ( ALEPH )
41209658 ( OCLC )

Downloads

This item has the following downloads:


Full Text











SEED DISPERSAL AND POST-DISPERSAL SEED FATE OF FOUR TREE SPECIES
IN A NEOTROPICAL CLOUD FOREST







By

DANIEL G. WENNY


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


1998






















To my grandfather, Donald G. Baker,
who led me on hikes in the woods of New Hampshire,


and to my kids, Malia and Jack,

with whom I can barely keep up.















ACKNOWLEDGMENTS


First and foremost I thank my advisor, Doug Levey, for his advice,

encouragement, patience, and help with all aspects of this project. His input was

instrumental in its success, and I will always be grateful for his help. He found a few

seeds even though the sunbittern eluded him. If I discover a new species, I will name it

after him in appreciation. I also thank my committee, Karen Bjomrndal, Colin Chapman,

Jack Putz, and Buzz Holling (also Frank Bonaccorso and Pete Feinsinger at an earlier

stage) for help developing the research proposal, clarifying the direction of the dissertation,

and editing the resulting manuscripts. I feel grateful they pointed me in the right direction

and let me take a few steps. Other members of the faculty were helpful with advice on

experimental design, statistical analysis, seminar presentations, and general reality checks:

Rich Kiltie, Dave Steadman, Mike Miyamoto, Frank Nordlie, Collette St. Mary, Brian

McNab, Larry McEdward, John Anderson, and Craig Osenberg. Nat Wheelwright,

George Powell, Carlos Guindon, Nalini Nadkamrni, Jack Longino, and Greg Murray also

provided encouragement and advice.

This research would not have been possible without valuable field assistance from

Victor Per6z, Ricardo Guindon, Jason Bennett, and Wendy Gibbons. I thank them

sincerely for their hard work under less than ideal conditions and sometimes with vague
instructions. Many other people helped with the project, either directly or indirectly. The

Monteverde community is a unique and wonderful place, and the two and a half years I

spent there were among the highlights of my life. I especially thank the Guindon,

Campbell, and Rockwell families for making me (and Wendy) feel so welcome there.

Special thanks go to Benito, Carlos, and Tomds Guindon for sharing their knowledge of







the area, and to Greg Murray, Mauricio Garcia, Rodrigo Solano, Bill Haber, Willow

Zuchowski, Frank Joyce, Alan Pounds, and Alan and Karen Masters for patiently

answering my endless questions. Finally, thanks go to Judy Poe and Bruce Pack for

renting us their beautiful octagonal house on the cliff edge, and to Leyn Rockwell for

emergency repairs.

Crucial financial support at the beginning of this project was received from the

Organization for Tropical Studies: a pilot study grant (Pew Charitable Trust), and a

Tropical Fellowship (Jesse Smith Noyes Foundation). Thanks go to Lucinda McDade and

Shaun Bennett for facilitating OTS support. Other early funding from Sigma Xi Grants-in-

Aid of Research and American Museum of Natural History (Frank M. Chapman Memorial

Fund) is gratefully acknowledged. The bulk of the fieldwork was supported by a

Dissertation Improvement Grant from the National Science Foundation (DEB 93-21553).

Indirect financial support was also contributed by the Gibbons and Wenny families. I am

also grateful for logistic support from the Organization for Tropical Studies, Monteverde

Institute (e-mail access), and the Tropical Science Center (lab space). Most of all, I greatly

appreciate the many years of support from the UF Department of Zoology.

I thank my extended family for encouragement, visits to Costa Rica during the three

years of fieldwork, and for not asking too many times what exactly I was doing. Most

important was support from my wife, Wendy Gibbons, who helped with fieldwork, edited

drafts of proposals and papers, and was generally nice when I needed a friend. Thanks go

to everyone who spent a day or two in the field in exchange for a tour. I apologize to my

mother, who helped carry equipment 3 miles to the study site after my promise of a quetzal,

which I was unable to find that day or on a second visit a year later.

Finally, I would like to note my appreciation of the zoology, botany, and wildlife
graduate students who made my tenure here more enjoyable. I thank them for discussions,

editing manuscripts, statistical advice, suffering through practice seminars, scheming up

side projects, and saying they liked the Ozric Tentacles tapes I made for them: Seth







Bigelow, Scot Duncan, Ron Edwards, Susan Moegenburg, Jay O'Sullivan, Greg Pryor,

Rafael Samudio, Lenny Santesteban, Mark Spritzer, Markus Tellkamp, and... Special

thanks go to Susan Moegenburg for being the ideally cheerful officemate and to Jason

Evert for being the ideally invisible officemate.















TABLE OF CONTENTS


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

ABSTRACT ....................................................................................... viii

CHAPTERS

1 GENERAL INTRODUCTION ................................................................. 1

2 SEED DISPERSAL, SEED PREDATION, AND SEEDLING
RECRUITMENT OF OCOTEA ENDRESIANA (LAURACEAE)

Introduction ........................................................................ 6
Study Site .......................................................................... 9
Study Species ..................................................................... 11
Methods ........................................................................... 15
Results ............................................................................. 23
Discussion ......................................................................... 47
Conclusion ........................................................................ 64


3 SEED DISPERSAL OF A HIGH-QUALITY FRUIT BY SPECIALIZED
FRUGIVORES: HIGH-QUALITY DISPERSAL?

Introduction ....................................................................... 66
Study Site ......................................................................... 70
Study Species ..................................................................... 71
Methods ........................................................................... 73
Results ............................................................................. 76
Discussion ......................................................................... 87

4 TWO-STAGE DISPERSAL OF TWO SPECIES OF GUAREA (MELIACEAE)

Introduction ....................................................................... 93
Study Site ......................................................................... 95
Study Species ..................................................................... 95
Methods ........................................................................... 96
Results ............................................................................. 99
Discussion ....................................................................... 107







5 ADVANTAGES OF SEED DISPERSAL: A RE-EVALUATION
OF DIRECTED DISPERSAL

Introduction ..................................................................... 110
Advantages of Dispersal ....................................................... 112
The Diffuse Mutualism Paradigm ............................................ 118
Classic Examples of Directed Dispersal ..................................... 120
Other Examples of Possible Directed Dispersal............................. 124
Conclusion: Going the Distance, and Beyond .............................. 134
APPENDIX OCOTEA ENDRESIANA SITE LOCATIONS .............................. 138

REFERENCES ................................................................................... 152

BIOGRAPHICAL SKETCH ................................................................... 178














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
SEED DISPERSAL AND POST-DISPERSAL SEED FATE OF FOUR TREE SPECIES
IN A NEOTROPICAL CLOUD FOREST

By

Daniel G. Wenny

May, 1998

Chairman: Douglas J. Levey
Major Department: Zoology

Seed dispersal by animals is thought to be important in tropical forests because

most trees in these forests produce fleshy or arillate fruits adapted for consumption by birds

and mammals. Very little is known about what happens to seeds after dispersal, yet post-

dispersal fate is important in determining patterns of plant recruitment I studied dispersal

patterns and post-dispersal seed fates of four tree species in Monteverde, Costa Rica. The

four species studied included two species of Lauraceae (Ocotea endresiana and

Beilschmiediapendula) and two Meliaceae (Guarea glabra and G. kunthiana). I determined

locations of dispersed seeds by following birds until they regurgitated or defecated seeds

and by systematically searching the study site for recently dispersed seeds. For each of the

four species, approximately 75% of the seeds dispersed by birds were deposited within 25
m of the parent trees. Ocotea endresiana showed a bimodal pattern in which bellbirds

(Procnias tricarunculata) dispersed many seeds under song perches and thus produced a

second peak in seed rain in addition to the peak near parent trees. Marked seeds were used

to determine if post-dispersal removal resulted in seed predation or secondary dispersal.

Removal rates were high for Ocotea (99% removed) and Guarea (85-90%), and low for







Beilschmiedia (17%). Secondary dispersal was observed for both Guarea species but not

for the Lauraceae species. Secondary dispersal of Guarea resulted in a slight increase in

overall dispersal distance, and a shift in microhabitat.

The spatial pattern of seedling recruitment was bimodal for Ocotea, reflecting the

initial pattern of dispersal. Beilschmiedia recruitment was not bimodal and was highest in

the zone approximately 10 m from the parental crowns. Guarea recruitment was not

studied in the field, but in growing house experiments, buried G. glabra seeds germinated

and established seedlings at a higher rate than seeds on the soil surface, suggesting a

benefit to secondary dispersal. The results of this study indicate that determining dispersal

pattern and post-dispersal fate are important to assess the influence of seed dispersers on

patterns of seedling recruitment













CHAPTER 1
GENERAL INTRODUCTION

"With plants there is a vast destruction of seeds. Seedlings, also, are destroyed in vast
numbers by various enemies." Charles Darwin (1859:78).

"Plants are not charitable beings." William J. Beal (1898:84).


Seed dispersal by animals is widespread and recognized as important by virtue of

the large number of plant species with fruits presumably adapted for animal consumption.

In tropical forests in particular, 70-90% of the trees and shrubs are thought to be dispersed

by animals (Stiles 1985, Howe 1986, Willson et al. 1989, Jordano 1992). This seed

dispersal interaction is characterized as a diffuse mutualism because most plants have

multiple dispersal agents and most seed dispersing animals disperse several to many

different plant species. Thus, fruit-frugivore coevolution generally occurs among groups

of plants and groups of dispersers rather than at the species-specific level seen in many

pollination systems (Wheelwright and Orians 1982). The advantage of this mutualism to

animals is a nutritive reward, whereas the advantages to plants are escape from density-

dependent mortality near the parent plant and arrival in suitable sites for establishment.

Arrival in suitable locations can be either a random process, in which widespread

dissemination of seeds would increase the chance of colonizing a good site (colonization

hypothesis), or a nonrandom process in which seeds are directed to suitable sites by

attracting certain dispersal agents (directed dispersal hypothesis). Escape and colonization

are thought to be the main advantages of seed dispersal for most plants (Howe 1986)

although few studies have examined these hypotheses in detail.







One of the main factors limiting our understanding of the advantages of dispersal is

that most previous studies have examined aspects of dispersal without considering

subsequent stages that may also influence recruitment. Dispersal is part of the plant

recruitment process, which is composed of several stages, and most previous studies have

been stage-specific. Zoologists have focused on fruit removal, foraging behavior, and gut

treatment of seeds, while botanists have studied germination and seedling growth (Howe

1993b). Seed predation has attracted the attention of both zoologists and botanists, but few

studies have integrated all the stages of recruitment in one study (but see Herrera et al.

1994). Our understanding of the importance of dispersal by animals is limited because we

know so little about where animals take seeds and what happens to them after dispersal.

Furthermore, many studies assume that when seeds are dispersed they will either live or die

in that place. Recent studies, however, indicate that many seeds experience a second stage

of dispersal by ants, rodents, dung beetles, and other animals (Estrada and Coates-Estrada

1991, Levey and Byrne 1993, Fragoso 1997). Thus, conclusions based on the pattern of

initial dispersal may be misleading.

The goal of this study was to examine the stages of recruitment, including seed

dispersal, seed predation, secondary dispersal, germination, and seedling establishment, to

link patterns of dispersal with patterns of recruitment. This study took place in

Monteverde, Costa Rica, which has been the site of some of the key studies in the seed

dispersal literature (e.g., Wheelwright and Orians 1982, Murray 1988, Nadkamrni and

Wheelwright in press). In Chapters 2 and 3,1 present studies of two tree species in the

plant family Lauraceae, Ocotea endresiana and Beilschmiediapendula. Lauraceous species

are an important component of Neotropical montane forests in terms of species richness

(Gentry 1990, Haber et al. 1996) and the variety of birds that are at least partially

dependent on their fruits (Wheelwright et al. 1984). In Chapter 4, I present a study of two

species of Meliaceae, Guarea glabra and G. kunthiana. The main objective for all four

species was to find the locations where seeds are dispersed and follow the fate of those







seeds after dispersal to link patterns of dispersal with patterns of recruitment. All four of

these tree species are common in the Monteverde forests, and understanding their dispersal

and recruitment patterns should shed light on how the forest works in replacing itself or

perhaps in changing over time.

For the study presented in Chapter 2,1 followed birds from fruiting trees and

systematically searched the ground for dispersed Ocotea endresiana seeds. I protected the

located seeds in situ with wire mesh cages to determine germination rates in the absence of

seed predation by mammals. After germination I removed the cages to assess seedling

establishment as a function of habitat characteristics. To study seed predation, I placed

marked seeds next to the caged seeds and recorded their fate over regular intervals. The

fates of caged and marked seeds were also studied for seeds placed at random locations.

These data were used to determine if dispersers take seeds to a nonrandom set of sites with

characteristics predictably favorable for seedling survival. The results of this study show

that most seeds dispersed by birds landed within 25 m of parent trees, but approximately

12% are dispersed to gaps by male bellbirds (Procnias tricarunculata) which have habitual

song perches on exposed branches. Post-dispersal seed predation is very high but did not

differ between seeds in gaps and in the forest understory. Secondary dispersal was not

observed. Of the seeds that survived seed predators, most germinated, and seedlings in

gaps grew taller and survived at a higher rate than seedlings in the forest understory. This

study is apparently the first example of directed dispersal by birds of a tropical tree.

In Chapter 3, 1 present data on the dispersal and post-dispersal fate of

Beilschmiediapendula. In this study I determined dispersal locations solely by searching

the ground because the long retention times of these large seeds (12 g) precluded following

the birds long enough to find the seeds. Otherwise the methods were similar to those used

for Ocotea. The distribution of dispersed seeds was similar to that of Ocotea except that

few seeds were dispersed to gaps. Seed predation was very low, and again, secondary

dispersal was not observed. Habitat characteristics were not useful in predicting seedling







survival The highest probability of seedling recruitment was 10-20 m from the edge of

parent tree crowns. I discuss the results in terms of dispersal quality and suggest that for

this species dispersal within 20 m of the crown is relatively high-quality dispersal and

approximately 20% of the seeds receive such treatment.

In Chapter 4,1 present data on the dispersal systems of two species of Guarea

(Meliaceae) that produced fruit at a time when Ocotea and Beilschmiedia did not. As in

the Beilschmiedia study, dispersed seeds were located by systematic ground searches.

The distribution of dispersed seeds was similar to the previous two species, but 30-45% of

the seeds had a second stage of dispersal, presumably by agoutis (Dasyprocta punctata).

Agoutis buried many of the seeds they removed in shallow surface caches. Thus,

secondary dispersal resulted in a rearrangement of the initial seed shadow.

In Chapter 5,1 further explore the general idea of dispersal directed

disproportionately to sites suitable for survival. The implication of directed dispersal is a

disproportionate effect on plant recruitment, but examples of directed dispersal are thought

to be rare and unusual (Howe 1986). Nevertheless, directed dispersal is well established in

a least three systems: scatterhoarding pine seeds by nutcrackers and jays, dispersal of

mistletoes by small birds, and dispersal of understory herbaceous plants by ants. I review

the literature and present evidence that directed dispersal may be more common than

previously believed, but is likely to be subtle in many instances. Possible examples include

dispersal by animals that have habitual perches or defecation sites, dispersal to nurse plants

in arid ecosystems, dispersal to perches in successional landscapes, and secondary

dispersal by ants, rodents, and dung beetles. Our understanding of directed dispersal is

limited by a lack of studies that integrate dispersal patterns and post-dispersal fate.

The main theme of this dissertation is that determining patterns of dispersal

generated by animals is a crucial research need. This type of data is difficult and time-

consuming to collect, but should yield great insight into the role of dispersal in ecosystem

function, as well as the evolution of plant-animal interactions. In addition, a greater






5

understanding of seed dispersal and subsequent post-dispersal fates may be useful in

conservation and restoration ecology.














CHAPTER 2
SEED DISPERSAL, SEED PREDATION, AND SEEDLING RECRUITMENT OF A
NEOTROPICAL MONTANE TREE



"No one has succeeded in measuring actual patterns of seed dispersal produced by different
bird species for any bird-dispersed plant species. Even if we could determine where birds
dropped all seeds, it would still be difficult to rank different bird species according to
dispersal quality because so little is known about seedling and sapling microhabitat
requirements." Nathaniel T. Wheelwright (1988:832)


Introduction

Seed dispersal determines the spatial arrangement and physical environment of

seeds from which the next cohort of seedlings is selected. The links between seed

dispersal and seedling recruitment, however, are poorly understood. Dispersal is the first

of a series of events, one or more of which may be important in limiting recruitment. For

vertebrate-dispersed plants, these stages include fruit removal, seed dissemination, post-

dispersal seed predation, potentially secondary dispersal, germination, and seedling

establishment. Previous studies have generally focused on only one or a few of these

stages (Herrera et al. 1994, Schupp and Fuentes 1995).

Despite many studies on fruit consumption and seed-handling by birds and

mammals, seed shadows (spatial patterns of dispersed seeds) generated by animals are

poorly known (Willson 1993). While concentrations of seeds under or near fruiting trees
have been noted and, indeed, are expected (Janzen 1970, Howe 1989, Willson 1993), the

pattern of seed distribution farther from fruiting trees is likely to be the most important part

of the seed shadow for plant fitness and population ecology (Portnoy and Willson 1993,

Schupp and Fuentes 1995). In particular, the tail of the distribution is often heterogeneous







as a result of disperser behavior (McDonnell and Stiles 1983, Hoppes 1988, Chavez-

Ramirez and Slack 1994, Julliot 1996). Fruit-eating vertebrates differ in foraging behavior

(Cruz 1981, Santana C. and Milligan 1984, Trainer and Will 1984, Moermond and

Denslow 1985), fruit removal rates (Howe and Vande Kerckhove 1980, Bronstein and

Hoffman 1987, Englund 1993), seed-handling techniques (Janzen 1983a, Levey 1987,

Corlett and Lucas 1990, Stiles and Rosselli 1993), and effects on seed germination

(Krefting and Roe 1949, Compton et al 1996, Traveset and WiUson 1997), but few

studies have compared seed shadows generated by different dispersers (but see Thomas et

al. 1988, Chavez-Ramirez and Slack 1994), or the consequences of dispersal pattern on

recruitment (but see Howe 1989, 1990a, Herrera et al. 1994, Schupp 1995, Schupp and

Fuentes 1995).

After dispersal, seed predation and seedling mortality are often extensive for

tropical trees (DeSteven and Putz 1984, Howe et al. 1985, Chapman 1989a, Hammond

1995, Cintra and Homrna 1997, Peres et al. 1997) and may be important in influencing the

spatial pattern of recruitment (ConneUll 1970, Janzen 1970, Harper 1977, Hubbell 1980,

Clark and Clark 1984, Becker et al. 1985, McCanny 1985, Howe 1989). In addition,

studies on forest dynamics have suggested that canopy gaps are often crucial recruitment

sites for tree seedlings (Hartshorn 1978, Denslow 1987, Swaine and Whitmore 1988,

Swaine 1996). Few studies, however, have examined seed predation as a link between

seed dispersal and seedling recruitment. In fact, aside from research on Virola in Panama

(Howe and Vande Kerckhove 1980, Howe et al. 1985, Howe 1990b, 1993a), Phillyrea in

Spain (Herrera et al. 1994, Jordano and Herrera 1995), and Ficus in Borneo (Laman 1995,

1996a, b), it is difficult to find studies that consider the various stages leading to

recruitment for any species of fleshy-fruited, vertebrate-dispersed plant. Furthermore,

some plant species typically considered to be bird or mammal-dispersed, may have a

second (or even third) stage of dispersal by an entirely different dispersal vector (Roberts

and Heithaus 1986, Clifford and Monteith 1989, Forget and Milleron 1991, Estrada et al.







1993, Levey and Byrne 1993, Nogales et al. 1996). To understand the relative importance

of seed dispersers and seed predators in forest dynamics, it is necessary to study each stage

leading to recruitment (and, ideally, to reproductive age).

The main factor that limits our understanding of the link between seed dispersers
and seedling recruitment is the difficulty of finding dispersed seeds (Harper 1977, Janzen

1983a, Schupp and Fuentes 1995). Thus, most studies on vertebrate-dispersed plants have

assessed the importance for plant fitness of different dispersers by the amount of fruit in the

diet and gut treatment of seeds and have not considered post-dispersal fate of seeds

(reviewed by Howe 1986, Stiles 1989, Jordano 1992, Willson 1992, Levey et al. 1994).

Similarly, studies on seed predation and seedling recruitment have relied on experimentally

dispersed seeds, usually without data on the actual patterns of seed rain (Price and Jenldkins

1986, Crawley 1992, Hulme 1993, Schupp 1995).

The objective of this study was to link patterns of seed dispersal and seed predation
with those of seedling recruitment, by determining the locations of naturally dispersed

seeds, and then following the fate of those seeds after dispersal. I studied a common tree,

Ocotea endresiana Mez (Lauraceae), in old growth montane forest in Costa Rica, to answer

the following questions: 1) Do the five main fruit consumers of 0. endresiana generate

different patterns of seed rain, and subsequently influence germination and the pattern of

seedling recruitment? 2) What proportion of the dispersed seeds are subsequently killed by

seed predators, and what animals are the most important seed predators? 3) Do

scatterhoarding rodents provide a second stage of dispersal? 4) What proportion of the

dispersed seeds germinate and establish as seedlings under different microhabitat

conditions? To answer these questions, I used a combination of natural and manipulative
experiments to assess the types of sites to which seeds are dispersed and how the habitat

characteristics and spatial arrangement of those sites relative to the parent trees influence

post-dispersal survival.







A secondary objective was to assess three hypothesized advantages of dispersal: 1)

escape from high mortality caused by distance- or density-dependent factors near

conspecifics (escape hypothesis); 2) colonization of rare, unpredictable, ephemeral sites,

such as treefall gaps (colonization hypothesis); and 3) directed dispersal to particular

microhabitats suitable for survival (directed dispersal hypothesis) (Howe and Smallwood

1982, Howe 1986, Willson 1992). A fourth advantage of dispersal, gene flow (Levin and

Kerster 1974, Hamrick and Nason 1996), was beyond the scope of this study. These

hypotheses are not mutually exclusive and can be assessed only if dispersal sites and post-

dispersal fates are known. Evidence for the escape hypothesis should include higher rates

of seed predation and/or seedling mortality closer to the parent tree, or where seeds and

seedlings occur in dense concentrations. The difference between colonization and directed

dispersal is whether seeds arrive and survive in specific habitats more often than expected

by chance (Howe and Smallwood 1982, Howe 1986, Murray 1986, 1988, Schupp et al.

1989, Willson 1992). In particular, if seeds are dispersed to sites with random

microhabitat characteristics then the colonization hypothesis is supported. If certain species
predictably disperse seeds to a nonrandom subset of the available microhabitats and post-

dispersal survival is predictably higher in those microhabitats than in random sites, then the

directed dispersal hypothesis is supported. Escape and colonization are generally thought

to be the main advantages for tropical trees dispersed by fruit-eating birds and mammals

(Howe 1986, Murray 1988).


Study Site
This study was conducted from May 1993 to July 1996 in the Monteverde Cloud

Forest Preserve (100 12'N, 84042W) in the Cordillera de Tilaran, northern Costa Rica.

This 10,000 ha preserve is adjacent to other protected lands encompassing one of the

largest areas of relatively unbroken forest in Costa Rica. The study area contains the full

complement of birds and mammals, including top predators, that have historically occurred







in the area (Young and McDonald in press), although certainly the abundances of some
species have changed (Fogden 1993). Other characteristics of the fauna are described by

Nadkamrni and Wheelwright (in press).

The study area was in relatively undisturbed lower montane rain forest (Hartshorn

1983) along the continental divide at 1600 m elevation. A 5 ha area, 500 m from the

beginning of the Valley Trail (Sendero El Valle), was mapped and marked into 10 x 10 m

quadrats with PVC tubing at every grid point. The site's vegetation is classified as leeward

cloud forest by Lawton and Dryer (1980). The canopy is 25-30 m tall and dominated by

several species of Lauraceae, Sapium oligoneuron (Euphobiaceae), Ficus crassiuscula

(Moraceae), Inga sp. (Leguminosae) and Pouteria viridis (Sapotaceae) (Lawton and Dryer

1980). The relatively dense understory (compared to lowland rainforest) is dominated by

Rubiaceae, Acanthaceae, Gesneriaceae, and Heliconiaceae. The vegetation of the area is

described in more detail by Lawton and Dryer (1980) and Nadkamrni et al. (1995).

Canopy gaps caused by falling trees and branches are common and may be an

important habitat for recruitment of some plants in the study area (Lawton and Putz 1988,

Lawton 1990). Although gaps may be formed at any time of year, most are formed during

the dry season when strong winds are more frequent (K.G. Murray, personal

communication). Gaps are also formed under standing dead trees, especially Sapium,

which shed large limbs for several years before the remainder of the trunk falls (D. Wenny,

personal observation, C. Guindon, personal communication). Approximately 5.3% of the

study area was in gaps >10 m2 with vegetation <2 m tall (sensu Brokaw 1982a), including

two Sapium gaps. Average canopy cover (measured with a spherical densiometer 1.2 m

above the ground) in gaps and gap edges was 90.7% ( 9.6, range 60-92, N = 78), while

closed canopy forest averaged 96.2% ( 4.3, range 91-100, N = 234) cover. These values

are similar to those at other Neotropical forest sites (Howe et al. 1985, Levey 1988b, Clark

1994).







Most woody plants at Monteverde are animal-dispersed, as is typical of Neotropical

forests in general and tropical montane forests in particular (Gentry 1982, Howe and

SmaUllwood 1982, Tanner 1982, Stiles 1985, Levey and Stiles 1994). Bird dispersal

predominates: 77% of the understory trees and shrubs and 63% of the canopy trees have

fruit morphology suggesting dispersal by birds (Stiles 1985). At least 70 resident and

migrant bird species feed on the fruits of over 150 species of trees and shrubs

(Wheelwright et al. 1984). Most of these birds regurgitate or defecate seeds intact

(Wheelwright et al. 1984, Murray 1988).

The average annual rainfall at 1520 m on the Pacific slope about 3 km from the

study site is approximately 2500 mm, with most of the precipitation occurring between

May and November. Actual rainfall in the study site was probably greater than 2500 mm,

but the seasonal pattern was similar (Nadkamrni and Wheelwright in press). Range gauges

underestimate the amount of precipitation from mist and cloud interception, which

contribute up to 50% of the precipitation in some Neotropical montane forests (Cavelier

1996). Temperatures recorded at the study site during this project ranged from 15 to

220C.



Study Species
The Lauraceae is an important family in Neotropical forests in terms of species

richness, a food resource for birds, and economic value (Wheelwright 1983, 1985a, 1991,

Burger and van der Werff 1990, Gentry 1990, Martfnez-Ramos and Soto-Castro 1993,

Guindon 1996). Members of the Lauraceae are also important dietary components for

frugivorous birds in Africa, Southeast Asia, and Australia (Crome 1975, Snow 1981, Sun

et al. 1997). Ocotea endresiana Mez (listed as 0. austinii in Wheelwright et al. 1984,

Wheelwright 1985a, 1986) is a common canopy tree in montane forests in central and

northwestern Costa Rica from 1100-2300 m (Burger and van der Werff 1990). In the

Monteverde area it occurs from 1550-1700 m along the continental divide. Twenty to 30







other species of Lauraceae occur in the same area (Wheelwright 1985a, 1986, Haber

1991). Ocotea endresiana begins flowering in August, the mid-rainy season, and is

pollinated by small flies and other insects. Fruits ripen the following May and June, the

early rainy season. Ripe fruits (18 x 9 mm, 1.2 g) have blue-black skin, lipid-rich pulp,

and are held in a shallow reddish receptacle, typical of the Lauraceae genera Ocotea and

Nectandra (Wheelwright et al. 1984, Burger and van der Werff 1990). Most of the volume

of the fruit is a single seed (15 x 8 mm, 0.75 g, fresh weight) composed of a small embryo

and two large cotyledons surrounded by a thin (0.3 mm) seed coat. Although small

compared to some other Lauraceae species, 0. endresiana fruits and seeds are among the

largest at Monteverde (Wheelwright et al. 1984). Thirty-eight 0. endresiana trees were in

the study area; most data in this study are from 21 trees in the core 5 ha study site (Fig. 2-

1). Fruits and seeds for some experiments were collected from trees on the periphery of

the main study area. Large fruit crops (2110 1495 fruits/tree) were produced in 1993 and

1995, but not in 1994. Each 0. endresiana tree was an average of 36.5 m ( 12.1) from

the three closest conspecific adults.

The most common avian visitors to fruiting Lauraceae include the largest and most
frugivorous species at Monteverde (Wheelwright et al. 1984). Ocotea endresiana fruits are

eaten primarily by five species of birds: emerald toucanet (Ramphastidae: Aulacorhynchus

prasinus), resplendent quetzal (Trogonidae: Pharomachrus nwmocinno), three-wattled bellbird

(Cotingidae: Procnias tricarunculata), mountain robin (Turdidae: Turdus plebejus) and

black guan (Cracidae: Chamaepetes unicolor), all of which breed in the study site during

the fruiting season. The first four species swallow fruits intact and regurgitate seeds in

viable condition. Seeds in this size range are generally regurgitated 15-30 minutes after
ingestion (Wheelwright 1991). Guan defecate seeds in viable condition (Wheelwright

1991). These bird species (especially quetzals) typically remain in a fruiting tree after

eating several fruits, and often regurgitate seeds under the same tree or nearby

(Wheelwright 1983, 1991). Seed processing times for guans are unknown, but they


























Figure 2-1. Map of the study site showing locations of fruiting Ocotea endresiana trees
(dark circles with numbers) and bellbird song perches (X). Thick lines indicate streams
and thin lines are trails. Grid lines are 50 m apart. North is at the top. Data on seed
locations for each of the trees numbered on this map are presented in the Appendix.







14






















co














CC3




A N
cr)S








generally leave a fruiting tree before defecating the seeds from that foraging bout (D.

Wenny, personal observation). Several other bird species (Wheelwright et al. 1984), as

well as spider monkeys (Ateles geoffroyi), occasionally eat 0. endresiana fruits and

probably disperse viable seeds (e.g., Chapman 1989a).


Methods
A combination of observational and experimental data was collected to determine

the probability of seedling establishment and one year survival for seeds dispersed by

birds. In particular, the focus of this project was to find the locations of seeds dispersed

naturally and to compare the post-dispersal fate of those seeds with experimentally placed

seeds to determine the influence of dispersers on seedling recruitment. The locations of

seeds naturally dispersed by birds were classified in two categories according to their

position relative to the parental tree crown to reflect the prevailing assumption that seed

deposition under parental crowns leads to a low probability of survival (Janzen 1970,

Howe et al. 1985, Hulme 1997). Sites were classified as non-dispersed if directly under

the crown of a fruiting 0. endresiana tree or as dispersed if not under such a tree. Even

though all dispersed and non-dispersed seeds in this study had been regurgitated or

defecated by birds (and therefore "dispersed" sensu Janzen 1983a), it is generally believed

that seeds deposited under the parent trees have very little chance of survival (e.g., Janzen

1970, Howe et al. 1985). Therefore, the term "dispersed" in this paper always refers to

seeds that are not deposited directly under the crown of a fruiting conspecific, and "non-

dispersed" refers to seeds that are regurgitated or defecated by birds directly under the

crown of a fruiting tree. For convenience, I will refer to sites of dispersed seeds as

dispersed sites and of non-dispersed seeds as non-dispersed sites, even though the sites

themselves were not capable of being dispersed.







Seed Dispersal

Seeds were located by following birds until they dropped, regurgitated, or

defecated seeds, and by systematic ground searches. These ground searches were started

at the base of a fruiting tree and proceeded along 10 m wide transects to 50-60 m from the

trunk. It was impossible to search the entire site with equal intensity, but an effort was

made to cover the entire site at least once every two weeks, so that over the course of the

two-month season of fruit ripening each 10 x 10 m plot was checked at least four times.

Bird observations began whenever one of the five major consumers of 0. endresiana fruits

was located, and continued as long as the bird was visible or until the bird regurgitated or

defecated seeds. A total of 193 hr was spent following birds.

In addition to these dispersed and non-dispersed seeds, other seeds were

experimentally placed at randomly selected sites to determine if birds deposited seeds in

sites with specific or random characteristics. Random numbers generated on a hand

calculator were used as coordinates of sites. The post-dispersal fate of seeds at these

random locations was compared to the fate of seeds at the dispersed and non-dispersed

locations.

Microhabitat characteristics. For all sites and seeds I measured seed characteristics

and microhabitat variables I thought might influence seed predation, germination, or

seedling survival Seed length and width were measured with dial calipers, and seed mass

was measured with a 5 or 10 g spring balance. Canopy cover was estimated with a

spherical densiometer (Lemmon 1957). Leaf litter was the number of leaves pierced by a

metal stake thrust into the soil at the site. Vegetation density was the number of stems

within a 50 cm radius of the site. The distances to the nearest woody stem > 1 m in height,

tree > 10 cm DBH (diameter at breast height), trunk of fruiting Ocotea tree, and fallen log

were measured with a fiberglass measuring tape. These variables were selected based on

their demonstrated importance in previous studies. Seed size may influence the probability

of seed predation (Price and Jenkins 1986, Hulme 1993) or seedling size (Howe and








Richter 1982). Canopy cover (light availability) is known to be an important factor for

germination and tropical seedling growth (Howe et al. 1985, Mulkey et al. 1996, Swaine

1996). Vegetation density and distance to objects may influence rodent activity and seed

predation (Smythe 1978, Kiltie 1981, Kitchings and Levey 1981). Finally, leaf litter may

influence seed predation or germination (Schupp 1988a, Molofsky and Augspurger 1992,

Myster and Pickett 1993).


Post-dispersal Seed Predation

At each dispersed, non-dispersed, and random site, a marked seed was used to

assess rate of seed predation, to identify seed predators, and to determine if secondary

dispersal occurred. For this treatment I used seeds that were regurgitated by birds and

collected under fruiting trees adjacent to the study site. Seeds were marked by gluing 50-

75 cm of unwaxed dental floss to the seed, and tying about 50 cm of flagging tape to the

distal end of the floss. Because the glue held best on seeds with a dry seed coat, seeds

were taken inside a lab room and allowed to dry for 1-3 h before gluing. Each marked seed

was placed at a site the next morning. To determine if presence of dental floss and flagging

tape influenced seed removal, I conducted a pilot study in May 1993. Fifty marked seeds

and 50 unmarked seeds were placed singly at random locations and censused one week

later. Removal of marked and unmarked seeds was not significantly different (49 and 47

seeds removed, respectively; X2 = 0.047, df = 1, P > 0.10). Thus, the marking of seeds

was assumed to have no effect on seed removal.

Note that these marked seeds were placed next to the cages protecting the original

seeds deposited by birds. The cages may have served as cues for visually-searching seed

predators, but I believe that is unlikely for two reasons. First, at the beginning of the study
in 1993, cages were novel items and would not have been associated with seeds by the

seed predators. Second, because most marked seeds were removed, but the cages were left

at the sites for several months, most cages were not good indicators of seed availability.








Olfactory cues left by marking and handling the seeds are another possible confounding

factor, but the frequent rains probably diminished these (Whelan et al. 1994).

In 1993, all marked seeds were censused once each week for three weeks and then

at weeks 5 and 10. Because most seeds were removed during the first week after dispersal

in 1993, marked seeds in 1995 were checked more often: on days 1,3, and 7, and once

each week afterwards until week 10. If a marked seed was removed, the surrounding area

was searched to find the flagging tape-dental floss assembly. The end of the floss where

the seed had been attached was examined to determine the fate of the seed. If a seed was

entirely removed or if a piece of the seed coat remained attached to the floss, the seed was

classified as killed by seed predators because captive Peromyscus treated seeds with dental

floss in that manner when they consumed 0. endresiana seeds (D. Wenny, unpublished

data). The distance from the dispersal site to the predation site was measured and each site

of a removed seed was classified in the following categories: in burrow; in, under, or near

a fallen log (> 10 cm diameter); at the base of a large tree; in dense vegetation (defined as at

least 50% cover of plants less than 50 cm tall within 50 cm radius of site, assessed

visually); under or beside a clump of fallen branches, or on the leaf litter. If a marked seed

was removed but not eaten, it was left in the new location and included in subsequent

censuses.

Distance effect Two other experiments (one in 1995 and one in 1996) were
conducted with marked seeds to determine the effect of distance on seed removal, and

whether most seed removal occurs during the day or at night. In 1995 one seed was placed

at the base of each of 12 fruiting trees and at 20,40, 60, and 80 m from the base of the

tree. The orientation of each transect was carefully selected to avoid approaching within 80

m of other fruiting 0. endresiana trees. Seeds at these sites were censused on the same

schedule as the other marked sites in 1995. In 1996 a similar experiment was conducted,

except the seeds were 5 and 40 m from each of 22 trees. Shorter and fewer distances were

used so that more trees could be included, and because data from the previous year, as well








as other studies (e.g., Howe et al. 1985) indicated that the distance effect can be detected

beyond 20 m from conspecific adult In this case, the compass directions were randomly

selected although some directions were discarded if the 40 m treatment was not at least 40

m from all fruiting conspecifics. This experiment was run twice, once in the early fruiting

season (late May) and once two weeks later during the peak fruiting season (mid-June).

During the second of these trials, all sites were checked at dusk and dawn for two

consecutive days to determine whether most seed removal occurred during the day or night

Removal during the day likely could be attributed to the diurnal agouti, while removal at

night likely would be from nocturnal species, such as Peromyscus or other small rodents.

Exclosures. To determine the amount of seed predation attributable to large or small

mammals, 38 sets of exclosures were established in early 1995. Each set had three

treatments, each 1 m2: (1) "no rodents" exclosures made of 1 cm2 galvanized wire mesh

0.9 m tall; (2) "small rodents only" exclosures made of chicken wire mesh with 6 cm holes,

0.9 or 1.35 m tall; and (3) "all rodents" control plots with only wooden stakes marking the

comers. The bottom edges of the wire exclosures were buried 5 cm below ground and

held with two or three 25 cm metal stakes on each side. The comers were supported by

1.25 m lengths of 10 mm thick metal stakes. The tops were open to allow normal

accumulation of fallen leaves. Each set was located where convenient (avoiding trees and

fallen logs) within randomly located 10 x 10 m quadrats in the study site. One regurgitated

0. endresiana seed was placed in each exclosure and control plot in late June 1995. Each

seed was checked after 2, 4, and 8 d, 4 mo, and 1 yr.


Germination and Seedling Survival

The original regurgitated or defecated seed at each dispersed and non-dispersed site,

and each seed placed at a randomly located site was protected by a 4 x 4 x 2 cm cage made

of 3 mm galvinized wire mesh held in place by two metal stakes. Caged seeds were used

to determine germination rates and insect predation rates in the absence of mammalian seed








predators. Each site was checked weekly for at least 12 weeks and monthly thereafter until

June of the following year. Germination was defined as the splitting of the seed coat and

spreading of the cotyledons. Typically, the radicle had emerged by the next census after

germination, and a week later the stem was visible. As each seed germinated and the shoot

began to grow, the cage was removed to allow normal seedling growth. The seedling

location was marked with one of the stakes from the cage. Causes of seedling mortality

were classified as mammal, insect, fungal pathogen, physical, or unknown. Mammals

either ate the seed and left the damaged shoot behind (seed predators) or removed the entire

shoot (herbivores). Some seeds that appeared to have germinated were killed by beetle

larvae (Heilipus sp., Curculionidae) developing in the seed. Insect-killed seeds frequently

developed a root but never had a shoot >2 cm tall. Seedlings killed by fungal pathogens

were characterized by a wilted and discolored shoot above about 4 cm (Augspurger 1990).

Physical damage consisted of trampling by peccaries and other large mammals, or damage

from falling trees, branches, or large leaves (Clark and Clark 1991). Unknown causes of

mortality included cases that fit more than one category where the sequence of events could

not be determined, and cases that did not clearly fall into any category. A seed was

considered alive if the seed remained firm, even if the shoot had been eaten or otherwise

damaged. Such seeds resprouted repeatedly (D. Wenny, personal observation). Seeds

from 1995 that had not germinated or resprouted after one year were cut open and classified

as insect-killed if filled with frass, or viable if the embryo and cotyledons were not

discolored or mealy in appearance.

Germination trials. To determine the effect on 0. endresiana germination of

ingestion by birds and seed burial (which I initially thought would apply to seeds taken by

mammals), trials were conducted in a greenhouse constructed of nylon window screen over

a 12 x 6 m wooden frame on a concrete foundation set 25 cm into the ground. The roof

was corrugated plastic. Seeds were planted individually in cardboard milk cartons and cut

plastic bottles of various sizes. All containers were washed with hot soapy water and dried








in the sun before use. Soil was collected from a nearby secondary forest and mixed 3:1

with sand.

Three seed treatments were compared: (1) naturally regurgitated seeds collected

under fruiting trees or along the trail to the study site; (2) seeds removed by hand from ripe

fruits collected under fruiting trees; and (3) entire ripe fruits, either fallen or dropped.

Seeds or fruits that were misshapen, diseased, or had signs of insect infestation were not

used. Each treatment had 35 seeds or fruits, placed on top of the soil. Additionally, 10

seeds were planted 5 cm deep to determine if seeds could germinate after burial by

scatterhoarding rodents. Containers were watered as necessary to keep the soil moist.

Seeds were planted in June 1995 and checked weekly until early November 1995 when

seeds that had not germinated were cut open and assessed for viability as described above.

Seedling and sapling plots. To determine if 0. endresiana seedlings and saplings

are more likely to recruit under or away from conspecifics, seedlings and saplings (up to 3

m tall) were measured and mapped in paired 10 x 10 m plots. For each of 10 trees, one

plot was located near the tree with approximately half of the plot directly under the crown.

The second plot was located 10 to 20 m from the edge of the first plot and at least 15 m

from the crown edge. No plots were located in recent (< 5 yr) canopy gaps or along

streams.


Statistical Analyses

The influence of the microhabitat variables on removal of marked seeds at 1 day

(1995 only) and 2 weeks, and on germination, seedling establishment and 1-year survival

of caged seeds were examined with multiple logistic regression using SAS JMP (SAS

Institute 1989) For each of the seven binomial (live or dead) response variables mentioned

above, the model was run first with all predictor variables, and then with and without each

predictor variable to determine if the deletion of a given variable had a significant effect on

the amount of variation explained. This deletion procedure was repeated until only








significant predictors were retained (Trexler and Travis 1993). Models were selected

manually to avoid the problems of automatic stepwise procedures (James and McCulloch

1990). The predictor variables included three measures of seed size (length, width, and

mass), eight microhabitat characteristics (leaf litter, canopy cover, number of stems, and

distances to nearest caged seed, herbaceous stem, woody stem, 10 cm tree, parent tree

trunk, and fallen log), date of dispersal (Julian date), and tree number. Only single terms

were included in the models as lack-of-fit tests indicated that the single term models were

adequate and thus interaction terms were not required (SAS Institute 1989).

The relationships of the microhabitat variables within and among the dispersed,
non-dispersed, and random sites were examined with principal components analysis

(PROC FACTOR) from the SAS statistical package (SAS Institute 1988). The average

loadings for the three types of sites were compared with one-way analysis of variance from

Super Anova (Abacus Concepts 1989). Type III sums of squares were used to

compensate for the unequal sample sizes (Shaw and Mitchell-Olds 1993). Removal rates

were compared among treatments with survival analysis and Gehan-Wilcoxon tests from

SAS JMP (SAS Institute 1989) The Gehan-Wilcoxon test places greater weight on early

events than later events and was deemed appropriate for this study because most seeds

were removed during the first week (Pyke and Thompson 1986).

Parametric tests were used unless the data violated the assumptions of normality

and equal variance, in which case nonparametric procedures were used. Data in the form

of proportions were arcsin square-root transformed before analysis. Where multiple

comparisons on a data set were involved, the alpha value was adjusted according to the

number of comparisons planned (Bonferroni technique, Holm 1979). Data from the two

years were very similar and were combined for analyses in which sample sizes would have

been low otherwise. Throughout this paper mean values are followed by 1 SD.









Seed Dispersal

In 1993,284 seeds regurgitated or defecated by birds were found: 155 dispersed

and 129 non-dispersed, while in 1995, 234 dispersed and 171 non-dispersed seeds were

found. Ninety-five and 100 random sites were established in 1993 and 1995, respectively.

Twenty-four percent of the 1993 seeds and 28% of the 1995 seeds were found by

following or observing birds (N = 184), and the rest (N = 505) by searching the ground.

The dispersed seeds were most common within 10m of the crown edge (1993:

55%, 1995: 45%), but some seeds were as far as 70 m away. In rare cases (2.8%),

dispersed seeds were within 5 m of an 0. endresiana trunk (Fig. 2-2). Despite their close

proximity to the parent, they were classified as dispersed because they were not directly

under the parent crown, as a result of asymmetrical crown geometry or a tilted trunk (or

both). Random sites were more evenly distributed than dispersed or non-dispersed sites,

but showed a peak at 20-25 m in both years (Fig. 2-2). With dispersed and non-dispersed

sites combined, the seed distribution generated by birds in both years is best described by a

logarithmic function of distance from a fruiting conspecific (Fig. 2-2). Most seeds landed

within 20 m of the fruiting trees, but the tail of the seed distribution curve beyond 20 m

accounted for 18% of the sites in 1993 and 21% in 1995. Seed rain was highly variable

when averaged among trees and years (Fig. 2-3). In particular, note that seed distributions

showed secondary peaks at 45-50 and 60-65 m in both years (Fig. 2-2). These peaks

corresponded to habitual song perches used by bellbirds (see below) which were located

40-70 mi from fruiting trees in the years this study was conducted (Fig. 2-1). More

information on the locations of seeds is presented in the Appendix.

Microhabitat characteristics. The locations of dispersed and non-dispersed seeds,
and seeds at random sites differed significantly in both years with regard to canopy cover,

distance to parent, and number of stems (Table 2-1). Dispersed seed sites had lower

average canopy cover than non-dispersed or random sites in both years (post-hoc SNK












80

1993
60 D Dispersed (N=155)
wu Non-dispersed (N=129)
P,-
; *El Random (N=95)
U.
0 40-

m
0 20
z


0 "
z 0-:31 1.11 .1 nnn-nnn nn --

5 10 15 20 25 30 35 40 45 50 55 60 65 70


100
1995
80- E Dispersed (N=234)
U 1] Non-dispersed (N=171)
I- [ Random (N=100)
Uo 60
U.
0
w 40
m

Z 20-

o" 1E 0000ll.000.,n
5 10 15 20 25 30 35 40 45 50 55 60 65 70
DISTANCE FROM PARENT (m)




Figure 2-2. The distribution of seeds naturally regurgitated or defecated by birds directly
under the crown of a fruiting tree (non-dispersed), and away from fruiting trees
(dispersed); and the distribution of random sites at different distance categories from trunks
of fruiting trees in 1993 (above) and 1995 (below). Values on the x-axis represent the
maximum distance for each category (5 = 0-5 m). The seed shadow is truncated at 70 m
because the abundance of Ocotea endresiana trees in the study site make sites > 70 very
rare.
























15 '
Ul
LU
w
w
I-
M: 10
LU,
w
CL

wU 5-



0
0 5 10 152025303540455055606570

DISTANCE FROM CROWN EDGE (m)











Figure 2-3. The average number (+ 1 SD) of naturally regurgitated or defecated seeds per
tree (N = 21) for both years combined at different distances from the closest fruiting tree.
Distance class 0 includes all non-dispersed seeds directly below parental crowns. The
other distance categories (5 = 0-5 m, etc.) are measured from the edge of the crown of the
closest fruiting Ocotea endresiana tree. The values for each distance category represents the
number of seeds at a particular distance, averaged among all trees, for the entire study site.









Table 2-1. Summary of one-way analysis of variance tests for each habitat variable
compared among sites of dispersed (D), non-dispersed (N) seeds, and randomly located
sites (R) in 1993 and 1995. ANOVA tests had a Bonferroni-adjusted alpha value = 0.006.
Mean values are shown for each treatment for each variable. Standard deviations are in
parentheses below each mean. Leaf Ulitter was the number of leaves pierced by a metal stake
thrust into the soil at the site. Canopy cover was estimated with a spherical densiometer.
Vegetation density was estimated as the number of stems within a 50 cm radius of the site.
The last four variables are distances to closest stem, closest woody stem (> 1 m height),
closest tree (> 10 cm DBH), and closest log (> 10 cm). Within each year and variable for
which a significant difference among treatments was detected (indicated with after the F-
value), results of post-hoc tests (Student-Newman-Kuels tests) are indicated with a letter
after the mean. Means followed by different letters are significantly different (P < 0.05).

1993 1995
Variable D N R F2,375 D N R F2,502

leaf litter (#) 2.1 1.9 2.4 2.70 2.3 2.3 2.1 0.91
(1.3) (1.4) (1.4) (1.2) (1.2) (1.2)

canopy 94.3a 95.6b 95.8b 13.40* 94.5a 96.7b 96.5b 12.40*
cover (%) (3.6) (1.5) (1.7) (6.5) (1.6) (3.1)

vegetation 38.9a 32.1b 40.7a 7.98* 26.8a 19.9b 21.6b 15.50*
density (17.3) (16.2) (20.4) (15.1) (10.2) (10.9)

closest 10.8 12.2 10.9 2.22 11.6a 14.3b 13.3ab 5.46*
stem (cm) (5.5) (6.2) (6.2) (6.4) (7.4) (12.0)
woody 61.9 58.9 66.7 0.90 88.6 75.5 75.9 2.85
stem (cm) (45.7) (39.7) (42.3) (62.5) (60.4) (55.8)
closest 1.50 1.68 1.66 1.53 1.88 1.92 1.84 0.18
tree (min) (0.87) (0.97) (1.00) (1.08) (1.04) (1.08)
closest 1.47 1.30 1.28 0.80 1.97 1.78 1.83 0.77
log(m) (1.31) (1.37) (1.25) (1.74) (1.45) (1.66)
*P<0.006







tests; PF s < 0.01), while the non-dispersed and random sites were similar (P > 0.05). In

1993, dispersed and random sites had similarly (P > 0.05) high numbers of stems

compared to non-dispersed sites (P < 0.01). In 1995, dispersed sites had more stems than

both the other sites (P < 0.01), while random and non-dispersed sites had similar numbers

of stems (P > 0.05). The only other variable that differed among the treatments was

distance to closest stem in 1995. Dispersed sites were closer to stems than non-dispersed

sites (P < 0.01), while random sites did not differ from the other two treatments (P >

0.05).

In a principal component analysis of the 1995 data, the first two components

explained 63.8% of total variance, while for 1993 the first two components accounted for

45.5% of total variance (Table 2-2). The relatively low amount of variance explained is

due, in part, to the inclusion of random sites (Jackson 1993) which appeared as a spherical

cloud of points centered on the origin (Fig. 2-4). Removal on random sites, however,

increases the total variance explained by the first two components by only 2-3%. The

pattern was similar in both years: the first multivariate axis (PC 1) was characterized by

high negative loading of canopy cover, and high positive loadings of leaf litter depth and

closest 10 cm DBH tree (Table 2-2). The second axis (PC 2) was characterized by positive

loading of distance to closest stem in both years and negative loading of vegetation density

in 1995 but positive loading in 1993 (Table 2-2). The average loadings for PC 1 differed
significantly among the dispersed, non-dispersed, and random sites in 1993 (F2,369 =

45.99, P < 0.0001), and 1995 (F2,502 = 31.61, P < 0.0001). Each type of site differed

from the others (Fisher's LSD tests, all P's < 0.002). Loadings for PC 2 differed only
between dispersed and non-dispersed sites in 1995 (F2,502 = 6.07, P = 0.0025). This

analysis shows that site characteristics of dispersed and non-dispersed seeds were

nonrandom.

When the three treatments (dispersed, non-dispersed, and random) are identified on

a plot of the first two principal components, they showed broad overlap, except for some













1993
2 o
1 D

*+
S0 00 a 0 DISPERSED
0 1 + g + NON-DISPERSED
C3 1. + RANDOM
*2- ,1+1 i D O
+ ~++ +M9+ a a+
-3- 13 ," +

"4 I II I
-4 -2 0 2 4 6

PC 1


4-
1995

,r
2- al a DD

+ DISPERSED
0+ aNON-DISPERSED
S -2 a + + RANDOM
00 D O 3
0 0


-6 I I
-1 0 1 2 3
PC 1



Figure 2-4. Plot of sites with dispersed (open squares), non-dispersed (solid circles), and
randomly located seeds (crosses) from 1993 (above) and 1995 (below) on the first two
axes determined by principal components analysis of habitat variables. Axis 1 is primarily
a function of canopy cover leaf litter, and closest 10 cm DBH tree, and Axis 2 is primarily a
function of vegetation density and closest stem (see Table 2-2).










Table 2-2. Loadings (eigenvectors) of each habitat variable on the first two components
determined by principal component analysis of the correlation matrix. The proportion of
total variance explained by each component is listed in the last row.


1993


PC1


1995


PC2


PC1


PC2


leaf litter
canopy cover
#stems
closest stem
woody stem
10 cm tree
closest log


variation


0.5427 -0.2248 0.8885 0.2718
-0.5973 0.4796 -0.8117 0.2453
-0.4831 0.5784 0.1359 -0.8443
0.5856 0.5393 -0.2315 0.7122
0.3973 0.5065 0.2958 -0.3067
0.5034 -0.2890 0.7280 -0.0947
0.2300 0.3720 0.8854 0.2925


24.69% 20.76% 41.74%


22.04%







dispersed sites that have higher values for the first component These correspond to sites

with low canopy cover, relatively far from the parent trees (Fig. 2-4). Seeds in the sites

with values > 2.0 in 1993 and > 1.0 in 1995 for PC 1 were all dispersed by bellbirds into

large gaps surrounding song perches. Additionally, two of the 1993 seeds were dispersed

by Black Guans onto logs in another smaller gap. Bellbird-dispersed seeds, seeds

dispersed by other species, and random sites had significantly different PC 1 loadings in
1993 (F2,161 = 7.9, P = 0.006), and 1995 (F2,212 = 9.1, P < 0.001). Bellbird sites had

higher factor loadings than other species' sites and random sites (P's < 0.002), but sites of

the other species and random sites did not differ (P = 0.2).

Combining the seeds from both years, and using only seeds for which the disperser

was known, the average canopy cover at sites of seeds dispersed by bellbirds was

significantly lower than of sites of seeds dispersed by all the other species (89% and 96%

respectively; Kruskal-Wallis test: X2 = 42.7, df = 1, P < 0.001; Fig. 2-5). Similarly, the

average distance from the closest parent tree of bellbird sites was greater than any of the

other four species, (Kruskal-Wallis X2 = 66.06, df = 1, P < 0.001), but the difficulty in

following the birds (especially robins which tended to fly above the canopy) biases the

results in favor of bellbirds. Nevertheless, the data show that bellbirds tend not to drop

many seeds near parent trees (Fig. 2-6). Considering that approximately 5.3% of the study

area was in gaps (see Study Site), overall seed arrival in gaps (12%) occurred more often

than expected by chance (X2 = 36.45, df = 1, P < 0.001), whereas close to the expected

number of random sites (4.1%) was in gaps (X2 = 0.557, df = 1 P > 0.05, see Fig. 2-4).

Furthermore, bellbirds dispersed significantly more seeds to gaps than did the other species

(X2 = 39.3, df = 1, P < 0.001). Note that bellbirds dispersed seeds to only 2 of 29 gaps in

the study area.




















100



95


0
o go
a.
0
z
85-



80"
BELLBIRD GUAN QUETZAL ROBIN TOUCANET

DISPERSER SPECIES

Figure 2-5. Mean ( 1 SD) canopy cover (%) of sites of seed dispersed by bellbirds,
guans, quetzals, robins, and toucanets for both years combined. Sample size for each
species shown at the bottom of each bar.




















60o
0 TOUCANET
50- E RBCN
(0) -QUETZAL
' 40- GUJAN
w BELLBIRD
Co
0 30-
U.
to 20-

z
10


0 10 20 30 40 50 60 70

DISTANCE FROM CROWN EDGE (m)


Figure 2-6. The number of seeds dispersed at different distances from parental trunks by
bellbirds, guans, quetzals, robins, and toucanets for both years combined (N = 184).







Post-dispersal Seed Predation

More than 50% of marked seeds were removed within one week in both years (Fig.

2-7). In 1993, seeds at dispersed, non-dispersed, and random sites were removed at

similar rates with dispersed seeds showing a nonsignificant trend for slower removal than

seeds at non-dispersed or random sites (Wilcoxon X2 = 5.36, df = 2, P = 0.068). In

1995, dispersed seeds were removed more slowly than those at non-dispersed and random

sites (Wilcoxon X2 = 9.09, df = 2, P = 0.011). Ultimately, only 2 of 923 marked seeds

(one dispersed and one random site, both from 1995, and both ; 30 m from conspecific

trees) in the entire study survived to June 1996 (and germinated), for an overall predation

rate of 99.7%.

I found 75% of marked seeds after removal in 1993, and 94% in 1995. In all

cases, the seed was killed, and in most cases (96%) entirely consumed. Few seeds were

eaten (13%) but not moved. The general pattern was that seeds were taken a short distance

(range 0-21 m, N = 761) to a site with cover, presumably where the seed could be eaten in

safety. On average, removed seeds (not counting seeds eaten but not moved) were taken

2.5 m ( 3.8, N = 286) in 1993 and 2.1 m ( 2.9, N = 475) in 1995. About 50% of those

removed were found on the leaf litter, but some were found in small burrows (< 6 cm

diameter), in or under fallen logs, or in crevices at the bases of large trees (Fig. 2-8). Most

seeds taken to burrows, trees, and logs were clearly taken by small animals, probably

rodents, because the small diameter of the opening would exclude larger species such as

agoutis (Dasyprocta punctata). I found no evidence of scatterhoarding or secondary

dispersal.

Significant predictors of survival of marked seeds to 2 weeks in 1993 included

canopy cover, leaf litter depth, and dispersal date (Table 2-3). Seeds in sites with more

open canopy cover, deeper leaf litter, and later dispersal date were more likely to survive.

In 1995, 1-d survival was significantly predicted by greater seed mass, earlier dispersal

date, and the particular parent tree. At 2 wk, seeds farther from the parent tree, dispersed

















-0-
U----- ----


1993

Dispersed (N=148)
Non-dispersed (N=120)
Random (N=95)


0 20 40 60 80


1
--U-


100


1995
Dispersed (N=234)
Non-dispersed (N=171)
Random (N=100)


0 20 40 60


DAYS AFTER DISPERSAL







Figure 2-7. Removal of marked seeds from sites of dispersed (open squares), non-
dispersed (solid squares), and randomly located (crosses) seeds in 1993 (above) and 1995
(below). Sample sizes are shown in parentheses.



















BASE OF TREE

BURROW

LOG

FALLEN BRANCHES

DENSE VEGETATION

LEAF LITTER


E3 1993(N=286)
N 1995(N=475)


I p I I *


PERCENTAGE OF SITES
















Figure 2-8. Percentage of marked seeds found in different locations after removal by
animals. See text (Methods: post-dispersal seed predation) for definitions of categories.


I^^










Table 2-3. Results of logistic regressions of post-dispersal survival of marked seeds
against habitat variables. Only significant effects are listed.

Response r2 -21og likelihood predictorsa
(X2)

1993
2 wk survival 0.311 54.19*** + leaf litter***
canopy cover*
dispersal date***
1995
1 day survival 0.23 133.3*** seed mass**
+ dispersal date***
tree number*

2 wk survival 0.31 144.5*** + distance to parent***
dispersal date***
tree number*
*P < 0.05, **P < 0.01, ***P < 0.001.
sign in front of each predictor variable indicates positive or negative correlation with the
response variable.







later in the season, and at certain parent trees were more likely to survive. Logistic

regression models of survival past 2 wk were not significant because so few seeds

survived that long.

Distance effect Seed removal from transects radiating from fruiting 0. endresiana
trees was slower for seeds farther from the parent trees in both 1995 and 1996. The 1995

seed removal experiment, which ran for 10 weeks, showed that the effect of distance on

seed removal declined over time and that seed predation eventually approached 100% at all

distances (Fig. 2-9a). Only 1 seed (at 80 m) survived 10 wk. Removal of seeds was
significantly faster for seeds within 40m of parents than for seeds placed 60 or 80 m away

(Wilcoxon X2 = 4.5, df = 1, P = 0.03). The 1996 experiment (Fig. 2-9b) showed that

seed removal was significantly faster for seeds experimentally dispersed during the peak of

fruiting than before the fruiting peak, but only for seeds 5 m (under the crown) from

parents (X2 = 6.87, df =1, P = 0.009) and not for seeds 40 m away (X2 = 0.21, df =1, P

= 0.64). Removal rates did not differ between seeds placed 5 and 40 m from parents at
either time in the fruiting season (early: X2 = 0.65, df = 1, P = 0.42; middle: X2 = 3.17, df

= 1, P = 0.20).
The 1996 sample of 44 seeds was censused at dusk (18:00-18:30) and dawn

(05:30-06:00) for two consecutive days. Five seeds had been removed by the first dusk

census, two of which clearly were taken by small rodents into small holes in fallen logs.
By the next dawn, 27 more seeds had been removed. One seed was taken during that day

and four more by the next dawn. Thus, most seed removal (73%) took place at night (X2

= 25.9, df = 1, P < 0.001), probably by small rodents.

Exclosures. Removal rates differed significantly among the three treatments; seeds

accessible to all rodents were taken faster than the other two treatments (Wilcoxon X2 =
16.7, df = 2, P = 0.007). After 8 d, three seeds had been removed from the small mesh

exclosures, which were designed to exclude all rodents (Fig. 2-10). Two of these
exclosures had burrows inside that were probably dug during the three months between













100-
a

at 80-
80 Om
0 20 m
z
Z 60 o 40 m
60 m
Lu 80 m
re 40

a
W 20-
w
o o
tut

0 14 28 42 56 70



100 .
b

80- --0-- 5m EARLY
08
-0- 40m EARLY

(.-^ --- 5m PEAK
z
z 60 40m PEAK
z
w
| 40

a
Wl 20


0
w ^-


0 2 4 6 8 10 12 14

DAYS






Figure 2-9. (A) Removal of marked seeds experimentally dispersed 0, 20,40, 60, and 80
m from trunks of the fruiting Ocotea endresiana trees starting in June 1995. N = 12 at each
distance. (B) Removal of marked seeds at 5 and 40 m from parent trees in the early (late
May) and at the peak (mid-June) of fruiting season, 1996. N = 24 for each category. All
seeds at 0 and 5 m were under the parental crowns.



















100 NORODENTS


^ 80-


E 60 SMALL RODENTS
z
40-

w 2V
40-
W 20
C ALL RODENTS


0 2 4 6 8

DAYS








Figure 2-10. Percentage of seeds remaining in three types of exclosures after 8 days. For
each treatment N = 38. The 'no rodent' exclosure was designed to exclude all terrestrial
animals, 'small rodent' exclosures excluded large animals, but allowed access to small
animals through 6 cm mesh, and 'all rodent' plots were controls with no exclosure.







construction of exclosures and the beginning of the experiment The third exclosure was

hit by a fallen tree, thus creating access. Except for those three, seeds were not removed

from "no rodent" exclosures. Several other exclosures were damaged by falling branches

or breached by peccaries foraging for Inga pods, but these incidents either occurred after

the seed had been removed or did not lead to seed removal. After 8 d all the seeds from the

control plots had been removed, while 52% of the seeds from the "small rodents only"

exclosures had been removed (Fig. 2-10). After 4 mo, 82% of the seeds from the "small

rodents only" exclosure had been removed. Thus, although large mammals may take some

seeds, small mammals (presumably rodents) will eventually find and eat most seeds.


Germination and Seedling Survival

Ocotea endresiana seeds began germinating about 6 weeks after dispersal. The

germination rate of caged seeds ranged from 70 to 95% in both years and did not differ

among the seeds at dispersed, non-dispersed, and random sites (Fig. 2-11). Of the seeds

that germinated in 1993 and 1995,35% and 27% respectively, survived one year as

seedlings. Overall, 28% and 22% of seeds in 1993 and 1995 respectively, germinated and

survived one year. For the 1993 caged seeds, annual survival for the first three years

averaged 31% (Table 2-4). Seeds at four dispersed and five random sites from 1993

survived until June 1996, for an overall 3-year survival rate of 2% (Fig. 2-11). Within

each stage (germination, 1 yr, 2 yr and 3 yr survival), the three location treatments had

similar survival rates except for one-year survival in 1993 (F2,27 = 11.84, P = 0.0002).

Seeds at dispersed sites had significantly higher one-year survival than those at non-

dispersed and random sites (P < 0.001). Random and non-dispersed sites had similar one-

year survival (P = 0.35). Germination success of seeds defecated by guans did not differ

from that of seeds regurgitated by the other four species (X2 = 0.98, df = 1, P > 0.05).

None of the variables in the logistic regression models predicted germination

success in either year. Significant predictors of 1-year survival included lower canopy






















05 -.. r .. -
Z 0.80- Random
_>
00.60)

z
0.40b
M b
0
0. -
0 0.20-

0.00"
germ 1 yr 2yr 3yr germ lyr
1993 SITES 1995 SITES





Figure 2-11. Germination success, and annual seedling survival of dispersed and non-
dispersed seeds and seeds at randomly located sites (mean + SD per tree) in 1993 (left) and
1995 (right). Within each stage where a difference among sites was detected, bars with
different letters above are significantly different (P < 0.005; Bonferroni-adjusted alpha =
0.008).










Table 2-4. Results of logistic regressions of post-dispersal survival of seedlings
predicted by habitat variables. These seeds were protected from seed predators
with cages until germination. Only significant effects are listed. No variables
significantly predicted germination success (not shown).

Response r2 -2log likelihood predictorsa
X2


1993
1 yr survival 0.10 43.25*** canopy cover*
+ dispersal date*

1995
1 yr survival 0.13 67.52*** canopy cover*
+ dispersal date**
tree number*

*P < 0.05, **P < 0.01, ***P < 0.001.
sign in front of each predictor variable indicates positive or negative correlation
with the response variable.







cover and later dispersal date in both 1993 and 1995 (Table 2-4). Additionally, in 1995, 1-

yr survival was heterogeneous with respect to which particular fruiting tree was closest,

although no clearcut pattern was discernible in terms of crop size, tree size, or proximity to

other fruiting trees.

The causes of mortality of caged seeds and seedlings during the first year differed

among dispersed, non-dispersed, and random sites in both 1993 (X2 = 29.14, df = 6, P <

0.001) and 1995 (X2 = 33.59, df = 6, P < 0.001; Fig. 2-12). Many seedlings were killed

when seed predators (presumably rodents) removed the seed and severed the connection

between the shoot and root. Mammalian herbivores also killed seedlings by entirely

removing the leaves and shoot. Mortality by the combination of mammalian seed predators

and herbivores was significantly different among the three location treatments in 1995
(F2,37 = 4.98, P = 0.012), but not 1993 (F2,24 = 3.56, P = 0.044). In 1995, mammals

killed more seeds and seedlings at dispersed than at random sites (P < 0.05), but mortality

caused by mammals at non-dispersed sites did not differ from that at either dispersed or

random sites (P > 0.05; Fig. 2-12). Mortality caused by beetle larvae (Heilipus sp. and at

least one other unidentified species) differed among the three treatments in both years
(1993: F2,23 = 4.85, P = 0.012, 1995: F2,37 = 6.64, P = 0.003). In both years, more

non-dispersed than dispersed seeds were killed by beetles (P < 0.05; Fig. 2-12). Mortality

caused by fungal pathogens and physical damage (falling branches, trampling) did not

differ among treatments or years. For the three treatments, combined levels of mortality by

each cause were similar between the two years, except for insect predation which was

significantly higher in 1993 than 1995 (Student's t-test; t = 1.97, df= 52, P < 0.05) and

mortality by mammals which was significantly higher in 1995 than in 1993 (t = 2.53, df=

67, P < 0.01).

Average canopy cover for seedlings that survived one year at dispersed sites

(94.3% 3.6, N = 86) was significantly lower than for seedlings that died (95.2% 2.3,

N = 303), whereas canopy cover did not differ with one-year survival for non-dispersed or












0.8-
1993
[0 Dispersed (N=116)
3 0.6 -U Non-dispersed (N=112)
b D Random (N=73)

Z
o 0.4 a
a-a
0
0.
o 0.2-
0_


0.0 "
mammals insects fungal physical


a
0.80

b 1995
S 0 b Dispersed (N=186)
LU 0.60- b
-J Non-dispersed (N=134)
b Random (N=--82)
Z
0 0.40
P
ab
0
0. a
0 0.20



0.00- -
mammals insects fungal physical





Figure 2-12. Average (+ SD) proportion per tree of caged dispersed and non-dispersed
seeds and randomly located seeds killed by different causes during the first year for seeds
from 1993 (above) and 1995 (below). Cages were removed after germination and shoot
development. Sample sizes of seeds or seedlings killed shown in parentheses. Within
each cause of mortality where a difference among sites was detected, bars with different
letters above are significantly different (Bonferroni-adjusted alpha = 0.0125). See text
(Methods: Germination and seedling survival) for definitions of types of mortality.







random sites (2-way ANOVA, F2,877 = 3.98, P = 0.019). Denser canopy cover was also

a significant predictor of higher mortality by fungal pathogens (logistic regression, P =

0.0013). Seedlings that survived one year were significantly taller at sites with less dense

canopy cover (linear regression, r2 = 0.284, P < 0.001; Fig. 2-13). Seedlings from seeds

dispersed by bellbirds were significantly taller than those at other species' sites (t-test =

2.37, N = 38, P = 0.02).

Germination trials. Regurgitated seeds and seeds cleaned of pulp by hand had

higher germination rates than the seeds in intact fruits (X2 = 43.94, df = 2, P < .001). Of

30 seeds in each of the first two treatments, all but one seed germinated, and that one failed

because it was infested with a beetle larva. In contrast, seeds in intact fruits tended to

mold; 8 germinated and 22 appeared mealy and no longer viable after 12 wk. The 8 seeds

in fruits that germinated never developed a shoot outside the fruit pulp and eventually

rotted. Thus, ingestion of seeds by frugivores was beneficial in terms of pulp removal but

probably did not otherwise affect germination. Two of the 10 buried seeds started to

germinate (the seed coat appeared cracked) and seemed viable but the other eight seeds

rotted. After 4 mo, neither of the apparently germinating seeds had any sign of root

development, nor had the cotyledons begun to separate. Six of 10 buried seeds had

whitish mold on the seed coat. Thus, few buried seeds appear to establish as seedlings.

Seedling and sapling plots. Seedlings and saplings were equally common in 10 x

10 m plots under (6.6 1.5 individuals) and away (4.9 3.5) from the parent trees (paired

t-test; t = 1.15, P > 0.05). Individuals in plots away from the parent trees, however, were

taller (79.9 14.9 cm) than those near the parent trees (47.4 7.2; Mann-Whitney U test,

P = 0.009).


Recruitmnen
Based on the experiment with marked seeds, the probability of recruitment is

determined mostly by post-dispersal seed predators. Assuming the experiment reflects the






















16

*

oE 12



wZ 8 rm- mu
12 U m u

0 a a a a a



z m mm
a w- -
aa


w 4 m mum





70 80 90 100

CANOPY COVER (%)


Figure 2-13. Seedling height (N = 140) plotted as a function of percent canopy cover for
both years combined (height = 28.93 0.233x; r2 = 0.284, P < 0.001). Values for canopy
cover below approximately 92% are associated with canopy gaps.







absolute level of seed predation, and all seeds that escape predation germinate and survive,

then recruitment was 0% in 1993 and 0.4% in 1995. On the other hand, if seed removal at

2 wk (75-93%) reflects the ultimate pattern of survival (e.g., Lieberman 1996) then it is

possible to calculate the influence of stage-specific mortality patterns on the relative
probability of recruitment (Table 2-5, Fig. 2-14). The initial pattern of seed rain shows a
strong peak near the parent trees and secondary peaks beyond 40 m that correspond with

bellbird song perches (Fig. 2-14a). The bellbird perches encompass a range of canopy
cover values from forest/gap edge (=92%) to gap centers (,70%) and thus seed rain is

spread across a wider range of canopy conditions with increasing distance from the parent
trees. Two-week seed survival is higher for seeds > 30 m from fruiting trees, but relatively

consistent across the range of canopy cover (Fig. 2-14b). Germination is high for all seeds

regardless of distance from parent or canopy cover (Fig. 2-14c). One-year seedling

survival is highest for sites > 40 m from parents and < 90% canopy cover (Fig. 2-14d).
The cumulative probability of recruitment (the product of seed rain, 2 wk seed survival,

germination, and one year survival) shows a peak near the parent trees and secondary

peaks corresponding to the bellbird perches (Fig. 2-14e). Note that the probability of a

seed dispersed by birds surviving one year is < 0.2% at any location.



Discussion
The results illustrate four main points: (1) bellbirds dispersed seeds in a different
pattern than the other species; (2) post-dispersal seed predation is highest near the parents,

but is also high everywhere; (3) the pattern of recruitment is bimodal with a peak near the
parent tree in closed canopy forest and another in gaps corresponding to bellbird perches;

and, (4) all three advantages of dispersal (escape, colonization, and directed dispersal) are

supported. These points will be discussed in detail below.











Table 2-5. Summary of post-dispersal stages leading to seedling recruitment. Values for
stage-specific probabilities represent the average proportion of individuals per tree that
entered and survived each stage. Cumulative probabilities are the product of the current
and previous stage-specific probabilities. Data for 2- and 3-yr survival of the 1995 cohort
are unavailable.

Stage Stage-specific probability Cumulative probability

1993 1995 1993 1995
2wkseed 0.094 +0.113 0.137 +0.167 0.094 0.137
germination 0.759 + 0.188 0.855 0.122 0.0713 0.1171
1 yr seedling 0.309 + 0.237 0.212 0.185 0.0220 0.0248
2 yr seedling 0.245 +0.181 --- 0.0054 ---
3 yr seedling 0.360 + 0.115 --- 0.0019 ---



























Figure 2-14. Average seed rain (a), 2 wk seed survival (b), germination (c), 1 yr seedling
survival (d), and cumulative probability of recruitment (e) as functions of canopy cover and
distance from parent for both years combined and averaged among 21 trees. For a-d, each
bar represents the percentage of the seeds or seedlings that survived that stage in each
distance/canopy cover category, averaged among trees. Categories with fewer than three
seeds were not included and are blank, while categories with a flat rectangle are zero. Seed
rain determined from the original distribution of seeds naturally dispersed by birds. Two-
week seed survival estimated from the removal of marked seeds. Germination and seedling
survival estimated from the seeds caged until after germination. Recruitment (the
cumulative probability of surviving every stage) is calculated as the product (a x b x c x d)
for each distance/cover category. For clarity, standard deviations are not shown.










SEED RAIN


C-)


Figure 2-14









2-WK SEED SURVIVAL






1O00


00



40'

O0,0


Figure 2-14 -continued


C)









GERMINATION




200











Z~ Z"-Z ~


Figure 2-14 -continued





53




1-YR SEEDLING SURVIVAL


Figure 2-14 --continued


C)








RECRUITMENT



"0.1

.-0.o0
"0.06


"0.0 4





9 50
V C:- s 1-.. 6'
e:_z "o '


Figure 2-14 --continued







Seed Disperal
Most 0. endresiana seeds handled by birds land under or just beyond the crowns of

parent trees. This distribution of seeds is typical for vertebrate-dispersed plants (Dirzo and

Domfnguez 1986, Hoppes 1988, Willson 1993, Laman 1996a). The abundance of 0.

endresiana adults in the study area makes dispersal beyond 80 m from any conspecific tree

virtually impossible. Considering the restricted elevational range of this species, longer-

distance dispersal may lead to arrival in habitats unsuitable for establishment (Wheelwright

1988). Thus, while regurgitating seeds just outside the crown of a fruiting tree may seem

like poor quality dispersal (McKey 1975), the shear abundance of seeds in that region may

lead to a small peak in recruitment near the parent tree (Fig. 2-14e) as predicted by Hubbell

(1980) and Condit et al. (1992). The long tail of the seed distribution, however, is

apparently important for recruitment because seeds dispersed farther from the parent trees

have a greater chance of arriving in a habitat where the probability of seedling survival is

higher (i.e., a safe site), as will be shown below.

Of the five species of birds that dispersed most of the seeds in this study, all but

one deposited most seeds in closed canopy forest. Only male bellbirds frequently

dispersed seeds to treefall gaps. In my study site, bellbird song perches were in dead

Sapium oligoneuron trees, bordering large treefall gaps. Such sites are typical for

bellbirds: Snow (1977) reported that bellbird song perches at lower elevations in the

Monteverde area (where the forest is fragmented) are usually on dead branches in tall trees

on the forest edge. In other Neotropical forests, three other species of bellbirds (Procnias

spp.) exhibit similar behavior in habitual use of tall, exposed song perches, often on dead

branches (Snow 1961, 1970, 1973a). Snags used by bellbirds in my site had several

branches and the birds made use of many different branches within each tree, such that

seeds under them were scattered over approximately 25 m2, including a gradient of site

conditions from gap to forest understory. Male bellbirds also dispersed seeds in forest as

do the other species, but they typically spend 80-95% of the day in the vicinity of the song








perch (Snow 1977) and probably disperse a similar percentage of seeds they process under

song perches. Female bellbirds were rarely seen at the song perches, and then only for a

few minutes.

The influence of bellbirds on the pattern of seed fall is shown by the slight increase

in the number of seeds 40-65 m from the parent trees (Fig. 2-3). Bellbird perches in my

site happened to be far (> 40 m) from any fruiting 0. endresiana trees. The other four bird

species that disperse 0. endresiana seeds occasionally perch on the edges of gaps, (D.

Wenny, personal observation) although whether they do so more often than expected based

on perch availability is unknown. Because birds do not (cannot?) regurgitate in flight, and

guans apparently do not defecate in flight (personal observation), the most likely way an 0.

endresiana seed can arrive in a gap is via male bellbirds. I never observed the other four

species on bellbird perches. Thus, 0. endresiana has two types of dispersers that deposit

seeds in two different but predictable patterns: male bellbirds depositing large numbers

(52%) of the seeds they disperse in and near gaps, and the other four species depositing

seeds in forest, most (75%) within 25 m of parent trees and fewer more than 25 m away.

Additionally, on several occasions I followed quetzals and guans from one fruiting 0.

endresiana to another, where they deposited seeds before eating fruits in the second tree.

Although I had no method of identifying the source tree of the seeds, it is likely that

dispersal of seeds from one tree to another conspecific occurs regularly (Wheelwright

1991, Gibson and Wheelwright 1995).

The patterns of seed distribution described here may be similar for other trees in

cloud forests where these disperser species occur during the breeding season, but may be

very different during the rest of the year. The five species of birds that disperse 0.

endresiana are the most important fruit consumers of canopy trees in the area (Wheelwright

et al. 1984). At least 20 other species of trees are dispersed by these birds and probably

have similar patterns of seedfall as 0. endresiana, with a peak near the parent tree and a

second peak at bellbird perches. However, all five dispersers (except perhaps toucanets)







undertake elevational shifts that correspond with fruit availability (Wheelwright 1983,

Levey and Stiles 1992, Powell and Bjork 1995, Guindon 1996) and leave the study area in

late July or early August (In 1994 they left in June when 0. endresiana failed to fruit).

Male bellbirds may not provide such a pronounced peak in seed dispersal to song perches

after the breeding season because they seem to use many different perches in the non-

breeding season and tend to use subcanopy perches that are not exposed (Stiles and Skutch

1989). Thus, seed fall for the community as a whole is likely temporally as well as

spatially heterogeneous.

The implications of seed dispersal to habitual perches deserves further study. Do

habitual perches represent foci of plant recruitment or of seed predation? Seed dispersal to

perches has been frequently studied in successional landscapes (Livingston 1972,

McDonnell 1986, Holthuijzen et al. 1987, Guevara and Laborde 1993, McClanahan and

Wolfe 1993, Robinson and Handel 1993, Vieira et al. 1994, Nepstad et al. 1996, Duncan

1997). In these systems, directed dispersal may play a key role in the restoration of

abandoned pastures, logged forests, and other degraded lands. It is possible that directed

dispersal is more common in intact forests than previously expected. For example,

manakins (Pipridae) and cocks-of-the-rock (Rupicola: Cotingidae) choose lek perches that

are more sunlit than average understory perches (Endler and Thery 1996), and such sites

may provide growth advantages for seedlings. Snow (1961, 1970, 1973b) notes that

other species of bellbirds (Procnias) preferentially select perches on dead trees or branches,

or in sparsely-vegetated trees. For these species, competition among males for females

drives them to be as conspicuous as possible. The fortuitous outcome of this behavior may

be a disproportionate effect on plant recruitment in the vicinity of their display areas (Th6ry

and Larpin 1993). Whether habitual perches represent foci of seedling recruitment

(McDonnell and Stiles 1983, McDonnell 1986) or lead to density-dependent seed and

seedling mortality (Wheelwright 1988), needs to be examined in more detail.







Seed Predation

Removal of marked seeds after dispersal in this study always resulted in predation.
Although seeds were taken into burrows and logs, I never found any treatment of seeds

indicative of scatterhoarding, such as burial in the soil (Smythe 1978, Hallwachs 1986,

Forget 1990, 1993) or under piles of leaf litter (Forget 1991). In addition, the failure of

buried seeds to germinate and survive in the greenhouse experiment suggests that

scatterhoarding would not be advantageous for 0. endresiana recruitment. Only two

species of mammals known to scatterhoard seeds occur in the study site: spiny pocket mice

(Heteromys desmarestianus) and agoutis (Dasyprocta punctata). Both species are much

less common here than at lower elevations (D. Wenny personal observation, K.G. Murray,

personal communication). Although some species of squirrels (Sciurus) and deer mice

(Peromyscus) are known to cache seeds in other regions of the world (Vander Wall 1990),

information on tropical species (in this site S. deppei, and P. mexicanus) is lacking

(Emmons 1990).

The main seed predators in this study were probably rodents, particularly the deer

mouse Peromyscus mexicanus, by far the most common terrestrial rodent in the study area

(Anderson 1982, Langtimm 1992). Based on the exclosure experiment, however, it is

possible that species larger than mice (i.e., agouti, paca, peccary, wood-quail) are also

important seed predators. Because removal of seeds from the open control plots was about

twice that of the "small rodents only" plots, agoutis and other large mammals may have

been responsible for as much as 50% of 0. endresiana seed predation. Nevertheless, the

exclosures also showed that in the absence of agoutis, small rodents will find and eat most

of the seeds, albeit over a longer time period. Additionally, the finding that most marked

seeds from the removal experiment were taken at night and removal of seeds always

resulted in predation rather than scatterhoarding suggest that agoutis, which are mostly

diurnal and known scatterhoarders (Smythe 1978, Hallwachs 1986), may not be taking







many seeds (although few data exist on the extent to which agoutis eat but do not
scatterhoard certain species, e.g., Smythe 1978).

The high degree of post-dispersal seed predation reported here is not unusual
(Harper 1977, Cavers 1983). Recent reviews have found 90-100% losses of seeds to
predators in 22 of 62 studies (Crawley 1992, Hulme 1993). Estimates of seed predation

rates for Neotropical trees range from 40-50% for the canopy tree Brosimum alicastrum
(Burkey 1994), 36-98% for the canopy emergent tree Dipteryxpanamensis (DeSteven and

Putz 1984), 87% for Dipteryx micrantha (Cintra and Homrna 1997), 60-90% for the
subcanopy tree Faramea occidentalis (Schupp 1988b), 50-90% for the subcanopy palm
Welfia georgii (Schupp and Frost 1989), 97% for the canopy palm Astrocaryumwn murumuru

(Cintra and Horna 1997), and 96% for the canopy tree Virola nobilis (Howe et al. 1985).

The study by Burkey (1994) was short-term (21 days) and may underestimate total

predation. Also, in the study by DeSteven and Putz (1984), a site with only 36% predation
had low populations of seed predators as a result of hunting. Previous studies on seed

predation of Lauraceae found 96-100% predation (Wheelwright 1988, Chapman 1989a,

Holl and Lulow 1997).

The high level of seed predation in my study site could be explained by dense

vegetation and cloudy conditions, providing small rodents protection from predators

(Bowers and Dooley 1993, Visquez 1996). High predation is probably not a consequence
of high rodent populations due to lack of predators because potential predators of mice such
as owls, forest-falcons, and cats are relatively common (D. Wenny, personal observation).
During the day, light levels in the study area are often reduced as a result of cloudy weather

(Cavelier 1996, Chazdon et al. 1996) as well as dense vegetation. Indeed, on a few
occasions mice were seen during the day (especially Peromyscus and Scotinomys) and
some marked seeds were removed by small rodents during the day (see Post-dispersal seed
predation: distance effect). Also, local naturalists report seeing more mice on misty or

rainy days and nights than during clear weather (T. Guindon, personal communication).







Thus, the high levels of predation for 0. endresiana could be a result of longer activity

patterns by small rodents and may be typical for many large-seeded tree species in cloud

forests.


Seedling Survival and Recruitment

Most of the seeds protected from rodents germinated. High germination rates for

Lauraceae are apparently typical (Wheelwright 1985b). Some seeds that germinated were

eventually killed by beetle larvae (Heilipus sp. Curculionidae) developing in the

cotyledons. After the cages were removed from the growing seedlings, the seeds as well

as the seedlings were susceptible to predation by mammals. It is difficult to determine the

role of each species involved but the suite of species that kill seedlings is probably larger

than the suite of seed predators, and could include pacas (Agoutipaca), brocket deer

(Mazama americana), tapir (Tapirus bairdiz), and peccaries (Tayassu tajacu), in addition to

agoutis and smaller rodents.

Insects were most likely to kill non-dispersed seeds or seedlings while mammals

killed all types of seeds (dispersed, non-dispersed, and random). This finding is consistent

with Howe (1993a) and Terborgh et al. (1993), who found that insect-caused mortality

was distance-dependent, but mortality by mammals was not. In New Guinea, Merg (1994)

found the opposite pattern: insects tended to kill dispersed seeds and mammals killed non-

dispersed seeds. In contrast to insects and mammals, fungal pathogens killed

proportionately more 0. endresiana seedlings in closed-canopy forest than in gaps.

Augspurger (1984) also showed fungal pathogens of seedlings were less prevalent in gaps

than forest understory for nine species of tropical trees.

The greater height of seedlings and saplings in plots >20 m from the parent trees

than in plots beneath the crowns, suggests that seedling mortality was higher closer to the

parent trees. Seedling ages are not known, but this difference suggests that longer-term

survival several years after dispersal is higher for seedlings >20 m from the crown of








parent trees than for seedlings beneath crowns. Thus, seed dispersal away from the crown

of a conspecific appears to increase the chance of recruitment over the long term (Clark and
Clark 1984, Li et aL 1996).

The overall pattern of recruitment with respect to distance from parent trees and

canopy cover was bimodal. This pattern was caused by the combination of distance-

dependent seed predation (Fig. 2-14b) and the influence of canopy cover on seedling

survival (Fig. 2-14d), despite the fact that most seeds landed close to the parent trees in

closed-canopy forest (Fig. 2-14a). This pattern is an example of spatial discordance caused

by the lack of congruity among the stages leading to recruitment (Herrera et al. 1994,

Jordano 1995). The occurrence of such discordance emphasizes the importance of stage-

specific survival patterns on patterns of recruitment and the need for data on the sequential

stages of plant reproduction rather than on only one or a few stages. Bellbirds are clearly

an important part of the dispersal system of 0. endresiana, as over half (52%) of the seeds

they dispersed landed in a zone of higher recruitment.


Synthesis: Advantages of Dispersal

Despite nearly complete predation, all three advantages of dispersal receive some

support for Ocotea endresiana. With respect to the escape hypothesis, mortality for both

seeds and seedlings had a component of density and/or distance dependence. Post-

dispersal seed predation was higher for seeds directly under the parental crowns than for

seeds dispersed away from the crowns (Fig. 2-8). Seed predation was higher for seeds

closer to the trees (Fig. 2-9, 2-12). Also, of seeds initially protected from mammalian seed

predators, more dispersed than non-dispersed seeds established as seedlings and survived
one year (Fig. 2-12). Overall, seeds dispersed away from the parent trees had a greater

probability of survival and recruitment. These results support the general predictions of the

Janzen/Connell model of higher seed and seedling mortality near the parent tree (Connell

1970, Janzen 1970). However, removal of experimentally dispersed seeds showed that







although distance is important, its effect decreased over time because removal approached

100% at all distances. Only two marked seeds were not removed by seed predators: one at

30 m and one at 70 m. Thus, although dispersal is advantageous for removal from the

zone of highest seed predation and seedling mortality near the parent tree, predation is so

high everywhere that dispersal beyond approximately 25 m is apparently not effective in

providing an additional escape advantage. Furthermore, the occurrence of seedlings in the

crown edge plots suggests that some recruitment can take place near the parents. Indeed, a

peak in recruitment occurred within 10 m of the crown edge (Fig. 2-14e), assuming the

pattern of seed predation at 2 wk (Fig. 2-14b). This peak is the one predicted by Janzen

(1970), and it occurs where Hubbell (1980) suggested it would, as a result of high seed

rain close to the parent trees and incomplete density-dependent mortality (see also Condit et

al. 1992).

Thus, factors in addition to escape from seed predation may be important in

recruitment. During a tree's lifetime, recruitment may be episodic, which if true, could

explain why 0. endresiana seedlings are fairly common, despite the high levels of seed

predation observed in the two years of this study. Perhaps stochastic events combine to

result in low populations of seed predators and an increase in seedling recruitment only a

few times during a tree's reproductive years. Long-term studies are necessary to test

HubbeUll's (1980) hypothesis that the variance in seed predation among years may be more

important than the average in determining the probability of recruitment

Dispersal of 0. endresiana by guans, quetzals, robins, and toucanets is consistent
with the colonization hypothesis. Typically, the colonization hypothesis applies to pioneer

species that gain an advantage in random dissemination of seeds to increase the chance that

a few will arrive in a suitable location, such as treefall gaps. Many such species have seeds

capable of dormancy and are incorporated into the soil seed bank, where they wait for a gap

to form above them. Shade tolerant trees also benefit from colonization through

suppressed growth in the seedling and sapling stage rather than the seed stage, and







accordingly a "seedling bank" develops in the understory. My results are probably

representative of many shade-tolerant canopy trees in tropical wet forests (Howe 1993a,

Lieberman 1996, Whitmore 1996). Because most 0. endresiana seeds are dispersed to

sites within closed canopy forest, most surviving seedlings are part of the understory

seedling bank. During three years of intensive field work in this study site, I only found

two subadult 0. endresiana (5-15 m tall), and both were in or near old (> 5 yr) treefall

gaps.

The directed dispersal hypothesis is also supported. Species that have directed

dispersal gain an advantage in arrival at specific sites associated with a higher probability of

survival. The few examples of directed dispersal in bird-dispersed plants are for parasitic

epiphytes (mistletoes) that can establish only on branches (Davidar 1983, Reid 1989,

Sargent 1995, Larson 1996). For 0. endresiana, male bellbirds provide directed dispersal

to gaps, where seeds have a slightly lower probability of early predation by mammals

(Table 2-3), and a slightly higher chance of recruitment than in forest understory (Fig. 2-

13, 2-12e). One must take into account the fact that most seeds land a few meters beyond

the crowns of the parent trees, and therefore, this zone contributes the most individuals to

the seedling/sapling stage. Seeds that land in a gap edge zone (sensu Popma et al. 1988),

in this case often associated with bellbird song perches, gain an advantage in survival and

growth. Thus, although bellbird perches are not the only places seedlings can survive, a

seed dispersed by a bellbird to a gap has a slightly higher chance of germinating and

surviving than one in the forest understory. Additionally, because gaps in Monteverde tend

to expand when trees on the edge of the gap fall (Lawton and Putz 1988), seeds dispersed

by bellbirds that germinate and survive, are more likely to encounter a favorable growth

environment (i.e., higher light levels) in the future, whereas the prospects for seedlings in

the understory are unpredictable.

These results are consistent with other studies that show shade tolerant species to be

capable of recruitment under a wider range of conditions than shade-intolerant pioneer







species (Canham 1989, Denslow and Hartshorn 1994, Lieberman 1996). In montane

forests, shade-tolerant species may be better able to recruit in gaps than in lowland forests,

because cloudy conditions moderate the temperature fluctuations (Cavelier 1996). Thus,

seedlings of shade-tolerant species in montane forests may gain an advantage of higher

light levels in gaps without the increased risk of desiccation found in other rain forests

(Bazzaz and Pickett 1980, Veenendaal et al. 1995, Whitmore 1996). In addition, shade-

intolerant pioneer species are less common and require larger gaps with increasing elevation

in some forests (Whitmore 1996), suggesting that the seedlings of non-pioneers are more

common in montane forest gaps.


Conclusion
The overall conclusion of this study is that the pattern of recruitment of 0.

endresiana depends on the combined effects of seed dispersers, seed predators, and

seedling mortality. Selection on plant traits occurs during each stage, and selection during

sequential stages may be opposed (Wheelwright and Orians 1982, Herrera 1985, 1986,

Wheelwright 1988, Herrera et al. 1994). For example, high seed predation overall and the

slight preference for larger seeds by seed predators may select for smaller seeds or larger

seed crops, while seed dispersers may prefer larger fruits, which have larger seeds, but

occur in smaller fruit crops (but see Howe and Vande Kerckhove 1980, 1981,

Wheelwright 1991, Mazer and Wheelwright 1993). On the other hand, seed size did not

influence germination, seedling height, or seedling survival. In addition, the bimodal

spatial pattern of recruitment may represent disruptive selection on seedling traits. Some

trees were visited by all five species of dispersers, while the trees far from bellbird perches

tended not to be visited by bellbirds. Thus, seeds dispersed to gaps beneath bellbird

perches were mostly from a subset of the available trees. The extent to which such

differences in dispersal and subsequent recruitment affect gene flow is poorly understood

(Gibson and Wheelwright 1995, Hamrick and Nason 1996). Further studies that compare





65

dispersal patterns and the subsequent stages leading to recruitment at different sites, as well

as over longer time periods are especially needed.













CHAPTER 3
SEED DISPERSAL OF A HIGH-QUALITY FRUIT BY SPECIALIZED FRUGIVORES:
HIGH-QUALITY DISPERSAL?




"It is well known that most Lauraceae seeds are free ofpredation." Oscar C. Castro

(1993:67)


Introduction

In a seminal paper, McKey (1975) proposed that tropical trees producing nutrient-

rich fruits attract specialized frugivores that provide high-quality seed dispersal. He noted

that this strategy of high-quality fruits gaining high-quality dispersal by specialized

frugivores was one endpoint of a continuum of dispersal strategies. High-quality, or

specialist, plants produce large fruits with one or a few large seeds and lipid or protein-rich

pulp. At the other end of the continuum, generalist plants produce large crops of fruits

with many small seeds and pulp composed mostly of sugars and water. The many species

of small, opportunistic birds that eat generalist fruits McKey expected to be less reliable

dispersers because they were thought to be less dependent on fruit than are the specialized

frugivores. McKey's framework was based on three ideas. First, the general similarity of

fruit characteristics of plants dispersed by different types of dispersal agents (i.e., dispersal

syndromes; Ridley 1930, van der Piji 1972). Second, observations that fruit was an

important dietary component for many species, but few species were totally dependent on

fruit (Orians 1969, Morton 1973). Third, Snow's (1971) ideas about possible coevolution

between plants with large-seeded, nutritious fruits, and highly frugivorous birds that

dispersed the seeds. These ideas were further expanded by Howe and Estabrook (1977) to







incorporate crop size, phenology and visitation rates. Although components of the

specialist/generalist framework have been confirmed, data to fully test it are still lacking

(Howe 1993b, Schupp and Fuentes 1995). Howe (1993b) suggested that the main reason

the framework remains untested is because plant ecologists and zoologists study different

aspects of the multiple stages of the plant recruitment process. Quantifying dispersal

quality is key to testing the model, yet to examine dispersal quality, it is necessary to study

both the dispersal pattern and post-dispersal fate of seeds (Janzen 1983c, Howe 1993b,

Jordano and Herrera 1995, Schupp and Fuentes 1995).

Another reason that the specialist/generalist framework has not been adequately

tested is because species-specific coevolution between plants and dispersers is now

considered unlikely; diffuse coevolution (sensu Janzen 1980) between groups of plants and

groups of dispersers is thought to be more typical (Wheelwright and Orians 1982, Janzen

1983c, Howe 1984, Herrera 1985, Levey et al. 1994). The diffuse mutualism paradigm is

based on four main points. First, because plants offer a reward (the fruit pulp) to

dispersers for fruit removal, but no reward for seed dispersal to an appropriate site, plants

have little control over what happens to seeds ingested by potential dispersers

(Wheelwright and Orians 1982). Second, most large, highly frugivorous animals eat a

wide range of fruits including both high-quality and low-quality species (Wheelwright

1983, Wheelwright et al. 1984). Similarly, small-bodied frugivores can be highly selective

among the available fruits, and species of approximately the same body size may handle

fruits and seeds very differently (Moermond and Denslow 1983, 1985, Moermond et al.

1986, Levey 1986, 1987). Third, most plant species with fleshy fruits are eaten (and

presumably dispersed) by several to many species of animals of differing body size, seed

handling techniques, and movement patterns (Wheelwright et al. 1984, Bronstein and

Hoffman 1987, Jordano 1992). Fourth, plants probably evolve at different rates than the

animals that disperse them (Herrera 1986). Therefore, the opportunities for dispersal-

related plant and animal traits to coevolve at a species-specific level are minimal, and neither







body size nor amount of fruit in the diet can be used to make any consistent prediction

about the quality of dispersal provided by that species (Henrrera 1984b, 1986, Howe 1984,

Levey 1987).

Although many studies have addressed dispersal quality in terms of gut treatment of

seeds (Krefting and Roe 1949, Traveset and Willson 1997, Wahaj et al. in press), few

have examined dispersal quality in terms of suitability for growth and survival of sites

where seeds are dispersed (Schupp 1993). It is reasonable to expect seeds from an

individual plant to be dispersed to a variety of sites, and it is well known that all potential

dispersal sites are not equally suitable for seedling establishment or growth to maturity

(Grubb 1977, Harper 1977, Murray 1988, Bazzaz 1991, Schupp 1995). Thus, dispersal

quality may vary within and among conspecific trees, as well as among dispersers.

Many models of dispersal are based on wind-dispersed species (Green 1983, Geritz

et al. 1984, Greene and Johnson 1989, 1996, Okubo and Levin 1989, Andersen 1991)

because so little is known about where animals disperse seeds (Janzen 1983a, Willson

1993, Laman 1996a). Until recently, the difficulty in finding seeds dispersed by animals

has limited investigations of dispersal quality (in terms of the suitability of dispersal sites

for recruitment) to parasitic mistletoes, which have specific and easily quantifiable safe sites

(Davidar 1983, Reid 1989, Sargent 1995), and to ant-dispersed species that are dispersed

over relatively small scales (Horvitz and Schemske 1986b, Hanzawa et al. 1988).

Dispersal quality has been defined as the probability that a dispersed seed will survive to

reproductive age (Schupp 1993). The product of dispersal quality and dispersal quantity

(number of seeds dispersed) equals disperser effectiveness, which is defined as the

proportion of seedlings (Reid 1989) or ideally, adult plants (Schupp 1993) in a population

resulting from activities of a particular dispersal agent (see also Bustamante and Canals L.

1995). Most previous studies have experimentally examined seed and seedling survival

without considering the actual pattern of seed deposition generated by dispersers.

Therefore, dispersal quality cannot be assessed (Howe 1993b, Herrera et al. 1994, Schupp







and Fuentes 1995). Although dispersal quantity has been studied more intensively than

dispersal quality, preliminary estimates for Virola indicate that quality is more strongly

correlated with disperser effectiveness than is quantity (Schupp 1993).

One of the prime examples of a specialized dispersal system, and indeed an integral

part of the development of seed dispersal theory, is the plant family Lauraceae (Snow

1971, McKey 1975, Wheelwright and Orians 1982). Most lauraceous trees produce large

one-seeded, lipid-rich fruits, and are dispersed by large highly frugivorous birds such as

bellbirds (Cotingidae), trogons (Trogonidae), and toucans (Ramphastidae) throughout the

Neotropics (Snow 1981, Wheelwright 1983, Avila H. et al. 1996). Species of Lauraceae

are also important for frugivorous birds in the Paleotropics (Crome 1975, Sun et al. 1997).

In this study I examined seed dispersal and seedling establishment of a Neotropical

Lauraceae (Beilschmiedia pendula; hereafter Beilschmiedia), dispersed by four species of

highly frugivorous birds. Beilschmiedia is one of the largest bird-dispersed seeds

throughout much of its geographic range. In the montane forests of northwestern Costa

Rica, it is the largest bird-dispersed seed (range = 5-20g, mean SD = 12.9 3.6 g;

(Wheelwright et al. 1984, Burger and van der Werff 1990, Haber et al. 1996).

Beilschmiedia produces fruits that clearly fit the pattern of high-quality fruits. They are

large, have low pulp/seed ratios, and are high in lipids relative to other species of fruits

(Snow 1971, Wheelwright et al. 1984, Moermond and Denslow 1985). Fruits are

produced prior to and during the breeding seasons of the main dispersers, at a time of year

when the number of tree species in fruit is intermediate (Haber et al. 1996). Within the

study site, Beilschmiedia fruit availability overlaps with few other Lauraceae species

(Wheelwright 1985a). The four main seed dispersers are large, deposit intact seeds, and

are reliable consumers in the sense that all four species are relatively common and regularly

eat Beilschmiedia fruits (Wheelwright 1983, Wheelwright et al. 1984, 1985b, Guindon

1996). Naturalist guides in the Monteverde Cloud Forest Preserve note that fruiting

Beilschmiedia trees are predictable spots to find large frugivorous birds. In years or areas







when the trees do not fruit, quetzals delay breeding or move elsewhere (T. Guindon, R.

Guindon, A. Villegas, personal communications). The influence of these dispersers on

recruitment of Beilschmiedia, however, is not well known.

The goal of this study was to estimate the quality of seed dispersal provided by
birds for Beilschmiedia. I examined dispersal pattern, seed predation, germination, and 1-

yr seedling survival to determine the influence of each stage on seedling recruitment,

because each stage may be important in limiting recruitment and few studies have integrated

dispersal patterns with their consequences (but see Herrera et al. 1994, Schupp and

Fuentes 1995). First, I determined the spatial distribution of dispersed seeds in relation to

parent trees by searching for regurgitated or defecated seeds beneath and beyond the

canopy of fruiting Beilschmiedia trees. At each site where I found a seed, I protected the

seed from vertebrate seed predators, measured microhabitat variables, and assessed

germination and 1-yr seedling survival to determine which variables were associated with

the highest probability of recruitment in the absence of predation. Then, I examined the

roles of seed predation and possible secondary dispersal in altering the seed shadow

produced by birds by recording the fate of marked seeds placed at the same sites as the

protected seeds. Another way of documenting quality of dispersal is to compare success at

sites where seeds are dispersed to success at random sites. Thus, microhabitat

characteristics, seed predation, germination, and seedling survival were also examined for

seeds placed at randomly located sites. Finally, I compared the spatial distribution of

naturally established saplings to that of dispersed seeds in relation to parent trees to estimate

the effects of longer-term patterns of mortality.



Study Site
This study was conducted January 1995 to June 1996 in the Monteverde Cloud
Forest Preserve (10012'N, 84042W) in the Cordillera de Tilaran, Costa Rica. This

10,000 ha preserve is administered by the Tropical Science Center of San Jose, Costa Rica.







The study area was in relatively undisturbed lower montane rain forest (Hartshorn 1983)
along the continental divide at 1600 m elevation. A 5 ha area, 500 m from the beginning of

the Valley Trail (Sendero El Valle), was mapped and marked into 10 x 10 m quadrats with

PVC tubing at every grid point. The site's vegetation is classified as leeward cloud forest
by Lawton and Dryer (1980) The canopy is 25-30 m tall and dominated by several species

of Lauraceae, Sapium oligoneuron (Euphobiaceae), Ficus crassiuscula (Moraceae), Inga

spp. (Leguminosae) and Pouteria viridis (Sapotaceae). The understory is dominated by

Rubiaceae, Acanthaceae, Gesneriaceae, Heliconiaceae, and Aracaceae. The vegetation of

the area is described in more detail by Lawton and Dryer (1980) and Nadkamrni et al.
(1995). Other characteristics of the Monteverde area are described by Nadkamrni and

Wheelwright (in press).

The average annual rainfall at 1520 m on the Pacific slope about 3 km from the

study site is approximately 2500 mm, with most of the precipitation occurring between
May and November. Actual rainfall in the study site was probably greater than 2500 mm

(due to the higher elevation of the study site relative to the rain gauge), but the seasonal

pattern was similar (Nadkami and Wheelwright in press). Range gauges underestimate the

amount of precipitation from mist and cloud interception, which contribute up to 50% of

the precipitation in some Neotropical montane forests (Cavelier 1996). Temperatures

recorded at the study site during this project ranged from 15 to 220C.


Study Spci
Beilschmiedia pendula [(Sw.) Hemsley] is a common canopy tree in Costa Rican
montane forests from 600-2000 m (Burger and van der Werff 1990). In the Monteverde

area it occurs from 1500-1600 m (Haber et al. 1996). Beilschmiedia begins flowering in
the late dry season (March) and is pollinated by small flies and other insects. Ripe fruits

are available from mid-January through late April. Fruits have black skin and lipid-rich

pulp (Wheelwright et al. 1984, Burger and van der Werff 1990). Most of the volume of







the fruit is a single seed (length: 48.55 + 7.65 mm, width: 20.70 1.78 mm, mass: 12.89

3.60 g; mean SD, N = 293) that is composed of a small embryo and two large
cotyledons (for additional measurements see Mazer and Wheelwright 1993). Seed size

varies from 5.1 to 21.7 g. Compared to other genera of Lauraceae, Beilschminedia has a

relatively thick (2.5 mm) endocarp, and the pulp is more tightly attached to the seed. Seven

Beilschmiedia trees were in the 5 ha study site. Data for two trees that had adjacent

crowns were pooled for analyses. Fruits and seeds for some experiments were collected

from trees on the periphery of the main study area. Large fruit crops (1837 1024

fruits/tree) were produced by the seven trees in 1995, but not in 1994 or 1996 (see also

Wheelwright 1986).

Beilschmiedia fruits are eaten primarily by four species of birds: emerald toucanet

(Ramphastidae: Aulacorhynchus prasinus), resplendent quetzal (Trogonidae: Pharomachrus

mocinno), three-wattled bellbird (Cotingidae: Procnias tricarunculata), and black guan

(Cracidae: Chamaepetes unicolor). The first three species typically remain in a fruiting tree

after eating several fruits, and often regurgitate viable seeds 30-70 min later under the same

tree or nearby (Wheelwright 1983, 1991). Guans defecate seeds in viable condition, and

they generally leave a fruiting tree before defecating the seeds from that foraging bout.

Although seed retention times by guans were not recorded in this study, Guix and Ruiz

(1997) reported a median retention time of 6.2 hr for a larger cracid (Penelope obscura).

All four species can be considered fruit specialists in the sense that they depend on fruit for

most if not all of their dietary requirements, at least at some times of year (Wheelwright

1983, Avila H. et al. 1996). However, quetzals and toucanets also eat large insects, and

small vertebrates (Skutch 1967, Wheelwright 1983, Riley and Smith 1992, Avila H. et al.

1996), while guans also eat leaves (Haber et al. 1996). Bellbirds apparently eat only fruits

although diets of female bellbirds are poorly known (Snow 1982).

Dropped or fallen fruits are eaten by agoutis (Dasyprocta punctata) and possibly

pacas (Agoutipaca) and other rodents. Agoutis chew off pulp and sometimes the







endocarp, and leave the seeds under the trees, but do not eat or bury seeds. Thus, these

species probably do not provide significant dispersal for Beilschmiedia.


Methods

Seed Dispersal

Seeds were located by systematically searching the ground for freshly regurgitated

or defecated seeds from late January through mid-April. The ground searches started at the

base of a fruiting tree and proceeded along 10 m wide transects (delineated by the PVC

markers) to 100 m from each Beilschmiedia trunk. Additional sites away from the focal

trees were also searched; some were randomly selected and others were chosen due to bird

activity in the area. It was impossible to search the entire site with equal intensity, but an

effort was made to cover the entire site at least once every three weeks, so that over the

course of the three-month fruiting season each 10 x 10 m plot was checked at least four

times. Defecated seeds were assumed to be dispersed by black guans because guans are

the only avian disperser in the area that defecates large seeds. Sites were classified as non-

dispersed if directly under the crown of a fruiting Beilschmiedia tree, or as dispersed if not

under such a tree. All the dispersed and non-dispersed seeds, as well as the seeds placed at

random sites (see below) were covered with wire mesh cages to protect them from

vertebrate seed predators. By doing so, I was able to calculate accurate estimates of

germination and seedling establishment in the absence of seed predation by mammals.

In addition to these naturally dispersed and non-dispersed seeds, other seeds were

placed at 50 randomly selected sites to compare seed predation, germination and survival at

dispersed, non-dispersed, and random sites. These sites were selected with random

numbers generated by a hand calculator and used as coordinates within the study site. The

post-dispersal fate of seeds at these random locations was compared to the fate of seeds at

the dispersed and non-dispersed locations to determine how the probability of recruitment

differs among the three types of sites.









Germination and Seedling Survival

The original dispersed and non-dispersed seeds, and all seeds placed at randomly

located sites were protected by a 4 x 8 x 4 cm cage made of 1 cm galvanized wire mesh

held in place by two 25 cm metal stakes. Caged seeds were used to determine germination

rates and insect predation rates in the absence of mammalian seed predators. Each site was

checked weekly for at least 12 weeks and at 15-17 mo after dispersal (hereafter 1-yr

survival). Germination was defined as the splitting of the seed coat and spreading of the

cotyledons. Typically, the radicle had emerged by the census after germination, and a

week later the stem was visible. As each seed germinated and the shoot began to grow, the

cage was removed to allow normal seedling growth. The seedling location was marked

with one of the stakes from the cage. Causes of seedling mortality were classified as

mammal, insect, fungal pathogen, physical, or unknown. Mammals either ate the seed and

left the damaged shoot behind (seed predators) or removed the entire shoot (herbivores).

Some seeds that appeared to have germinated were eventually killed by beetles inside the

cotyledons. Insect-killed seeds frequently developed a root but never had a shoot >2 cm

tall. Seedlings killed by fungal pathogens were characterized by a wilted and discolored

shoot (Augspurger 1990). A seed was considered alive as long as it remained firm, even if

the shoot had been eaten or otherwise damaged. Such seeds resprouted repeatedly.


Post-dispersal Seed Fate

At each dispersed, non-dispersed, and random site, a marked seed was used to

assess rates of seed predation, to identify seed predators, and to determine if secondary

dispersal occurred. Seed predation and secondary dispersal could alter any pattern

generated by dispersers and thus be more important in determining the probability of

recruitment than primary dispersal (Wheelwright 1988, Herrera et al. 1994). For this

treatment I used regurgitated seeds collected under fruiting trees adjacent to the study site.







Seeds were marked by gluing 50-75 cm of unwaxed dental floss to the seed, and tying

about 50 cm of flagging tape to the distal end of the floss. Because the glue held best on

seeds with a dry seed coat, seeds were taken inside and allowed to dry for 1-3 h before

gluing. Each marked seed was placed at a site the next morning.

All marked seeds were censused on days 1, 3, and 7, and once each week

afterwards until week 5. If a marked seed was removed, the surrounding area was

searched to find the flagging tape-dental floss assembly. The end of the floss where the

seed had been attached was examined to determine the fate (present or absent) of the seed.

In the few cases where seeds were partially consumed, teeth or bill marks were examined

to identify the consumer.


Seedling and Sapling Plots

To determine if Beilschmiedia seedlings and saplings are more likely to recruit

under or away from conspecifics, seedlings and saplings (up to 5 m height) were measured

and mapped in paired 10 x 10 m plots. For each of five trees, one plot was located near the

tree with approximately half of the plot directly under the crown. The second plot was

located at least 20-40 m from the crown edge. Beilschmiedia did not fruit in the study site

in 1994, so most individuals were presumably at least 2 yr old. Such seedlings can be

distinguished from new seedlings by the lack of a seed, woody stem, and presence of

epiphylls.


Microhabitat Characteristics

For all sites and seeds I measured seed characteristics and microhabitat variables

which might influence seed predation, germination, or seedling survival. Seed length and

width were measured with dial calipers and seed mass was measured with a hand-held

digital scale accurate to 0.01g. Canopy cover was estimated with a spherical densiometer

(Lemmon 1957). Leaf litter depth was measured as the number of leaves pierced by a







metal stake thrust into the soil at the site. Vegetation density was the number of stems

within a 50 cm radius of the site. The distances to the nearest woody stem, tree > 10 cm

DBH (diameter at breast height), crown edge of fruiting Beilschmiedia tree, and fallen log

were measured with a fiberglass measuring tape.


Statistical Analyses

Data were analyzed with tests from SAS JMP (SAS Institute 1989). Parametric
tests were used unless data violated the assumptions of normality and equal variance, in

which case nonparametric procedures were used. Survival proportions were arcsin square-

root transformed before analysis. Throughout this paper mean values are followed by 1

SD.


Results
Over the three-month fruiting season, 217 regurgitated and 27 defecated seeds were

found. One hundred twenty-nine (53%) of these were deposited beyond the crowns of

fruiting trees (dispersed), while 115 (47%) were directly under the trees (non-dispersed).

Of the dispersed seeds, 67% landed within 20 m of crown edge, but some seeds were

found up to 70 m away (Fig. 3-1). Dispersal distance (for dispersed seeds) was not

correlated with seed mass (Spearman's Rho = 0.024, P = 0.77), length (Rho = -0.073, P

= 0,39), or width (Rho = 0.014, P = 0.87).

Seeds defecated (presumably by guans) averaged lighter (Wilcoxon X2 = 6.71, P =
0.009), and shorter (X2 = 9.86, P = 0.002), but not narrower (X2 = 0.77, P = 0.23) than

seeds regurgitated by the other three species (Table 3-1). Defecated seeds were dispersed

significantly farther (39.6 17.2 m) from the parental crowns than regurgitated seeds

(14.3 13.4 m; Wilcoxon X2 = 30.49, P < 0.001).

Very few marked seeds (17%) were eaten or removed after dispersal (or placement

at random sites), and removal rates did not differ among dispersed, non-dispersed, or























30

I111
w
w
cc
I-
20
w
a.
CL

iw 10



0
0 5 10152025303540455055606570

DISTANCE FROM CROWN EDGE (m)



Figure 3-1. The average number of regurgitated or defecated Beilschmiedia seeds (+ 1 SD)
at 5 m intervals from crown edges of six fruiting Beilschmiedia trees. The first category
(distance 0) includes all the non-dispersed seeds, while the remaining categories include
only dispersed seeds.










Table 3-1. Average ( 1 SD) mass, length, and width of Beilschmiedia seeds
defecated (presumably by black guans) or regurgitated by quetzals, toucanets and
bellbirds. Sizes of regurgitated and defecated seeds were compared with Wilcoxon
rank sum test. Measurements for all seeds, including seeds placed at random sites
(N = 294), are shown for comparison.
Defecated Regurgitated All Seeds
N= 27 217 294

Mass (g) 11.25 (2.82) 13.04 (3.54)** 12.89 (3.60)
Length (mm) 43.08 (5.53) 48.18 (7.84)** 48.55 (7.65)
Width (mm) 20.40 (1.50) 20.82 (1.20)ns 20.70 (1.78)

** P < 0.01
ns P = 0.23








random sites (survival analysis; Wilcoxon X2 = 1.2, df= 2, P > 0.1; Fig. 3-2). The most

conspicuous post-dispersal seed predators were two species of beetles, one of which

(Nitidulae) buried the seeds and fed upon the rotting cotyledonary reserves. The second

beetle species (Curculionidae) consumed the seed by tunneling through the endocarp to

reach the seed. None of the seeds buried by beetles (N = 24, 8.2%) produced seedlings

(although a few germinated), and thus the beetles likely do not act as secondary dispersers.

Small rodents removed eight seeds (2.7%) and took them into burrows at least 50 cm deep.

Such a depth would preclude seedling establishment. Therefore, rodents probably are not

secondary dispersers of Beilschmiedia. Black-breasted wood-quail (Odontophorus

leucolaemus) pecked apart seven seeds (based on one direct observation and bill marks left

on pieces of seeds). An additional 11 marked seeds were removed but were not found.

Neither width, length, or mass were significant effects in logistic regression models

predicting 5-wk seed survival (Table 3-2). Seeds at dispersed and randomly located sites

were more likely to survive 5 wk than seeds at non-dispersed sites (Wald X2 = 6.96, P =

0.031) and similarly, seed survival was positively correlated with dispersal distance (Wald

X2 = 7.33, P = 0.026; Table 3-2). Seed survival was also positively correlated with

amount of leaf litter (Wald X2 = 5.97, P = 0.051), vegetation density (Wald X2 = 11.57,

P = 0.003), and date of dispersal (Wald X2 = 14.00, P < 0.001).

Virtually all caged seeds initiated germination (98%) and the majority (68%) had

established seedlings by late July, three to five months after dispersal. The average

proportion per tree of seeds that germinated, established seedlings, or survived one year,

did not differ among the three treatments (Kruskal-Wallis tests, Ps > 0.05; Fig. 3-3).

Overall, however, the number of non-dispersed seeds that established seedlings was

greater than expected if all three treatments had equal survival (X2 = 9.39, df= 2, P =
0.009).

Seed mass was not correlated with time to germination after dispersal (r = -0.086, P

= 0.15), but shoots from larger seeds grew faster after germination (r = -0.24, P < 0.001).





























S DISPERSED
S NON-DISPERSED
-U- RANDOM


DAYS AFTER DISPERSAL


Figure 3-2. Post-dispersal predation of seeds from dispersed (N = 129), non-dispersed (N
= 115), and random (N = 50) locations. Removal rates did not differ among the three
treatments.










Table 3-2. Results of logistic regressions of post-dispersal survival of marked seeds after 5
weeks, and one-year survival of seeds initially caged against habitat variables. Only
significant effects are listed. Nonsignificant variables included seed mass, seed length,
seed width, canopy cover, distance to log, distance to woody stem, and distance to 10 cm
tree.

Response R2 -2 log likelihood predictorsa
(X2)


5 wk seed survival 0.251 72.67*** + leaf litter*
+ dispersal date***
+ vegetation density***
+ distance from parent*
type (U
1 yr seedling survival 0.12 21.24*** + 3-mo seedling height**
dispersal date***

*P<.0.05, **P<0.01, ***P<0.001.
sign in front of each predictor indicates positive or negative correlation with the response



















t

R At












GERMINATION


o DISPERSED
* NON-DISPERSED
0 RANDOM


ESTABLISHMENT


1-YR SURVIVAL
1-YR SURVIVAL


Figure 3-3. Proportion of seeds germinating, establishing seedlings 3 mo after dispersal,
and surviving one year, at dispersed (N = 129), non-dispersed (N = 115), and random (N
= 50) locations. The average proportion surviving each stage did not differ among the
three treatments (Kruskal-Wallis tests, P's > 0.05).


0.8-

0.6-


0.4-

0.2-


0.0 --







Seeds that established seedlings by July 1995 averaged significantly heavier (13.5 3.5 g,

N = 165) than seeds that did not establish seedlings (12.1 3.6 g, N = 127; t = 3.2, P <

0.002), but seedlings that survived one year (13.1 3.6 g, N = 143) were not from larger

seeds than those seedlings that did not survive (12.7 3.5 g, N = 149; t = 0.87, P =

0.38). Of the seedlings that survived one year, seedling height was positively correlated

with initial seed mass (N = 143, r = 0.41, P < 0.001).

The predominant source of seedling mortality was fungal pathogens, although

many seeds resprouted several times even after fungal attack. Fungal pathogens killed

more seedlings directly beneath the parental crowns (non-dispersed) than seedlings at

dispersed or random locations (X2 = 20.29, df= 4, P < 0.001). As for the marked seeds,

some caged seeds (6.8%) were buried or eaten by beetles. Herbivory by insects and

mammals was probably the primary cause of seedling mortality after seedling

establishment, but the effect of herbivory was difficult to quantify because the seedlings

were examined too infrequently.

Seedling survival at one year was predicted by only two variables in the logistic

regression analysis (Table 3-2). In contrast to seed survival at 5 weeks, one-year seedling

survival was negatively correlated with dispersal date (X2 = 9.69, P = 0.002). Seedling

survival was positively correlated with seedling height at 3 mo (X2 = 6.59, P = 0.01; Table

3-2).

The abundance of seedlings and saplings (up to 5 m height) was higher in plots

close to the fruiting trees (paired t-test = 3.44, df= 4, P = 0.005), but the median height of

individuals was greater in the plots 20-40 m away from the trees (Wilcoxon Rank Sums;

X2 = 5.08, df= 1, P = 0.024; Fig. 3-4). Most seedlings (58%) in the plots close to

fruiting trees were less than 50 cm in height, while most individuals (70%) in the plots far

from adult trees were over 50 cm and ranged up to 5 m. Because the Beilschmiedia trees in

the study site did not produce fruits in 1994, the youngest individuals were at least 2 yr

old.





84













15 *
250
U)
< 200
=10
03 150
z
ILL""

E FA=
Ltl
w
~50

z
011 01
CLOSE FAR CLOSE FAR




Figure 3-4. Average number and height of seedlings and saplings up to 5 m tall in paired
10 x 10 m plots (N = 5 pairs) under and 20-40 m away from Beilschmiedia crowns.
Asterisks indicate significant differences (**P = 0.005, *P = 0.024).



























Figure 3-5. Average (+ 1 SD) proportion of seeds surviving post-dispersal seed predators
(a), germinating (b), establishing seedlings (c), and surviving one year (d) as functions of
distance from the crown edge. Seeds deposited directly beneath the parental crowns (non-
dispersed) are included in distance 0. The probability of recruitment (e) was calculated as
the product of each of the four previous stages for each tree. Different letters above bars
indicate significantly different means (P < 0.05) based on ANOVA and Fisher's LSD on
arcsin-transformed data. Panels with no letters had no significant differences.












GERMINATION


1.00-

0.75-

0.50-

0.25-

0.00-


S3I
0 10 20 30 40 50


ESTABLISHMENT


b





0 10 20 30 40 50


1.00-

0.75-

0.50-

0.25-

0.00-


0 1.25-
z
- 1.00-

0.75-
z



0
O 0.50 -

O 0.25 -

m 0.00-
I.


RECRUITMENT

b E



a ac
Ta L




0 10 20 30 40 50
DISTANCE (m)


0 10 20 30 40 50

1-YEAR SURVIVAL
a D
ab ab abc





0 10 20 30 40 50bc


0 10o 20 30 40 50


0
z
1.00.

0 0.75
V)
z
O 0.50
F-
O 025'
0.
0
- 0.00,
IL


1.00-

-- 0.75-

< 0.50-
m
0
C 0.25-
0.00-
0.00-


SEED SURVIVAL








The probability of recruitment per seed as a function of distance from parent trees

was calculated as the product of the probabilities of surviving seed predation, germination,

seedling establishment, and 1-yr seedling survival. Seed predation and germination were

not as important in limiting recruitment, as were seedling establishment and 1-yr survival

(Fig. 3-5). Establishment and 1 yr survival were higher for seeds under the crown and up

to 20 m from the crown than for seeds dispersed 30-40 m from the crown (Fig. 3-5). The

overall probability of recruitment varied with distance (One-way ANOVA on arcsin
transformed data, F5,20 = 5.75, P = 0.002). Recruitment probability was higher 10-20 m

from the crown than within 10 m (Fisher's LSD, P's = 0.01) or beyond 20 m (P's <

0.008; Fig. 3-5E). Recruitment was higher in the zone 20-30 m than 10-20 m from adults,

but the difference was marginally significant (P = 0.0517). Note that these probabilities do

not include the number of seeds per distance interval (Fig. 3-1), because that is a

component of dispersal quantity not quality.



Discussion
The observed spatial pattern of dispersed Beilschmiedia seeds in relation to parent

trees was similar to that of other vertebrate-dispersed trees (Howe 1993a, Laman 1996a).

The vast majority of the seeds disseminated by birds landed under or within 20 m of the

crowns of parent trees. One of the four disperser species, black guan, dispersed seeds

over greater distances than the other disperser species. If distance alone were the most

important factor in determining the probability of recruitment (i.e., high quality dispersal),

then guans might provide higher quality dispersal than the other species. Guans, however,

pass seeds through the digestive tract, and dispersed smaller seeds than the other species,

and small seeds were less likely to establish as seedlings than were large seeds. Thus,

large seeds are more likely to land near the parent trees where they face the possibility of

density-dependent mortality from beetles and fungal pathogens, and competition from

siblings. Wheelwright (1985b) also noted that quetzals, bellbirds, and toucanets are highly







selective with regard to fruit size, and tend to drop large fruits and consume smaller fruits.

These dropped fruits are not dispersed any appreciable distance by terrestrial animals, and

the seeds are not eaten very often by vertebrates.

Both germination and seedling survival are relatively high. A seed has about a 40%

chance of surviving one year regardless of dispersal distance or microhabitat Larger seeds

produced larger seedlings initially, and larger seedlings were more likely to survive one

year. The large seeds of Beilschmiedia enable the seedlings to resprout several times. The

lack of a correlation between seedling survival at one year and initial seed mass is probably

because large seeds are more likely to land under or near the parent where fungal attack and

seed predation were higher than for seeds dispersed farther.

In Howe's (1993b:4) summary of the characteristics of specialized dispersal

systems, the only post-dispersal component listed that could be construed as dispersal

quality is "seed dispersal away from parents (is) critical for recruitment." For

Beilschmiedia, seeds dispersed 30 to 50 m fared no better, and in some cases a little worse,

during the first year than seeds dispersed up to 20 m from the crowns. Many seedlings

survived more than one year under parent trees. Thus, it is unclear from this study if

dispersal of seeds away from parent trees is critical in the sense that it is absolutely

required, or critical in the sense that in leads to a higher probability of recruitment compared

to seeds directly below parents. But clearly, dispersal beyond parental crowns leads to a

higher probability of recruitment than beneath crowns.

Quality of dispersal is defined as the probability that a dispersed seed will survive to

reproductive age (Schupp 1993). Distance from parent trees is the only variable that seems

important for one-year survival of Beilschmiedia. Seed size and the measured habitat

variables had minimal effects at best. The data on one-year survival presented above

combined with the distribution of older seedlings and saplings suggest that dispersal a short

distance away from the parent tree leads to a higher probability of survival than either
longer dispersal or deposition directly below the fruiting tree. Density-dependent survival







is shown by the higher survival from seed predators and fungal pathogens with distance

and suggested by the taller saplings farther from adult trees. The resulting probability of

recruitment of one-year seedlings reaches a mode 10-20 m from the parent trees.

Approximately 20% of the seeds receive high quality dispersal by being deposited in this

zone. Whether or not seedlings in this zone retain the highest probability recruitment to

reproductive age (to satisfy Schupp's definition of dispersal quality), awaits longer-term

studies. Unfortunately, long term studies on recruitment of long-lived trees are logistically

difficult.

Even with some component of high-quality dispersal, the majority of seeds receives

no or low quality dispersal. Seed mass of Beilschmiedia ranges from 5 to 22 g, and the

largest seeds are frequently dropped by the current dispersers (Wheelwright 1985b). These

apparently too-large seeds suggest either an alternative dispersal strategy or a missing

disperser. Although the concept of determining extinct dispersers is fraught with

assumptions (Janzen and Martin 1982, Howe 1985, Hunter 1989, Witmer and Cheke

1991), Wheelwright (1985b) speculated that bare-necked umbrellabirds (Cephalopterus

glabricoUlis) may be missing from the disperser assemblage of Lauraceae in the Monteverde

continental divide area (see also Fogden 1993). Umbrellabirds were observed in the study

site on several occasions but are not known to breed there currently. Umbrellabirds could

easily swallow the largest Beilschnmiedia fruits. The addition of dispersal by umbrellabirds

would considerably alter the results presented in this study. Tapirs (Tapirus bairdii) may

also be a missing disperser. Tapir populations in Costa Rica have been greatly reduced by

hunting (Janzen 1983d). They do occur rarely in the study site, but most observations

were in the wet season when Beilschmiedia was not fruiting. Although tapir tracks were

not observed under fruiting Beilschmiedia trees, the structure of the fruit and seed is similar

to other species eaten and dispersed by tapirs (Janzen 1982a, Bodmer 1991, Fragoso

1997).








Another explanation for the large seeds avoided by birds is that the large number of

dropped or uneaten fruits that fall beneath parent trees is an alternative dispersal strategy not

adequately addressed in this study. Agoutis and possibly other rodents eat the pulp from

fallen fruits, but apparently do not eat or scatterhoard the seeds. The removal of fallen

fruits was not quantified in this study, and it is possible that rodents remove some seeds

from under parental crowns, thus providing dispersal equivalent to birds in terms of

distance. However, rodents often chew some or all of the protective endocarp in addition

to the pericarp, which may render the seeds more susceptible to disease, desiccation, or

seed predation by insects. In any case, a Beilschmiedia seed directly under the parental

crown does have a chance of survival (albeit lower than a seed dispersed 10-20 m), in

contrast to other large-seeded species (Clark and Clark 1984, Howe et al. 1985, Crawley

1992, but see Chapman and Chapman 1995).

It is generally believed that Lauraceae seeds have secondary compounds that protect

them from seed predators (Castro 1993), either by making the seeds toxic or unpalatable,

or by limiting the number of seeds an individual seed predator can physiologically tolerate

during a given day. Although the chemistry of Beilschmiedia seeds has not been

examined, the resinous fragrance of cut seeds and the low predation rates suggest that the

seeds are chemically defended. These defenses are likely an important part of the trade-offs

involved in the Beilschmiedia dispersal strategy. Large seed size is thought to foster

recruitment in the shaded understory (Foster and Janson 1985, but see Kelly and Purvis

1993, Hammond and Brown 1995). Because suitable sites for Beilschmiedia germination

and short-term survival are predicted only by a relatively short distance from parent trees,

rather than specific habitat characteristics, the large seeds may enable recruitment in the

shaded understory despite high conspecific density if they are chemically defended and can

avoid density dependent seed predators and fungal pathogens (e.g., Kitajima 1996).

Although seed predation by beetles and rodents, and seedling mortality by fungal

pathogens showed some indication of density dependence, the overall levels of mortality







were not high enough to preclude one-year survival of many seedlings under and near the

parent trees.

The data for Beilschmiedia suggest no advantage of dispersal over long distances.

Although theoretically dispersal is always advantageous for gene flow and to increase the

chance of occupying vacant sites (Hamilton and May 1977), the probability of

Beilschmiedia recruitment drops considerably 30-50 m from the parent trees, and few

seeds were observed dispersed beyond 50 m. Perhaps Beilschmiedia recruitment is

limited by edaphic factors or other habitat requirements not measured in this study. One

factor important for many tropical species is mutualistic association with mycorrhizal fungi.

Beilschmiedia seedlings increased relative growth rates after experimental inoculation

(Lovelock et al. 1996). Short-distance dispersal may increase the chance of encountering

mycorrhizal fungi similar to those associated with the parent trees, although mycorrhizae

are not known to be highly species specific (Wilkinson 1997b). Another explanation for

the advantage of short over long distance dispersal is that conditions for growth and

establishment become more unpredictable, or at least predictably not better than close to the

parent with increasing distance. The lack adaptations for long-distant dispersal in many

desert plants may be explained by the harsh environment (Ellner and Shmida 1981). For

these plants, investment in dispersal structures is wasted energy if the chances of survival

are low everywhere, and determined largely by stochastic processes. Perhaps the large

investment per seed in Beilschmiedia limits dispersal to areas near, and thus in similar

habitats to, the parent trees, but favors persistent shade-tolerant seedlings. A similar trade-

off between dispersal and seed survival has been noted for some African trees (Chapman

and Chapman 1996).

Dispersal quality for tropical trees has received little attention from ecologists

despite insights it may provide about forest dynamics (Howe 1993b, Schupp 1993), not to

mention the theoretical importance it is assumed to have (McKey 1975, Venable and Brown

1993). The results from this study suggest that dispersal quality is more a characteristic of




Full Text
SEED DISPERSAL AND POST-DISPERSAL SEED FATE OF FOUR TREE SPECIES
IN A NEOTROPICAL CLOUD FOREST
By
DANIEL G. WENNY
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
1998

To my grandfather, Donald G. Baker,
who led me on hikes in the woods of New Hampshire,
and to my kids, Malia and Jack,
with whom I can barely keep up.

ACKNOWLEDGMENTS
First and foremost I thank my advisor, Doug Levey, for his advice,
encouragement, patience, and help with all aspects of this project His input was
instrumental in its success, and I will always be grateful for his help. He found a few
seeds even though the sunbittem eluded him. If I discover a new species, I will name it
after him in appreciation. I also thank my committee, Karen Bjomdal, Colin Chapman,
Jack Putz, and Buzz Holling (also Frank Bonaccorso and Pete Feinsinger at an earlier
stage) for help developing the research proposal, clarifying the direction of the dissertation,
and editing the resulting manuscripts. I feel grateful they pointed me in the right direction
and let me take a few steps. Other members of the faculty were helpful with advice on
experimental design, statistical analysis, seminar presentations, and general reality checks:
Rich Kiltie, Dave Steadman, Mike Miyamoto, Frank Nordlie, Collette St. Mary, Brian
McNab, Larry McEdward, John Anderson, and Craig Osenberg. Nat Wheelwright,
George Powell, Carlos Guindon, Nalini Nadkami, Jack Longino, and Greg Murray also
provided encouragement and advice.
This research would not have been possible without valuable field assistance from
Victor Peréz, Ricardo Guindon, Jason Bennett, and Wendy Gibbons. I thank them
sincerely for their hard work under less than ideal conditions and sometimes with vague
instructions. Many other people helped with the project, either directly or indirectly. The
Monteverde community is a unique and wonderful place, and the two and a half years I
spent there were among the highlights of my life. I especially thank the Guindon,
Campbell, and Rockwell families for making me (and Wendy) feel so welcome there.
Special thanks go to Benito, Carlos, and Tomás Guindon for sharing their knowledge of
in

the area, and to Greg Murray, Mauricio Garcia, Rodrigo Solano, Bill Haber, Willow
Zuchowski, Frank Joyce, Alan Pounds, and Alan and Karen Masters for patiently
answering ray endless questions. Finally, thanks go to Judy Poe and Bruce Pack for
renting us their beautiful octagonal house on the cliff edge, and to Leyn Rockwell for
emergency repairs.
Crucial financial support at the beginning of this project was received from the
Organization for Tropical Studies: a pilot study grant (Pew Charitable Trust), and a
Tropical Fellowship (Jesse Smith Noyes Foundation). Thanks go to Lucinda McDade and
Shaun Bennett for facilitating OTS support. Other early funding from Sigma Xi Grants-in-
Aid of Research and American Museum of Natural History (Frank M. Chapman Memorial
Fund) is gratefully acknowledged. The bulk of the fieldwork was supported by a
Dissertation Improvement Grant from the National Science Foundation (DEB 93-21553).
Indirect financial support was also contributed by the Gibbons and Wenny families. I am
also grateful for logistic support from the Organization for Tropical Studies, Monteverde
Institute (e-mail access), and the Tropical Science Center (lab space). Most of all, I greatly
appreciate the many years of support from the UF Department of Zoology.
I thank my extended family for encouragement, visits to Costa Rica during the three
years of fieldwork, and for not asking too many times what exactly I was doing. Most
important was support from my wife, Wendy Gibbons, who helped with fieldwork, edited
drafts of proposals and papers, and was generally nice when I needed a friend. Thanks go
to everyone who spent a day or two in the field in exchange for a tour. I apologize to my
mother, who helped carry equipment 3 miles to the study site after my promise of a quetzal,
which I was unable to find that day or on a second visit a year later.
Finally, I would like to note my appreciation of the zoology, botany, and wildlife
graduate students who made my tenure here more enjoyable. I thank them for discussions,
editing manuscripts, statistical advice, suffering through practice seminars, scheming up
side projects, and saying they liked the Ozric Tentacles tapes I made for them: Seth
IV

Bigelow, Scot Duncan, Ron Edwards, Susan Moegenburg, Jay O'Sullivan, Greg Pryor,
Rafael Samudio, Lenny Santesteban, Mark Spritzer, Markus Tellkamp, and... Special
thanks go to Susan Moegenburg for being the ideally cheerful officemate and to Jason
Evert for being the ideally invisible officemate.
v

TABLE OF CONTENTS
ACKNOWLEDGMENTS iü
ABSTRACT viii
CHAPTERS
1 GENERAL INTRODUCTION 1
2 SEED DISPERSAL, SEED PREDATION, AND SEEDLING
RECRUITMENT OF OCOTEA ENDRESIANA (LAURACEAE)
Introduction 6
Study Site 9
Study Species 11
Methods 15
Results 23
Discussion 47
Conclusion 64
3SEED DISPERSAL OF A HIGH-QUALITY FRUIT BY SPECIALIZED
FRUGIVORES: HIGH-QUALITY DISPERSAL?
Introduction 66
Study Site 70
Study Species 71
Methods 73
Results 76
Discussion 87
4TWO-STAGE DISPERSAL OF TWO SPECIES OF GUAREA (MELIACEAE)
Introduction 93
Study Site 95
Study Species 95
Methods 96
Results 99
Discussion 107
vi

5 ADVANTAGES OF SEED DISPERSAL: A RE-EVALUATION
OF DIRECTED DISPERSAL
Introduction 110
Advantages of Dispersal 112
The Diffuse Mutualism Paradigm 118
Classic Examples of Directed Dispersal 120
Other Examples of Possible Directed Dispersal 124
Conclusion: Going the Distance, and Beyond 134
APPENDIX OCOTEAENDRESIANA SITE LOCATIONS 138
REFERENCES 152
BIOGRAPHICAL SKETCH 178
vn

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
SEED DISPERSAL AND POST-DISPERSAL SEED FATE OF FOUR TREE SPECIES
IN A NEOTROPICAL CLOUD FOREST
By
Daniel G. Wenny
May, 1998
Chairman: Douglas J. Levey
Major Department: Zoology
Seed dispersal by animals is thought to be important in tropical forests because
most trees in these forests produce fleshy or añílate fruits adapted for consumption by birds
and mammals. Very little is known about what happens to seeds after dispersal, yet post¬
dispersal fate is important in determining patterns of plant recruitment I studied dispersal
patterns and post-dispersal seed fates of four tree species in Monteverde, Costa Rica. The
four species studied included two species of Lauraceae (Ocotea endresiana and
Beilschmiedia péndula) and two Meliaceae (Guarea glabra and G. kunthiana). I determined
locations of dispersed seeds by following birds until they regurgitated or defecated seeds
and by systematically searching the study site for recently dispersed seeds. For each of the
four species, approximately 75% of the seeds dispersed by birds were deposited within 25
m of the parent trees. Ocotea endresiana showed a bimodal pattern in which bellbirds
(Procnias tricarunculata) dispersed many seeds under song perches and thus produced a
second peak in seed rain in addition to the peak near parent trees. Marked seeds were used
to determine if post-dispersal removal resulted in seed predation or secondary dispersal.
Removal rates were high for Ocotea (99% removed) and Guarea (85-90%), and low for
viii

Beilschmiedia (17%). Secondary dispersal was observed for both Guarea species but not
for the Lauraceae species. Secondary dispersal of Guarea resulted in a slight increase in
overall dispersal distance, and a shift in microhabitat.
The spatial pattern of seedling recruitment was bimodal for Ocotea, reflecting the
initial pattern of dispersal. Beilschmiedia recruitment was not bimodal and was highest in
the zone approximately 10 m from the parental crowns. Guarea recruitment was not
studied in the field, but in growing house experiments, buried G. glabra seeds germinated
and established seedlings at a higher rate than seeds on the soil surface, suggesting a
benefit to secondary dispersal. The results of this study indicate that determining dispersal
pattern and post-dispersal fate are important to assess the influence of seed dispersers on
patterns of seedling recruitment
IX

CHAPTER 1
GENERAL INTRODUCTION
"With plants there is a vast destruction of seeds. Seedlings, also, are destroyed in vast
numbers by various enemies." Charles Darwin (1859:78).
"Plants are not charitable beings." William J. Beal (1898:84).
Seed dispersal by animals is widespread and recognized as important by virtue of
the large number of plant species with fruits presumably adapted for animal consumption.
In tropical forests in particular, 70-90% of the trees and shrubs are thought to be dispersed
by animals (Stiles 1985, Howe 1986, Willson et al. 1989, Jordano 1992). This seed
dispersal interaction is characterized as a diffuse mutualism because most plants have
multiple dispersal agents and most seed dispersing animals disperse several to many
different plant species. Thus, fruit-frugivore coevolution generally occurs among groups
of plants and groups of dispersers rather than at the species-specific level seen in many
pollination systems (Wheelwright and Orians 1982). The advantage of this mutualism to
animals is a nutritive reward, whereas the advantages to plants are escape from density-
dependent mortality near the parent plant and arrival in suitable sites for establishment
Arrival in suitable locations can be either a random process, in which widespread
dissemination of seeds would increase the chance of colonizing a good site (colonization
hypothesis), or a nonrandom process in which seeds are directed to suitable sites by
attracting certain dispersal agents (directed dispersal hypothesis). Escape and colonization
are thought to be the main advantages of seed dispersal for most plants (Howe 1986)
although few studies have examined these hypotheses in detail.
1

2
One of the main factors limiting our understanding of the advantages of dispersal is
that most previous studies have examined aspects of dispersal without considering
subsequent stages that may also influence recruitment. Dispersal is part of the plant
recruitment process, which is composed of several stages, and most previous studies have
been stage-specific. Zoologists have focused on fruit removal, foraging behavior, and gut
treatment of seeds, while botanists have studied germination and seedling growth (Howe
1993b). Seed predation has attracted the attention of both zoologists and botanists, but few
studies have integrated all the stages of recruitment in one study (but see Herrera et al.
1994). Our understanding of the importance of dispersal by animals is limited because we
know so little about where animals take seeds and what happens to them after dispersal.
Furthermore, many studies assume that when seeds are dispersed they will either live or die
in that place. Recent studies, however, indicate that many seeds experience a second stage
of dispersal by ants, rodents, dung beetles, and other animals (Estrada and Coates-Estrada
1991, Levey and Byme 1993, Fragoso 1997). Thus, conclusions based on the pattern of
initial dispersal may be misleading.
The goal of this study was to examine the stages of recruitment, including seed
dispersal, seed predation, secondary dispersal, germination, and seedling establishment, to
link patterns of dispersal with patterns of recruitment. This study took place in
Monteverde, Costa Rica, which has been the site of some of the key studies in the seed
dispersal literature (e.g., Wheelwright and Orians 1982, Murray 1988, Nadkami and
Wheelwright in press). In Chapters 2 and 3,1 present studies of two tree species in the
plant family Lauraceae, Ocotea endresiana and Beilschmiedia péndula. Lauraceous species
are an important component of Neotropical montane forests in terms of species richness
(Gentry 1990, Haber et al. 1996) and the variety of birds that are at least partially
dependent on their fruits (Wheelwright et al. 1984). In Chapter 4,1 present a study of two
species of Meliaceae, Guarea glabra and G. kunthiana. The main objective for all four
species was to find the locations where seeds are dispersed and follow the fate of those

3
seeds after dispersal to link patterns of dispersal with patterns of recruitment. All four of
these tree species are common in the Monteverde forests, and understanding their dispersal
and recruitment patterns should shed light on how the forest works in replacing itself or
perhaps in changing over time.
For the study presented in Chapter 2,1 followed birds from fruiting trees and
systematically searched the ground for dispersed Ocotea endresiana seeds. I protected the
located seeds in situ with wire mesh cages to determine germination rates in the absence of
seed predation by mammals. After germination I removed the cages to assess seedling
establishment as a function of habitat characteristics. To study seed predation, I placed
marked seeds next to the caged seeds and recorded their fate over regular intervals. The
fates of caged and marked seeds were also studied for seeds placed at random locations.
These data were used to determine if dispersers take seeds to a nonrandom set of sites with
characteristics predictably favorable for seedling survival. The results of this study show
that most seeds dispersed by birds landed within 25 m of parent trees, but approximately
12% are dispersed to gaps by male bellbirds (Procnias tricarunculata) which have habitual
song perches on exposed branches. Post-dispersal seed predation is very high but did not
differ between seeds in gaps and in the forest understory. Secondary dispersal was not
observed. Of the seeds that survived seed predators, most germinated, and seedlings in
gaps grew taller and survived at a higher rate than seedlings in the forest understory. This
study is apparently the first example of directed dispersal by birds of a tropical tree.
In Chapter 3,1 present data on the dispersal and post-dispersal fate of
Beilschmiedia péndula. In this study I determined dispersal locations solely by searching
the ground because the long retention times of these large seeds (12 g) precluded following
the birds long enough to find the seeds. Otherwise the methods were similar to those used
for Ocotea. The distribution of dispersed seeds was similar to that of Ocotea except that
few seeds were dispersed to gaps. Seed predation was very low, and again, secondary
dispersal was not observed. Habitat characteristics were not useful in predicting seedling

4
survival. The highest probability of seedling recruitment was 10-20 m from the edge of
parent tree crowns. I discuss the results in terms of dispersal quality and suggest that for
this species dispersal within 20 m of the crown is relatively high-quality dispersal and
approximately 20% of the seeds receive such treatment
In Chapter 4,1 present data on the dispersal systems of two species of Guarea
(Meliaceae) that produced fruit at a time when Ocotea and Beilschmiedia did not. As in
the Beilschmiedia study, dispersed seeds were located by systematic ground searches.
The distribution of dispersed seeds was similar to the previous two species, but 30-45% of
the seeds had a second stage of dispersal, presumably by agoutis (Dasyprocta punctata).
Agoutis buried many of the seeds they removed in shallow surface caches. Thus,
secondary dispersal resulted in a rearrangement of the initial seed shadow.
In Chapter 5,1 further explore the general idea of dispersal directed
disproportionately to sites suitable for survival. The implication of directed dispersal is a
disproportionate effect on plant recruitment, but examples of directed dispersal are thought
to be rare and unusual (Howe 1986). Nevertheless, directed dispersal is well established in
a least three systems: scatterhoarding pine seeds by nutcrackers and jays, dispersal of
mistletoes by small birds, and dispersal of understory herbaceous plants by ants. I review
the literature and present evidence that directed dispersal may be more common than
previously believed, but is likely to be subtle in many instances. Possible examples include
dispersal by animals that have habitual perches or defecation sites, dispersal to nurse plants
in arid ecosystems, dispersal to perches in successional landscapes, and secondary
dispersal by ants, rodents, and dung beetles. Our understanding of directed dispersal is
limited by a lack of studies that integrate dispersal patterns and post-dispersal fate.
The main theme of this dissertation is that determining patterns of dispersal
generated by animals is a crucial research need. This type of data is difficult and time-
consuming to collect, but should yield great insight into the role of dispersal in ecosystem
function, as well as the evolution of plant-animal interactions. In addition, a greater

5
understanding of seed dispersal and subsequent post-dispersal fates may be useful in
conservation and restoration ecology.

CHAPTER 2
SEED DISPERSAL, SEED PREDATION, AND SEEDLING RECRUITMENT OF A
NEOTROPICAL MONTANE TREE
"No one has succeeded in measuring actual patterns of seed dispersal produced by different
bird species for any bird-dispersed plant species. Even if we could determine where birds
dropped all seeds, it would still be difficult to rank different bird species according to
dispersal quality because so little is known about seedling and sapling microhabitat
requirements." NathanielT. Wheelwright (1988:832)
Introduction
Seed dispersal determines the spatial arrangement and physical environment of
seeds from which the next cohort of seedlings is selected. The links between seed
dispersal and seedling recruitment, however, are poorly understood. Dispersal is the first
of a series of events, one or more of which may be important in limiting recruitment. For
vertebrate-dispersed plants, these stages include fruit removal, seed dissemination, post¬
dispersal seed predation, potentially secondary dispersal, germination, and seedling
establishment Previous studies have generally focused on only one or a few of these
stages (Herrera et al. 1994, Schupp and Fuentes 1995).
Despite many studies on fruit consumption and seed-handling by birds and
mammals, seed shadows (spatial patterns of dispersed seeds) generated by animals are
poorly known (Willson 1993). While concentrations of seeds under or near fruiting trees
have been noted and, indeed, are expected (Janzen 1970, Howe 1989, Willson 1993), the
pattern of seed distribution farther from fruiting trees is likely to be the most important part
of the seed shadow for plant fitness and population ecology (Portnoy and Willson 1993,
Schupp and Fuentes 1995). In particular, the tail of the distribution is often heterogeneous
6

7
as a result of disperser behavior (McDonnell and Stiles 1983, Hoppes 1988, Chavez-
Ramirez and Slack 1994, Julliot 1996). Fruit-eating vertebrates differ in foraging behavior
(Cruz 1981, Santana C. and Milligan 1984, Trainer and Will 1984, Moermond and
Denslow 1985), fruit removal rates (Howe and Vande Kerckhove 1980, Bronstein and
Hoffman 1987, Englund 1993), seed-handling techniques (Janzen 1983a, Levey 1987,
Corlett and Lucas 1990, Stiles and Rosselli 1993), and effects on seed germination
(Krefting and Roe 1949, Compton et al. 1996, Traveset and Willson 1997), but few
studies have compared seed shadows generated by different dispersers (but see Thomas et
al. 1988, Chavez-Ramirez and Slack 1994), or the consequences of dispersal pattern on
recruitment (but see Howe 1989,1990a, Herrera et al. 1994, Schupp 1995, Schupp and
Fuentes 1995).
After dispersal, seed predation and seedling mortality are often extensive for
tropical trees (DeSteven and Putz 1984, Howe et al. 1985, Chapman 1989a, Hammond
1995, Cintra and Homa 1997, Peres et al. 1997) and may be important in influencing the
spatial pattern of recruitment (Connell 1970, Janzen 1970, Harper 1977, Hubbell 1980,
Clark and Clark 1984, Becker et al. 1985, McCanny 1985, Howe 1989). In addition,
studies on forest dynamics have suggested that canopy gaps are often crucial recruitment
sites for tree seedlings (Hartshorn 1978, Denslow 1987, Swaine and Whitmore 1988,
Swaine 1996). Few studies, however, have examined seed predation as a link between
seed dispersal and seedling recruitment In fact, aside from research on Virola in Panama
(Howe and Vande Kerckhove 1980, Howe et al. 1985, Howe 1990b, 1993a), Phillyrea in
Spain (Herrera et al. 1994, Jordano and Herrera 1995), and Ficus in Borneo (Laman 1995,
1996a, b), it is difficult to find studies that consider the various stages leading to
recruitment for any species of fleshy-fruited, vertebrate-dispersed plant. Furthermore,
some plant species typically considered to be bird or mammal-dispersed, may have a
second (or even third) stage of dispersal by an entirely different dispersal vector (Roberts
and Heithaus 1986, Clifford and Monteith 1989, Forget and Milleron 1991, Estrada et al.

8
1993, Levey and Byrne 1993, Nogales et al. 1996). To understand the relative importance
of seed dispersers and seed predators in forest dynamics, it is necessary to study each stage
leading to recruitment (and, ideally, to reproductive age).
The main factor that limits our understanding of the link between seed dispersers
and seedling recruitment is the difficulty of finding dispersed seeds (Harper 1977, Janzen
1983a, Schupp and Fuentes 1995). Thus, most studies on vertebrate-dispersed plants have
assessed the importance for plant fitness of different dispersers by the amount of fruit in the
diet and gut treatment of seeds and have not considered post-dispersal fate of seeds
(reviewed by Howe 1986, Stiles 1989, Jordano 1992, Willson 1992, Levey et al. 1994).
Similarly, studies on seed predation and seedling recruitment have relied on experimentally
dispersed seeds, usually without data on the actual patterns of seed rain (Price and Jenkins
1986, Crawley 1992, Hulme 1993, Schupp 1995).
The objective of this study was to link patterns of seed dispersal and seed predation
with those of seedling recruitment, by determining the locations of naturally dispersed
seeds, and then following the fate of those seeds after dispersal. I studied a common tree,
0cotea endresiana Mez (Lauraceae), in old growth montane forest in Costa Rica, to answer
the following questions: 1) Do the five main fruit consumers of O. endresiana generate
different patterns of seed rain, and subsequently influence germination and the pattern of
seedling recruitment? 2) What proportion of the dispersed seeds are subsequently killed by
seed predators, and what animals are the most important seed predators? 3) Do
scatterhoarding rodents provide a second stage of dispersal? 4) What proportion of the
dispersed seeds germinate and establish as seedlings under different microhabitat
conditions? To answer these questions, I used a combination of natural and manipulative
experiments to assess the types of sites to which seeds are dispersed and how the habitat
characteristics and spatial arrangement of those sites relative to the parent trees influence
post-dispersal survival.

9
A secondary objective was to assess three hypothesized advantages of dispersal: 1)
escape from high mortality caused by distance- or density-dependent factors near
conspecifics (escape hypothesis); 2) colonization of rare, unpredictable, ephemeral sites,
such as treefall gaps (colonization hypothesis); and 3) directed dispersal to particular
microhabitats suitable for survival (directed dispersal hypothesis) (Howe and Smallwood
1982, Howe 1986, Willson 1992). A fourth advantage of dispersal, gene flow (Levin and
Kerster 1974, Hamrick and Nason 1996), was beyond the scope of this study. These
hypotheses are not mutually exclusive and can be assessed only if dispersal sites and post¬
dispersal fates are known. Evidence for the escape hypothesis should include higher rates
of seed predation and/or seedling mortality closer to the parent tree, or where seeds and
seedlings occur in dense concentrations. The difference between colonization and directed
dispersal is whether seeds arrive and survive in specific habitats more often than expected
by chance (Howe and Smallwood 1982, Howe 1986, Murray 1986, 1988, Schupp et al.
1989, Willson 1992). In particular, if seeds are dispersed to sites with random
microhabitat characteristics then the colonization hypothesis is supported. If certain species
predictably disperse seeds to a nonrandom subset of the available microhabitats and post-
dispersal survival is predictably higher in those microhabitats than in random sites, then the
directed dispersal hypothesis is supported. Escape and colonization are generally thought
to be the main advantages for tropical trees dispersed by fruit-eating birds and mammals
(Howe 1986, Murray 1988).
Study Site
This study was conducted from May 1993 to July 1996 in the Monteverde Cloud
Forest Preserve (10°12’N, 84°42W) in the Cordillera de Tilaran, northern Costa Rica.
This 10,000 ha preserve is adjacent to other protected lands encompassing one of the
largest areas of relatively unbroken forest in Costa Rica. The study area contains the full
complement of birds and mammals, including top predators, that have historically occurred

10
in the area (Young and McDonald in press), although certainly the abundances of some
species have changed (Fogden 1993). Other characteristics of the fauna are described by
Nadkami and Wheelwright (in press).
The study area was in relatively undisturbed lower montane rain forest (Hartshorn
1983) along the continental divide at 1600 m elevation. A 5 ha area, 500 m from the
beginning of the Valley Trail (Sendero El Valle), was mapped and marked into 10 x 10 m
quadrats with PVC tubing at every grid point. The site's vegetation is classified as leeward
cloud forest by Lawton and Dryer (1980). The canopy is 25-30 m tall and dominated by
several species of Lauraceae, Sapiurn oligoneuron (Euphobiaceae), Ficus crassiuscula
(Moraceae), Inga sp. (Leguminosae) and Pouteria viridis (Sapotaceae) (Lawton and Dryer
1980). The relatively dense understory (compared to lowland rainforest) is dominated by
Rubiaceae, Acanthaceae, Gesneriaceae, and Heliconiaceae. The vegetation of the area is
described in more detail by Lawton and Dryer (1980) and Nadkami et al. (1995).
Canopy gaps caused by falling trees and branches are common and may be an
important habitat for recruitment of some plants in the study area (Lawton and Putz 1988,
Lawton 1990). Although gaps may be formed at any time of year, most are formed during
the dry season when strong winds are more frequent (K.G. Murray, personal
communication). Gaps are also formed under standing dead trees, especially Sapiurn,
which shed large limbs for several years before the remainder of the trunk falls (D. Wenny,
personal observation, C. Guindon, personal communication). Approximately 5.3% of the
study area was in gaps ^10 m^ with vegetation <2 m tall (sensu Brokaw 1982a), including
two Sapiurn gaps. Average canopy cover (measured with a spherical densiometer 1.2 m
above the ground) in gaps and gap edges was 90.7% (± 9.6, range 60-92, N = 78), while
closed canopy forest averaged 96.2% (± 4.3, range 91-100, N = 234) cover. These values
are similar to those at other Neotropical forest sites (Howe et al. 1985, Levey 1988b, Clark
1994).

11
Most woody plants at Monteverde are animal-dispersed, as is typical of Neotropical
forests in general and tropical montane forests in particular (Gentry 1982, Howe and
Smallwood 1982, Tanner 1982, Stiles 1985, Levey and Stiles 1994). Bird dispersal
predominates: 77% of the understory trees and shrubs and 63% of the canopy trees have
fruit morphology suggesting dispersal by birds (Stiles 1985). At least 70 resident and
migrant bird species feed on the fruits of over 150 species of trees and shrubs
(Wheelwright et al. 1984). Most of these birds regurgitate or defecate seeds intact
(Wheelwright et al. 1984, Murray 1988).
The average annual rainfall at 1520 m on the Pacific slope about 3 km from the
study site is approximately 2500 mm, with most of the precipitation occurring between
May and November. Actual rainfall in the study site was probably greater than 2500 mm,
but the seasonal pattern was similar (Nadkami and Wheelwright in press). Range gauges
underestimate the amount of precipitation from mist and cloud interception, which
contribute up to 50% of the precipitation in some Neotropical montane forests (Cavelier
1996). Temperatures recorded at the study site during this project ranged from 15° to
22°C.
Study Species
The Lauraceae is an important family in Neotropical forests in terms of species
richness, a food resource for birds, and economic value (Wheelwright 1983,1985a, 1991,
Burger and van der Werff 1990, Gentry 1990, Martinez-Ramos and Soto-Castro 1993,
Guindon 1996). Members of the Lauraceae are also important dietary components for
frugivorous birds in Africa, Southeast Asia, and Australia (Crome 1975, Snow 1981, Sun
et al. 1997). Ocotea endresiana Mez (listed as O. austinii in Wheelwright et al. 1984,
Wheelwright 1985a, 1986) is a common canopy tree in montane forests in central and
northwestern Costa Rica from 1100-2300 m (Burger and van der Werff 1990). In the
Monteverde area it occurs from 1550-1700 m along the continental divide. Twenty to 30

12
other species of Lauraceae occur in the same area (Wheelwright 1985a, 1986, Haber
1991). Ocotea endresiana begins flowering in August, the mid-rainy season, and is
pollinated by small flies and other insects. Fruits ripen the following May and June, the
early rainy season. Ripe fruits (18x9 mm, 1.2 g) have blue-black skin, lipid-rich pulp,
and are held in a shallow reddish receptacle, typical of the Lauraceae genera Ocotea and
Nectandra (Wheelwright et al. 1984, Burger and van der Werff 1990). Most of the volume
of the fruit is a single seed (15 x 8 mm, 0.75 g, fresh weight) composed of a small embryo
and two large cotyledons surrounded by a thin (0.3 mm) seed coat. Although small
compared to some other Lauraceae species, O. endresiana fruits and seeds are among the
largest at Monteverde (Wheelwright et al. 1984). Thirty-eight O. endresiana trees were in
the study area; most data in this study are from 21 trees in the core 5 ha study site (Fig. 2-
1). Fruits and seeds for some experiments were collected from trees on the periphery of
the main study area. Large fruit crops (2110 ± 1495 fruits/tree) were produced in 1993 and
1995, but not in 1994. Each O. endresiana tree was an average of 36.5 m (± 12.1) from
the three closest conspecific adults.
The most common avian visitors to fruiting Lauraceae include the largest and most
frugivorous species at Monteverde (Wheelwright et al. 1984). Ocotea endresiana fruits are
eaten primarily by five species of birds: emerald toucanet (Ramphastidae: Aulacorhynchus
prasinus), resplendent quetzal (Trogonidae: Pharomachrus mocinno), three-wattled bellbird
(Cotingidae: Procnias tricarunculata), mountain robin (Turdidae: Turdus plebejus) and
black guan (Cracidae: Chamaepetes unicolor), all of which breed in the study site during
the fruiting season. The first four species swallow fruits intact and regurgitate seeds in
viable condition. Seeds in this size range are generally regurgitated 15-30 minutes after
ingestion (Wheelwright 1991). Guan defecate seeds in viable condition (Wheelwright
1991). These bird species (especially quetzals) typically remain in a fruiting tree after
eating several fruits, and often regurgitate seeds under the same tree or nearby
(Wheelwright 1983, 1991). Seed processing times for guans are unknown, but they

Figure 2-1. Map of the study site showing locations of fruiting Ocotea endresiana trees
(dark circles with numbers) and bellbird song perches (X). Thick lines indicate streams
and thin lines are trails. Grid lines are 50 m apart North is at the top. Data on seed
locations for each of the trees numbered on this map are presented in the Appendix.

14

15
generally leave a fruiting tree before defecating the seeds from that foraging bout (D.
Wenny, personal observation). Several other bird species (Wheelwright et al. 1984), as
well as spider monkeys (Ateles geoffroyi), occasionally eat O. endresiana fruits and
probably disperse viable seeds (e.g., Chapman 1989a).
Methods
A combination of observational and experimental data was collected to determine
the probability of seedling establishment and one year survival for seeds dispersed by
birds. In particular, the focus of this project was to find the locations of seeds dispersed
naturally and to compare the post-dispersal fate of those seeds with experimentally placed
seeds to determine the influence of dispersers on seedling recruitment The locations of
seeds naturally dispersed by birds were classified in two categories according to their
position relative to the parental tree crown to reflect the prevailing assumption that seed
deposition under parental crowns leads to a low probability of survival (Janzen 1970,
Howe et al. 1985, Hulme 1997). Sites were classified as non-dispersed if directly under
the crown of a fruiting O. endresiana tree or as dispersed if not under such a tree. Even
though all dispersed and non-dispersed seeds in this study had been regurgitated or
defecated by birds (and therefore "dispersed" sensu Janzen 1983a), it is generally believed
that seeds deposited under the parent trees have very little chance of survival (e.g., Janzen
1970, Howe et al. 1985). Therefore, the term "dispersed" in this paper always refers to
seeds that are not deposited directly under the crown of a fruiting conspecific, and "non-
dispersed" refers to seeds that are regurgitated or defecated by birds directly under the
crown of a fruiting tree. For convenience, I will refer to sites of dispersed seeds as
dispersed sites and of non-dispersed seeds as non-dispersed sites, even though the sites
themselves were not capable of being dispersed.

16
Seed Dispersal
Seeds were located by following birds until they dropped, regurgitated, or
defecated seeds, and by systematic ground searches. These ground searches were started
at the base of a fruiting tree and proceeded along 10 m wide transects to 50-60 m from the
trunk. It was impossible to search the entire site with equal intensity, but an effort was
made to cover the entire site at least once every two weeks, so that over the course of the
two-month season of fruit ripening each 10 x 10 m plot was checked at least four times.
Bird observations began whenever one of the five major consumers of O. endresiana fruits
was located, and continued as long as the bird was visible or until the bird regurgitated or
defecated seeds. A total of 193 hr was spent following birds.
In addition to these dispersed and non-dispersed seeds, other seeds were
experimentally placed at randomly selected sites to determine if birds deposited seeds in
sites with specific or random characteristics. Random numbers generated on a hand
calculator were used as coordinates of sites. The post-dispersal fate of seeds at these
random locations was compared to the fate of seeds at the dispersed and non-dispersed
locations.
Microhabitat characteristics. For all sites and seeds I measured seed characteristics
and microhabitat variables I thought might influence seed predation, germination, or
seedling survival. Seed length and width were measured with dial calipers, and seed mass
was measured with a 5 or 10 g spring balance. Canopy cover was estimated with a
spherical densiometer (Lemmon 1957). Leaf litter was the number of leaves pierced by a
metal stake thrust into the soil at the site. Vegetation density was the number of stems
within a 50 cm radius of the site. The distances to the nearest woody stem > 1 m in height,
tree > 10 cm DBH (diameter at breast height), trunk of fruiting Ocotea tree, and fallen log
were measured with a fiberglass measuring tape. These variables were selected based on
their demonstrated importance in previous studies. Seed size may influence the probability
of seed predation (Price and Jenkins 1986, Hulme 1993) or seedling size (Howe and

17
Richter 1982). Canopy cover (light availability) is known to be an important factor for
germination and tropical seedling growth (Howe et al. 1985, Mulkey et al. 1996, Swaine
1996). Vegetation density and distance to objects may influence rodent activity and seed
predation (Smythe 1978, Kiltie 1981, Kitchings and Levey 1981). Finally, leaf litter may
influence seed predation or germination (Schupp 1988a, Molofsky and Augspurger 1992,
Myster and Pickett 1993).
Post-dispersal Seed Predation
At each dispersed, non-dispersed, and random site, a marked seed was used to
assess rate of seed predation, to identify seed predators, and to determine if secondary
dispersal occurred. For this treatment I used seeds that were regurgitated by birds and
collected under fruiting trees adjacent to the study site. Seeds were marked by gluing 50-
75 cm of unwaxed dental floss to the seed, and tying about 50 cm of flagging tape to the
distal end of the floss. Because the glue held best on seeds with a dry seed coat, seeds
were taken inside a lab room and allowed to dry for 1-3 h before gluing. Each marked seed
was placed at a site the next morning. To determine if presence of dental floss and flagging
tape influenced seed removal, I conducted a pilot study in May 1993. Fifty marked seeds
and 50 unmarked seeds were placed singly at random locations and censused one week
later. Removal of marked and unmarked seeds was not significantly different (49 and 47
seeds removed, respectively; X2 = 0.047, df = 1, P > 0.10). Thus, the marking of seeds
was assumed to have no effect on seed removal.
Note that these marked seeds were placed next to the cages protecting the original
seeds deposited by birds. The cages may have served as cues for visually-searching seed
predators, but I believe that is unlikely for two reasons. First, at the beginning of the study
in 1993, cages were novel items and would not have been associated with seeds by the
seed predators. Second, because most marked seeds were removed, but the cages were left
at the sites for several months, most cages were not good indicators of seed availability.

18
Olfactory cues left by marking and handling the seeds are another possible confounding
factor, but the frequent rains probably diminished these (Whelan et al. 1994).
In 1993, all marked seeds were censused once each week for three weeks and then
at weeks 5 and 10. Because most seeds were removed during the first week after dispersal
in 1993, marked seeds in 1995 were checked more often: on days 1,3, and 7, and once
each week afterwards until week 10. If a marked seed was removed, the surrounding area
was searched to find the flagging tape-dental floss assembly. The end of the floss where
the seed had been attached was examined to determine the fate of the seed. If a seed was
entirely removed or if a piece of the seed coat remained attached to the floss, the seed was
classified as killed by seed predators because captive Peromyscus treated seeds with dental
floss in that manner when they consumed O. endresiana seeds (D. Wenny, unpublished
data). The distance from the dispersal site to the predation site was measured and each site
of a removed seed was classified in the following categories: in burrow; in, under, or near
a fallen log (> 10 cm diameter); at the base of a large tree; in dense vegetation (defined as at
least 50% cover of plants less than 50 cm tall within 50 cm radius of site, assessed
visually); under or beside a clump of fallen branches, or on the leaf litter. If a marked seed
was removed but not eaten, it was left in the new location and included in subsequent
censuses.
Distance effect Two other experiments (one in 1995 and one in 1996) were
conducted with marked seeds to determine the effect of distance on seed removal, and
whether most seed removal occurs during the day or at night. In 1995 one seed was placed
at the base of each of 12 fruiting trees and at 20,40,60, and 80 m from the base of the
tree. The orientation of each transect was carefully selected to avoid approaching within 80
m of other fruiting O. endresiana trees. Seeds at these sites were censused on the same
schedule as the other marked sites in 1995. In 1996 a similar experiment was conducted,
except the seeds were 5 and 40 m from each of 22 trees. Shorter and fewer distances were
used so that more trees could be included, and because data from the previous year, as well

19
as other studies (e.g., Howe et al. 1985) indicated that the distance effect can be detected
beyond 20 ra from conspecific adult In this case, the compass directions were randomly
selected although some directions were discarded if the 40 m treatment was not at least 40
m from all fruiting conspecifics. This experiment was run twice, once in the early fruiting
season (late May) and once two weeks later during the peak fruiting season (mid-June).
During the second of these trials, all sites were checked at dusk and dawn for two
consecutive days to determine whether most seed removal occurred during the day or night.
Removal during the day likely could be attributed to the diurnal agouti, while removal at
night likely would be from nocturnal species, such as Peromyscus or other small rodents.
Exclosures. To determine the amount of seed predation attributable to large or small
mammals, 38 sets of exclosures were established in early 1995. Each set had three
treatments, each 1 m^: (1) "no rodents" exclosures made of 1 cm^ galvanized wire mesh
0.9 m tall; (2) "small rodents only" exclosures made of chicken wire mesh with 6 cm holes,
0.9 or 1.35 m tall; and (3) "all rodents" control plots with only wooden stakes marking the
comers. The bottom edges of the wire exclosures were buried 5 cm below ground and
held with two or three 25 cm metal stakes on each side. The comers were supported by
1.25 m lengths of 10 mm thick metal stakes. The tops were open to allow normal
accumulation of fallen leaves. Each set was located where convenient (avoiding trees and
fallen logs) within randomly located 10 x 10 m quadrats in the study site. One regurgitated
O. endresiana seed was placed in each exclosure and control plot in late June 1995. Each
seed was checked after 2,4, and 8 d, 4 mo, and 1 yr.
Germination and Seedling Survival
The original regurgitated or defecated seed at each dispersed and non-dispersed site,
and each seed placed at a randomly located site was protected by a 4 x 4 x 2 cm cage made
of 3 mm galvinized wire mesh held in place by two metal stakes. Caged seeds were used
to determine germination rates and insect predation rates in the absence of mammalian seed

20
predators. Each site was checked weekly for at least 12 weeks and monthly thereafter until
June of the following year. Germination was defined as the splitting of the seed coat and
spreading of the cotyledons. Typically, the radicle had emerged by the next census after
germination, and a week later the stem was visible. As each seed germinated and the shoot
began to grow, the cage was removed to allow normal seedling growth. The seedling
location was marked with one of the stakes from the cage. Causes of seedling mortality
were classified as mammal, insect, fungal pathogen, physical, or unknown. Mammals
either ate the seed and left the damaged shoot behind (seed predators) or removed the entire
shoot (herbivores). Some seeds that appeared to have germinated were killed by beetle
larvae (Heilipus sp., Curculionidae) developing in the seed. Insect-killed seeds frequently
developed a root but never had a shoot >2 cm tall. Seedlings killed by fungal pathogens
were characterized by a wilted and discolored shoot above about 4 cm (Augspurger 1990).
Physical damage consisted of trampling by peccaries and other large mammals, or damage
from falling trees, branches, or large leaves (Clark and Clark 1991). Unknown causes of
mortality included cases that fit more than one category where the sequence of events could
not be determined, and cases that did not clearly fall into any category. A seed was
considered alive if the seed remained firm, even if the shoot had been eaten or otherwise
damaged. Such seeds resprouted repeatedly (D. Wenny, personal observation). Seeds
from 1995 that had not germinated or resprouted after one year were cut open and classified
as insect-killed if filled with frass, or viable if the embryo and cotyledons were not
discolored or mealy in appearance.
Germination trials. To determine the effect on 0. endresiana germination of
ingestion by birds and seed burial (which I initially thought would apply to seeds taken by
mammals), trials were conducted in a greenhouse constructed of nylon window screen over
a 12 x 6 m wooden frame on a concrete foundation set 25 cm into the ground. The roof
was corrugated plastic. Seeds were planted individually in cardboard milk cartons and cut
plastic bottles of various sizes. All containers were washed with hot soapy water and dried

21
in the sun before use. Soil was collected from a nearby secondary forest and mixed 3:1
with sand.
Three seed treatments were compared: (1) naturally regurgitated seeds collected
under fruiting trees or along the trail to the study site; (2) seeds removed by hand from ripe
fruits collected under fruiting trees; and (3) entire ripe fruits, either fallen or dropped.
Seeds or fruits that were misshapen, diseased, or had signs of insect infestation were not
used. Each treatment had 35 seeds or fruits, placed on top of the soil. Additionally, 10
seeds were planted 5 cm deep to determine if seeds could germinate after burial by
scatterhoarding rodents. Containers were watered as necessary to keep the soil moist.
Seeds were planted in June 1995 and checked weekly until early November 1995 when
seeds that had not germinated were cut open and assessed for viability as described above.
Seedling and sapling plots. To determine if O. endresiana seedlings and saplings
are more likely to recruit under or away from conspecifics, seedlings and saplings (up to 3
m tall) were measured and mapped in paired 10 x 10 m plots. For each of 10 trees, one
plot was located near the tree with approximately half of the plot directly under the crown.
The second plot was located 10 to 20 m from the edge of the first plot and at least 15 m
from the crown edge. No plots were located in recent (< 5 yr) canopy gaps or along
streams.
Statistical Analyses
The influence of the microhabitat variables on removal of marked seeds at 1 day
(1995 only) and 2 weeks, and on germination, seedling establishment and 1-year survival
of caged seeds were examined with multiple logistic regression using S AS JMP (SAS
Institute 1989) For each of the seven binomial (live or dead) response variables mentioned
above, the model was run first with all predictor variables, and then with and without each
predictor variable to determine if the deletion of a given variable had a significant effect on
the amount of variation explained. This deletion procedure was repeated until only

22
significant predictors were retained (Trexler and Travis 1993). Models were selected
manually to avoid the problems of automatic stepwise procedures (James and McCulloch
1990). The predictor variables included three measures of seed size (length, width, and
mass), eight microhabitat characteristics (leaf litter, canopy cover, number of stems, and
distances to nearest caged seed, herbaceous stem, woody stem, 10 cm tree, parent tree
trunk, and fallen log), date of dispersal (Julian date), and tree number. Only single terms
were included in the models as lack-of-fit tests indicated that the single term models were
adequate and thus interaction terms were not required (SAS Institute 1989).
The relationships of the microhabitat variables within and among the dispersed,
non-dispersed, and random sites were examined with principal components analysis
(PROC FACTOR) from the SAS statistical package (SAS Institute 1988). The average
loadings for the three types of sites were compared with one-way analysis of variance from
Super Anova (Abacus Concepts 1989). Type HI sums of squares were used to
compensate for the unequal sample sizes (Shaw and Mitchell-Olds 1993). Removal rates
were compared among treatments with survival analysis and Gehan-Wilcoxon tests from
SAS JMP (SAS Institute 1989) The Gehan-Wilcoxon test places greater weight on early
events than later events and was deemed appropriate for this study because most seeds
were removed during the first week (Pyke and Thompson 1986).
Parametric tests were used unless the data violated the assumptions of normality
and equal variance, in which case nonparametric procedures were used. Data in the form
of proportions were arcsin square-root transformed before analysis. Where multiple
comparisons on a data set were involved, the alpha value was adjusted according to the
number of comparisons planned (Bonferroni technique, Holm 1979). Data from the two
years were very similar and were combined for analyses in which sample sizes would have
been low otherwise. Throughout this paper mean values are followed by ± 1 SD.

23
Results
Seed Dispersal
In 1993,284 seeds regurgitated or defecated by birds were found: 155 dispersed
and 129 non-dispersed, while in 1995,234 dispersed and 171 non-dispersed seeds were
found. Ninety-five and 100 random sites were established in 1993 and 1995, respectively.
Twenty-four percent of the 1993 seeds and 28% of the 1995 seeds were found by
following or observing birds (N = 184), and the rest (N = 505) by searching the ground.
The dispersed seeds were most common within 10m of the crown edge (1993:
55%, 1995: 45%), but some seeds were as far as 70 m away. In rare cases (2.8%),
dispersed seeds were within 5 m of an 0. endresiana trunk (Fig. 2-2). Despite their close
proximity to the parent, they were classified as dispersed because they were not directly
under the parent crown, as a result of asymmetrical crown geometry or a tilted trunk (or
both). Random sites were more evenly distributed than dispersed or non-dispersed sites,
but showed a peak at 20-25 m in both years (Fig. 2-2). With dispersed and non-dispersed
sites combined, the seed distribution generated by birds in both years is best described by a
logarithmic function of distance from a fruiting conspecific (Fig. 2-2). Most seeds landed
within 20 m of the fruiting trees, but the tail of the seed distribution curve beyond 20 m
accounted for 18% of the sites in 1993 and 21% in 1995. Seed rain was highly variable
when averaged among trees and years (Fig. 2-3). In particular, note that seed distributions
showed secondary peaks at 45-50 and 60-65 m in both years (Fig. 2-2). These peaks
corresponded to habitual song perches used by bellbirds (see below) which were located
40-70 m from fruiting trees in the years this study was conducted (Fig. 2-1). More
information on the locations of seeds is presented in the Appendix.
Microhabitat characteristics. The locations of dispersed and non-dispersed seeds,
and seeds at random sites differed significantly in both years with regard to canopy cover,
distance to parent, and number of stems (Table 2-1). Dispersed seed sites had lower
average canopy cover than non-dispersed or random sites in both years (post-hoc SNK

24
CO
ID
<0
IL
O
DC
Ul
m
5
3
DISTANCE FROM PARENT (m)
Figure 2-2. The distribution of seeds naturally regurgitated or defecated by birds directly
under the crown of a fruiting tree (non-dispersed), and away from fruiting trees
(dispersed); and the distribution of random sites at different distance categories from trunks
of fruiting trees in 1993 (above) and 1995 (below). Values on the x-axis represent the
maximum distance for each category (5 = 0-5 m). The seed shadow is truncated at 70 m
because the abundance of Ocotea endresiana trees in the study site make sites > 70 very
rare.

25
UJ
ui
cc
cc
UJ
Q.
to
Q
UJ
UJ
(O
0 5 10 1 5 20 25 30 35 40 45 50 55 60 65 70
DISTANCE FROM CROWN EDGE (m)
Figure 2-3. The average number (+ 1 SD) of naturally regurgitated or defecated seeds per
tree (N = 21) for both years combined at different distances from the closest fruiting tree.
Distance class 0 includes all non-dispersed seeds directly below parental crowns. The
other distance categories (5 = 0-5 m, etc.) are measured from the edge of the crown of the
closest fruiting Ocotea endresiana tree. The values for each distance category represents the
number of seeds at a particular distance, averaged among all trees, for the entire study site.

26
Table 2-1. Summary of one-way analysis of variance tests for each habitat variable
compared among sites of dispersed (D), non-dispersed (N) seeds, and randomly located
sites (R) in 1993 and 1995. ANOVA tests had a Bonferroni-adjusted alpha value = 0.006.
Mean values are shown for each treatment for each variable. Standard deviations are in
parentheses below each mean. Leaf litter was the number of leaves pierced by a metal stake
thrust into the soil at the site. Canopy cover was estimated with a spherical densiometer.
Vegetation density was estimated as the number of stems within a 50 cm radius of the site.
The last four variables are distances to closest stem, closest woody stem (> 1 m height),
closest tree (> 10 cm DBH), and closest log (> 10 cm). Within each year and variable for
which a significant difference among treatments was detected (indicated with * after the F-
value), results of post-hoc tests (Student-Newman-Kuels tests) are indicated with a letter
after the mean. Means followed by different letters are significantly different (P < 0.05).
1993
1995
Variable
D
N
R
^2,375
D
N
R
^2,502
leaf litter (#)
2.1
(1.3)
1.9
(1.4)
2.4
(1.4)
2.70
2.3
(1.2)
2.3
(1.2)
2.1
(1-2)
0.91
canopy
cover (%)
94.3a
(3.6)
95.6b
(1.5)
95.8b
(1.7)
13.40*
94.5a
(6.5)
96.7b
(1.6)
96.5b
(3.1)
12.40*
vegetation
density
38.9a
(17.3)
32.1b
(16.2)
40.7a
(20.4)
7.98*
26.8a
(15.1)
19.9b
(10.2)
21.6b
(10.9)
15.50*
closest
stem (cm)
10.8
(5.5)
12.2
(6.2)
10.9
(6.2)
2.22
11.6a
(6.4)
14.3b
(7.4)
13.3ab
(12.0)
5.46*
woody
stem (cm)
61.9
(45.7)
58.9
(39.7)
66.7
(42.3)
0.90
88.6
(62.5)
75.5
(60.4)
75.9
(55.8)
2.85
closest
tree (m)
1.50
(0.87)
1.68
(0.97)
1.66
(1.00)
1.53
1.88
(1.08)
1.92
(1.04)
1.84
(1.08)
0.18
closest
log (m)
1.47
(1.31)
1.30
(1.37)
1.28
(1.25)
0.80
1.97
(1.74)
1.78
(1.45)
1.83
(1.66)
0.77
*P<0.006

27
tests; F s < 0.01), while the non-dispersed and random sites were similar (P > 0.05). In
1993, dispersed and random sites had similarly (P > 0.05) high numbers of stems
compared to non-dispersed sites (P < 0.01). In 1995, dispersed sites had more stems than
both the other sites (P < 0.01), while random and non-dispersed sites had similar numbers
of stems (P > 0.05). The only other variable that differed among the treatments was
distance to closest stem in 1995. Dispersed sites were closer to stems than non-dispersed
sites (P < 0.01), while random sites did not differ from the other two treatments (P >
0.05).
In a principal component analysis of the 1995 data, the first two components
explained 63.8% of total variance, while for 1993 the first two components accounted for
45.5% of total variance (Table 2-2). The relatively low amount of variance explained is
due, in part, to the inclusion of random sites (Jackson 1993) which appeared as a spherical
cloud of points centered on the origin (Fig. 2-4). Removal on random sites, however,
increases the total variance explained by the first two components by only 2-3%. The
pattern was similar in both years: the first multivariate axis (PC 1) was characterized by
high negative loading of canopy cover, and high positive loadings of leaf litter depth and
closest 10 cm DBH tree (Table 2-2). The second axis (PC 2) was characterized by positive
loading of distance to closest stem in both years and negative loading of vegetation density
in 1995 but positive loading in 1993 (Table 2-2). The average loadings for PC 1 differed
significantly among the dispersed, non-dispersed, and random sites in 1993 (F2369 =
45.99, P < 0.0001), and 1995 (^2,502 = 31.61, P < 0.0001). Each type of site differed
from the others (Fisher's LSD tests, all P's < 0.002). Loadings for PC 2 differed only
between dispersed and non-dispersed sites in 1995 (^2,502 = 6.07, P = 0.0025). This
analysis shows that site characteristics of dispersed and non-dispersed seeds were
nonrandom.
When the three treatments (dispersed, non-dispersed, and random) are identified on
a plot of the first two principal components, they showed broad overlap, except for some

PC 2 PC 2
28
â–¡ DISPERSED
• NON-DISPERSED
+ RANDOM
PC 1
â–¡ DISPERSED
• NON-DISPERSED
+ RANDOM
PC 1
Figure 2-4. Plot of sites with dispersed (open squares), non-dispersed (solid circles), and
randomly located seeds (crosses) from 1993 (above) and 1995 (below) on the first two
axes determined by principal components analysis of habitat variables. Axis 1 is primarily
a function of canopy cover leaf litter, and closest 10 cm DBH tree, and Axis 2 is primarily a
function of vegetation density and closest stem (see Table 2-2).

29
Table 2-2. Loadings (eigenvectors) of each habitat variable on the first two components
determined by principal component analysis of the correlation matrix. The proportion of
total variance explained by each component is listed in the last row.
1993
1995
PC 1
PC 2
PC 1
PC 2
leaf Utter
0.5427
-0.2248
0.8885
0.2718
canopy cover
-0.5973
0.4796
-0.8117
0.2453
#stems
-0.4831
0.5784
0.1359
-0.8443
closest stem
0.5856
0.5393
-0.2315
0.7122
woody stem
0.3973
0.5065
0.2958
-0.3067
10 cm tree
0.5034
-0.2890
0.7280
-0.0947
closest log
0.2300
0.3720
0.8854
0.2925
variation
24.69%
20.76%
41.74%
22.04%

30
dispersed sites that have higher values for the first component These correspond to sites
with low canopy cover, relatively far from the parent trees (Fig. 2-4). Seeds in the sites
with values > 2.0 in 1993 and > 1.0 in 1995 for PC 1 were all dispersed by bellbirds into
large gaps surrounding song perches. Additionally, two of the 1993 seeds were dispersed
by Black Guans onto logs in another smaller gap. Bellbird-dispersed seeds, seeds
dispersed by other species, and random sites had significantly different PC 1 loadings in
1993 (*2,161 = 7.9, P = 0.006), and 1995 (F2,212 = 9.1, P < 0.001). Bellbird sites had
higher factor loadings than other species' sites and random sites (P's < 0.002), but sites of
the other species and random sites did not differ (P = 0.2).
Combining the seeds from both years, and using only seeds for which the disperser
was known, the average canopy cover at sites of seeds dispersed by bellbirds was
significantly lower than of sites of seeds dispersed by all the other species (89% and 96%
respectively; Kruskal-Wallis test: .X2 = 42.7, df = 1, P < 0.001; Fig. 2-5). Similarly, the
average distance from the closest parent tree of bellbird sites was greater than any of the
other four species, (Kruskal-Wallis X2 = 66.06, df = 1, P < 0.001), but the difficulty in
following the birds (especially robins which tended to fly above the canopy) biases the
results in favor of bellbirds. Nevertheless, the data show that bellbirds tend not to drop
many seeds near parent trees (Fig. 2-6). Considering that approximately 5.3% of the study
area was in gaps (see Study Site), overall seed arrival in gaps (12%) occurred more often
than expected by chance (X2 = 36.45, df = 1, P < 0.001), whereas close to the expected
number of random sites (4.1%) was in gaps (X2 = 0.557, df = 1 P > 0.05, see Fig. 2-4).
Furthermore, bellbirds dispersed significantly more seeds to gaps than did the other species
(X2 = 39.3, df = 1, P < 0.001). Note that bellbirds dispersed seeds to only 2 of 29 gaps in
the study area.

CANOPY COVER (%)
31
100 n
BELLBIRD GUAN QUETZAL ROBIN TOUCANET
DISPERSER SPECIES
Figure 2-5. Mean (± 1 SD) canopy cover (%) of sites of seed dispersed by bellbirds,
guans, quetzals, robins, and toucanets for both years combined. Sample size for each
species shown at the bottom of each bar.

NUMBER OF SEEDS
32
DISTANCE FROM CROWN EDGE (m)
Figure 2-6. The number of seeds dispersed at different distances from parental trunks by
bellbirds, guans, quetzals, robins, and toucanets for both years combined (N = 184).

33
Post-dispersal Seed Predation
More than 50% of marked seeds were removed within one week in both years (Fig.
2-7). In 1993, seeds at dispersed, non-dispersed, and random sites were removed at
similar rates with dispersed seeds showing a nonsignificant trend for slower removal than
seeds at non-dispersed or random sites (Wilcoxon X2 = 5.36, df = 2, P = 0.068). In
1995, dispersed seeds were removed more slowly than those at non-dispersed and random
sites (Wilcoxon X2 = 9.09, df = 2, P = 0.011). Ultimately, only 2 of 923 marked seeds
(one dispersed and one random site, both from 1995, and both ^ 30 m from conspecific
trees) in the entire study survived to June 1996 (and germinated), for an overall predation
rate of 99.7%.
I found 75% of marked seeds after removal in 1993, and 94% in 1995. In all
cases, the seed was killed, and in most cases (96%) entirely consumed. Few seeds were
eaten (13%) but not moved. The general pattern was that seeds were taken a short distance
(range 0-21 m, N = 761) to a site with cover, presumably where the seed could be eaten in
safety. On average, removed seeds (not counting seeds eaten but not moved) were taken
2.5 m (± 3.8, N = 286) in 1993 and 2.1 m (± 2.9, N = 475) in 1995. About 50% of those
removed were found on the leaf litter, but some were found in small burrows (< 6 cm
diameter), in or under fallen logs, or in crevices at the bases of large trees (Fig. 2-8). Most
seeds taken to burrows, trees, and logs were clearly taken by small animals, probably
rodents, because the small diameter of the opening would exclude larger species such as
agoutis (Dasyprocta punctata). I found no evidence of scatterhoarding or secondary
dispersal.
Significant predictors of survival of marked seeds to 2 weeks in 1993 included
canopy cover, leaf litter depth, and dispersal date (Table 2-3). Seeds in sites with more
open canopy cover, deeper leaf litter, and later dispersal date were more likely to survive.
In 1995, 1-d survival was significantly predicted by greater seed mass, earlier dispersal
date, and the particular parent tree. At 2 wk, seeds farther from the parent tree, dispersed

34
o
<
s
UJ
a.
V)
o
LU
UJ
c/>
DAYS AFTER DISPERSAL
Figure 2-7. Removal of marked seeds from sites of dispersed (open squares), non-
dispersed (solid squares), and randomly located (crosses) seeds in 1993 (above) and 1995
(below). Sample sizes are shown in parentheses.

35
PERCENTAGE OF SITES
Figure 2-8. Percentage of marked seeds found in different locations after removal by
animals. See text (Methods: post-dispersal seed predation) for definitions of categories.

Table 2-3. Results of logistic regressions of post-dispersal survival of marked seeds
against habitat variables. Only significant effects are listed.
Response
r2
-21og likelihood
(X2)
predictors3
1993
2 wk survival
0.311
54.19***
+ leaf litter***
- canopy cover*
- dispersal date***
1995
1 day survival
0.23
133 3***
- seed mass**
+ dispersal date***
tree number*
2 wk survival
0.31
144.5***
+ distance to parent***
- dispersal date***
tree number*
*P < 0.05, **P < 0.01, ***P < 0.001.
asign in front of each predictor variable indicates positive or negative correlation with the
response variable.

37
later in the season, and at certain parent trees were more likely to survive. Logistic
regression models of survival past 2 wk were not significant because so few seeds
survived that long.
Distance effect Seed removal from transects radiating from fruiting O. endresiana
trees was slower for seeds farther from the parent trees in both 1995 and 1996. The 1995
seed removal experiment, which ran for 10 weeks, showed that the effect of distance on
seed removal declined over time and that seed predation eventually approached 100% at all
distances (Fig. 2-9a). Only 1 seed (at 80 m) survived 10 wk. Removal of seeds was
significantly faster for seeds within 40m of parents than for seeds placed 60 or 80 m away
(Wilcoxon X2 = 4.5, df = 1, P = 0.03). The 1996 experiment (Fig. 2-9b) showed that
seed removal was significantly faster for seeds experimentally dispersed during the peak of
fruiting than before the fruiting peak, but only for seeds 5 m (under the crown) from
parents (X2 = 6.87, df =1, P = 0.009) and not for seeds 40 m away (X2 = 0.21, df =1, P
= 0.64). Removal rates did not differ between seeds placed 5 and 40 m from parents at
either time in the fruiting season (early: X2 = 0.65, df = 1, P = 0.42; middle: X2 = 3.17, df
= 1, P = 0.20).
The 1996 sample of 44 seeds was censused at dusk (18:00-18:30) and dawn
(05:30-06:00) for two consecutive days. Five seeds had been removed by the first dusk
census, two of which clearly were taken by small rodents into small holes in fallen logs.
By the next dawn, 27 more seeds had been removed. One seed was taken during that day
and four more by the next dawn. Thus, most seed removal (73%) took place at night (X2
= 25.9, df = 1, P < 0.001), probably by small rodents.
Exclosures. Removal rates differed significantly among the three treatments; seeds
accessible to all rodents were taken faster than the other two treatments (Wilcoxon X2 =
16.7, df = 2, P = 0.007). After 8 d, three seeds had been removed from the small mesh
exclosures, which were designed to exclude all rodents (Fig. 2-10). Two of these
exclosures had burrows inside that were probably dug during the three months between

38
DAYS
Figure 2-9. (A) Removal of marked seeds experimentally dispersed 0, 20,40, 60, and 80
m from trunks of the fruiting Ocotea endresiana trees starting in June 1995. N = 12 at each
distance. (B) Removal of marked seeds at 5 and 40 m from parent trees in the early (late
May) and at the peak (mid-June) of fruiting season, 1996. N = 24 for each category. All
seeds at 0 and 5 m were under the parental crowns.

SEEDS REMAINING (%)
39
Figure 2-10. Percentage of seeds remaining in three types of exclosures after 8 days. For
each treatment N = 38. The 'no rodent' exclosure was designed to exclude all terrestrial
animals, 'small rodent' exclosures excluded large animals, but allowed access to small
animals through 6 cm mesh, and 'all rodent' plots were controls with no exclosure.

40
construction of exclosures and the beginning of the experiment. The third exclosure was
hit by a fallen tree, thus creating access. Except for those three, seeds were not removed
from "no rodent" exclosures. Several other exclosures were damaged by falling branches
or breached by peccaries foraging for Inga pods, but these incidents either occurred after
the seed had been removed or did not lead to seed removal. After 8 d all the seeds from the
control plots had been removed, while 52% of the seeds from the "small rodents only"
exclosures had been removed (Fig. 2-10). After 4 mo, 82% of the seeds from the "small
rodents only" exclosure had been removed. Thus, although large mammals may take some
seeds, small mammals (presumably rodents) will eventually find and eat most seeds.
Germination and Seedling Survival
Ocotea endresiana seeds began germinating about 6 weeks after dispersal. The
germination rate of caged seeds ranged from 70 to 95% in both years and did not differ
among the seeds at dispersed, non-dispersed, and random sites (Fig. 2-11). Of the seeds
that germinated in 1993 and 1995, 35% and 27% respectively, survived one year as
seedlings. Overall, 28% and 22% of seeds in 1993 and 1995 respectively, germinated and
survived on¿ year. For the 1993 caged seeds, annual survival for the first three years
averaged 31% (Table 2-4). Seeds at four dispersed and five random sites from 1993
survived until June 1996, for an overall 3-year survival rate of 2% (Fig. 2-11). Within
each stage (germination, 1 yr, 2 yr and 3 yr survival), the three location treatments had
similar survival rates except for one-year survival in 1993 (Fiji = 11.84, P = 0.0002).
Seeds at dispersed sites had significantly higher one-year survival than those at non-
dispersed and random sites (P < 0.001). Random and non-dispersed sites had similar one-
year survival (P = 0.35). Germination success of seeds defecated by guans did not differ
from that of seeds regurgitated by the other four species (X2 = 0.98, df = 1, P > 0.05).
None of the variables in the logistic regression models predicted germination
success in either year. Significant predictors of 1-year survival included lower canopy

PROPORTION SURVIVING
41
1.00 H
0.80-
0.60-
0.40-
0.20-
0.00
O Dispersed
I Non-dispersed
â–¡ Random
. l. J, r1!
germ 1 y r
1993 SITES
2yr 3yr germ 1 y r
1995 SITES
Figure 2-11. Germination success, and annual seedling survival of dispersed and non-
dispersed seeds and seeds at randomly located sites (mean + SD per tree) in 1993 (left) and
1995 (right). Within each stage where a difference among sites was detected, bars with
different letters above are significantly different (P < 0.005; Bonferroni-adjusted alpha =
0.008).

42
Table 2-4. Results of logistic regressions of post-dispersal survival of seedlings
predicted by habitat variables. These seeds were protected from seed predators
with cages until germination. Only significant effects are listed. No variables
significantly predicted germination success (not shown).
Response
r2
-21og likelihood
*2
predictors3
1993
1 yr survival
0.10
43.25***
- canopy cover*
+ dispersal date*
1995
1 yr survival
0.13
67.52***
- canopy cover*
+ dispersal date**
tree number*
*P < 0.05, **P < 0.01, ***/> < 0.001.
^ign in front of each predictor variable indicates positive or negative correlation
with the response variable.
r

43
cover and later dispersal date in both 1993 and 1995 (Table 2-4). Additionally, in 1995,1-
yr survival was heterogeneous with respect to which particular fruiting tree was closest,
although no clearcut pattern was discernible in terms of crop size, tree size, or proximity to
other fruiting trees.
The causes of mortality of caged seeds and seedlings during the first year differed
among dispersed, non-dispersed, and random sites in both 1993 (X2 = 29.14, df = 6, P <
0.001) and 1995 (X2 = 33.59, df = 6, P < 0.001; Fig. 2-12). Many seedlings were killed
when seed predators (presumably rodents) removed the seed and severed the connection
between the shoot and root Mammalian herbivores also killed seedlings by entirely
removing the leaves and shoot. Mortality by the combination of mammalian seed predators
and herbivores was significantly different among the three location treatments in 1995
(P2.37 = 4.98, P = 0.012), but not 1993 (F2.24 = 3.56, P = 0.044). In 1995, mammals
killed more seeds and seedlings at dispersed than at random sites (P < 0.05), but mortality
caused by mammals at non-dispersed sites did not differ from that at either dispersed or
random sites (P > 0.05; Fig. 2-12). Mortality caused by beetle larvae (Heilipus sp. and at
least one other unidentified species) differed among the three treatments in both years
(1993; F2,23 = 4.85, P = 0.012, 1995: ¿*2,37 = 6.64, P = 0.003). In both years, more
non-dispersed than dispersed seeds were killed by beetles (P < 0.05; Fig. 2-12). Mortality
caused by fungal pathogens and physical damage (falling branches, trampling) did not
differ among treatments or years. For the three treatments, combined levels of mortality by
each cause were similar between the two years, except for insect predation which was
significantly higher in 1993 than 1995 (Student's t-test; t = 1.97, df = 52, P < 0.05) and
mortality by mammals which was significantly higher in 1995 than in 1993 (t = 2.53, df =
67, P < 0.01).
Average canopy cover for seedlings that survived one year at dispersed sites
(94.3% + 3.6, N = 86) was significantly lower than for seedlings that died (95.2% ± 2.3,
N = 303), whereas canopy cover did not differ with one-year survival for non-dispersed or

44
o
LU
2
o
h
CC
O
a
O
DC
a
0.8 -i
mammals insects
1993
O Dispersed (N=116)
| Non-dispersed (N=112)
â–¡ Random (N=73)
fungal physical
a
0.80 i -r
1995
â–¡ Dispersed (N= 186)
H Non-dispersed (N=134)
â–¡ Random (N=82)
X
-'ÉS£
fungal physical
Figure 2-12. Average (+ SD) proportion per tree of caged dispersed and non-dispersed
seeds and randomly located seeds killed by different causes during the first year for seeds
from 1993 (above) and 1995 (below). Cages were removed after germination and shoot
development Sample sizes of seeds or seedlings killed shown in parentheses. Within
each cause of mortality where a difference among sites was detected, bars with different
letters above are significantly different (Bonferroni-adjusted alpha = 0.0125). See text
(Methods: Germination and seedling survival) for definitions of types of mortality.

45
random sites (2-way ANOVA, F^in = 3.98, P = 0.019). Denser canopy cover was also
a significant predictor of higher mortality by fungal pathogens flogistic regression, P =
0.0013). Seedlings that survived one year were significantly taller at sites with less dense
canopy cover (linear regression, r2 = 0.284, P < 0.001; Fig. 2-13). Seedlings from seeds
dispersed by bellbirds were significantly taller than those at other species' sites (t-test =
2.37, N = 38, P = 0.02).
Germination trials. Regurgitated seeds and seeds cleaned of pulp by hand had
higher germination rates than the seeds in intact fruits (X2 = 43.94, df = 2, P < .001). Of
30 seeds in each of the first two treatments, all but one seed germinated, and that one failed
because it was infested with a beetle larva. In contrast, seeds in intact fruits tended to
mold; 8 germinated and 22 appeared mealy and no longer viable after 12 wk. The 8 seeds
in fruits that germinated never developed a shoot outside the fruit pulp and eventually
rotted. Thus, ingestion of seeds by frugivores was beneficial in terms of pulp removal but
probably did not otherwise affect germination. Two of the 10 buried seeds started to
germinate (the seed coat appeared cracked) and seemed viable but the other eight seeds
rotted. After 4 mo, neither of the apparently germinating seeds had any sign of root
development, nor had the cotyledons begun to separate. Six of 10 buried seeds had
whitish mold on the seed coat. Thus, few buried seeds appear to establish as seedlings.
Seedling and sapling plots. Seedlings and saplings were equally common in 10 x
10 m plots under (6.6 ±1.5 individuals) and away (4.9 ± 3.5) from the parent trees (paired
t-test; t = 1.15, P > 0.05). Individuals in plots away from the parent trees, however, were
taller (79.9 ± 14.9 cm) than those near the parent trees (47.4 ± 7.2; Mann-Whitney U test,
P = 0.009).
Recruitment
Based on the experiment with marked seeds, the probability of recruitment is
determined mostly by post-dispersal seed predators. Assuming the experiment reflects the

SEEDLING HEIGHT (cm)
46
Figure 2-13. Seedling height (N = 140) plotted as a function of percent canopy cover for
both years combined (height = 28.93 - 0.233x; r2 = 0.284, P < 0.001). Values for canopy
cover below approximately 92% are associated with canopy gaps.

47
absolute level of seed predation, and all seeds that escape predation germinate and survive,
then recruitment was 0% in 1993 and 0.4% in 1995. On the other hand, if seed removal at
2 wk (75-93%) reflects the ultimate pattern of survival (e.g., Lieberman 1996) then it is
possible to calculate the influence of stage-specific mortality patterns on the relative
probability of recruitment (Table 2-5, Fig. 2-14). The initial pattern of seed rain shows a
strong peak near the parent trees and secondary peaks beyond 40 m that correspond with
bellbird song perches (Fig. 2-14a). The bellbird perches encompass a range of canopy
cover values from forest/gap edge (=92%) to gap centers (=70%) and thus seed rain is
spread across a wider range of canopy conditions with increasing distance from the parent
trees. Two-week seed survival is higher for seeds > 30 m from fruiting trees, but relatively
consistent across the range of canopy cover (Fig. 2-14b). Germination is high for all seeds
regardless of distance from parent or canopy cover (Fig. 2-14c). One-year seedling
survival is highest for sites > 40 m from parents and < 90% canopy cover (Fig. 2-14d).
The cumulative probability of recruitment (the product of seed rain, 2 wk seed survival,
germination, and one year survival) shows a peak near the parent trees and secondary
peaks corresponding to the bellbird perches (Fig. 2-14e). Note that the probability of a
seed dispersed by birds surviving one year is < 0.2% at any location.
Discussion
The results illustrate four main points: (1) bellbirds dispersed seeds in a different
pattern than the other species; (2) post-dispersal seed predation is highest near the parents,
but is also high everywhere; (3) the pattern of recruitment is bimodal with a peak near the
parent tree in closed canopy forest and another in gaps corresponding to bellbird perches;
and, (4) all three advantages of dispersal (escape, colonization, and directed dispersal) are
supported. These points will be discussed in detail below.

48
Table 2-5. Summary of post-dispersal stages leading to seedling recruitment Values for
stage-specific probabilities represent the average proportion of individuals per tree that
entered and survived each stage. Cumulative probabilities are the product of the current
and previous stage-specific probabilities. Data for 2- and 3-yr survival of the 1995 cohort
are unavailable.
Stage
Stage-specific probability
Cumulative probability
1993
1995
1993
1995
2 wk seed
0.094 + 0.113
0.137 + 0.167
0.094
0.137
germination
0.759 + 0.188
0.855 ±0.122
0.0713
0.1171
1 yr seedling
0.309 + 0.237
0.212 ±0.185
0.0220
0.0248
2 yr seedling
0.245 + 0.181
—
0.0054
—
3 yr seedling
0.360 + 0.115
—
0.0019
—

Figure 2-14. Average seed rain (a), 2 wk seed survival (b), germination (c), 1 yr seedling
survival (d), and cumulative probability of recruitment (e) as functions of canopy cover and
distance from parent for both years combined and averaged among 21 trees. For a-d, each
bar represents the percentage of the seeds or seedlings that survived that stage in each
distance/canopy cover category, averaged among trees. Categories with fewer than three
seeds were not included and are blank, while categories with a flat rectangle are zero. Seed
rain determined from the original distribution of seeds naturally dispersed by birds. Two-
week seed survival estimated from the removal of marked seeds. Germination and seedling
survival estimated from the seeds caged until after germination. Recruitment (the
cumulative probability of surviving every stage) is calculated as the product (a x b x c x d)
for each distance/cover category. For clarity, standard deviations are not shown.

50
SEED RAIN
Figure 2-14
SEED RAIN (%)

51
2-WK SEED SURVIVAL
Figure 2-14 —continued
SEED SURVIVAL (%)

52
GERMINATION
Figure 2-14 --continued
GERMINATION (%)

53
1-YR SEEDLING SURVIVAL
Figure 2-14 —continued
1-YR SURVIVAL (%)

54
RECRUITMENT
Figure 2-14 --continued
(Pr) RECRUITMENT

55
Seed Dispersal
Most O. endresiana seeds handled by birds land under or just beyond the crowns of
parent trees. This distribution of seeds is typical for vertebrate-dispersed plants (Dirzo and
Dominguez 1986, Hoppes 1988, Willson 1993, Laman 1996a). The abundance of O.
endresiana adults in the study area makes dispersal beyond 80 m from any conspecific tree
virtually impossible. Considering the restricted elevational range of this species, longer-
distance dispersal may lead to arrival in habitats unsuitable for establishment (Wheelwright
1988). Thus, while regurgitating seeds just outside the crown of a fruiting tree may seem
like poor quality dispersal (McKey 1975), the shear abundance of seeds in that region may
lead to a small peak in recruitment near the parent tree (Fig. 2-14e) as predicted by Hubbell
(1980) and Condit et al. (1992). The long tail of the seed distribution, however, is
apparently important for recruitment because seeds dispersed farther from the parent trees
have a greater chance of arriving in a habitat where the probability of seedling survival is
higher (i.e., a safe site), as will be shown below.
Of the five species of birds that dispersed most of the seeds in this study, all but
one deposited most seeds in closed canopy forest. Only male bellbirds frequently
dispersed seeds to treefall gaps. In my study site, bellbird song perches were in dead
Sapium oligoneuron trees, bordering large treefall gaps. Such sites are typical for
bellbirds: Snow (1977) reported that bellbird song perches at lower elevations in the
Monteverde area (where the forest is fragmented) are usually on dead branches in tall trees
on the forest edge. In other Neotropical forests, three other species of bellbirds (Procnias
spp.) exhibit similar behavior in habitual use of tall, exposed song perches, often on dead
branches (Snow 1961, 1970, 1973a). Snags used by bellbirds in my site had several
branches and the birds made use of many different branches within each tree, such that
seeds under them were scattered over approximately 25 m^, including a gradient of site
conditions from gap to forest understory. Male bellbirds also dispersed seeds in forest as
do the other species, but they typically spend 80-95% of the day in the vicinity of the song

56
perch (Snow 1977) and probably disperse a similar percentage of seeds they process under
song perches. Female bellbirds were rarely seen at the song perches, and then only for a
few minutes.
The influence of bellbirds on the pattern of seed fall is shown by the slight increase
in the number of seeds 40-65 m from the parent trees (Fig. 2-3). Bellbird perches in my
site happened to be far (> 40 m) from any fruiting O. endresiana trees. The other four bird
species that disperse O. endresiana seeds occasionally perch on the edges of gaps, (D.
Wenny, personal observation) although whether they do so more often than expected based
on perch availability is unknown. Because birds do not (cannot?) regurgitate in flight, and
guans apparently do not defecate in flight (personal observation), the most likely way an O.
endresiana seed can arrive in a gap is via male bellbirds. I never observed the other four
species on bellbird perches. Thus, O. endresiana has two types of dispersers that deposit
seeds in two different but predictable patterns: male bellbirds depositing large numbers
(52%) of the seeds they disperse in and near gaps, and the other four species depositing
seeds in forest, most (75%) within 25 m of parent trees and fewer more than 25 m away.
Additionally, on several occasions I followed quetzals and guans from one fruiting O.
endresiana to another, where they deposited seeds before eating fruits in the second tree.
Although I had no method of identifying the source tree of the seeds, it is likely that
dispersal of seeds from one tree to another conspecific occurs regularly (Wheelwright
1991, Gibson and Wheelwright 1995).
The patterns of seed distribution described here may be similar for other trees in
cloud forests where these disperser species occur during the breeding season, but may be
very different during the rest of the year. The five species of birds that disperse O.
endresiana are the most important fruit consumers of canopy trees in the area (Wheelwright
et al. 1984). At least 20 other species of trees are dispersed by these birds and probably
have similar patterns of seedfall as O. endresiana, with a peak near the parent tree and a
second peak at bellbird perches. However, all five dispersers (except perhaps toucanets)

57
undertake elevational shifts that correspond with fruit availability (Wheelwright 1983,
Levey and Stiles 1992, Powell and Bjork 1995, Guindon 1996) and leave the study area in
late July or early August (In 1994 they left in June when O. endresiana failed to fruit).
Male bellbirds may not provide such a pronounced peak in seed dispersal to song perches
after the breeding season because they seem to use many different perches in the non-
breeding season and tend to use subcanopy perches that are not exposed (Stiles and Skutch
1989). Thus, seed fall for the community as a whole is likely temporally as well as
spatially heterogeneous.
The implications of seed dispersal to habitual perches deserves further study. Do
habitual perches represent foci of plant recruitment or of seed predation? Seed dispersal to
perches has been frequently studied in successional landscapes (Livingston 1972,
McDonnell 1986, Holthuijzen et al. 1987, Guevara and Laborde 1993, McClanahan and
Wolfe 1993, Robinson and Handel 1993, Vieira et al. 1994, Nepstad et al. 1996, Duncan
1997). In these systems, directed dispersal may play a key role in the restoration of
abandoned pastures, logged forests, and other degraded lands. It is possible that directed
dispersal is more common in intact forests than previously expected. For example,
manakins (Pipridae) and cocks-of-the-rock (Rupicola: Cotingidae) choose lek perches that
are more sunlit than average understory perches (Endler and Théry 1996), and such sites
may provide growth advantages for seedlings. Snow (1961,1970, 1973b) notes that
other species of bellbirds (Procnias) preferentially select perches on dead trees or branches,
or in sparsely-vegetated trees. For these species, competition among males for females
drives them to be as conspicuous as possible. The fortuitous outcome of this behavior may
be a disproportionate effect on plant recruitment in the vicinity of their display areas (Théry
and Larpin 1993). Whether habitual perches represent foci of seedling recruitment
(McDonnell and Stiles 1983, McDonnell 1986) or lead to density-dependent seed and
seedling mortality (Wheelwright 1988), needs to be examined in more detail.

58
Seed Predation
Removal of marked seeds after dispersal in this study always resulted in predation.
Although seeds were taken into burrows and logs, I never found any treatment of seeds
indicative of scatterhoarding, such as burial in the soil (Smythe 1978, Hallwachs 1986,
Forget 1990,1993) or under piles of leaf litter (Forget 1991). In addition, the failure of
buried seeds to germinate and survive in the greenhouse experiment suggests that
scatterhoarding would not be advantageous for O. endresiana recruitment Only two
species of mammals known to scatterhoard seeds occur in the study site: spiny pocket mice
(Heteromys desmarestianus) and agoutis (Dasyproctapunctata). Both species are much
less common here than at lower elevations (D. Wenny personal observation, K.G. Murray,
personal communication). Although some species of squirrels (Sciurus) and deer mice
(Peromyscus) are known to cache seeds in other regions of the world (Vander Wall 1990),
information on tropical species (in this site S. deppei, and P. mexicanus) is lacking
(Emmons 1990).
The main seed predators in this study were probably rodents, particularly the deer
mouse Peromyscus mexicanus, by far the most common terrestrial rodent in the study area
(Anderson 1982, Langtimm 1992). Based on the exclosure experiment, however, it is
possible that species larger than mice (i.e., agouti, paca, peccary, wood-quail) are also
important seed predators. Because removal of seeds from the open control plots was about
twice that of the "small rodents only" plots, agoutis and other large mammals may have
been responsible for as much as 50% of O. endresiana seed predation. Nevertheless, the
exclosures also showed that in the absence of agoutis, small rodents will find and eat most
of the seeds, albeit over a longer time period. Additionally, the finding that most marked
seeds from the removal experiment were taken at night and removal of seeds always
resulted in predation rather than scatterhoarding suggest that agoutis, which are mostly
diurnal and known scatterhoarders (Smythe 1978, Hallwachs 1986), may not be taking

59
many seeds (although few data exist on the extent to which agoutis eat but do not
scatterhoard certain species, e.g., Smythe 1978).
The high degree of post-dispersal seed predation reported here is not unusual
(Harper 1977, Cavers 1983). Recent reviews have found 90-100% losses of seeds to
predators in 22 of 62 studies (Crawley 1992, Hulme 1993). Estimates of seed predation
rates for Neotropical trees range from 40-50% for the canopy tree Brosimum alicastrum
(Burkey 1994), 36-98% for the canopy emergent tree Dipteryxpanamensis (DeSteven and
Putz 1984), 87% for Dipteryx micrantha (Cintra and Homa 1997), 60-90% for the
subcanopy tree Faramea occidentals (Schupp 1988b), 50-90% for the subcanopy palm
Welfia georgii (Schupp and Frost 1989), 97% for the canopy palm Astrocaryum murumuru
(Cintra and Homa 1997), and 96% for the canopy tree Virola nobilis (Howe et al. 1985).
The study by Burkey (1994) was short-term (21 days) and may underestimate total
predation. Also, in the study by DeSteven and Putz (1984), a site with only 36% predation
had low populations of seed predators as a result of hunting. Previous studies on seed
predation of Lauraceae found 96-100% predation (Wheelwright 1988, Chapman 1989a,
Holl and Lulow 1997).
The high level of seed predation in my study site could be explained by dense
vegetation and cloudy conditions, providing small rodents protection from predators
(Bowers and Dooley 1993, Vásquez 1996). High predation is probably not a consequence
of high rodent populations due to lack of predators because potential predators of mice such
as owls, forest-falcons, and cats are relatively common (D. Wenny, personal observation).
During the day, light levels in the study area are often reduced as a result of cloudy weather
(Cavelier 1996, Chazdon et al. 1996) as well as dense vegetation. Indeed, on a few
occasions mice were seen during the day (especially Peromyscus and Scotinomys) and
some marked seeds were removed by small rodents during the day (see Post-dispersal seed
predation: distance effect). Also, local naturalists report seeing more mice on misty or
rainy days and nights than during clear weather (T. Guindon, personal communication).

60
Thus, the high levels of predation for O. endresiana could be a result of longer activity
patterns by small rodents and may be typical for many large-seeded tree species in cloud
forests.
Seedling Survival and Recruitment
Most of the seeds protected from rodents germinated. High germination rates for
Lauraceae are apparently typical (Wheelwright 1985b). Some seeds that germinated were
eventually killed by beetle larvae (Heilipus sp. Curculionidae) developing in the
cotyledons. After the cages were removed from the growing seedlings, the seeds as well
as the seedlings were susceptible to predation by mammals. It is difficult to determine the
role of each species involved but the suite of species that kill seedlings is probably larger
than the suite of seed predators, and could include pacas (Agouti paca), brocket deer
(Mamma americana), tapir (Tapirus bairdii), and peccaries (Tayassu tajacu), in addition to
agoutis and smaller rodents.
Insects were most likely to kill non-dispersed seeds or seedlings while mammals
killed all types of seeds (dispersed, non-dispersed, and random). This finding is consistent
with Howe (1993a) and Terborgh et al. (1993), who found that insect-caused mortality
was distance-dependent, but mortality by mammals was not. In New Guinea, Merg (1994)
found the opposite pattern: insects tended to kill dispersed seeds and mammals killed non-
dispersed seeds. In contrast to insects and mammals, fungal pathogens killed
proportionately more O. endresiana seedlings in closed-canopy forest than in gaps.
Augspurger (1984) also showed fungal pathogens of seedlings were less prevalent in gaps
than forest understory for nine species of tropical trees.
The greater height of seedlings and saplings in plots >20 m from the parent trees
than in plots beneath the crowns, suggests that seedling mortality was higher closer to the
parent trees. Seedling ages are not known, but this difference suggests that longer-term
survival several years after dispersal is higher for seedlings >20 m from the crown of

61
parent trees than for seedlings beneath crowns. Thus, seed dispersal away from the crown
of a conspecific appears to increase the chance of recruitment over the long term (Clark and
Clark 1984, Li et al. 1996).
The overall pattern of recruitment with respect to distance from parent trees and
canopy cover was bimodal. This pattern was caused by the combination of distance-
dependent seed predation (Fig. 2-14b) and the influence of canopy cover on seedling
survival (Fig. 2-14d), despite the fact that most seeds landed close to the parent trees in
closed-canopy forest (Fig. 2-14a). This pattern is an example of spatial discordance caused
by the lack of congruity among the stages leading to recruitment (Herrera et al. 1994,
Jordano 1995). The occurrence of such discordance emphasizes the importance of stage-
specific survival patterns on patterns of recruitment and the need for data on the sequential
stages of plant reproduction rather than on only one or a few stages. Bellbirds are clearly
an important part of the dispersal system of O. endresiana, as over half (52%) of the seeds
they dispersed landed in a zone of higher recruitment
Synthesis: Advantages of Dispersal
Despite nearly complete predation, all three advantages of dispersal receive some
support for Ocotea endresiana. With respect to the escape hypothesis, mortality for both
seeds and seedlings had a component of density and/or distance dependence. Post¬
dispersal seed predation was higher for seeds directly under the parental crowns than for
seeds dispersed away from the crowns (Fig. 2-8). Seed predation was higher for seeds
closer to the trees (Fig. 2-9,2-12). Also, of seeds initially protected from mammalian seed
predators, more dispersed than non-dispersed seeds established as seedlings and survived
one year (Fig. 2-12). Overall, seeds dispersed away from the parent trees had a greater
probability of survival and recruitment These results support the general predictions of the
Janzen/Connell model of higher seed and seedling mortality near the parent tree (Connell
1970, Janzen 1970). However, removal of experimentally dispersed seeds showed that

62
although distance is important, its effect decreased over time because removal approached
100% at all distances. Only two marked seeds were not removed by seed predators: one at
30 m and one at 70 m. Thus, although dispersal is advantageous for removal from the
zone of highest seed predation and seedling mortality near the parent tree, predation is so
high everywhere that dispersal beyond approximately 25 m is apparently not effective in
providing an additional escape advantage. Furthermore, the occurrence of seedlings in the
crown edge plots suggests that some recruitment can take place near the parents. Indeed, a
peak in recruitment occurred within 10 m of the crown edge (Fig. 2-14e), assuming the
pattern of seed predation at 2 wk (Fig. 2-14b). This peak is the one predicted by Janzen
(1970), and it occurs where Hubbell (1980) suggested it would, as a result of high seed
rain close to the parent trees and incomplete density-dependent mortality (see also Condit et
al. 1992).
Thus, factors in addition to escape from seed predation may be important in
recruitment. During a tree's lifetime, recruitment may be episodic, which if true, could
explain why O. endresiana seedlings are fairly common, despite the high levels of seed
predation observed in the two years of this study. Perhaps stochastic events combine to
result in low populations of seed predators and an increase in seedling recruitment only a
few times during a tree's reproductive years. Long-term studies are necessary to test
Hubbell's (1980) hypothesis that the variance in seed predation among years may be more
important than the average in determining the probability of recruitment
Dispersal of O. endresiana by guans, quetzals, robins, and toucanets is consistent
with the colonization hypothesis. Typically, the colonization hypothesis applies to pioneer
species that gain an advantage in random dissemination of seeds to increase the chance that
a few will arrive in a suitable location, such as treefall gaps. Many such species have seeds
capable of dormancy and are incorporated into the soil seed bank, where they wait for a gap
to form above them. Shade tolerant trees also benefit from colonization through
suppressed growth in the seedling and sapling stage rather than the seed stage, and

63
accordingly a "seedling bank" develops in the understory. My results are probably
representative of many shade-tolerant canopy trees in tropical wet forests (Howe 1993a,
Lieberman 1996, Whitmore 1996). Because most O. endresiana seeds are dispersed to
sites within closed canopy forest, most surviving seedlings are part of the understory
seedling bank. During three years of intensive field work in this study site, I only found
two subadult O. endresiana (5-15 m tall), and both were in or near old (> 5 yr) treefall
gaps.
The directed dispersal hypothesis is also supported. Species that have directed
dispersal gain an advantage in arrival at specific sites associated with a higher probability of
survival. The few examples of directed dispersal in bird-dispersed plants are for parasitic
epiphytes (mistletoes) that can establish only on branches (Davidar 1983, Reid 1989,
Sargent 1995, Larson 1996). For O. endresiana, male bellbirds provide directed dispersal
to gaps, where seeds have a slightly lower probability of early predation by mammals
(Table 2-3), and a slightly higher chance of recruitment than in forest understory (Fig. 2-
13,2-12e). One must take into account the fact that most seeds land a few meters beyond
the crowns of the parent trees, and therefore, this zone contributes the most individuals to
the seedling/sapling stage. Seeds that land in a gap edge zone (sensu Popma et al. 1988),
in this case often associated with bellbird song perches, gain an advantage in survival and
growth. Thus, although bellbird perches are not the only places seedlings can survive, a
seed dispersed by a bellbird to a gap has a slightly higher chance of germinating and
surviving than one in the forest understory. Additionally, because gaps in Monteverde tend
to expand when trees on the edge of the gap fall (Lawton and Putz 1988), seeds dispersed
by bellbirds that germinate and survive, are more likely to encounter a favorable growth
environment (i.e., higher light levels) in the future, whereas the prospects for seedlings in
the understory are unpredictable.
These results are consistent with other studies that show shade tolerant species to be
capable of recruitment under a wider range of conditions than shade-intolerant pioneer

64
species (Canham 1989, Denslow and Hartshorn 1994, Lieberman 1996). In montane
forests, shade-tolerant species may be better able to recruit in gaps than in lowland forests,
because cloudy conditions moderate the temperature fluctuations (Cavelier 1996). Thus,
seedlings of shade-tolerant species in montane forests may gain an advantage of higher
light levels in gaps without the increased risk of desiccation found in other rain forests
(Bazzaz and Pickett 1980, Veenendaal et al. 1995, Whitmore 1996). In addition, shade-
intolerant pioneer species are less common and require larger gaps with increasing elevation
in some forests (Whitmore 1996), suggesting that the seedlings of non-pioneers are more
common in montane forest gaps.
Conclusion
The overall conclusion of this study is that the pattern of recruitment of O.
endresiana depends on the combined effects of seed dispersers, seed predators, and
seedling mortality. Selection on plant traits occurs during each stage, and selection during
sequential stages may be opposed (Wheelwright and Orians 1982, Herrera 1985,1986,
Wheelwright 1988, Herrera et al. 1994). For example, high seed predation overall and the
slight preference for larger seeds by seed predators may select for smaller seeds or larger
seed crops, while seed dispersers may prefer larger fruits, which have larger seeds, but
occur in smaller fruit crops (but see Howe and Vande Kerckhove 1980,1981,
Wheelwright 1991, Mazer and Wheelwright 1993). On the other hand, seed size did not
influence germination, seedling height, or seedling survival. In addition, the bimodal
spatial pattern of recruitment may represent disruptive selection on seedling traits. Some
trees were visited by all five species of dispersers, while the trees far from bellbird perches
tended not to be visited by bellbirds. Thus, seeds dispersed to gaps beneath bellbird
perches were mostly from a subset of the available trees. The extent to which such
differences in dispersal and subsequent recruitment affect gene flow is poorly understood
(Gibson and Wheelwright 1995, Hamrick and Nason 1996). Further studies that compare

65
dispersal patterns and the subsequent stages leading to recruitment at different sites, as well
as over longer time periods are especially needed.

CHAPTER 3
SEED DISPERSAL OF A HIGH-QUALITY FRUIT BY SPECIALIZED FRUGIVORES:
HIGH-QUALITY DISPERSAL?
"It is well known that most Lauraceae seeds are free of predation." Oscar C. Castro
(1993:67)
Introduction
In a seminal paper, McKey (1975) proposed that tropical trees producing nutrient-
rich fruits attract specialized frugivores that provide high-quality seed dispersal. He noted
that this strategy of high-quality fruits gaining high-quality dispersal by specialized
frugivores was one endpoint of a continuum of dispersal strategies. High-quality, or
specialist, plants produce large fruits with one or a few large seeds and lipid or protein-rich
pulp. At the other end of the continuum, generalist plants produce large crops of fruits
with many small seeds and pulp composed mostly of sugars and water. The many species
of small, opportunistic birds that eat generalist fruits McKey expected to be less reliable
dispersers because they were thought to be less dependent on fruit than are the specialized
frugivores. McKey’s framework was based on three ideas. First, the general similarity of
fruit characteristics of plants dispersed by different types of dispersal agents (i.e., dispersal
syndromes; Ridley 1930, van der Pijl 1972). Second, observations that fruit was an
important dietary component for many species, but few species were totally dependent on
fruit (Orians 1969, Morton 1973). Third, Snow's (1971) ideas about possible coevolution
between plants with large-seeded, nutritious fruits, and highly frugivorous birds that
dispersed the seeds. These ideas were further expanded by Howe and Estabrook (1977) to
66

67
incorporate crop size, phenology and visitation rates. Although components of the
specialist/generalist framework have been confirmed, data to fully test it are still lacking
(Howe 1993b, Schupp and Fuentes 1995). Howe (1993b) suggested that the main reason
the framework remains untested is because plant ecologists and zoologists study different
aspects of the multiple stages of the plant recruitment process. Quantifying dispersal
quality is key to testing the model, yet to examine dispersal quality, it is necessary to study
both the dispersal pattern and post-dispersal fate of seeds (Janzen 1983c, Howe 1993b,
Jordano and Herrera 1995, Schupp and Fuentes 1995).
Another reason that the specialist/generalist framework has not been adequately
tested is because species-specific coevolution between plants and dispersers is now
considered unlikely; diffuse coevolution (sensu Janzen 1980) between groups of plants and
groups of dispersers is thought to be more typical (Wheelwright and Orians 1982, Janzen
1983c, Howe 1984, Herrera 1985, Levey et al. 1994). The diffuse mutualism paradigm is
based on four main points. First, because plants offer a reward (the fruit pulp) to
dispersers for fruit removal, but no reward for seed dispersal to an appropriate site, plants
have little control over what happens to seeds ingested by potential dispersers
(Wheelwright and Orians 1982). Second, most large, highly frugivorous animals eat a
wide range of fruits including both high-quality and low-quality species (Wheelwright
1983, Wheelwright et al. 1984). Similarly, small-bodied frugivores can be highly selective
among the available fruits, and species of approximately the same body size may handle
fruits and seeds very differently (Moermond and Denslow 1983,1985, Moermond et al.
1986, Levey 1986, 1987). Third, most plant species with fleshy fruits are eaten (and
presumably dispersed) by several to many species of animals of differing body size, seed
handling techniques, and movement patterns (Wheelwright et al. 1984, Bronstein and
Hoffman 1987, Jordano 1992). Fourth, plants probably evolve at different rates than the
animals that disperse them (Herrera 1986). Therefore, the opportunities for dispersal-
related plant and animal traits to coevolve at a species-specific level are minimal, and neither

68
body size nor amount of fruit in the diet can be used to make any consistent prediction
about the quality of dispersal provided by that species (Herrera 1984b, 1986, Howe 1984,
Levey 1987).
Although many studies have addressed dispersal quality in terms of gut treatment of
seeds (Krefting and Roe 1949, Traveset and Willson 1997, Wahaj et al. in press), few
have examined dispersal quality in terms of suitability for growth and survival of sites
where seeds are dispersed (Schupp 1993). It is reasonable to expect seeds from an
individual plant to be dispersed to a variety of sites, and it is well known that all potential
dispersal sites are not equally suitable for seedling establishment or growth to maturity
(Grubb 1977, Harper 1977, Murray 1988, Bazzaz 1991, Schupp 1995). Thus, dispersal
quality may vary within and among conspecific trees, as well as among dispersers.
Many models of dispersal are based on wind-dispersed species (Green 1983, Geritz
et al. 1984, Greene and Johnson 1989, 1996, Okubo and Levin 1989, Andersen 1991)
because so little is known about where animals disperse seeds (Janzen 1983a, Willson
1993, Laman 1996a). Until recently, the difficulty in finding seeds dispersed by animals
has limited investigations of dispersal quality (in terms of the suitability of dispersal sites
for recruitment) to parasitic mistletoes, which have specific and easily quantifiable safe sites
(Davidar 1983, Reid 1989, Sargent 1995), and to ant-dispersed species that are dispersed
over relatively small scales (Horvitz and Schemske 1986b, Hanzawa et al. 1988).
Dispersal quality has been defined as the probability that a dispersed seed will survive to
reproductive age (Schupp 1993). The product of dispersal quality and dispersal quantity
(number of seeds dispersed) equals disperser effectiveness, which is defined as the
proportion of seedlings (Reid 1989) or ideally, adult plants (Schupp 1993) in a population
resulting from activities of a particular dispersal agent (see also Bustamante and Canals L.
1995). Most previous studies have experimentally examined seed and seedling survival
without considering the actual pattern of seed deposition generated by dispersers.
Therefore, dispersal quality cannot be assessed (Howe 1993b, Herrera et al. 1994, Schupp

69
and Fuentes 1995). Although dispersal quantity has been studied more intensively than
dispersal quality, preliminary estimates for Virola indicate that quality is more strongly
correlated with disperser effectiveness than is quantity (Schupp 1993).
One of the prime examples of a specialized dispersal system, and indeed an integral
part of the development of seed dispersal theory, is the plant family Lauraceae (Snow
1971, McKey 1975, Wheelwright and Orians 1982). Most lauraceous trees produce large
one-seeded, lipid-rich fruits, and are dispersed by large highly frugivorous birds such as
bellbirds (Cotingidae), trogons (Trogonidae), and toucans (Ramphastidae) throughout the
Neotropics (Snow 1981, Wheelwright 1983, Avila H. et al. 1996). Species of Lauraceae
are also important for frugivorous birds in the Paleotropics (Crome 1975, Sun et al. 1997).
In this study I examined seed dispersal and seedling establishment of a Neotropical
Lauraceae (Beilschmiedia péndula; hereafter Beilschmiedia), dispersed by four species of
highly frugivorous birds. Beilschmiedia is one of the largest bird-dispersed seeds
throughout much of its geographic range. In the montane forests of northwestern Costa
Rica, it is the largest bird-dispersed seed (range = 5-20g, mean ± SD = 12.9 ± 3.6 g;
(Wheelwright et al. 1984, Burger and van der Werff 1990, Haber et al. 1996).
Beilschmiedia produces fruits that clearly fit the pattern of high-quality fruits. They are
large, have low pulp/seed ratios, and are high in lipids relative to other species of fruits
(Snow 1971, Wheelwright et al. 1984, Moermond and Denslow 1985). Fruits are
produced prior to and during the breeding seasons of the main dispersers, at a time of year
when the number of tree species in fruit is intermediate (Haber et al. 1996). Within the
study site, Beilschmiedia fruit availability overlaps with few other Lauraceae species
(Wheelwright 1985a). The four main seed dispersers are large, deposit intact seeds, and
are reliable consumers in the sense that all four species are relatively common and regularly
eat Beilschmiedia fruits (Wheelwright 1983, Wheelwright et al. 1984, 1985b, Guindon
1996). Naturalist guides in the Monteverde Cloud Forest Preserve note that fruiting
Beilschmiedia trees are predictable spots to find large frugivorous birds. In years or areas

70
when the trees do not fruit, quetzals delay breeding or move elsewhere (T. Guindon, R.
Guindon, A. Villegas, personal communications). The influence of these dispersers on
recruitment of Beilschmiedia, however, is not well known.
The goal of this study was to estimate the quality of seed dispersal provided by
birds for Beilschmiedia. I examined dispersal pattern, seed predation, germination, and 1-
yr seedling survival to determine the influence of each stage on seedling recruitment,
because each stage may be important in limiting recruitment and few studies have integrated
dispersal patterns with their consequences (but see Herrera et al. 1994, Schupp and
Fuentes 1995). First, I determined the spatial distribution of dispersed seeds in relation to
parent trees by searching for regurgitated or defecated seeds beneath and beyond the
canopy of fruiting Beilschmiedia trees. At each site where I found a seed, I protected the
seed from vertebrate seed predators, measured microhabitat variables, and assessed
germination and 1-yr seedling survival to determine which variables were associated with
the highest probability of recruitment in the absence of predation. Then, I examined the
roles of seed predation and possible secondary dispersal in altering the seed shadow
produced by birds by recording the fate of marked seeds placed at the same sites as the
protected seeds. Another way of documenting quality of dispersal is to compare success at
sites where seeds are dispersed to success at random sites. Thus, microhabitat
characteristics, seed predation, germination, and seedling survival were also examined for
seeds placed at randomly located sites. Finally, I compared the spatial distribution of
naturally established saplings to that of dispersed seeds in relation to parent trees to estimate
the effects of longer-term patterns of mortality.
Study Site
This study was conducted January 1995 to June 1996 in the Monteverde Cloud
Forest Preserve (10°12'N, 84°42'W) in the Cordillera de Tilaran, Costa Rica. This
10,000 ha preserve is administered by the Tropical Science Center of San Jose, Costa Rica.

71
The study area was in relatively undisturbed lower montane rain forest (Hartshorn 1983)
along the continental divide at 1600 m elevation. A 5 ha area, 500 m from the beginning of
the Valley Trail (Sendero El Valle), was mapped and marked into 10 x 10 m quadrats with
PVC tubing at every grid point. The site's vegetation is classified as leeward cloud forest
by Lawton and Dryer (1980) The canopy is 25-30 m tall and dominated by several species
of Lauraceae, Sapium oligoneuron (Euphobiaceae), Ficus crassiuscula (Moraceae), Inga
spp. (Leguminosae) and Pouteria viridis (Sapotaceae). The understory is dominated by
Rubiaceae, Acanthaceae, Gesneriaceae, Heliconiaceae, and Aracaceae. The vegetation of
the area is described in more detail by Lawton and Dryer (1980) and Nadkami et al.
(1995). Other characteristics of the Monteverde area are described by Nadkami and
Wheelwright (in press).
The average annual rainfall at 1520 m on the Pacific slope about 3 km from the
study site is approximately 2500 mm, with most of the precipitation occurring between
May and November. Actual rainfall in the study site was probably greater than 2500 mm
(due to the higher elevation of the study site relative to the rain gauge), but the seasonal
pattern was similar (Nadkami and Wheelwright in press). Range gauges underestimate the
amount of precipitation from mist and cloud interception, which contribute up to 50% of
the precipitation in some Neotropical montane forests (Cavelier 1996). Temperatures
recorded at the study site during this project ranged from 15° to 22°C.
Study Species
Beilschmiedia péndula [(Sw.) Hemsley] is a common canopy tree in Costa Rican
montane forests from 600-2000 m (Burger and van der Werff 1990). In the Monteverde
area it occurs from 1500-1600 m (Haber et al. 1996). Beilschmiedia begins flowering in
the late dry season (March) and is pollinated by small flies and other insects. Ripe fruits
are available from mid-January through late April. Fruits have black skin and lipid-rich
pulp (Wheelwright et al. 1984, Burger and van der Werff 1990). Most of the volume of

72
the fruit is a single seed Qength: 48.55 ± 7.65 mm, width: 20.70 + 1.78 mm, mass: 12.89
± 3.60 g; mean ± SD, N = 293) that is composed of a small embryo and two large
cotyledons (for additional measurements see Mazer and Wheelwright 1993). Seed size
varies from 5.1 to 21.7 g. Compared to other genera of Lauraceae, Beilschmiedia has a
relatively thick (2.5 mm) endocarp, and the pulp is more tightly attached to the seed. Seven
Beilschmiedia trees were in the 5 ha study site. Data for two trees that had adjacent
crowns were pooled for analyses. Fruits and seeds for some experiments were collected
from trees on the periphery of the main study area. Large fruit crops (1837 ± 1024
fruits/tree) were produced by the seven trees in 1995, but not in 1994 or 1996 (see also
Wheelwright 1986).
Beilschmiedia fruits are eaten primarily by four species of birds: emerald toucanet
(Ramphastidae: Aulacorhynchus prasinus), resplendent quetzal (Trogonidae: Pharomachrus
mocirmo), three-wattled bellbird (Cotingidae: Procnias tricarunculata), and black guan
(Cracidae: Chamaepetes unicolor). The first three species typically remain in a fruiting tree
after eating several fruits, and often regurgitate viable seeds 30-70 min later under the same
tree or nearby (Wheelwright 1983,1991). Guans defecate seeds in viable condition, and
they generally leave a fruiting tree before defecating the seeds from that foraging bout.
Although seed retention times by guans were not recorded in this study, Guix and Ruiz
(1997) reported a median retention time of 6.2 hr for a larger cracid {Penelope obscura).
All four species can be considered fruit specialists in the sense that they depend on fruit for
most if not all of their dietary requirements, at least at some times of year (Wheelwright
1983, Avila H. et al. 1996). However, quetzals and toucanets also eat large insects, and
small vertebrates (Skutch 1967, Wheelwright 1983, Riley and Smith 1992, Avila H. et al.
1996), while guans also eat leaves (Haber et al. 1996). Bellbirds apparently eat only fruits
although diets of female bellbirds are poorly known (Snow 1982).
Dropped or fallen fruits are eaten by agoutis {Dasyprocta punctata) and possibly
pacas {Agouti paca) and other rodents. Agoutis chew off pulp and sometimes the

73
endocarp, and leave the seeds under the trees, but do not eat or bury seeds. Thus, these
species probably do not provide significant dispersal for Beilschmiedia.
Methods
Seed Dispersal
Seeds were located by systematically searching the ground for freshly regurgitated
or defecated seeds from late January through mid-April. The ground searches started at the
base of a fruiting tree and proceeded along 10 m wide transects (delineated by the PVC
markers) to 100 m from each Beilschmiedia trunk. Additional sites away from the focal
trees were also searched; some were randomly selected and others were chosen due to bird
activity in the area. It was impossible to search the entire site with equal intensity, but an
effort was made to cover the entire site at least once every three weeks, so that over the
course of the three-month fruiting season each 10 x 10 m plot was checked at least four
times. Defecated seeds were assumed to be dispersed by black guans because guans are
the only avian disperser in the area that defecates large seeds. Sites were classified as non-
dispersed if directly under the crown of a fruiting Beilschmiedia tree, or as dispersed if not
under such a tree. All the dispersed and non-dispersed seeds, as well as the seeds placed at
random sites (see below) were covered with wire mesh cages to protect them from
vertebrate seed predators. By doing so, I was able to calculate accurate estimates of
germination and seedling establishment in the absence of seed predation by mammals.
In addition to these naturally dispersed and non-dispersed seeds, other seeds were
placed at 50 randomly selected sites to compare seed predation, germination and survival at
dispersed, non-dispersed, and random sites. These sites were selected with random
numbers generated by a hand calculator and used as coordinates within the study site. The
post-dispersal fate of seeds at these random locations was compared to the fate of seeds at
the dispersed and non-dispersed locations to determine how the probability of recruitment
differs among the three types of sites.

74
Germination and Seedling Survival
The original dispersed and non-dispersed seeds, and all seeds placed at randomly
located sites were protected by a 4 x 8 x 4 cm cage made of 1 cm galvanized wire mesh
held in place by two 25 cm metal stakes. Caged seeds were used to determine germination
rates and insect predation rates in the absence of mammalian seed predators. Each site was
checked weekly for at least 12 weeks and at 15-17 mo after dispersal (hereafter 1-yr
survival). Germination was defined as the splitting of the seed coat and spreading of the
cotyledons. Typically, the radicle had emerged by the census after germination, and a
week later the stem was visible. As each seed germinated and the shoot began to grow, the
cage was removed to allow normal seedling growth. The seedling location was marked
with one of the stakes from the cage. Causes of seedling mortality were classified as
mammal, insect, fungal pathogen, physical, or unknown. Mammals either ate the seed and
left the damaged shoot behind (seed predators) or removed the entire shoot (herbivores).
Some seeds that appeared to have germinated were eventually killed by beetles inside the
cotyledons. Insect-killed seeds frequently developed a root but never had a shoot >2 cm
tall. Seedlings killed by fungal pathogens were characterized by a wilted and discolored
shoot (Augspurger 1990). A seed was considered alive as long as it remained firm, even if
the shoot had been eaten or otherwise damaged. Such seeds resprouted repeatedly.
Post-dispersal Seed Fate
At each dispersed, non-dispersed, and random site, a marked seed was used to
assess rates of seed predation, to identify seed predators, and to determine if secondary
dispersal occurred. Seed predation and secondary dispersal could alter any pattern
generated by dispersers and thus be more important in determining the probability of
recruitment than primary dispersal (Wheelwright 1988, Herrera et al. 1994). For this
treatment I used regurgitated seeds collected under fruiting trees adjacent to the study site.

75
Seeds were marked by gluing 50-75 cm of unwaxed dental floss to the seed, and tying
about 50 cm of flagging tape to the distal end of the floss. Because the glue held best on
seeds with a dry seed coat, seeds were taken inside and allowed to dry for 1-3 h before
gluing. Each marked seed was placed at a site the next morning.
All marked seeds were censused on days 1, 3, and 7, and once each week
*
afterwards until week 5. If a marked seed was removed, the surrounding area was
searched to find the flagging tape-dental floss assembly. The end of the floss where the
seed had been attached was examined to determine the fate (present or absent) of the seed.
In the few cases where seeds were partially consumed, teeth or bill marks were examined
to identify the consumer.
Seedling and Sapling Plots
To determine if Beilschmiedia seedlings and saplings are more likely to recruit
under or away from conspecifics, seedlings and saplings (up to 5 m height) were measured
and mapped in paired 10 x 10 m plots. For each of five trees, one plot was located near the
tree with approximately half of the plot directly under the crown. The second plot was
located at least 20-40 m from the crown edge. Beilschmiedia did not fruit in the study site
in 1994, so most individuals were presumably at least 2 yr old. Such seedlings can be
distinguished from new seedlings by the lack of a seed, woody stem, and presence of
epiphylls.
Microhabitat Characteristics
For all sites and seeds I measured seed characteristics and microhabitat variables
which might influence seed predation, germination, or seedling survival. Seed length and
width were measured with dial calipers and seed mass was measured with a hand-held
digital scale accurate to O.Olg. Canopy cover was estimated with a spherical densiometer
(Lemmon 1957). Leaf litter depth was measured as the number of leaves pierced by a

76
metal stake thrust into the soil at the site. Vegetation density was the number of stems
within a 50 cm radius of the site. The distances to the nearest woody stem, tree > 10 cm
DBH (diameter at breast height), crown edge of fruiting Beilschmiedia tree, and fallen log
were measured with a fiberglass measuring tape.
Statistical Analyses
Data were analyzed with tests from SAS JMP (SAS Institute 1989). Parametric
tests were used unless data violated the assumptions of normality and equal variance, in
which case nonparametric procedures were used. Survival proportions were arcsin square-
root transformed before analysis. Throughout this paper mean values are followed by ± 1
SD.
Results
Over the three-month fruiting season, 217 regurgitated and 27 defecated seeds were
found. One hundred twenty-nine (53%) of these were deposited beyond the crowns of
fruiting trees (dispersed), while 115 (47%) were directly under the trees (non-dispersed).
Of the dispersed seeds, 67% landed within 20 m of crown edge, but some seeds were
found up to 70 m away (Fig. 3-1). Dispersal distance (for dispersed seeds) was not
correlated with seed mass (Spearman's Rho = 0.024, P = 0.77), length (Rho = -0.073, P
= 0.39), or width (Rho = 0.014, P = 0.87).
Seeds defecated (presumably by guans) averaged lighter (Wilcoxon X^ = 6.71, P =
0.009), and shorter (X^ = 9.86, P = 0.002), but not narrower (X^ = 0.77, P = 0.23) than
seeds regurgitated by the other three species (Table 3-1). Defecated seeds were dispersed
significantly farther (39.6 ± 17.2 m) from the parental crowns than regurgitated seeds
(14.3 ± 13.4 m; Wilcoxon X^ = 30.49, P < 0.001).
Very few marked seeds (17%) were eaten or removed after dispersal (or placement
at random sites), and removal rates did not differ among dispersed, non-dispersed, or

77
ui
iU
DC
DC
LU
a
co
a
LU
Ui
in
30 -
20 -
0 5 1 0 15 20 25 30 35 40 45 50 55 60 65 70
DISTANCE FROM CROWN EDGE (m)
Figure 3-1. The average number of regurgitated or defecated Beibchmiedia seeds (+ 1 SD)
at 5 m intervals from crown edges of six fruiting Beibchmiedia trees. The first category
(distance 0) includes all the non-dispersed seeds, while the remaining categories include
only dispersed seeds.

78
Table 3-1. Average (± 1 SD) mass, length, and width of Beilschmiedia seeds
defecated (presumably by black guans) or regurgitated by quetzals, toucanets and
bellbirds. Sizes of regurgitated and defecated seeds were compared with Wilcoxon
rank sum test Measurements for all seeds, including seeds placed at random sites
(N = 294), are shown for comparison.
Defecated
N = 27
Regurgitated
217
All Seeds
294
Mass (g)
Length (mm)
Width (mm)
11.25 (2.82)
43.08 (5.53)
20.40 (1.50)
13.04 (3.54)** 12.89 (3.60)
48.18 (7.84)** 48.55 (7.65)
20.82 (1.20)ns 20.70 (1.78)
** P < 0.01
ns P = 0.23

79
random sites (survival analysis; Wilcoxon X2 = 1.2, df = 2, P > 0.1; Fig. 3-2). The most
conspicuous post-dispersal seed predators were two species of beetles, one of which
(Nitidulae) buried the seeds and fed upon the rotting cotyledonary reserves. The second
beetle species (Curculionidae) consumed the seed by tunneling through the endocarp to
reach the seed. None of the seeds buried by beetles (N = 24, 8.2%) produced seedlings
(although a few germinated), and thus the beetles likely do not act as secondary dispersers.
Small rodents removed eight seeds (2.7%) and took them into burrows at least 50 cm deep.
Such a depth would preclude seedling establishment Therefore, rodents probably are not
secondary dispersers of Beilschmiedia. Black-breasted wood-quail (Odontophorus
leucolaemus) pecked apart seven seeds (based on one direct observation and bill marks left
on pieces of seeds). An additional 11 marked seeds were removed but were not found.
Neither width, length, or mass were significant effects in logistic regression models
predicting 5-wk seed survival (Table 3-2). Seeds at dispersed and randomly located sites
were more likely to survive 5 wk than seeds at non-dispersed sites (Wald X2 = 6.96, P =
0.031) and similarly, seed survival was positively correlated with dispersal distance (Wald
X¿ = 7.33, P = 0.026; Table 3-2). Seed survival was also positively correlated with
amount of leaf litter (Wald X?- = 5.97, P = 0.051), vegetation density (Wald X^ = 11.57,
P = 0.003), and date of dispersal (Wald X^ = 14.00, P < 0.001).
Virtually all caged seeds initiated germination (98%) and the majority (68%) had
established seedlings by late July, three to five months after dispersal. The average
proportion per tree of seeds that germinated, established seedlings, or survived one year,
did not differ among the three treatments (Kruskal-Wallis tests, P's > 0.05; Fig. 3-3).
Overall, however, the number of non-dispersed seeds that established seedlings was
greater than expected if all three treatments had equal survival (X^ = 9.39, df = 2, P =
0.009).
Seed mass was not correlated with time to germination after dispersal (r = -0.086, P
= 0.15), but shoots from larger seeds grew faster after germination (r = -0.24, P < 0.001).

PROPORTION REMAINING
80
Figure 3-2. Post-dispersal predation of seeds from dispersed (N = 129), non-dispersed (N
= 115), and random (N = 50) locations. Removal rates did not differ among the three
treatments.

81
Table 3-2. Results of logistic regressions of post-dispersal survival of marked seeds after 5
weeks, and one-year survival of seeds initially caged against habitat variables. Only
significant effects are listed. Nonsignificant variables included seed mass, seed length,
seed width, canopy cover, distance to log, distance to woody stem, and distance to 10 cm
tree.
Response
R2
-2 log likelihood
(x2)
predictorsa
5 wk seed survival
0.251
72.67***
+ leaf litter*
+ dispersal date***
+ vegetation density***
+ distance from parent*
type (U < D = R)*
1 yr seedling survival
0.12
21.24***
+ 3-mo seedling height**
- dispersal date***
*P^0.05, **P^0.01,
***P<0.001.
asign in front of each predictor indicates positive or negative correlation with the response

PROPORTION SURVIVING
82
Figure 3-3. Proportion of seeds germinating, establishing seedlings 3 mo after dispersal,
and surviving one year, at dispersed (N = 129), non-dispersed (N =115), and random (N
= 50) locations. The average proportion surviving each stage did not differ among the
three treatments (Kruskal-Wallis tests, P's > 0.05).

83
Seeds that established seedlings by July 1995 averaged significantly heavier (13.5 ± 3.5 g,
N = 165) than seeds that did not establish seedlings (12.1 ± 3.6 g, N = 127; t = 3.2, P <
0.002), but seedlings that survived one year (13.1 ± 3.6 g, N = 143) were not from larger
seeds than those seedlings that did not survive (12.7 ± 3.5 g, N = 149; t = 0.87, P =
0.38). Of the seedlings that survived one year, seedling height was positively correlated
with initial seed mass (N = 143, r = 0.41, P < 0.001).
The predominant source of seedling mortality was fungal pathogens, although
many seeds resprouted several times even after fungal attack. Fungal pathogens killed
more seedlings directly beneath the parental crowns (non-dispersed) than seedlings at
dispersed or random locations (X^ = 20.29, df=4,P< 0.001). As for the marked seeds,
some caged seeds (6.8%) were buried or eaten by beetles. Herbivory by insects and
mammals was probably the primary cause of seedling mortality after seedling
establishment, but the effect of herbivory was difficult to quantify because the seedlings
were examined too infrequently.
Seedling survival at one year was predicted by only two variables in the logistic
regression analysis (Table 3-2). In contrast to seed survival at 5 weeks, one-year seedling
survival was negatively correlated with dispersal date (X^ = 9.69, P = 0.002). Seedling
survival was positively correlated with seedling height at 3 mo (X2 = 6.59, P = 0.01; Table
3-2).
The abundance of seedlings and saplings (up to 5 m height) was higher in plots
close to the fruiting trees (paired i-test = 3.44, df= 4, P = 0.005), but the median height of
individuals was greater in the plots 20-40 m away from the trees (Wilcoxon Rank Sums;
X2 = 5.08, df= 1, P = 0.024; Fig. 3-4). Most seedlings (58%) in the plots close to
fruiting trees were less than 50 cm in height, while most individuals (70%) in the plots far
from adult trees were over 50 cm and ranged up to 5 m. Because the Beilschmiedia trees in
the study site did not produce fruits in 1994, the youngest individuals were at least 2 yr
old.

84
CLOSE FAR CLOSE FAR
Figure 3-4. Average number and height of seedlings and saplings up to 5 m tall in paired
10 x 10 m plots (N = 5 pairs) under and 20-40 m away from Beilschmiedia crowns.
Asterisks indicate significant differences (**P = 0.005, *P = 0.024).

Figure 3-5. Average (+ 1 SD) proportion of seeds surviving post-dispersal seed predators
(a), germinating (b), establishing seedlings (c), and surviving one year (d) as functions of
distance from the crown edge. Seeds deposited directly beneath the parental crowns (non-
dispersed) are included in distance 0. The probability of recruitment (e) was calculated as
the product of each of the four previous stages for each tree. Different letters above bars
indicate significantly different means {P < 0.05) based on ANOVA and Fisher's LSD on
arcsin-transformed data. Panels with no letters had no significant differences.

PROPORTION SURVIVING PROPORTION SURVIVING
86
SEED SURVIVAL
1.25
1.00
0.75
0.50
0.25
0.00
0 10 20 30 40 50
GERMINATION
B
0 10 20 30 40 50
ESTABLISHMENT
1-YEAR SURVIVAL
1.00
0.75
0.50
0.25
0.00
0 10 20 30 40 50
RECRUITMENT
DISTANCE (m)

87
The probability of recruitment per seed as a function of distance from parent trees
was calculated as the product of the probabilities of surviving seed predation, germination,
seedling establishment, and 1-yr seedling survival. Seed predation and germination were
not as important in limiting recruitment, as were seedling establishment and 1-yr survival
(Fig. 3-5). Establishment and 1 yr survival were higher for seeds under the crown and up
to 20 m from the crown than for seeds dispersed 30-40 m from the crown (Fig. 3-5). The
overall probability of recruitment varied with distance (One-way ANOVA on arcsin
transformed data, F520 = 5.75, P = 0.002). Recruitment probability was higher 10-20 m
from the crown than within 10 m (Fisher's LSD, P's = 0.01) or beyond 20 m (P's <
0.008; Fig. 3-5E). Recruitment was higher in the zone 20-30 m than 10-20 m from adults,
but the difference was marginally significant (P = 0.0517). Note that these probabilities do
not include the number of seeds per distance interval (Fig. 3-1), because that is a
component of dispersal quantity not quality.
Discussion
The observed spatial pattern of dispersed Beilschmiedia seeds in relation to parent
trees was similar to that of other vertebrate-dispersed trees (Howe 1993a, Laman 1996a).
The vast majority of the seeds disseminated by birds landed under or within 20 m of the
crowns of parent trees. One of the four disperser species, black guan, dispersed seeds
over greater distances than the other disperser species. If distance alone were the most
important factor in determining the probability of recruitment (i.e., high quality dispersal),
then guans might provide higher quality dispersal than the other species. Guans, however,
pass seeds through the digestive tract, and dispersed smaller seeds than the other species,
and small seeds were less likely to establish as seedlings than were large seeds. Thus,
large seeds are more likely to land near the parent trees where they face the possibility of
density-dependent mortality from beetles and fungal pathogens, and competition from
siblings. Wheelwright (1985b) also noted that quetzals, bellbirds, and toucanets are highly

88
selective with regard to fruit size, and tend to drop large fruits and consume smaller fruits.
These dropped fruits are not dispersed any appreciable distance by terrestrial animals, and
the seeds are not eaten very often by vertebrates.
Both germination and seedling survival are relatively high. A seed has about a 40%
chance of surviving one year regardless of dispersal distance or microhabitat. Larger seeds
produced larger seedlings initially, and larger seedlings were more likely to survive one
year. The large seeds of Beilschmiedia enable the seedlings to resprout several times. The
lack of a correlation between seedling survival at one year and initial seed mass is probably
because large seeds are more likely to land under or near the parent where fungal attack and
seed predation were higher than for seeds dispersed farther.
In Howe's (1993b:4) summary of the characteristics of specialized dispersal
systems, the only post-dispersal component listed that could be construed as dispersal
quality is "seed dispersal away from parents (is) critical for recruitment." For
Beilschmiedia, seeds dispersed 30 to 50 m fared no better, and in some cases a little worse,
during the first year than seeds dispersed up to 20 m from the crowns. Many seedlings
survived more than one year under parent trees. Thus, it is unclear from this study if
dispersal of seeds away from parent trees is critical in the sense that it is absolutely
required, or critical in the sense that in leads to a higher probability of recruitment compared
to seeds directly below parents. But clearly, dispersal beyond parental crowns leads to a
higher probability of recruitment than beneath crowns.
Quality of dispersal is defined as the probability that a dispersed seed will survive to
reproductive age (Schupp 1993). Distance from parent trees is the only variable that seems
important for one-year survival of Beilschmiedia. Seed size and the measured habitat
variables had minimal effects at best The data on one-year survival presented above
combined with the distribution of older seedlings and saplings suggest that dispersal a short
distance away from the parent tree leads to a higher probability of survival than either
longer dispersal or deposition directly below the fruiting tree. Density-dependent survival

89
is shown by the higher survival from seed predators and fungal pathogens with distance
and suggested by the taller saplings farther from adult trees. The resulting probability of
recruitment of one-year seedlings reaches a mode 10-20 m from the parent trees.
Approximately 20% of the seeds receive high quality dispersal by being deposited in this
zone. Whether or not seedlings in this zone retain the highest probability recruitment to
reproductive age (to satisfy Schupp's definition of dispersal quality), awaits longer-term
studies. Unfortunately, long term studies on recruitment of long-lived trees are logistically
difficult
Even with some component of high-quality dispersal, the majority of seeds receives
no or low quality dispersal. Seed mass of Beilschmiedia ranges from 5 to 22 g, and the
largest seeds are frequently dropped by the current dispersers (Wheelwright 1985b). These
apparently too-large seeds suggest either an alternative dispersal strategy or a missing
disperser. Although the concept of determining extinct dispersers is fraught with
assumptions (Janzen and Martin 1982, Howe 1985, Hunter 1989, Witmer and Cheke
1991), Wheelwright (1985b) speculated that bare-necked umbrellabirds (Cephalopterus
glabricollis) may be missing from the disperser assemblage of Lauraceae in the Monteverde
continental divide area (see also Fogden 1993). Umbrellabirds were observed in the study
site on several occasions but are not known to breed there currently. Umbrellabirds could
easily swallow the largest Beilschmiedia fruits. The addition of dispersal by umbrellabirds
would considerably alter the results presented in this study. Tapirs (Tapirus bairdii) may
also be a missing disperser. Tapir populations in Costa Rica have been greatly reduced by
hunting (Janzen 1983d). They do occur rarely in the study site, but most observations
were in the wet season when Beilschmiedia was not fruiting. Although tapir tracks were
not observed under fruiting Beilschmiedia trees, the structure of the fruit and seed is similar
to other species eaten and dispersed by tapirs (Janzen 1982a, Bodmer 1991, Fragoso
1997).

90
Another explanation for the large seeds avoided by birds is that the large number of
dropped or uneaten fruits that fall beneath parent trees is an alternative dispersal strategy not
adequately addressed in this study. Agoutis and possibly other rodents eat the pulp from
fallen fruits, but apparently do not eat or scatterhoard the seeds. The removal of fallen
fruits was not quantified in this study, and it is possible that rodents remove some seeds
from under parental crowns, thus providing dispersal equivalent to birds in terms of
distance. However, rodents often chew some or all of the protective endocarp in addition
to the pericarp, which may render the seeds more susceptible to disease, desiccation, or
seed predation by insects. In any case, a Beilschmiedia seed directly under the parental
crown does have a chance of survival (albeit lower than a seed dispersed 10-20 m), in
contrast to other large-seeded species (Clark and Clark 1984, Howe et al. 1985, Crawley
1992, but see Chapman and Chapman 1995).
It is generally believed that Lauraceae seeds have secondary compounds that protect
them from seed predators (Castro 1993), either by making the seeds toxic or unpalatable,
or by limiting the number of seeds an individual seed predator can physiologically tolerate
during a given day. Although the chemistry of Beilschmiedia seeds has not been
examined, the resinous fragrance of cut seeds and the low predation rates suggest that the
seeds are chemically defended. These defenses are likely an important part of the trade-offs
involved in the Beilschmiedia dispersal strategy. Large seed size is thought to foster
recruitment in the shaded understory (Foster and Janson 1985, but see Kelly and Purvis
1993, Hammond and Brown 1995). Because suitable sites for Beilschmiedia germination
and short-term survival are predicted only by a relatively short distance from parent trees,
rather than specific habitat characteristics, the large seeds may enable recruitment in the
shaded understory despite high conspecific density if they are chemically defended and can
avoid density dependent seed predators and fungal pathogens (e.g., Kitajima 1996).
Although seed predation by beetles and rodents, and seedling mortality by fungal
pathogens showed some indication of density dependence, the overall levels of mortality

91
were not high enough to preclude one-year survival of many seedlings under and near the
parent trees.
The data for Beilschmiedia suggest no advantage of dispersal over long distances.
Although theoretically dispersal is always advantageous for gene flow and to increase the
chance of occupying vacant sites (Hamilton and May 1977), the probability of
Beilschmiedia recruitment drops considerably 30-50 m from the parent trees, and few
seeds were observed dispersed beyond 50 m. Perhaps Beilschmiedia recruitment is
limited by edaphic factors or other habitat requirements not measured in this study. One
factor important for many tropical species is mutualistic association with mycorrhizal fungi.
Beilschmiedia seedlings increased relative growth rates after experimental inoculation
(Lovelock et al. 1996). Short-distance dispersal may increase the chance of encountering
mycorrhizal fungi similar to those associated with the parent trees, although mycorrhizae
are not known to be highly species specific (Wilkinson 1997b). Another explanation for
the advantage of short over long distance dispersal is that conditions for growth and
establishment become more unpredictable, or at least predictably not better than close to the
parent with increasing distance. The lack adaptations for long-distant dispersal in many
desert plants may be explained by the harsh environment (Ellner and Shmida 1981). For
these plants, investment in dispersal structures is wasted energy if the chances of survival
are low everywhere, and determined largely by stochastic processes. Perhaps the large
investment per seed in Beilschmiedia limits dispersal to areas near, and thus in similar
habitats to, the parent trees, but favors persistent shade-tolerant seedlings. A similar trade¬
off between dispersal and seed survival has been noted for some African trees (Chapman
and Chapman 1996).
Dispersal quality for tropical trees has received little attention from ecologists
despite insights it may provide about forest dynamics (Howe 1993b, Schupp 1993), not to
mention the theoretical importance it is assumed to have (McKey 1975, Venable and Brown
1993). The results from this study suggest that dispersal quality is more a characteristic of

92
the plants than of the dispersers per se. Although dispersers are important, many are
interchangeable for a given plant species, or treat different plant species differently
(Wheelwright and Orians 1982, Levey 1986). Because dispersal quality is the most
difficult aspect of dispersal to quantify, it may be most profitable to study plant species that
fall at the endpoints of the fruit quality continuum noted by McKey (1975). Specifically,
these include small fruits with many small seeds and pulp low in lipids (e.g., Solanaceae,
Melastomataceae) on one end of the spectrum, and large, single-seeded, nutrient-rich fruits
(or arils) such as Lauraceae, Palmae, or Myristaceae on the other. Much of the attention by
researchers has been on species that are unusual and not representative of most plants (e.g.
Ficus, Cecropia, mistletoes). Ficus includes many species of hemiepiphytes (strangler
figs) and mistletoes are obligate stem parasites. The fruit-like structures of both Ficus
(synconia) and Cecropia (spadices) are relatively large with many small seeds and are eaten
pieces rather than swallowed whole by many potential dispersers, including birds (e.g.,
Levey 1986). Families or genera with wide geographic distributions, either pantropical
(e.g., Lauraceae, Palmae) or cosmopolitan (e.g., Solarium, Juniperus, Prunus) are likely to
lead to fruitful comparative analyses that address the underlying evolutionary factors
involved in seed dispersal systems.

CHAPTER 4
TWO-STAGE DISPERSAL OF TWO SPECIES OF GUAREA (MELIACEAE)
It is well established that animals...bury nuts, and it is widely accepted that these animals
are the primary dispersers of many ...trees and shrubs. However, exactly what happens to
the nuts after they have been buried is poorly understood " Stephen B. Vander Wall
(1990:198)
Introduction
Seed dispersal systems involving two stages with different dispersal agents have
been reported in a variety of tropical plants. Typically, the first stage of dispersal is by
birds or mammals that deposit intact seeds on the ground, and the second stage is either by
ants that discard seeds in refuse piles or nests (Roberts and Heithaus 1986, Kaspari 1993,
Levey and Byrne 1993), or caviomorph rodents (Dasyproctidae) that scatterhoard seeds
and fail to retrieve some of them (Forget and Milleron 1991, Forget 1993, Fragoso 1997).
Dung beetles also act as secondary seed dispersers when they carry off and bury dung
containing seeds (Estrada and Coates-Estrada 1991, Shepherd and Chapman in press), and
similarly, carnivores may occasionally disperse seeds ingested by their prey species (Hall
1987, Dean and Milton 1988, Nogales et al. 1996). In addition to these examples,
VanderWall (1992) reported two-stage dispersal of a temperate pine involving wind and
small rodents, and Clifford and Monteith (1989) described three-stage dispersal of an
Australian shrub involving birds (emus), ants, and explosive dispersal. Examples of two-
stage abiotic dispersal have also been described (Redbo-Torstensson and Telenius 1995,
Fischer 1997, Greene and Johnson 1997). Finally, post-dispersal seed dormancy could be
viewed as a second stage of dispersal through time rather than space (e.g., Fenner 1985,
Murray 1988).
93

94
The key component of all these systems is that secondary dispersers rearrange the
pattern of seed distribution (seed shadow) produced by the first stage of dispersal, and in
doing so, place some seeds in a different microhabitat that may provide better chances of
seed survival. For example, dispersal by ants is often to nutrient-rich sites (Beattie and
Culver 1983, Horvitz and Schemske 1986a, but see Rice and Westoby 1986, Hanzawa et
al. 1988, Bond and Stock 1989), while seed dispersal by scatterhoarding rodents decreases
the probability of seed predation by other species of seed predators even though many if
not most seeds are retrieved and eaten by the scatterhoarder (Hallwachs 1986, Smythe
1989). Thus, secondary dispersal may have a great impact on plant recruitment and may
counteract or accentuate the patterns generated during the first stage of dispersal (e.g.,
Herrera et al. 1994).
In many Neotropical forests, the most important scatterhoarding rodents are agoutis
(Dasyprocta spp.) and acouchis (Myoprocta). They bury seeds 1-3 cm in the soil rather
than simply caching seeds under leaf litter as does Proechimys (Forget 1991), and seeds
under leaf litter are more accessible to other seed predators, such as peccaries and insects,
than are buried seeds (Smythe 1989, Forget 1991). Many other small rodents cache seeds
in deep burrows (larderhoarding) or on branches (Emmons 1990, Vander Wall 1990)
which are probably not suitable sites for seedling establishment
Although many species of plants and seed-eating animals are involved in this
interaction, the effects of scatterhoarding have been studied in only a few species
(Hallwachs 1986, Smythe 1989, Forget 1993, Peres et al. 1997), and thus the influence of
secondary dispersers in rearranging the initial seed shadow is poorly known. In addition,
the importance of scatterhoarding as secondary dispersal for some plants is in dispute
(Larson and Howe 1987, Forget and Milleron 1991).
The purpose of this study was to follow the post-dispersal fate for naturally
dispersed seeds of two species of congeneric trees (Guarea glabra and G. kunthiana). The
specific objectives were to determine the pattern of dispersal generated by birds and

95
compare it to the pattern after seed predation and secondary dispersal, to determine if
secondary dispersal resulted in a shift in microhabitat and overall dispersal distance. Post¬
dispersal fates of the two species were compared to assess the influence of seed size on
removal rates. Experiments in a simple greenhouse were also conducted with G. glabra to
examine the effects of seed burial on germination and seedling establishment.
Study Site
This study was conducted May to August 1994 in the Monteverde Cloud Forest
Preserve (10°12'N, 84°42'W) in the Cordillera de Tilaran in northern Costa Rica. This
preserve is administered by the Tropical Science Center of San Jose, Costa Rica. The
average annual rainfall is about 2500 mm, with most occurring between May and
November. Additional precipitation from mist and cloud interception is probably
substantial (e.g., Cavelier and Goldstein 1989, Lawton 1990) but has not been well
quantified at Monteverde. The study area is in relatively undisturbed lower montane rain
forest (Hartshorn 1983) along the continental divide at 1600 m elevation. A 5 ha area 500
m from the beginning of the Valley Trail (Sendero El Valle) was mapped and marked into
10 m x 10 m quadrats with the grid points marked with PVC pipes. The vegetation of the
area is described by Lawton and Dryer (1980) and Nadkami and Wheelwright (in press).
Study Species
The genus Guarea is widespread in the Neotropics and a few species occur in
Africa (Gentry 1993). The taxonomy and relationships of Costa Rican Guarea species are
not well known (Haber et al. 1996). Guarea glabra (Vahl) is an understory tree ranging 8-
17 m in height, with crown radii approximately 2-5 m, while G. kunthiana (Adr. Juss.) is a
tree of the canopy and upper understory 12-25 m in height, with approximately 8-12 m
crown radii. Both species produce dehiscent capsules with 1 to 8 (usually 4-6) añílate
seeds per fruit. Fresh seed mass averaged 4.5 g (± 1.9) for G. kunthiana, and 0.8 (± 0.2)

96
for G. glabra (N = 50 for each species). The seeds have a distinctive seed coat and
irregular shapes. Along the long axis of the seed, one side is smooth with a lighter-colored
dot near the center, while the other side has a pitted surface from which fibers emerge and
converge into an attachment that holds the arillate seed to the capsule. The red to orange
aril covers all of the smooth side and gets progressively thinner on the other side closer to
the attachment. The seed coat of G. glabra is 0.5 mm (± 0.03) thick, while that of G.
kunthiana is much thicker (1.1 ± 0.1 mm; N = 10 for each species).
Based on 14 hr of watches at fruiting trees and casual observations, the main
consumers of G. kunthiana and G. glabra arils are birds, primates, and squirrels. The
animals most frequently seen feeding in these trees were toucanets (Aulacorhynchus
prasinus) and guans (Chamaepetes unicolor). In addition, G. glabra is dispersed by robins
(Turdus plebejus), and both tree species are occasionally (2-5 observations) dispersed by
mantled howler (Alouatta palliata) and white-fronted capuchin (Cebus capucinus) monkeys,
and rarely (1 observation) by spider monkeys (Ateles geoffroyi). Squirrels (Sciurus
deppei) chew the arils but drop seeds beneath the parent trees. Detailed foraging
observations were not undertaken but additional dispersal agents are expected, especially
for the smaller-seeded G. glabra (e.g., Howe and DeSteven 1979). At lower elevations on
the Pacific slope, Masked Tityra (Tityra semifasciata), Golden-olive Woodpeckers (Piculus
rubiginosus), Olive-striped Flycatchers (Mionectes olivaceus), and Black-faced Solitaires
(Myadestes melanops), in addition to guans, toucanets, and robins, are known to consume
arils of Guarea species (Wheelwright et al. 1984). Five fruiting G. kunthiana, and eight
fruiting G. glabra occurred in the study site.
Methods
Dispersed seeds were located in June - August 1994 by systematically searching the
ground for freshly regurgitated, dropped, or defecated seeds. The searches started at the

97
base of a fruiting tree and proceeded along 10 m wide transects (delineated by the PVC
markers) 50-60 m from the trunk. It was impossible to search the entire site with equal
intensity, but an effort was made to cover the entire site at least once every two weeks, so
that over the course of the 10 wk fruiting season each 10 x 10 m plot was checked at least
four times. Seeds with a damaged seed coat (by gnawing of squirrels or other animals)
directly under the fruiting trees were not included. When fresh, the seeds are light
yellowish-tan in color and become darker in 2-3 days as they are exposed to air. Thus,
recently deposited seeds could be distinguished from older seeds. Seeds were marked by
gluing 50-75 cm of unwaxed dental floss to the seed, and tying about 50 cm flagging tape
to the floss. Seeds were allowed to air dry for 1-3 hours in a lab room before gluing.
Similar marking procedures have been used with no evidence of effects on seed removal in
several other Neotropical sites (Schupp 1988a, Forget 1996, Peres and Baider 1997), as
well as with other species in this same study site (Chapter 2). Each marked seed was
returned to its original location the next morning.
The distance from each dispersed seed to the trunk of the closest fruiting
conspecific (hereafter referred to as the "parent" tree) was measured with a tape measure (if
less than 50 m) or estimated from the study site map if more than 50 m. Seeds were
classified as non-dispersed if directly below the parental crown, and as dispersed if beyond
the crown, but the distance from the edge of the crown was not measured. For all sites I
measured microhabitat variables I thought might influence seed removal. Canopy cover
was estimated with a spherical densiometer (Lemmon 1957). Leaf litter depth was
estimated as the number of leaves pierced by a metal stake thrust into the soil at the site.
Vegetation density was estimated as the number of stems within a 50 cm radius of the site.
The distances to the nearest tree trunk over 10 cm DBH (diameter at breast height) and
fallen log over 5 cm diameter were measured with a fiberglass measuring tape. The same
variables were measured at the sites of buried seeds after scatterhoarding. In addition, the
distance from the initial dispersal location to the secondary dispersal location was

98
measured. These variables were selected based on their importance in previous studies.
Canopy cover, vegetation density, and distance to objects may influence rodent foraging
patterns (Kiltie 1981, Kitchings and Levey 1981, Bowers and Dooley 1993, Vásquez
1996). Leaf litter may influence seed predation or germination (Molofsky and Augspurger
1992, Myster and Pickett 1993, Cintra 1997a).
All sites were censused on days 1, 3, 7, and once each week afterwards for a total
of 12 weeks. If a marked seed was removed, the surrounding area was searched until the
flagging tape-dental floss assembly was found. The end of the floss where the seed was
attached was examined to determine the fate of the seed. If a seed was removed from the
floss and pieces of the seed coat remained, the seed was classified as eaten. If a seed was
buried in the soil (not just under leaf litter) it was classified as cached by agoutis
(Dasyprocta punctata). Seed predation by peccaries (Tayassu tajacu) was determined on
the basis of tracks and overturned leaf litter, typical of peccary foraging packs (Kiltie
1981). At the end of the study, all remaining seeds were examined and classified as either
viable if the seed was solid and not mealy inside, or not viable if it was infested with
insects or was mealy, discolored, or rotten inside.
Experiments were conducted in a screened greenhouse to determine the effects of
seed burial on germination and establishment. In June 1995, seeds were collected from
beneath fruiting trees outside of the study site. Seeds were examined carefully to exclude
those with signs of insect infestation. All seeds were weighed and measured, as
intraspecific seed size variation has been shown to be important in germination and seedling
growth in other studies (Howe and Richter 1982, Moegenburg 1996). In each of 19 plastic
containers (approximately 10 X 10X5 cm) with drain holes, four seeds were planted; two
were buried 2 cm deep and two were placed on the soil surface, for a total sample of 38
buried seeds and 38 surface seeds. The containers were watered frequently to keep the soil
moist (as it would be in the forest at that time of year). They were checked at least once
each week for seedling emergence and growth.

99
Statistical tests from SAS JMP (SAS Institute 1989) were used. Dispersal
distances were not normally distributed, even after appropriate transformations, so in most
cases nonparametric tests were used. Removal rates of the two species were compared
with survival analysis and the Wilcoxon test, which places greater emphasis on earlier
events (Pyke and Thompson 1986). For the greenhouse germination experiment, the
effects of seed size and burial treatment on germination success were analyzed with logistic
regression models (Trexler and Travis 1993). Microhabitat characteristics of initial
dispersal sites were compared with those at cache sites with paired ¿-tests with an alpha-
value adjusted for multiple comparisons (Holm 1979). Throughout this paper mean values
are followed by 1 SD.
Results
Dispersal bv Birds
The first stage of dispersal is largely by birds and occasionally by mammals.
During early June to mid-August, 120 G. kunthiana and 404 G. glabra seeds were found.
Twenty (16.7%) G. kunthiana and 44 (10.9%) G. glabra seeds were deposited under the
crowns (Fig. 4-1). For seeds dispersed beyond the crowns, G. kunthiana seeds were
dispersed farther from the trunk (median = 14 m, mean = 18.1 ± 11.5 m) than G. glabra
seeds (median = 10 m, mean = 13.9 ± 10.1 m; Wilcoxon rank sum test, X2 = 17.07, df=
1, P < 0.001). Approximately 2% (9 of 404) of G. glabra seeds had signs of pre¬
dispersal insect infestation, while 10% (12 of 120) of the G. kunthiana seeds were
unusually small and/or light and later determined to be hollow and inviable. Hollow seeds
of G. glabra were never found, and only 2 G. kunthiana seeds had obvious signs of pre¬
dispersal insect infestation (Table 4-1).

100
CO
Q
LU
Ul
CO
cc
111
m
3
Z
CO
Q
Hi
Ui
CO
CC
LU
00
2
3
DISTANCE FROM TRUNK (m)
Figure 4-1. Number of seeds regurgitated or defecated by birds and mammals (bars), and
the number of seeds buried by rodents (lines) at 10 m distance intervals from fruiting eight
Guarea glabra and five G. hmthiana trees. Distance class 0 includes all seeds directly
below the crowns, while the other distance classes are measured from the trunks. Crown
radii were approximately 2-5 m for G. glabra and 6-10 m for G. kunthiana.

101
Table 4-1. Post-dispersal fate of Guarea seeds after 12 weeks. Seeds regurgitated or
defecated by birds and mammals were marked with a dental floss and flagging tape
assembly (see Methods), and searched for if removed. Seeds were considered eaten by
mammals if removed from the floss and pieces of the seed coat remained, eaten by insects
if entry or exit holes were visible, cached if found buried in the soil, and inviable if
appearing mealy.
G. kunthiana G. glabra
Fate
N
(%)
N
(%)
removed, eaten
72
(60.0)
157
(38.9)
removed, cached, eaten
23
(19.1)
48
(11-9)
removed, cached
9
(7.5)
136
(33.7)
removed, missing
2
(1.7)
1
(0.2)
not removed, viable
0
(0)
20
(4.9)
not removed, insects
2
(1.7)
39
(9.7)
not removed, inviable
12
(10.0)
3
(0.007)
Total sample
120
404

102
Post-dispersal Seed Fate: Seed Predation and Scatterhoarding
The most striking difference in the post-dispersal fate between the two tree species
was that peccaries rapidly consumed many G. kunthiana seeds that fell under or were
dispersed near the parent trees, whereas they did not consume G. glabra. The combination
of seed predation by peccaries and rodents and scatterhoarding by agoutis left no viable
uncached G. kunthiana seeds of the original sample of 120 after 12 weeks. Agoutis buried
32 (27%) seeds singly 1-3 cm deep. Only 9 (28%) of these buried seeds had not been dug
up and eaten after 12 weeks (Table 4-1). Three cached seeds were dug up and reburied,
but all three were later eaten. In the descriptions of cache site microhabitats below, only the
characteristics of the first cache are included.
In contrast to G. kunthiana, 45% (184 of 404) of G. glabra seeds were found
buried and 136 of those (74%) remained buried after 12 weeks. In addition, 20 (5%)
apparently viable seeds were not eaten or cached (Table 4-1). Insects killed many seeds
(10%) beneath and near parent trees compared to G. kunthiana, where removal by
peccaries superseded insect seed predation.
The overall removal rates (seeds eaten plus seeds cached) were similar for the two
species (Fig. 4-2), but G. kunthiana seeds were more often eaten than cached (survival
analysis; Wilcoxon = 4.6, df = 1, P = 0.03). For both species, the highest proportion
of cached seeds occurred in the first three weeks (Fig. 4-2). After that time, the proportion
of cached G. kunthiana declined as most seeds were dug up and eaten, while the proportion
of G. glabra remained relatively constant as most cached seeds were not retrieved.
The fate of cached G. kunthiana seeds was influenced by distance to parent trees.
Cached seeds that were eventually dug up and eaten were closer to parent trunks than were
seeds that were not retrieved (Wilcoxon Rank Sum test, P < 0.001). Of 23 seeds that
were cached and later eaten, 19 were directly beneath the crowns of parent trees. For G.
glabra, distance from conspecifics did not influence the fate of cached seeds (Wilcoxon
Rank Sum test, P = 0.69).

103
c
®
o
i_
o
O.
100
75 -
50 -
25 -
Guarea alabra
â–¡ NOT REMOVED
ES BURIED
â–¡ EATEN
c
a>
o
w
©
o.
G. kunthlana
â–¡ NOT REMOVED
B BURIED
0 EATEN
WEEK
Figure 4-2. Cumulative percent of seeds eaten (diagonal lines), buried (stippled), or not
removed (open space at top) over the course of the fruiting season for G. glabra and G.
kunthiana. Note that time zero is not the same day for all seeds; the individual phenologies
have been adjusted for ease of comparison.

104
The distance between the initial dispersal site and the site of burial was greater for
G. kunthiana, (6.1 ± 4.0 m) than for G. glabra (1.2 ± 1.3 m; Wilcoxon rank Sums, P
<0.001). The net outcome of secondary dispersal was a slight increase in the average
distance from parent trees (Table 4-2). The seed shadow of cached seeds was also
truncated because most seeds directly under the parents as well as those far (> 30 m) from
parents were eaten rather than cached (Fig. 4-1). The initial distance from the trunk was
not correlated with how far seeds were moved during secondary dispersal for G.
kunthiana: (Spearman's Rho = 0.068 P = 0.354), or G. glabra: (Rho = 0.0584, P =
0.755).
G. kunthiana seeds were often cached near logs; secondary dispersal sites were
closer to logs than were the initial dispersal locations (paired t-test: t = 3.24, df = 31, P =
0.003; Table 4-2). For G. glabra, secondary dispersal sites tended to have less leaf litter (t
= 3.86, df = 183, P < 0.001) and lower vegetation density (f = 3.38, df = 183, P <
0.001) than the initial dispersal locations. The same trends were evident for G. kunthiana,
(leaf litter: t = 2.66, df = 31, P = 0.012; vegetation density: t = 2.69, df = 31, P = 0.011;
Table 4-2).
Germination
None of the marked G. glabra seeds germinated in the field. Of the cached seeds
that remained after 12 weeks, 79% (107 of 136) appeared mealy or rotten. Most of the rest
(14%) were infested with insects, while the remainder (7%) were solid and appeared
viable. The 9 cached G. kunthiana seeds remaining at the end of the study appeared
viable.
In the greenhouse, germination was higher for buried than for surface seeds (X2 =
22.03, df = 1, P < 0.001; Fig. 4-3). Seeds in the buried treatment established more
seedlings than surface seeds (X2 = 11.08, df = 1, P < 0.001; Fig. 4-3). In logistic
regression models including seed width, length, and mass in addition to a treatment effect

105
Table 4-2. Habitat characteristics (average ± SD) of Guarea kunthiana and G. glabra
dispersal locations before (primary dispersal site) and after scatterhoarding (secondary
dispersal site). N = 32 for G. kunthiana and 184 for G. glabra. Within each variable,
values for each site before and after scatterhoarding were compared with paired-t-tests with
an alpha value of 0.008 adjusted for multiple comparisons. See text for details of test
results.
G. kunthiana G. glabra
primary
secondary
primary
secondary
conspecific tree (m)
15.67
(11.85)
15.96**
(10.09)
12.75
(10.15)
13.99**
(8.67)
10 cm tree (m)
1.23
(0.91)
1.57
(0.85)
1.95
(1.05)
1.74*
(0.89)
log (m)
1.48
(1.38)
0.91**
(0.84)
1.93
(1.68)
1.82
(1.57)
leaf litter (#)
2.56
(1.29)
1.69*
(1.47)
2.55
(1.43)
2.01**
(1.13)
vegetation density
43.09
(18.70)
31.28*
(15.97)
26.35
(15.45)
21.63**
(11.17)
canopy cover (%)
95.28
(1-82)
95.50
(2.05)
96.05
(2.51)
96.30
(3.63)
*0.05 > P > 0.008 (marginally significant)
**P^ 0.008

NUMBER OF SEEDS
106
germinated not
Figure 4-3. Number of seeds buried 2 cm or placed on the soil surface that germinated and
established seedlings in a greenhouse. N = 38 seeds in each treatment.

107
(buried or surface), only treatment was a significant predictor of the probability of
germination (R2 = 0.23, X2 = 18.59, P < 0.001).
Discussion
Secondary Dispersal
Scatterhoarding in combination with seed predation resulted in a rearrangement of
the seed shadow, in terms of an increase in mean distance from parent trees, as well as a
shift in microhabitat. In particular, seeds directly below parent trees are more likely to be
eaten rather than cached, or retrieved if cached. On the other hand, the increase in distance
would be expected at random, especially for seeds close to the tree, because the possible
destinations in a circle centered on a seed range from 100% farther way from the tree
directly at the trunk, to approximately 50% way with increasing distance. For G.
kunthiana, it is apparent that peccaries focus on the seeds close to parent trees, and they
seemed to regularly check the trees several times during the fruiting season. Peccaries also
exploited the similarly clumped resources of Inga spp. later in the year. The faster removal
of G. kunthiana seeds is attributable to peccaries, which preferred them over G. glabra.
A shift in microhabitat is also a consequence of scatterhoarding, and perhaps more
important than a shift in distance. Guarea kunthiana seeds were often cached near fallen
logs, while G. glabra seeds were often cached in areas with less leaf litter and sparser
vegetation. It is possible that the process of scatterhoarding removes some leaf litter and
herbaceous vegetation, as the same trends were evident (but only marginally significant) for
G. kunthiana. While G. glabra was sometimes cached near logs, most often the caches
were not obviously associated with any object. In addition, G. kunthiana seeds tended to
be moved longer distances during secondary dispersal. These differences between the
species suggest that more than one animal species was involved in scatterhoarding, or that
one species of disperser treats the two seed species differently. Forget (1990) points out
that the only Neotropical species known to bury seeds in the soil are caviomorph rodents

108
(Dasyprocta, Myoproctd). If other species are involved in scatterhoarding, the likely
candidates in Monteverde are Sciurus deppei, Peromyscus nudipes, and Heteromys
desmarestianus. Species in all three of these genera are possible scatterhoarders in other
areas (VanderWall 1990). The possibility of scatterhoarding by non-caviomorph rodents in
the tropics deserves further study.
Germination
Lack of germination in the field was likely an experimental artifact The hypocotyl
of Guarea emerges from the middle of the long side of the seeds, rather than from the end.
Unfortunately, the most convenient place to glue the floss used to mark the seeds, was
directly on top of the point of emergence, which was not determined until later. Thus the
glue and line assembly may have interfered with germination and resulted in most seeds
rotting. Nevertheless, the germination experiments showed that burial is advantageous for
G. glabra. More buried seeds germinated and established seedlings than seeds on the soil
surface. The importance of seed burial has been shown in several other studies. The
clearest case is that of Astrocaryum palms in Panama, which Smythe (1989) showed were
eaten by peccaries and insects if not scatterhoarded by agoutis. In addition to seed
predators, seeds on the surface face greater fluctuations in temperature and moisture than
do buried seeds. Buried seeds may also have longer induced dormancy than surface seeds
(Fenner 1985).
Poor germination of G. glabra in the field may explain the low retrieval rates (26%)
of cached seeds compared to G. kunthiana (72%). With some species scatterhoarded by
agoutis the food reward is the developing seedling (Emmons 1990, Forget and Milleron
1991), particularly the fleshy cotyledons of species with epigeal germination (e.g.,
Garwood 1996). Plants may shift defensive compounds from cotyledons to leaves after
germination, thus rendering the cotyledon more palatable (Janzen 1981). If such

109
compounds in G. glabra limit seed consumption by agoutis, perhaps they cache seeds to
consume less-defended parts of seedlings after germination.
Conclusions
The alteration of the seed shadows during secondary dispersal is likely an important
stage in recruitment for Guarea, and other species as well. Very few studies have
examined seed dispersal by scatterhoarding rodents. Judging from those available (Smythe
1989, Forget 1993,1996, Cintra 1997b, Peres and Baider 1997) and the ubiquity of
caviomorph rodents in Neotropical forests (Wright et al. 1994), the effects of
scatterhoarding on forest composition could be considerable. Peres and Baider (1997)
suggested that scatterhoarding by agoutis can lead to clumped distributions of trees.
Fragoso (1997) also found a clumped distribution of palms that were dispersed primarily
by tapirs and secondarily by agoutis. He attributed the clumped distribution, however, to
tapirs, rather than to agoutis, because the tapirs deposit many seeds in habitual latrines up
to 2 km from existing palm patches, whereas the agoutis then space out the seeds from the
latrines (Fragoso 1997, see also Bodmer 1990). Other studies, however, have not found
clumped patterns of tree species scatterhoarded by caviomorph rodents (Forget 1997). The
relationships of primary dispersal, secondary dispersal, seed predation, habitat effects, and
seedling survival, are just beginning to be unraveled and require further study to better
understand the role of multi-stage dispersal in forest dynamics.

CHAPTER 5
ADVANTAGES OF DISPERSAL: A RE-EVALUATION OF DIRECTED DISPERSAL
"But the real importance of a large number of eggs or seeds is to make up for much
destruction at some period of life; and this period in the great majority of cases is an early
one." Charles Darwin (1859:77)
"I have no doubt that, per gram eaten, the seed predators have the largest impact on tropical
forest structure of any animal life form." Daniel H. Janzen (1983b:172)
"It matters who defecates what where." D. H. Janzen (1986:251)
Introduction
Seed dispersal has a major influence on plant fitness because it determines the
locations in which seeds, and subsequently seedlings, live or die. Theoretically, plants
could increase fitness if a higher proportion of seeds were dispersed to sites where
offspring have a predictably high probability of survival relative to random sites. This
process is known as directed dispersal. To do so, a plant must have a predictable dispersal
vector, either biotic or abiotic, that takes seeds disproportionately to suitable sites. Such a
pattern of dispersal likely will be retained in the distribution of adult plants (Schupp and
Fuentes 1995), in which case a particular dispersal agent has a disproportionate effect on
plant recruitment and demography. If true, then patterns of dispersal, not just dispersal,
may have a greater role in the structure and composition of plant communities than
currently believed.
To document directed dispersal, two conditions must be met: (1) nonrandom
movement of seeds and, (2) disproportionate arrival in sites with known characteristics that
are especially favorable for survival (Howe and Smallwood 1982, Howe 1986). Although
110

Ill
nonrandom movement patterns and habitat selection by animals, and variation in
microhabitat suitability for plant growth are well documented, examples of directed
dispersal are thought to be rare and unusual (Howe 1986). I suggest that directed dispersal
is more common and ecologically significant than previously believed, but has been
overlooked for several reasons. First, the current paradigm of diffuse coevolution between
plants and their dispersers does not predict the evolution of adaptations for directed
dispersal. Second, the alternative advantages of dispersal (escape, colonization, directed
dispersal; see below) are not mutually exclusive and can be difficult to distinguish; the
demonstration of one advantage does not preclude the others. In particular, the emphasis
of research on escape obscures the occurrence of directed dispersal. Third, few studies
have integrated patterns of seed distribution (seed shadows) and the consequences of such
patterns for plant recruitment in a way that can detect directed dispersal. Plant recruitment
is a multistage process that has most often been studied in a stage-specific manner (Houle
1995, Jordano and Herrera 1995, Schupp and Fuentes 1995). Zoologists have studied the
fruit removal and gut treatment stages, while botanists have focused on the seedling stages
(Howe 1993b). Fourth, finding dispersed seeds, especially those dispersed by vertebrates,
is difficult Identifying parents of dispersed seeds is also difficult, and so far has been
limited to studies of isolated trees (Gladstone 1979, Augspurger 1983, Augspurger and
Kitajima 1992), or indirectly by genetic studies (Gibson and Wheelwright 1995, Hamrick
and Nason 1996) Fifth, several different names have been used for the same or closely
related ideas. The term directed dispersal was introduced by Howe and Smallwood (1982)
at the same time the initial framework of seed dispersal theory was being critically
examined (Wheelwright and Orians 1982, Herrera 1985,1986), Thus, because directed
dispersal was not expected under the diffuse mutualism view, the term was not widely
adopted, and instead terms such as 'safe sites,' 'nurse plants,' 'recruitment foci,'
'succession facilitation,’ and 'targeted dispersal' were used for concepts related to directed
dispersal (Grubb 1977, Harper 1977, Stiles 1989, Bazzaz 1991, Vieira et al. 1994).

112
I review the literature on seed dispersal and argue that directed dispersal is not
inconsistent with the diffuse mutualism paradigm and although subtle, may be quite
common. After defining and describing the advantages of dispersal for plants, and
outlining the development of the diffuse mutualism paradigm, I will briefly outline three
classic examples of directed dispersal: scatterhoarding pine seeds and acorns by corvids,
dispersal by ants to nests or refuse piles, and dispersal of mistletoes by small birds. I then
discuss what may be further examples of directed dispersal to illustrate how widespread it
may be. Finally, I will suggest specific research needed to demonstrate directed dispersal.
The directed dispersal hypothesis can be tested with data on patterns of dispersal and post¬
dispersal fate.
Advantages of Dispersal
Plants display many morphological features associated with differing methods of
seed dispersal (e.g., Beal 1898, Ridley 1930, van der Pijl 1972). Considering the cost to
the plant of producing such dispersal-related structures, it is reasonable to expect some
advantage to dispersal (Hamilton and May 1977, Thompson and Willson 1978, Howe and
Smallwood 1982). Three hypotheses concerning the spatial aspects of dispersal have been
proposed: 1) escape from high mortality caused by distance- or density-dependent factors
near conspecifics (escape hypothesis), 2) colonization of rare, unpredictable, ephemeral
sites, such as treefall gaps (colonization hypothesis), and 3) directed dispersal to particular
microhabitats especially suitable for survival (Howe and Smallwood 1982, Howe 1986,
Willson 1992). Because these hypotheses are not mutually exclusive, they can be assessed
only if dispersal sites and post-dispersal fates are known.
In addition to the three hypotheses mentioned above, several other advantages of
dispersal have been proposed and will be mentioned here but not discussed in detail
because they are not concerned with the spatial distribution of seeds in relation to adult
trees. Gene flow is another advantage of dispersal (Levin and Kerster 1974). Gene flow

113
helps avoid inbreeding, and may occur during both pollination and dispersal (Hamrick and
Nason 1996). The contributions of pollination and dispersal to gene flow can be
distinguished by recently developed techniques, and future studies should be helpful in
evaluating the influence of different dispersal vectors on genetic structure of plant
populations (Hamrick and Loveless 1986, Hamrick et al. 1993, Gibson and Wheelwright
1995, Weiblen and Thomson 1995, Alvarez-Buylla et al. 1996). Note that inbreeding is
not necessarily a problem for all plant species, and in some cases short-distance dispersal is
favored over long-distance (Howe and Smallwood 1982). Also, the spatial distance among
conspecific trees is not necessarily a good indication of the genetic distance among
individuals, or the distance traveled during dispersal (e.g., Gibson and Wheelwright
1995).
Effects on germination have also been proposed as advantages of dispersal
(reviewed by Traveset and Willson 1997). Although animals may scarify seeds and
influence the rate or proportion of seeds germinating, the effects of gut passage have little
to do with dispersal per se (Howe and Smallwood 1982). Consumption of fleshy fruits by
vertebrates is generally advantageous for removal of the pulp, but often does not otherwise
affect germination (Chapter 2, Janzen 1982b, Howe et al. 1985, Jackson et al. 1988).
Escape. The escape hypothesis is expected to be an advantage for most plants and
is supported by numerous studies that have shown density- or distance-dependent mortality
near parent trees (e.g., Clark and Clark 1984, Howe 1986). Note that high levels of seed
predation or seedling mortality are not sufficient to demonstrate benefits of escape. If
99.9% of a plant's progeny is killed regardless of location, then escape is not an advantage
of dispersal. Instead, the evidence should be disproportionate mortality correlated
negatively with distance or positively with density (Howe 1986). Even if escape is shown
to be an advantage for a particular plant species, however, colonization or directed dispersal
could also be important for the seeds that do escape.

114
Colonization. The colonization hypothesis is most relevant when suitable sites for
establishment are unpredictable or randomly distributed, as is thought to be the case for
new treefall gaps in tropical forests (Hartshorn 1978, Brokaw 1985, Van der Meer and
Bongers 1996). For species that require canopy gaps to reach reproductive maturity, the
expected dispersal strategy is colonization via blanketing the understory with propagules
capable of dormancy or suppressed growth until a gap forms and increased light levels
induce germination or more rapid growth. Widespread dispersal of seeds in the area
around the parent plants will maximize the number of different sites occupied, and increase
the chance that some seeds land in sites that become suitable in the future. Colonization of
sites currently favorable, such as recent disturbances, should involve dispersal to such sites
in proportion to their abundance. Note that in most forest ecosystems, gaps vary greatly in
size and most gaps are small (Brokaw 1982b, Martinez-Ramos and Alvarez-Buylla 1986,
Murray 1986, Lawton and Putz 1988, Runkle 1990, Gray and Spies 1996). The
importance of small gaps and even subcanopy gaps is not well known (Connell et al.
1997).
Directed Dispersal. Alternatively, directed dispersal can result if dispersal agents
deposit seeds disproportionately in suitable locations. Thus, directed dispersal has two
components: non-random arrival and increased survival in predictable locations (Howe and
Smallwood 1982, Howe 1986, Schupp et al. 1989). To benefit from directed dispersal, a
plant must have fruits with characteristics that attract certain dispersers more than others, or
must have a morphology that enables the propagules to arrive in certain habitat patches
more often than expected by chance (Venable and Brown 1993). Note, however, that the
examples below show that plants need not be adapted specifically for directed dispersal as
was originally theorized. Instead, I argue that directed dispersal occurs in many situations

115
in which a plant has multiple dispersal agents, and some provide directed dispersal to
particular sites while others provide widespread dissemination to a variety of sites.
Distinguishing the Alternative Advantages. Many plants are likely to benefit from
more than one, if not all three advantages. Distinguishing the relative roles of each
advantage for different plants should implicate which specific dispersal agents are
particularly important in different ecosystems. If seeds are dispersed to sites with random
microhabitat characteristics (not including distance from parent), then the colonization
hypothesis is supported. If seeds disproportionately arrive in a nonrandom subset of the
available microhabitats and post-dispersal survival is predictably higher in those
microhabitats than in random sites, then the directed dispersal hypothesis is supported.
Escape is supported if dispersal decreases the probability of being killed by seed predators,
herbivores, or other mortality factors that act in a density dependent fashion. In practice
however, these three hypotheses can be difficult to distinguish (Howe 1986). I now
discuss the reasons for this difficulty and suggest potential solutions.
Most dispersal does not result in seedling establishment, even by "high quality"
dispersers, thus directed dispersal is likely to be subtle. Because most seeds die, research
on escape has overshadowed colonization and directed dispersal. Much larger sample sizes
are needed to detect factors correlated with seedling survival than for seed removal in a
natural setting (Howe 1990b). In addition, many studies either do not attempt to
distinguish the three hypotheses, combine colonization and directed dispersal into
colonization, or ignore directed dispersal altogether (Herrera and Jordano 1981, Dirzo and
Dominguez 1986, Martinez-Ramos and Alvarez-Buylla 1986, Levey et al. 1994).
Furthermore, all the data to evaluate directed dispersal are seldom collected in one study.
In particular, the ecological literature is dominated by stage-specific studies, whose results
cannot necessarily be used to distinguish among the hypotheses. Long-term studies that
document survival through sequential stages including dispersal, germination, seedling

116
establishment, and growth to maturity, are needed to evaluate the consequences of dispersal
and thereby tease apart the hypotheses (Herrera et al. 1994, Houle 1995, Schupp and
Fuentes 1995). Such studies are difficult for long-lived species.
In most cases, the distribution of dispersed seeds is highly leptokurtic and right-
skewed, with most seeds near the parent and progressively fewer farther away (Janzen
1970, Levin and Kerster 1974, Willson 1993, Herrera et al. 1994, Houle 1995, Schupp
and Fuentes 1995). The shape of the curve is best described by either negative exponential
or inverse power functions (Portnoy and Willson 1993, Willson 1993, Laman 1996a, but
see Murray 1988). Therefore, seed shadows are always non-random with respect to
distance from the parent plant Plotting the abundance of seeds over distance as a linear
function, however, can obscure the great heterogeneity within the tails of the distribution
(Janzen 1983a, Stiles 1989, Portnoy and Willson 1993, Willson 1993). It is important to
note that non-random dispersal associated with such heterogeneity is a necessary but not
sufficient condition for directed dispersal. For directed dispersal to occur, non-random
dispersal must be to suitable sites, so that the overall result is a positive, disproportionate
effect on recruitment Similarly, many studies show heterogeneity in seedling
distributions. Yet, these patterns could be caused by higher seed fall in some areas than
others, or by higher survival in some areas, or both (Hubbell 1980, Jordano and Herrera
1995, Schupp and Fuentes 1995). The important point is that on a per seed basis, directed
dispersal yields a higher probability of survival to maturity than does widespread
dissemination (i.e., colonization).
A key element for distinguishing directed dispersal from colonization is determining
where seeds actually land. Few studies have determined the seed shadows generated by
different dispersal agents. This aspect is particularly important for plant species dispersed
by many different dispersal agents. By attracting many disperser species with different
behaviors and dietary requirements, the seeds are likely dispersed to many different types
of sites. Such a pattern suggests that widespread dissemination is the main advantage of

117
dispersal (e.g., Laman 1996b). However, a subset of the disperser assemblage may be
responsible for dispersing most of the seeds that eventually establish. Even for plant
species with a restricted set of dispersers, it has been shown that disperser species differ in
the per seed probability of establishment (Reid 1989).
The other critical aspect necessary to assess directed dispersal is determining post¬
dispersal plant fate. It is well known that plant mortality is generally highest in the seed
and seedling stages (Darwin 1859, Harper 1977). Every potential recruit requires a space
in which to grow, in addition to water, nutrients, and light. Such spaces are rare in the
environment. Therefore, the probability of establishment for any seed is very low. It
follows then, that any survival benefit at early stages is likely to increase the chances of
further survival (Schupp and Fuentes 1995). As further illustration of distinguishing the
dispersal hypotheses, consider the following hypothetical situation. A tree or shrub with
small-seeded fruits attracts a variety of birds that disperse 100 seeds in the area around the
parent plant in a typical right-skewed distribution. Insects kill all but one of the seeds
within 10 m of the plant, and rodents kill all but ten of the seeds beyond that distance.
Nine of these seeds are taken by ants, eight of which are eaten and one of which is
discarded in viable condition on a refuse pile. The seed not taken by ants is incorporated
into the soil seed bank when a peccary steps on it and pushes it into the mud, where it can
remain dormant for 2 years. The other two seeds, one within 10 m of the parent plant and
one in the ant refuse pile both germinate and establish as seedlings. Because of the
nutrients in the refuse pile, that seedling grows larger. In this example, note that escape,
colonization, and directed dispersal are all supported. Of the three live offspring produced
by the plant which is most likely to reach maturity? The best choice is the largest seedling,
in which case the relatively rare event of directed dispersal contributes as much or more to
the plant's fitness as escape and colonization. In summary, more detailed studies on where
seeds are dispersed, how they get there, and what happens afterwards are necessary to

118
distinguish the hypotheses. I know of only one study that has attempted to examine the
three hypotheses (Masaki et al. 1994).
The Diffuse Mutualism Paradigm
The evolutionary aspects of the fruit-frugivore interaction noted by Ridley (1930)
and van der Pijl (1972), were first clearly developed by Snow (1965, 1971). These ideas
were later expanded into the specialist/generalist dichotomy of plant traits and dispersal
quality (McKey 1975, Howe and Estabrook 1977). The expectation was that large-seeded,
nutrient-rich fruits would attract dispersal agents that provided high quality dispersal in
terms of dispersal to sites appropriate for establishment. In contrast, small but abundant
nutrient-poor fruits attracted opportunistic foragers that provided lower quality dispersal,
but the sheer number of seeds dispersed resulted in a few landing in appropriate sites. In
other words, seeds in high-quality fruits were expected to benefit from directed dispersal,
while seeds in low quality fruits were expected to benefit from widespread dissemination.
Both benefited from escape. The implication was that fruit-frugivore evolution could be
relatively specific, involving adaptations between pairs or small sets of species, as seen in
flower-pollinator evolution (Wheelwright and Orians 1982). Over the next decade,
however, the view that emerged from more detailed studies (mainly on avian foraging
behavior) was that coevolution between plants and their dispersers was not species-
specific, but instead was diffuse, involving adaptations between large groups of plants and
groups of dispersers (Wheelwright and Orians 1982, Howe 1984, Herrera 1985, 1986,
Levey et al. 1994). Indeed, although plant species vary tremendously in the number of
potential dispersers they attract, and a few have very restricted sets of dispersal agents
(Janzen and Martin 1982, Chapman et al. 1992), there are no unequivocal examples of a
plant species entirely dependent one species of disperser (Witmer and Cheke 1991).
Some interpretations of the diffuse mutualism framework seem to exclude the
possibility of directed dispersal. Two of the main objections to the specialist/generalist

119
dichotomy are (1) that plants reward dispersers for removing fruits but not for dispersing
them to appropriate sites and thus can do little to influence where dispersers take seeds, and
(2) suitable sites for plant establishment (safe sites) are spatially and temporally
unpredictable (Wheelwright and Orians 1982, Janzen 1983c). Studies on pollination were
well ahead of those on seed dispersal when specialist/generalist framework was being
developed (Howe 1993b), and both the specialist/generalist dichotomy and the critiques
that followed were based on comparisons to pollination. This comparison to pollination is
exemplified by the statement "plants cannot direct the dispersal of seeds to a particular
location with a degree of exactness comparable to pollen dispersal, though possibly they
could favor animal vectors with particular foraging behaviors, habitat preferences, or
probabilistic patterns of seed dispersal" (Wheelwright and Orians 1982:406). By
implication, escape and colonization became the most likely benefits of dispersal, and
directed dispersal became unlikely as an adaptive strategy. Note, however, in the above
quotation that "particular foraging behaviors, habitat preferences, or probabilistic patterns
of seed dispersal" are exactly the conditions that result in non-random patterns of dispersal.
If these patterns of dispersal are coupled with arrival to suitable sites, then directed
dispersal is achieved. In fairness, the objections to the specialist/generalist theory raised by
Wheelwright and Orians (1982) and others, were not to reject directed dispersal per se but
rather to refute the idea of species-specific coevolution between plants and dispersers.
Interest in seed dispersal mutualisms paralleled that in factors promoting tropical
forest diversity, especially seed predation and seedling mortality in relation to adult
distributions (Connell 1970, Janzen 1970, Hubbell 1980). The escape hypothesis is also
referred to as the Janzen/Connell hypothesis, although the two are not equivalent (Howe
and Smallwood 1982, Clark and Clark 1984, Burkey 1994, Cintra 1997b). As noted
above, the escape hypothesis predicts disproportionate mortality as a function of distance
from parent plants, while the Janzen/Connell hypothesis states that such disproportionate
mortality for one species allows other species a higher probability of establishment, and

120
thus promotes species coexistence. Both hypotheses predict an advantage of local
dispersal. Subsequently, seed predation experiments have dominated the literature with
relatively little effort placed on documenting seed survival as a function of any other
variable beside distance from parents (but see Schupp 1988a, Kitajima and Augspurger
1989, Howe 1990b, Laman 1995, Cintra 1997a). In addition, because actual patterns of
seed dispersal are so poorly known, demonstration of the escape advantage predicts
nothing about colonization or directed dispersal, but suggests that one or both may also
occur.
Classic Examples of Directed Dispersal
Three systems are typically given as examples of directed dispersal: scatterhoarding
by corvids, mistletoe dispersal by small birds, and dispersal of elaisome-bearing seeds of
herbaceous plants by ants. Although these systems are often thought of as exceptions to
the general rule of diffuse coevolution (Wheelwright and Orians 1982, Levey et al. 1994),
they are actually consistent with the theory, as all three examples involve each species of
plant being dispersed by more than one species of disperser, which in turn disperse several
other species of plants (e.g., Restrepo 1987). They may be more specialized than most
other plant/disperser systems and perhaps illustrate one endpoint of a continuum of
dispersal strategies.
Corvids
Approximately 20 species of pines (Pinus; most in the subgenus Strobus) are
dispersed by jays and nutcrackers (Corvidae; Tomback and Linhart 1990, Benkman 1995).
Most other pines are wind-dispersed, although they may benefit from secondary dispersal
by scatterhoarding rodents (Vander Wall 1992). These pines occur in western North
America and across Eurasia, and are distinguished from the wind-dispersed species by
having large seeds that lack a well-developed winged appendage for wind dispersal, and

121
cones that do not release the seeds (Toraback and Linhart 1990). The most specialized
dispersers are the two species of nutcrackers (Nucifraga). Nutcrackers have a sublingual
pouch (unique among birds) in which they carry up to 90 seeds (Vander Wall and Baida
1977, Tomback 1982). They bury seeds in the ground in shallow surface caches of 1-4
seeds. Presumably they disperse seeds considerable distances (Lanner 1996). Each
nutcracker caches thousands of seeds each year, exceeding its dietary requirements 2-5
times (Vander Wall and Baida 1977, Tomback 1982). Several other corvids cache pine
seeds and are important dispersers, including pinyon jays (Gymnorhinus cyanocephalus),
western scrub jays (Aphelocoma califomica), and Steller's jays (Cyanocitta stelleri). These
birds have highly developed spatial memory (Baida and Kamil 1989, Kamil and Jones
1997), but do not retrieve all caches every year (Lanner 1996). Most seeds that are not
taken by birds and fall beneath adult Pinus monophylla trees are harvested by rodents,
which are less effective dispersers than corvids because they larderhoard many seeds in
nearby burrows (Vander Wall 1997). Most of the corvid-dispersed pine species are found
in xeric habitats, where burial protects seeds from desiccation (Lanner 1996). However,
both the corvids and the rodents scatterhoard some seeds in sites under bushes that provide
shade and are thus beneficial for germination and establishment (Vander Wall 1997). In
addition, some of the pines are early successional species (pioneers) and the birds,
nutcrackers and pinyon jays in particular, are known to make frequent caches in open areas
(Lanner 1996).
Corvids and squirrels also disperse oaks and beeches (Fagaceae) by
scatterhoarding. Although the presumed mutual adaptations in these species are less clear
than for nutcrackers and pines, in is clear that many seeds are cached in suitable sites for
germination and establishment (Bossema 1979, Darley-Hill and Johnson 1981, Sork 1983,
Johnson et al. 1997). In addition, dispersal by jays has been implicated in the rapid post-
Pleistocene northward spread of oaks and other plant species (Johnson and Webb 1989,
Wilkinson 1997a, Clark et al. 1998).

122
Mistletoes
Mistletoes (about 40 genera; Viscaceae, Loranthaceae, Eremolepidaceae) are
obligate stem-parasites that occur worldwide and are dispersed by passerine birds. The
pulp contains sticky viscin that is difficult to separate from the seed, even during gut
passage. The viscin enables the seeds to cling to a branch until germination and connection
with the host xylem. The viscin also allows several seeds to stick together and to the bird,
during defecation or regurgitation, and increases the chances that the string of seeds will hit
a branch or be wiped directly onto the branch by the bird. In Australasia, mistletoes are
dispersed by several species of flowerpeckers and mistletoebirds (Dicaeidae; Docters van
Leeuwen 1954, Reid 1991). In forested regions of Central America and South America
they are dispersed mainly by euphonias and other tanagers (Thraupidae) and to a lesser
extent flycatchers (e.g., Zimmerius vilissimus, Tyrannidae; Davidar 1983, Restrepo 1987,
Stiles and Skutch 1989, Monteiro et al. 1992, Sargent 1995). In a Chilean desert site,
mockingbirds {Mimas thenca; Mimidae) are the main dispersers of mistletoe (Martinez del
Rio et al. 1995,1996), while in North America, waxwings and phainopeplas
(Borabycillidae) are the main mistletoe dispersers (Walsberg 1975, Skeate 1987, Larson
1996). Mistletoe seeds in fruits not eaten by birds, and seeds dispersed to the ground have
no chance of survival.
Studies by Reid (1989) and Sargent (1995) illustrate directed dispersal of
mistletoes. Suitable establishment sites are not simply twigs and branches, but live twigs
of a certain size, depending on the species. In Australia, Amyema quandang mistletoes
established best on 1-6 mm diameter twigs (Reid 1989), while in Costa Rica,
Phoradendron robustissimum established best on 10-14 mm twigs (Sargent 1995). Larger
twigs and branches have thicker bark that may preclude establishment, and smaller twigs
are more likely to die than larger twigs after mistletoe infection (Sargent 1995). Sargent
(1994) showed that euphonias tend to perch on twigs of the appropriate size for

123
establishment, and therefore, within a tree establishment was on a non-random set of the
available twigs and branches. Reid (1989) similarly showed that mistletoebirds (Dicaeum
hiruninaceum) were more likely to deposit seeds on the appropriate-sized twigs than were
honeyeaters (Acanthagenys rufogularis). In both cases, a restricted set of dispersers is
entirely responsible for successful mistletoe establishment. Non-random distribution of
mistletoes among the available host plants (host preferences) have been shown in several
other studies (Monteiro et al. 1992, Martinez del Rio et al. 1995, Larson 1996).
Host preferences have been noted for other epiphytes (e.g., Ficus; Daniels 1991,
Laman 1996, Patel 1996, Putz 1989), but it is unclear if they are related to the pattern of
dispersal or of survival. Disproportionate establishment in certain sites does not imply
disproportionate dispersal to those sites (Schupp and Fuentes 1995), although August
(1981) did suggest that bats provided directed dispersal of fig seeds to palm trees in
Venezuela. Laman (1995, 1996b), however, proposed that host preferences of Bornean
Ficus are caused by differences in establishment requirements not by differential dispersal,
as Ficus seeds are dispersed by many species of birds and mammals and probably achieve
widespread dissemination.
Ants
A wide variety of tropical and temperate plants produce seeds with lipid-rich
elaisomes that attract ants (Beattie and Culver 1981, Horvitz and Schemske 1986b,
Ohkawara et al. 1997). This ant-plant mutualism is worldwide, occurring in habitats
ranging from deserts to tropical forests, and involving representatives of over 80 plant
families and numerous ant taxa (Beattie 1985). Typically, ants take the seeds back to their
nests and discard the seed after the nutrient-rich elaisome has been consumed. Seeds are
dispersed when placed on refuse piles or abandoned in nests (Beattie and Culver 1982).
Because the nests or refuse piles have higher concentrations of nutrients available to
seedlings compared to the surrounding soil, seedling growth or establishment is often

124
higher in such sites than in random sites (Davidson and Morton 1981, Horvitz 1981, Levey
and Byme 1993). In one of the few studies to document the fitness benefits of dispersal by
animals, Hanzawa et al. (1988) showed than a cohort of ant-dispersed Corydalis aurea
produced 90% more offspring that a control cohort. Arrival to nutrient-rich sites, however,
is apparently not the explanation for myrmecochory in some areas (Rice and Westoby
1986, Bond and Stock 1989). In California chaparral, ant nests are located in the gaps
between shrubs rather than under shrubs, so that seeds discarded in nests may face lower
competition from other plants (Boyd 1996).
In contrast to vertebrate frugivores which usually eat the reward (fruit pulp) before
dispersing seeds, ants typically take seeds back to their nests before eating the elaisome
(Culver and Beattie 1978, Horvitz 1981). In some species, removal of the elaisome
induces germination, and thus the ant-dispersal provides a predictable cue that a suitable
site has been reached (Horvitz 1981). Seeds not taken by ants are likely eaten by rodents
and other seed predators (O'Dowd and Hay 1980, Heithaus 1981, but see Smith et al.
1989), but many species release the seeds during the morning, when ants are active but
rodents are not (Gibson 1993, Ohkawara et al. 1997). Ants also act as secondary
dispersers of seeds in bird droppings (see below).
Other Examples of Possible Directed Dispersal
Abundant circumstantial evidence hints that directed dispersal is more common than
previously believed. Aside from the systems described above, few studies have attempted
to test the directed dispersal hypothesis. In this section I describe several possible
examples of directed dispersal that should be investigated in future studies. None of these
examples is unequivocal because seed shadows generated by individual disperser species
and post-dispersal seed fates are poorly known. But in all cases, one of the two
requirements of directed dispersal is met: either nonrandom patterns of post-dispersal
survival or disproportionate dispersal by a particular vector to certain sites.

125
Dispersal bv Wind and Birds to Gaps
Canopy gaps caused by dead or fallen trees or branches are crucial recruitment sites
in most forest ecosystems. Typically 1-2% of the forest canopy is opened annually, and
about 5% of the forest area is in gaps formed in the last 5 years (Lawton and Putz 1988,
Martinez-Ramos et al. 1988, Murray 1988). Because many plant species require gaps for
germination and seedling growth, seed arrival in gaps would be highly advantageous for
these species. Seeds can arrive via the seed bank (past dispersal) or via seed rain (current
dispersal). The former case would be colonization while the latter could be colonization if
arrival is in proportion to the area available, or directed dispersal if arrival is more often
than expected by chance.
Air currents around gaps may pull in wind-dispersed seeds from the surrounding
area (Schupp et al. 1989). In support of this hypothesis, at least three studies have found
disproportionate arrival in gaps of wind-dispersed species (Augspurger and Franson 1988,
Denslow and Gomez Diaz 1990, Loiselle et al. 1996). Similarly, Be tula lenta seeds
dispersed by wind across snow accumulated in indentations resulting in non-random seed
distributions (Matlack 1989). Because many wind-dispersed species are shade-intolerant
and require gaps for establishment, wind dispersal may often result in directed dispersal.
Many more wind-dispersed than animal-dispersed seeds arrived in gaps in all three
studies mentioned above, but overall seed rain was higher in forested sites. Nevertheless,
that does not preclude the possibility of directed dispersal to gaps by animals as well. All
three studies noted great heterogeneity among the samples, suggesting patterns in seed rain
at a scale not well quantified by paired gap/understory seed traps. One possible cause of
such heterogeneity in seed rain is that seed-dispersing animals may concentrate their
activities where resources are most abundant. It is well known that fruiting plants produce
larger crops in gaps (Levey 1988a, 1988b, 1990, Martinez-Ramos and Alvarez-Buylla
1995) and that frugivorous birds are also especially active in and around gaps even though

126
few species can be considered gap specialists (Schemske and Brokaw 1981, Willson et al.
1982, Levey 1988b). Based on these studies, Schupp et al. (1989) hypothesized that seed
rain of small-seeded animal-dispersed seeds would be higher in tropical forest gap edges,
but no studies have tested this idea. In Illinois, however, Hoppes (1988) did find such a
pattern; seed rain was often bimodal with most seeds landing near the parents and a second
smaller peak at gap edges. Overall, Hoppes (1988) estimated 50% of bird-dispersed seeds
landed in gaps and gap edges which together comprised only 16.8% of the study area. In
Japan, seed dispersal of Comus controversa by birds was not disproportionate to gaps,
but gap and gap edge sites were not distinguished (Masaki et al. 1994). Both Hoppes
(1987) and Denslow and Gomez-Diaz (1990), noted the likely occurrence of seed dispersal
from plants in one gap to another gap. The most likely plant species to benefit from
directed dispersal to gaps are those that can establish, not in new gaps, but in building-
phase gaps a few years old (Denslow 1987, Brandani et al. 1988).
Canopy gaps are formed when a tree or branch falls or when a tree dies standing.
Locations of new gaps are thought to be randomly distributed and unpredictable (Hartshorn
1978, Brokaw 1985, Martinez-Ramos et al. 1988, Van der Meer and Bongers 1996). The
fact that trees on gap edges are more likely to fall than are trees in closed canopy forest
(Lawton and Putz 1988, Young and Hubbell 1991, Van der Meer and Bongers 1996)
suggests that gap edges are predictably more favorable locations than random understory
locations for species that benefit from higher light levels. In addition, in some forests,
gaps are aggregated (Lawton and Putz 1988, Van der Meer and Bongers 1996). In
montane Costa Rica, Sapium oligoneuron trees often die standing and are frequently used
by Bellbirds as song perches (Chapter 2). Standing dead trees drop large branches for
several years before falling entirely. Thus, gaps may not always be as unpredictable as is
commonly believed. The possibility of directed dispersal to gaps or gap edges in tropical
forests remains to be examined.

127
Habitual Perches and Habitual Defecation Sites
Many frugivorous tropical birds have lek breeding systems in which the males have
display perches where they spend the majority of the day during the breeding season. Such
species range from manakins (Pipridae) to the much larger bellbirds, umbrellabirds, and
cocks-of-the-rock (Cotingdae) in Neotropical forests, and birds-of-paradise (Paradisaeidae)
in New Guinea. Many of these species are common and conspicuous. The males of two
species of manakins and two species of bellbirds spent 87-95% of the day at display
perches, leaving only for brief foraging bouts (D.W. Snow 1962b, 1962a, B.K. Snow
1970,1977). Most of the seeds dispersed by males of these species are probably deposited
in the vicinity of the leks or display perches. If perch or lek areas are also suitable
establishment sites, then these species may provide directed dispersal. For example three-
wattled bellbirds (Procnias tricarunculata) in Costa Rica typically display on tall exposed
perches such as dead branches or dead trees (Snow 1977). In a study on seed dispersal of
the canopy tree Ocotea endresiana (Lauraceae) 52% of the seeds dispersed by bellbirds
landed in gaps (under display perches) compared to only 3% of the seeds dispersed by four
other species (Chapter 2). Overall, including seeds whose dispersal agent was unknown,
1*1% of the seeds landed in gaps, which was much greater than the 5.3% expected based on
the area of the forest in gaps < 5 yr. Seedlings in gaps had almost twice the chance of
surviving one year compared to seedlings in closed canopy forest Thus, bellbirds
provided directed dispersal: disproportionate dispersal to especially suitable sites. Other
species of bellbirds and cotingas have similar behavior in perching on exposed limbs or in
trees with sparse foliage (Snow 1982), and considering the wide variety of fruits they eat,
bellbirds are likely to provide directed dispersal for other species as well.
Other lekking species such as manakins and cocks-of the-rock display in the forest
understory. Although this pattern may lead to colonization via accumulation of dormant
seeds in the soil (Krijger et al. 1997), and suppressed seedlings, Endler and Théry (1996)
have shown that Guianan cock-of-the-rock (Rupicola rupicola) and two species of

128
manakins (Corapipo gutturalis, Lepidothrix serena) select lek areas in specific light
environments in the forest understory so as to maximize conspicuousness during displays.
Some species of manakins even maintain their leks by removing leaves that obstruct the
view at lek height (Snow 1976). The importance of differences as small as 1-2% in light
availability has been demonstrated for some shade-tolerant seedlings (Howe et al. 1985,
Howe 1990b), and it is possible that some of the species dispersed by lekking species
benefit from higher light levels around leks. In cock-of-the-rock leks in French Guiana,
77% of the plants present as seedlings and saplings were likely dispersed by Rupicola,
while only 11.5% of the species in a forest understory site were shared with the lek (Théry
and Larpin 1993). Thus, the Rupicola leks can have a great impact on the vegetation
structure, and may contribute to clumped and patchy distributions of fruiting plants. In
contrast, manakins that disperse seeds of shade-intolerant Melastomataceae, but display in
the understory, contribute heavily to the soil seed bank but apparently not to the vegetation
around the lek (Krijger et al. 1997).
A similar situation has been found for sleeping trees of howler monkeys (Alouatta
seniculus) in French Guiana (Julliot 1996, 1997). Of six plant species studied, all were
more common as seedlings in the vicinity of sleeping roosts than on control sites. This
difference could be a result of higher seed input (as suggested by Julliot 1997) or by
differential survival of seedlings under roosts. Although the presence of seedlings
demonstrates successful establishment, the effects of seed predators and secondary
dispersal were not quantified. Habitual sleeping sites have been noted for several other
frugivorous primate species (Chapman 1989b, Heymann 1995) and forkinkajous (Potos
flavus; Julien-Laferriere 1993).
Lowland gorillas (Gorilla gorilla) in Gabon defecate around nest sites which are
typically located in open areas of the forest (Turin et al. 1991). Gorillas were the most
important disperser of the most common tree in the study area (Cola lizae: Sterculiaceae).
Cola seedling survival six months after dispersal at nest sites in open areas was 40%, but

129
only 1.1% in forest Although the seedlings in fecal clumps may face intense sibling
competition, the open canopy at nest sites provides a favorable growth environment (Tutin
et al. 1991). Gorillas eat fruits of at least 78 plant species, and may disperse many other
species in a similar pattern (Tutin et al. 1991).
Similarly, baboons (Papio anubis) in Ghana congregate daily on rocky outcrops
where they often defecate seeds. Lieberman et al. (1979) found seeds from 59 species in
dung collected from the outcrops. They noted that although the rocks themselves were not
good establishment sites, many seeds were likely washed by rain to the surrounding soil,
and the vegetation around the outcrops was dominated by the same plant species baboons
commonly dispersed.
Rhinoceros (Rhinoceros unicornis) in Nepal, and tapirs (Tapirus terrestris) in Brazil
have habitual latrine sites (Dinerstein and Wemmer 1988, Dinerstein 1991, Fragoso 1997).
Rhinoceros latrines in floodplain grasslands provide crucial recruitment sites for the shade
intolerant tree Trewia that dominates the riverine forest. Seeds in dung piles are not
secondarily dispersed by rodents but may be scattered by repeated use of latrines and
occasional floods (Dinerstein 1991). The seedlings may face interspecific competition but
are located in nutrient-rich sites. Seeds in tapir latrines are scatterhoarded by agoutis (see
below, Fragoso 1997).
Latrine use has also been documented for European badgers (Meles meles) in Italy
(Pignozzi 1992). Badgers have latrines around the periphery of their territories, and
dispersed the seeds of at least 10 plant species although seedling survival was not
documented. Other examples of seed dispersal concentrated at latrines or burrow sites of
highly frugivorous mammals deserve further study (e.g., palm civets, Paradoxurus
hermaphroditus; Joshi et al. 1995). In addition, species such as rabbits (Oryctolagus
cuniculus), are known to have latrines (Sneddon 1991), but their role as seed dispersers is
not well known (but see Nogales et aL 1995, Schupp et al. 1997a, b). In some cases,
dispersal concentrated in certain sites will no doubt lead to extremely high seed or seedling

130
mortality due to inappropriate growing conditions, intense seedling competition, or
attraction of predators or pathogens (Snow 1979, Howe 1989). Based on the studies
above, however, it is clear that concentrated seed dispersal is not necessarily
disadvantageous, even when mortality is high.
Nurse Plants
In a variety of arid ecosystems, shrubs and trees provide the critical recruitment
sites for many plants, and are referred to as nurse plants. In contrast to treefall gaps in
forests, the spaces between plants in arid lands are hostile to seedlings; water and heat
stress rather than light availability are often the factors limiting plant growth (Valiente-
Banuet et al. 1991, Fulbright et al. 1995, Callaway et al. 1996, Nabhan 1997). Soil under
nurse plants, especially nitrogen-fixing species, is sometimes higher in nutrients than the
surrounding soil (Barnes and Archer 1996), but nutrient availability is not necessarily the
main limiting factor (Franco-Pizaña et al. 1996). Whereas the lack of perches in recent
gaps may limit seed rain of animal-dispersed seeds in forests, nurse plants in arid and
semiarid ecosystems are often the only perches available and thus provide both high seed
input in a nonrandom pattern and a favorable microclimate for bird-dispersed species.
Several studies have shown birds leaving fruiting plants in non-random directions, and
higher seedling emergence under certain post foraging perches than others (Herrera and
Jordano 1981, Tester et al. 1987, Izhaki et al. 1991, Chavez-Ramirez and Slack 1994). In
some cases, seed predation under shrubs is lower than in the open (Callaway 1992), but
other studies have suggested the opposite result (Owens et al. 1995, Hulme 1996, 1997).
Dispersal by birds to nurse plants in arid and semiarid ecosystems presents a strong case
for directed dispersal, but virtually all studies have focused on seedling growth and
interspecific plant associations without documenting patterns of dispersal. Nurse plants
may be particularly important for maintaining populations of rare plant species (e.g., Cardel
et al. 1997).

131
¿jyirge Lggs
Similar to nurse plants, decomposing logs may serve as important recruitment sites
for some species. Whether seeds arrive at logs disproportionately has not been
documented, but the influence of nurse log abundance on forest composition has been
suggested (Scowcroft 1992, Hofgaard 1993). In addition, the propensity of certain bird
species to perch on fallen logs has been noted (e.g., Turdus albicolis; Charles-Dominique
1986). In a temperate woodland, 33% of the species on logs were probably ant-dispersed
and 25% were fleshy-fruited species typically dispersed by birds and mammals (Thompson
1980). As ant nests are often located in or near logs, and birds and mammals often perch
on logs, especially in treefall gaps, Thompson (1980) suggested that logs may differentially
accumulate seeds. Similarly, several Melastomataceae species have been noted in
association with logs in two tropical forests (Lawton and Putz 1988, Lack 1991). These
melastome species produce fleshy fruits typically dispersed by birds, and sometimes
removed from bird droppings and secondarily dispersed by ants (Levey and Byme 1993).
Some other species on fallen logs, however, clearly were established on the trees before
they fell (Lawton and Putz 1988). Another possibility for directed dispersal to nurse logs
is the dispersal of fungal spores by rodents (e.g., Cork and Kenagy 1989, Johnson 1996).
Perches in Successional Landscapes
Numerous studies have documented high rates of seed dispersal by animals to
perches in pastures, Oldfields, or other human-disturbed landscapes (Debussche et al.
1982, Guevara and Laborde 1993, McClanahan and Wolfe 1993, Robinson and Handel
1993, Da Silva et al. 1996, Ne'eman and Izhaki 1996). Such sites are referred to as
recruitment foci (McDonnell 1986, Myster and Pickett 1992) or succession facilitators
(Vieira et al. 1994). In these situations, the dispersal pattern is clearly nonrandom, as seed
rain beneath perches is often 1 or 2 orders of magnitude greater than in open areas (Willson

132
and Crome 1989, Robinson and Handel 1993, Nepstad et al. 1996, Duncan 1997).
Although the pattern of seed rain is similar to that for nurse plants in arid ecosystems, the
suitability of such sites for growth and survival is less clear. In active pastures, where
grazing animals would eat or trample seedlings in the open, dispersal beneath shrubs,
fencerows, or by rocks may represent the best recruitment sites within the pasture
(Livingston 1972, Herrera 1984a), even though more suitable sites may exist in the
adjacent habitat from which the seeds were dispersed.
This situation raises the possibility that directed dispersal may become more
common in managed or highly disturbed landscapes than in natural habitats. In managed
landscapes the distribution of safe sites may be more predictable than in natural areas. Safe
sites in managed landscapes are not necessarily rarer than in undisturbed areas, especially if
they are located near highly invasive species (De Pietri 1992, Vieira et al. 1994), or planted
shrubs or trees (Robinson and Handel 1993, Debussche and Isenmann 1994). If true, this
also illustrates the idea that plants may not be adapted for directed dispersal, but can take
advantage of it in some situations. In successional landscapes the species most likely to
benefit from directed dispersal are the smaller-seeded "generalist" plants that attract many
dispersers, rather than the larger-seeded fruits with a more restricted disperser assemblage
(Da Silva et al. 1996). It follows then, that the advantages of directed dispersal in
successional landscapes would diminish for the generalist plants as succession approaches
the climax stage, and then colonization of unpredictable sites increases in importance (see
also Debussche and Isenmann 1994).
Secondary Dispersal bv Agoutis. Ants, and Dung Beetles
Recent studies in tropical forests indicate that dispersal by vertebrates is often only
the first of two stages of dispersal. Seeds in frugivore defecations are removed by ants
(Roberts and Heithaus 1986, Kaspari 1993, Levey and Byrne 1993) and rodents (Janzen
1986, Moegenburg 1994, Fragoso 1997). In both cases, the secondary dispersers act

133
mainly as a seed predators, but the beneficial treatment of the seeds they do disperse may
lead to a large influence on plant recruitment Ants are ubiquitous in Neotropical forest
understory environments and rapidly remove small seeds from bird droppings. The ants
take seeds back to their nests and most seeds are eventually eaten. In one study, about 6%
of the seeds were discarded on refuse piles, but perhaps more importantly, nest twigs and
the seeds within them, were abandoned on average every three months (Levey and Byrne
1993). Seedlings had higher survival on experimental refuse piles under light levels typical
of small gaps. Thus, ants remove seeds from a high-density pattern in a frugivore
defecation where interspecific competition may be intense (Howe 1989, Loiselle 1990),
and place a few seeds in low-density, high-nutrient sites favorable for seedling growth.
Some plant species have been noted in association with rotting logs (Thompson 1980,
Lawton and Putz 1988, Lack 1991), and perhaps ants are the mechanism of seed arrival at
such nurse logs. Kauffman et al. (1991) suggested that some Ficus seeds have an
elaisome and are adapted for secondary dispersal by ants, but Byrne and Levey (1993)
found no obvious elaisomes on the Melastomataceae seeds commonly harvested by ants
from bird droppings.
Tapirs in Amazonian Brazil disperse palm seeds (Maximiliano, maripa) to latrine
sites, which are used repeatedly (Fragoso 1997). In terre firme forest, 96% of 25 tapir
latrines were located by buttresses of the emergent tree Couratari multiflora
(Lecythidaceae). Most seeds not dispersed by tapirs are killed by insects. The most
remarkable aspect of this system is that agoutis removed seeds from tapir latrines and
buried them in shallow surface caches. The density of seedlings and saplings <, 5 yr old
was significantly higher around latrine sites than around either parent trees or non-palm
control trees. Fragoso (1997) suggested that the combination of tapir and agouti dispersal
was responsible for the clumped distribution of the palms observed over large areas of the
forest Clumped distributions of other species of trees have been attributed to agoutis in
other Neotropical forests as well (Peres and B aider 1997). Scatterhoarding by agoutis can

134
be either primary or secondary dispersal, depending on the plant species, although the
benefits of scatterhoarding to seeds are similar in both cases; in addition to escape from
some insect and mammalian seed predators, burial may prevent desiccation and facilitate
germination and establishment (Hallwachs 1986, Smythe 1989, Forget 1990, Vander Wall
1990). Agoutis have been frequently noted caching seeds by logs or other objects (Smythe
1978, Hallwachs 1986, Forget 1990,1993, Chapter 4) but the extent to which these sites
represent suitable establishment sites has not been studied. In addition, Kiltie (1981) noted
that peccaries often forage near objects, perhaps to locate agouti caches.
Another example of secondary dispersal is removal of mammalian dung, and the
seeds it may contain, by dung beetles (Estrada and Coates-Estrada 1991, Estrada et al.
1993, Shepherd and Chapman in press). In this case, the dung beetles have no interest in
the seeds, and may even discard larger seeds (Andresen in press). The beetles bury small
balls of dung several cm deep. The females oviposit on the dung, the larvae consume most
of the dung ball, and leave the seeds behind. Estrada and Coates-Estrada (1991) found that
80% of the plant species dispersed by howler monkeys benefited from secondary dispersal
by dung beetles. They suggested that burial by beetles allow escape from rodent seed
predators, but the likely benefits of burial in nutrient-rich dung have not been examined.
Numerous species of dung beetles may be involved in secondary dispersal, and they vary
in the amount of dung taken, how far it is taken, and how deep it is buried (Estrada et al.
1993, Andresen in press).
Conclusions; Going the Distance, and Bevond
The examples above illustrate the potential for directed dispersal in a wide variety of
systems, and suggest its importance in restoration ecology, conservation, and in
understanding plant-animal interactions. A test of the directed dispersal hypothesis is
within reach, but the lack of detailed data on patterns of dispersal generated by particular

135
vectors, integrated with post-dispersal seed and seedling fate limits our understanding of
the role of directed dispersal in plant recruitment. In addition, seed rain into different
patches at the community level is still poorly known. Surprisingly, the models of Schupp
et al. (1989) have not been adequately tested. Testing their predictions of seed rain patterns
into gaps, gap edges, and forest understory should be fairly straightforward and should
identify certain plant species and their dispersers suitable for more detailed study that will
yield insight into plant community dynamics.
Tests of the escape hypothesis have shown that some mortality agents respond to
distance and/or density and other do not (Howe et al. 1985, Merg 1994). Now we need
studies that go beyond the distance factor and examine the characteristics of sites in which
seeds actually land, and the subsequent fate of those seeds. Experimental studies are useful
in determining causation, but they need to be based on some realistic initial pattern. Even
more important is the need for studies that examine the sequential stages of recruitment,
from dispersal through establishment (e.g., Herrera et al. 1994). Studies that ignore the
possibility of secondary dispersal are likely to lead to misleading conclusions about the
importance of dispersal patterns (Levey and Byrne 1993). Thus, another useful study
would be to examine the influence of areas of high seed rain, such as habitual perches,
leks, primate sleeping sites, large-mammal latrines, gap edges, or nurse plants, on plant
distributions and community composition. Are dense concentrations of seeds away from
the parent plants foci of recruitment for seedlings, seed predation, or secondary dispersal?
In addition to perches in successional landscapes, recruitment foci may be present within
forested areas as well. Emergent trees, standing dead trees, or perhaps trees with abundant
fruit crops (e.g., Ficus) may attract vertebrate seed-dispersers and have higher seed rain
than beneath trees nearby (Smith 1975, Coates-Estrada and Estrada 1986, Masaki et al.
1994). In a preliminary test, Fleming (1988) found no difference in the diversity or density
of vertebrate-dispersed plants around 5 Ficus spp. trees and 5 wind-dispersed

136
Calycophyllum candidissimum trees, but the scale (1600m2 plots), small sample size and
the different sizes of the trees are confounding factors.
Venable and Brown (1993) suggested that directed dispersal involves being
deposited in dung which acts as fertilizer, but high rates of post-dispersal seed removal
from defecations cast doubt on this idea (Janzen 1986, Chapman 1989a, Levey and Byrne
1993, Fragoso 1997). Nevertheless, Dinerstein and Wemmer (1988) and Dinerstein
(1991) showed that the distribution of Trema nudiflora in riparian forests in Nepal is
almost entirely attributable to dispersal by rhinoceros, and suggested that rhinoceros dung
piles provided nutrients necessary for the seedlings. The importance of dung for seeds
buried by dung beetles has not been examined and should be amenable to experimentation.
One of the difficulties in evaluating directed dispersal is that suitable locations for
establishment (i.e., safe sites) are often difficult to specify or are poorly known for most
species (Grubb 1977, Fowler 1988). In addition, safe sites for seeds may not be the same
as those for seedlings or adults (Smith 1984, Lamont et al. 1993, Schupp 1995), leading
some researchers to state that safe sites become more specific over time (Oswald and
Neuenschwander 1993). For example, sites occupied by saplings may a specific subset of
those occupied by seedlings, which in turn are a subset of those occupied by seeds
(Oswald and Neuenschwander 1993). For tropical forest species, the hypothesized
partitioning of gaps among seedlings in terms of gap age and position within the gap
suggests that recruitment conditions could be defined (Hartshorn 1978, Denslow 1987,
Brandani et al. 1988, Popma et al. 1988, Brown and Whitmore 1992). If so, the next step
would be to examine seed arrival to such sites (e.g., Schupp et al. 1989).
Another theme illustrated by the examples above is that directed dispersal may often
be a fortuitous outcome of disperser behavior rather than an adaptive strategy by plants.
Thus, in situations where disperser behavior can be predicted, dispersal can be manipulated
to increase the probability of directed dispersal. Regeneration of degraded land is often
limited by low seed input (Aide and Cavelier 1994, Duncan 1997). Accordingly,

137
manipulation of directed dispersal has potential to play a large role in restoration ecology
(De Pietri 1992, McClanahan and Wolfe 1993, Robinson and Handel 1993).
The implication of directed dispersal is a disproportionate effect on plant
recruitment. Therefore, the loss of a disperser species providing directed dispersal may
result in a marked decrease in fitness for that plant species. A corollary of this idea is that
in depauperate communities that have lost dispersers providing colonization, plant species
benefiting from directed dispersal will become more common. Similarly, directed dispersal
may play a role in facilitating encroachment of invasive plant species (Stiles 1982, Meyer
and Florence 1996, Hutchinson and Vankat 1997, Malo and Suarez 1997). For highly
invasive species with high seed/seedling survival due to lack of seed predators, pathogens,
or herbivores, the initial premise of the colonization hypothesis (i.e., rare and unpredictable
safe sites) may not be met In addition, the roles of introduced animals as seed dispersers
have been virtually ignored (Meyer and Florence 1996).
Abundant evidence hints at directed dispersal in both natural and disturbed
ecosystems. Examining movement and dispersal patterns of individual disperser species,
in combination with seed and seedling survival in different patch types into which seeds are
dispersed, should yield insight into the role of plant-animal interactions in ecosystem
function. When data are available to test the directed dispersal hypothesis, some of the
examples above may not be supported. Demonstrating directed dispersal, however, is not
the most important goal. By conducting the necessary studies we will learn a great deal
about the importance of dispersal patterns in community dynamics.

APPENDIX
OCOTEA ENDRESIANA LOCATION DATA
This appendix contains the locations of all the dispersed (type D) and non-dispersed (N)
Ocotea endresiana seeds found in 1993 and 1995. Tree numbers correspond with those in
Figure 2-1. Compass directions are given from the tree toward the site and distances are
from the trunk of the closest fruiting conspecific. This data set represents an estimated
0.7% of the total seed crop (approximately 100,000) produced in the study site in both
years combined.
year
tree
type
compass
distance
93
1
D
285
14.4
93
1
D
293
13.0
93
1
D
299
10.8
93
1
D
307
10.4
93
1
D
262
16.0
93
1
D
280
14.8
93
1
D
248
11.8
93
1
D
219
10.9
93
1
D
209
12.6
93
1
D
229
18.4
93
1
D
217
9.8
93
1
U
135
5.3
93
1
U
139
5.6
93
1
U
103
6.9
93
1
U
182
6.4
93
1
U
186
6
93
1
U
12
1.7
93
1
U
95
7.6
93
1
U
279
8
93
1
U
276
11.35
93
1
U
283
7.95
93
1
U
115
2.7
93
1
U
149
1.3
95
1
D
192
10.5
95
1
D
300
11
95
1
D
293
17.1
95
1
D
177
16.4
95
1
D
101
9.1
95
1
D
202
16.6
95
1
D
35
138

95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
139
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
D
173
10.9
D
4
21.2
D
24
15
U
316
9
U
320
4.3
U
253
5.15
U
296
2.95
u
206
2.25
u
214
7.3
u
308
5.6
u
279
5.7
u
272
6.2
u
245
7
u
296
9.1
u
282
8.4
u
10
3
u
268
4.15
u
318
2.2
u
279
7.15
u
233
5.55
u
152
2
u
308
2.05
u
210
3.7
u
118
2.6
u
95
4.9
u
187
6.3
u
46
4.5
u
112
2.2
u
104
2.2
u
328
5.05
u
318
2.4
u
5.3
u
10.5
D
229
8.7
D
23
10.1
D
232
12.3
D
237
12.3
D
241
14.6
D
230
14.1
D
216
16.4
D
249
14.2
D
225
16.6
D
213
12.8
D
306
12.9
D
322
11.3
D
191
14.2
D
198
12.7
D
194
12.3
D
189
12.0
D
178
11.3
D
172
11.1
D
192
12.6
U
266
6.3
U
215
3

93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
WOJt>JUJWU>WU>U>U>U>OJU)U)U>U>N>tON>tOtOtOtOtOtOtOtON>K>tOtON>tOMtOION>tON>tON>tOtOtON>tOtON>N)tOtOlON>tO
140
U
U
u
u
u
u
u
u
u
u
u
u
u
u
u
u
u
u
u
u
D
D
D
D
D
D
D
U
u
u
u
u
u
u
u
u
u
u
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
223
2.7
190
4.7
210
4.7
135
5.3
130
5.1
140
7.9
151
4.5
107
2.3
38
6.2
106
2.8
346
0.6
24
6.6
24
6.65
330
1.75
284
1.15
196
4.6
113
3.4
250
4.3
250
4.37
250
4.37
125
13.2
134
15
136
15.8
134
12.7
239
19.5
237
17.7
34
39
4
43
6
141
4.15
190
6
35
3.65
55
3.25
211
3.85
214
3.75
224
2.4
202
3.9
31
6.3
192
10.5
200
12.1
308
14.3
199
16.0
273
15.7
278
17.6
307
8.3
331
6.9
340
6.2
275
12.8
275
12.9
286
10.2
338
3.2
147
10.4
151
10.4
127
8.9

93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
95
95
95
95
95
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
141
D
129
8.5
D
156
10.3
D
129
8.9
D
143
10.0
D
264
8.4
D
284
6.6
D
308
4.9
D
341
3.0
D
343
3.1
D
356
3.8
D
283
2.2
D
272
5.3
D
126
8.5
U
83
4.2
u
109
3.25
u
177
2.8
u
248
5.6
u
161
1.6
u
215
2.2
u
113
3.4
u
119
4.7
u
113
5
u
113
5
u
147
2.85
u
145
2.15
u
139
1.8
u
220
1.7
u
240
1.3
u
170
2.95
u
145
5.2
u
115
5.1
u
175
4.7
D
280
12.7
D
9
4.4
D
244
11.25
D
248
11.8
D
249
12.15
D
18
7.0
D
8
5.7
D
11
12.4
D
342
13.2
D
320
5.6
D
269
5.9
D
291
7.4
D
135
14.4
D
75
11.3
D
116
14.2
D
115
15.1
D
151
10.6
D
149
10.8
D
160
9.4
U
161
5.4
U
158
5.15
U
183
3.7

93
93
93
93
93
93
93
93
95
95
95
95
95
95
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
95
95
95
95
95
95
95
95
142
U
U
U
u
u
u
u
u
D
D
D
U
u
u
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
U
u
u
u
u
u
u
u
u
u
u
u
u
u
u
D
D
D
D
D
D
D
D
195
5.3
139
7.1
171
5.5
171
5.4
165
5.75
150
6.1
118
4.3
157
3.9
42
6
46
5.75
36
157
1.9
297
3.55
303
2.3
203
10.4
214
10.5
95
11.3
110
10.9
260
11.5
267
9.1
110
19.7
123
23.8
198
19.9
223
18.2
239
18.3
257
18.2
152
11.3
140
14.5
196
17.7
192
10.7
209
17.7
186
5.05
198
4.1
94
5.4
127
6.9
131
6.6
142
5.5
202
3.2
280
5.9
180
4.05
240
3.4
241
2.6
329
4.6
323
4.3
359
3.1
83
2.3
62
10
200
8.4
187
9
84
9
82
9
72
9
150
8.6
163
7.3

95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
95
95
95
95
95
95
Os0sOs0\0sO\O\O\O\0\O\O\OsO\O\O\O\O\O\0\0sO\0\OsO\O\O\O\<7\(J\lJ\U\U\U\U\lJ\U\L/\U\V\L/\U\U\U\lJ\U\U\Lr\U\U\U\UiU\V\
143
u
301
3.8
u
207
4.75
u
163
6.9
u
242
2.5
u
270
5.85
u
290
4.5
u
285
2
u
295
2.1
u
301
2.95
u
271
4.7
u
267
3.4
u
294
3.3
u
272
2.65
u
333
2.5
u
81
5
u
104
3.7
u
56
4.1
u
200
5.5
u
229
3.5
u
226
1.55
u
106
1.65
u
93
5
u
40
2.8
u
175
4.6
D
270
6.3
D
274
5.3
D
275
5.2
D
283
5.0
D
245
6.7
D
259
7.1
D
269
9.7
D
279
8.2
D
292
10.6
D
293
11.7
D
292
11.5
D
294
14.2
D
307
14.0
D
291
13.2
D
301
11.3
D
303
10.8
U
144
1.9
U
174
4.4
U
242
3.2
u
276
3.85
u
284
3.6
u
283
3.3
u
284
2.9
u
170
4.1
D
148
10.8
D
134
11.15
D
129
12.1
D
131
9.7
D
134
25.75
D
140
18.8

95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
''D Os Os Os Os Os Os Os Os Os Os Os 0\ Os Os Os Os Os Os OsOsOsOsOsOsOsOsOsOsOsOsOsOsOsOsOsOOOsOOOOOsOsOsOsOsOsOsOsOsOsOsOs
144
D
140
12.4
D
134
10.3
D
117
9.8
D
139
15
D
135
31.25
D
131
10.55
D
131
12.5
D
142
14.3
D
135
17.8
D
133
19.65
D
141
25.1
D
172
10.1
D
72
D
72
D
197
28.7
D
70
D
131
13.7
D
123
13.75
D
40
10.3
D
127
14.9
D
141
10.95
D
142
19.5
D
134
13.1
D
135
28.3
D
167
8.95
D
163
10
D
210
10.1
D
207
10.2
D
243
10
D
122
27.9
D
131
11.15
D
300
18
D
135
26.1
D
149
13.6
D
217
7.7
D
238
8.8
D
180
11.05
D
38
U
30
2
U
124
9.75
U
117
9.25
U
101
6.2
U
212
1.5
U
214
2.7
U
248
6
U
111
8.3
U
130
6
U
130
4.3
u
281
4.15
u
186
3.75
u
186
1.25
D
52
D
42
D
38

95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
93
93
93
93
93
93
93
93
145
8
D
12
10
8
D
340
12.5
8
U
54
4.8
8
U
0
4
8
U
40
2.5
8
U
56
7.5
8
U
12
8
8
U
30
9
8
U
23
9
9
D
206
25
9
D
203
23.7
9
D
198
25.6
9
D
170
10.3
9
D
57.6
9
D
200
2.95
9
D
171
16.2
9
D
168
19.2
9
D
126
12.4
9
D
192
18
9
D
220
37.6
9
D
137
12.7
9
D
182
18.1
9
D
175
23.9
9
D
95
17.8
9
D
202
22.4
9
D
110
15
9
D
26
9
U
243
1.65
9
U
155
4.8
9
u
315
4.2
10
u
331
7.85
10
u
335
11.1
10
u
311
6
10
u
340
9.95
10
u
0
6
10
u
152
11.05
10
u
313
6.6
10
u
226
9.5
10
u
121
5
11
u
142
11.2
11
u
150
9.7
11
u
135
11.5
11
u
118
7.5
11
u
99
4
11
u
95
3.4
11
u
115
10.4
12
D
181
16.5
12
D
179
21.0
12
D
172
21.6
12
D
192
23.2
12
D
172
28.4
12
D
165
27.9
12
D
101
27.3
12
D
97
25.6

93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
146
12
D
103
25.6
12
D
74
23.5
12
D
56
25.6
12
D
53
24.9
12
D
56
22.7
12
D
113
48.9
12
D
113
48.8
12
D
115
46.5
12
D
115
46.9
12
D
115
49.5
12
D
134
51.5
12
D
97
46.5
12
D
94
46.0
12
D
54
25.6
12
D
123
44.0
12
U
168
5.6
12
U
129
12.1
12
U
182
2.4
12
u
243
3.2
12
u
347
2.65
12
u
0
12.2
12
u
90
11.75
12
u
62
16
12
u
21
2.45
12
u
140
6.9
12
u
150
8.25
12
u
147
13.45
12
D
30
12
D
30
12
D
133
27.6
12
D
35
12
D
35
12
D
35
12
D
38
12
D
62
12
D
62
12
D
50
12
D
65
12
D
117
26.2
12
D
95
23.6
12
D
101
28.75
12
D
65
12
D
65
12
D
60
12
D
45
12
D
45
12
D
45
12
D
50
12
D
48
12
D
48
12
D
48
12
D
38
12
D
140
16
12
D
178
21.2

95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
147
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
65
12
D
90
30.2
12
D
62
12
D
62
12
D
62
12
D
62
12
D
59
16
12
D
60
12
D
60
12
D
50
12
D
68
12
D
68
12
D
68
12
U
90
9.4
12
U
111
7.45
12
U
131
10.9
12
U
196
2.3
12
U
58
1.45
18
D
135
8.35
18
D
118
8.95
18
D
130
8.6
18
D
224
4.4
18
D
228
6.1
18
D
112
10.15
18
D
182
10.5
18
U
143
5.6
18
U
61
4.2
18
U
72
4.85
18
U
165
3.4
18
U
174
2.9
18
u
198
4.1
18
u
238
2.8
18
u
130
6.3
18
u
324
4.3
18
u
12
6.5
18
u
68
3.6
18
u
340
1.9
18
u
123
9.05
18
u
140
6.2
18
u
164
2.2
15
D
41
6.6

95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
95
95
95
95
95
148
15
D
15
14.15
15
D
39
14.2
15
D
42
13.75
15
D
45
13.1
15
D
51
16.15
15
D
15
15
D
20
15
D
50
14.65
15
D
35
16
15
D
42
15
D
4
24.2
15
D
25
15
D
44
15
D
40
15
D
28
13.9
15
D
31
15
15
D
1
23.15
15
D
359
22.1
15
D
43
15
D
28
15
D
2
22.4
15
D
25
16
D
35
10.4
16
D
38
11.6
16
D
28
10.2
16
D
36
11.9
16
D
36
12.9
16
D
116
12.8
16
D
107
13.1
16
D
52
13.0
16
D
120
16.2
16
D
296
12.0
16
U
3
2.8
16
U
12
3.35
16
U
5
6.2
16
u
37
6.95
16
u
68
5.8
16
u
84
2.15
16
u
150
4.4
16
u
164
3.3
16
u
264
4.85
16
u
121
0.9
16
u
264
1.1
16
u
0
1.1
16
u
153
4.8
16
u
69
5.1
16
u
65
7.3
16
u
233
1.9
16
u
228
2.15
16
u
90
6.1
16
u
340
1.8
16
u
70
7.5
16
u
124
5
16
u
245
1.7

95
95
95
95
95
95
95
95
95
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
93
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
93
93
149
16
U
255
34.5
17
D
206
19
17
D
219
14.8
17
D
230
13.3
17
D
214
13.4
17
U
263
3.2
17
U
265
5.9
17
U
196
14
17
U
201
12.6
18
D
207
15.8
18
D
207
15.8
18
D
177
9.2
18
D
141
19.4
18
D
182
8.7
18
D
27
12.9
18
D
27
14.5
18
D
332
14.9
18
D
67
15.0
18
D
321
16.5
18
D
90
17.1
18
D
207
15.8
18
U
142
10
18
U
142
10
18
U
169
12.1
18
U
135
11.2
18
u
142
12.3
18
u
149
17
18
u
84
11.2
18
u
18
u
87
9.6
18
u
87
8.5
18
u
95
10.5
18
u
90
12.2
18
u
99
12.5
18
u
99
12
18
D
172
20
18
D
149
15
18
D
167
18
18
D
68
7
18
D
165
19.6
18
D
353
6.5
18
D
103
12
18
U
142
8
18
U
126
2
18
U
0
4
18
U
104
8
18
U
58
7
19
D
114
12.7
19
U
202
7.1
19
u
164
8.4
19
u
245
2.7
19
u
196
3
20
D
321
10.8
20
D
304
11.0

93
93
93
93
93
93
93
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
95
150
20
D
270
8.9
20
D
35
14.5
20
U
59
4.7
20
U
28
5.3
20
U
23
4.1
20
U
272
1.6
20
u
278
1.52
20
D
234
16.2
20
D
25
20
D
34
20
U
173
2.2
20
u
248
11
20
u
350
5.5
20
u
336
3.8
20
u
328
1.2
20
u
217
1.14
20
u
40
2.2
17
D
170
6.1
17
U
105
5.3
17
D
139
19
17
u
179
6.1
24
D
45
24
D
30
24
U
154
7.2
24
u
100
7.3
24
u
146
10
27
D
159
7.8
27
D
143
15.6
11
D
230
26.5
38
D
336
8.6
38
U
282
5.05
38
D
54
13.5
38
D
54
21.8
38
D
89
19.85
38
D
42
15.4
38
D
89
18.4
38
D
17
1.75
38
D
49
19.3
38
D
49
18.15
38
D
357
2.8
38
D
90
6.7
38
D
122
16.9
38
D
122
19.25
38
D
8
14.15
38
U
190
3.2
40
D
83
15.6
40
D
89
15.6
40
D
105
14.5
40
D
83
15.5
40
D
78
11.1
40
D
249
13.4
40
D
250
15
40
D
55
21.4
40
D
89
10.1

95
95
95
95
95
95
95
95
95
95
95
151
40
D
73
18.6
40
D
60
16.1
40
D
57
16.2
40
D
59
19.25
40
D
67
17.5
40
D
57
15.4
40
U
102
5.6
40
U
289
5.1
40
U
285
6.5
40
U
273
5.8
8
D
124
27.5

REFERENCES
Abacus Concepts. 1989. SuperANOVA: accessible general linear modeling. Abacus
Concepts, Inc., Berkeley, California.
Aide, T. M. and J. Cavelier. 1994. Barriers to lowland tropical forest restoration in the
Sierra Nevada de Santa Marta, Colombia. Restoration Ecology 2:219-229.
Alvarez-Buylla, E. R., A. Chaos, D. Piñero, and A. A. Garay. 1996. Demographic
genetics of a pioneer tropical tree species: patch dynamics, seed dispersal, and seed
banks. Evolution 50:1155-1166.
Andersen, M. 1991. Mechanistic models for the seed shadows of wind-dispersed plants.
American Naturalist 137:476-497.
Anderson, S. D. 1982. Comparative population ecology of Peromyscus mexicanus in a
Costa Rican wet forest. Dissertation. University of Southern California, Los
Angeles, California.
Andresen, E. in press. Biotropica.
Augspurger, C. K. 1983. Seed dispersal of the tropical tree, Platypodium elegans, and the
escape of its seedlings from fungal pathogens. Journal of Ecology 71:759-771.
Augspurger, C. K. 1984. Seedling survival of tropical tree species: interactions of
dispersal distance, light-gaps, and pathogens. Ecology 65:1705-1712.
Augspurger, C. K. 1990. The potential impact of fungal pathogens on tropical plant
reproductive biology. Pages 237-245 in K. S. Bawa and M. Hadley, editors.
Reproductive ecology of tropical forest plants. UNESCO/Parthenon Publishing
Group, Paris/Camforth/Park Ridge.
Augspurger, C. K. and S. E. Franson. 1988. Input of wind-dispersed seeds into light-
gaps and forest sites in a Neotropical forest. Journal of Tropical Ecology 4:239-
252.
Augspurger, C. K. and K. Kitajima. 1992. Experimental studies of seedling recruitment
from contrasting seed distributions. Ecology 73:1270-1284.
August, P. V. 1981. Fig fruit consumption and seed dispersal by Artibeus jamaicensis in
the Llanos of Venezuela. Biotropica 13S:70-76.
Avila H., M. L., V. H. Hernández O., and E. Velarde. 1996. The diet of resplendent
quetzal (Pharomachrus mocinno mocinno: Trogonidae) in a Mexican cloud forest
Biotropica 28:720-727.
152

153
Balda, R. P. and A. C. Kamil. 1989. A comparative study of cache recovery by three
corvid species. Animal Behavior 38:486-495.
Barnes, P. W. and S. Archer. 1996. Influence of an overstorey tree (Prosopis glandulosd)
on associated shrubs in a savanna parkland: implications for patch dynamics.
Oecologia 105:493-500.
Bazzaz, F. A. 1991. Habitat selection in plants. American Naturalist 137:S116-S130.
Bazzaz, F. A. and S. T. A. Pickett. 1980. Physiological ecology of tropical succession: a
comparative review. Annual Review of Ecology and Systematics 11:287-310.
Beal, W. 1898. Seed dispersal. Ginn & Co., Boston, Massachusetts.
Beattie, A. J. 1985. The evolutionary ecology of ant-plant mutualisms. Cambridge
University Press, Cambridge.
Beattie, A. J. and D. C. Culver. 1981. The guild of myrmecochores in the herbaceous flora
of West Virginia forests. Ecology 62:107-115.
Beattie, A. J. and D. C. Culver. 1982. Inhumation: how ants and other invertebrates help
seeds. Nature 297:627.
Beattie, A. J. and D. C. Culver. 1983. The nest chemistry of two seed-dispersing ant
species. Oecologia 56:99-103.
Becker, P., L. W. Lee, E. D. Rothman, and W. D. Hamilton. 1985. Seed predation and
the coexistence of tree species: Hubbell's models revisited. Oikos 44:382-390.
Benkman, C. W. 1995. Wind dispersal capacity of pine seeds and the evolution of
different seed dispersal modes in pines. Oikos 73:221-224.
Bodmer, R. E. 1990. Fruit patch size and frugivory in the lowland tapir (Tapirus
terrestris). Journal of Zoology 222:121-128.
Bodmer, R. E. 1991. Strategies of seed dispersal and seed predation in Amazonian
ungulates. Biotropica 23:255-261.
Bond, W. J. and W. D. Stock. 1989. The costs of leaving home: ants disperse
myrmecochorous seeds to low nutrient sites. Oecologia 81:412-417.
Bossema, 1.1979. Jays and oaks: an ecoethological study of a symbiosis. Behaviour 70:1-
118.
Bowers, M. A. and J. L. Dooley. 1993. Predation hazard and seed removal by small
mammals: microhabitat versus patch scale effects. Oecologia 94:247-254.
Boyd, R. S. 1996. Ant-mediated seed dispersal of the rare chaparral shrub
Fremontodendron decumbens (Sterculiaceae). Madroño 43:299-315.
Brandani, A., G. S. Hartshorn, and G. H. Orians. 1988. Internal heterogeneity of gaps
and species richness in Costa Rican tropical wet forest. Journal of Tropical Ecology
4:99-119.

154
Brokaw, N. V. L. 1982a. The definition of treefall gap and its effect on measures of forest
dynamics. Biotropica 14:158-160.
Brokaw, N. V. L. 1982b. Treefalls: frequency, timing, and consequences. Pages 101-108
in E. G. Leigh, Jr., A. S. Rand, and D. M. Windsor, editors. The ecology of a
tropical forest: seasonal rhythms and long-term changes. Smithsonian Institution
Press, Washington, D.C.
Brokaw, N. V. L. 1985. Gap-phase regeneration in a tropical forest. Ecology 66:682-687.
Bronstein, J. L. and K. Hoffman. 1987. Spatial and temporal variation in frugivory at a
Neotropical fig, Ficus pertusa. Oikos 49:261-268.
Brown, N. D. and T. C. Whitmore. 1992. Do dipterocarp seedlings really partition tropical
rain forest gaps? Philosophical Transactions of the Royal Society, series B
335:369-378.
Burger, W. and H. van der Werff. 1990. Lauraceae. Flora Costaricensis 80. Fieldiana
Botany (New Series) 23:1-121.
Burkey, T. V. 1994. Tropical tree species diversity: a test of the Janzen-Connell model.
Oecologia 97:533-540.
Bustamante, R. O. and M. Canals L. 1995. Dispersal quality in plants: how to measure
efficiency and effectiveness of a seed disperser. Oikos 73:133-136.
Byrne, M. M. and D. J. Levey. 1993. Removal of seeds from frugivore defecations by
ants in a Costa Rican rain forest. Vegetatio 107/108:363-374.
Callaway, R. M. 1992. Effect of shrubs on recruitment of Quercus douglasii and Quercus
lobata in California. Ecology 73:2118-2128.
Callaway, R. M„ E. H. DeLucia, D. Moore, R. Nowak, and W. H. Schlesinger. 1996.
Competition and facilitation: contrasting effects of Artemisia tridentata on desert vs
montane pines. Ecology 77:2130-2141.
Canham, C. D. 1989. Different responses to gaps among shade-tolerant trees. Ecology
70:548-550.
Cardel, Y., V. Rico-Gray, J. G. García-Franco, and L. B. Thien. 1997. Ecological status
of Beaucamea gracilis, an endemic species of the semiarid Tehuacán Valley,
México. Conservation Biology 11:367-374.
Castro, O. C. 1993. Chemical and biological extractives of Lauraceae species in Costa
Rican tropical forests. Recent Advances in Phytochemistry 27:65-87.
Cavelier, J. 1996. Environmental factors and ecophysiological processes along altitudinal
gradients in wet tropical mountains. Pages 399-439 in S. S. Mulkey, R. L.
Chazdon, and A. P. Smith, editors. Tropical forest plant ecophysiology. Chapman
and Hall, New York, New York.
Cavelier, J. and G. Goldstein. 1989. Mist and fog interception in elfin cloud forests in
Colombia and Venezuela. Journal of Tropical Ecology 5:309-322.

155
Cavers, P. B. 1983. Seed demography. Canadian Journal of Botany 61:3578-3590.
Chapman, C. A. 1989a. Primate seed dispersal: the fate of dispersed seeds. Biotropica
21:148-154.
Chapman, C. A. 1989b. Spider monkey sleeping sites: use and availability. American
Journal of Primatology 18:53-60.
Chapman, C. A. and L. J. Chapman. 1995. Survival without dispersers: seedling
recruitment under parents. Conservation Biology 9:675-678.
Chapman, C. A. and L. J. Chapman. 1996. Frugivory and the fate of dispersed and non-
dispersed seeds of six African tree species. Journal of Tropical Ecology 12.
Chapman, L. J., C. A. Chapman, and R. W. Wrangham. 1992. Balanites wilsoniana:
elephant dependent dispersal? Journal of Tropical Ecology 8:275-283.
Charles-Dominique, P. 1986. Inter-relations between frugivorous vertebrates and pioneer
plants: Cecropia, birds and bats in French Guyana. Pages 119-135 in A. Estrada
and T. H. Fleming, editors. Frugivores and seed dispersal. Dr. W. Junk
Publishers, Dordrecht, Netherlands.
Chavez-Ramirez, F. and R. D. Slack. 1994. Effects of avian foraging and post-foraging
behavior on seed dispersal patterns of Ashe juniper. Oikos 71:40-46.
Chazdon, R. L., R. W. Pearcy, D. W. Lee, and N. Fetcher. 1996. Photosynthetic
responses of tropical forest plants to contrasting light environments. Pages 5-55 in
S. S. Mulkey, R. L. Chazdon, and A. P. Smith, editors. Tropical forest plant
ecophysiology. Chapman and Hall, New York, New York.
Cintra, R. 1997a. Leaf litter effects on seed and seedling predation of the palm
Astrocaryum murumuru and the legume tree Dipteryx micrantha in Amazonian
forest. Journal of Tropical Ecology 13:709-725.
Cintra, R. 1997b. A test of the Janzen-Connell model with two common tree species in
Amazonian forest. Journal of Tropical Ecology 13:641-658.
Cintra, R. and V. Homa. 1997. Seed and seedling survival of the palm Astrocaryum
murumuru and the legume tree Dipteryx micrantha in gaps in Amazonian forest
Journal of Tropical Ecology 13:257-277.
Clark, D. A. 1994. Plant demography. Pages 90-105 in L. A. McDade, K. S. Bawa, H.
A. Hespenheide, and G. S. Hartshorn, editors. La Selva: ecology and natural
history of a neotropical rainforest. University of Chicago Press, Chicago, Illinois.
Clark, D. A. and D. B. Clark. 1984. Spacing dynamics of a tropical tree: evaluation of the
Janzen-Connell model. American Naturalist 124:769-788.
Clark, D. B. and D. A. Clark 1991. The impact of physical damage on canopy tree
regeneration in tropical rain forest. Journal of Ecology 79:447-457.
Clark, J. S., C. Fastie, G. Hurtt, S. T. Jackson, C. Johnson, G. A. King, M. Lewis, J.
Lynch, S. Pacala, C. Prentice, E. W. Schupp, T. Webb HI, and P. Wycoff. 1998.

156
Reid's paradox of rapid plant migration: dispersal theory and interpretation of
paleoecological records. BioScience 48:13-24.
Clifford, H. T. and G. B. Monteith. 1989. A three phase seed dispersal mechanism in
Australian quinine bush (Peltostigma pubescens Domin). Biotropica 21:284-286.
Coates-Estrada, R. and A. Estrada. 1986. Fruiting and frugivores at a strangler fig in the
tropical rain forest of Los Tuxtlas, Mexico. Joumad of Tropical Ecology 2:349-357.
Compton, S. G., A. J. F. K. Craig, and I. W. R. Waters. 1996. Seed dispersal in an
African fig tree: birds as high quantity, low quality dispersers? Journal of
Biogeography 23:553-563.
Condit, R., S. P. Hubbell, and R. B. Foster. 1992. Recruitment near conspecific adults
and the maintenance of tree and shrub diversity in a neotropical forest American
Naturalist 140:261-286.
Connell, J. H. 1970. On the role of natural enemies in preventing competitive exclusion in
some marine animals and in rain forest trees. Pages 298-312 in P. J. den Boer and
G. R. Gradwell, editors. Dynamics of Populations. PUDOC, Wageningen, The
Netherlands.
Connell, J. H„ M. D. Lowman, and I. R. Noble. 1997. Subcanopy gaps in temperate and
tropical forests. Australian Journal of Ecology 22:163-168.
Cork, S. J. and G. J. Kenagy. 1989. Rates of gut passage and retention of hypogeous
fungal spores in two forest-dwelling rodents. Journal of Mammalogy 70:512-519.
Corlett, R. T. and P. W. Lucas. 1990. Alternative seed-handling strategies in primates:
seed-spitting by long-tailed macaques (Macaca fascicularis). Oecologia 82:166-
171.
Crawley, M. J. 1992. Seed predators and plant population dynamics. Pages 157-191 in M.
Fenner, editor. Seeds: the Ecology of Regeneration in Plant Communities. CAB
International, Wallingford, UK.
Crome, F. H. J. 1975. The ecology of fruit pigeons in tropical northern Queensland.
Australian Wildlife Research 2:155-185.
Cruz, A. 1981. Bird activity and seed dispersal of a montane forest tree (Dunalia
arborescens) in Jamaica. Biotropica 12S:34-44.
Culver, D. C. and A. J. Beattie. 1978. Myrmecochory in Viola: dynamics of seed-ant
interactions in some West Virginia species. Journal of Ecology 66:53-72.
Da Silva, J. M. C., C. Uhl, and G. Murray. 1996. Plant succession, landscape
management, and the ecology of frugivorous birds in abandoned Amazonian
pastures. Conservation Biology 10:491-503.
Darley-Hill, S. and W. C. Johnson. 1981. Acorn dispersal by the blue jay (Cyanocitta
cristata). Oecologia 50:231-232.
Darwin, C. 1859. The origin of species by means of natural selection. Doubleday, New
York.

157
Davidar, P. 1983. Birds and neotropical mistletoes: effects on seedling recruitment
Oecologia 60:271-273.
Davidson, D. W. and S. R. Morton. 1981. Myrmecochory in some plants (F.
chenopodiaceae) of the Australian arid zone. Oecologia 50:357-366.
De Pietri, D. E. 1992. Alien shrubs in a national park: can they help in the recovery of
natural degraded forest? Biological Conservation 62:127-130.
Dean, W. R. J. and S. J. Milton. 1988. Dispersal of seeds by raptors. African Journal of
Ecology 26:173-176.
Debussche, M„ J. Escarré, and J. Lepart. 1982. Omithochory and plant succession in
Mediterranean abandoned orchards. Vegetado 48:255-266.
Debussche, M. and P. Isenmann. 1994. Bird-dispersed seed rain and seedling
establishment in patchy Mediterranean vegetadon. Oikos 69:414-426.
Denslow, J. S. 1987. Tropical rainforest gaps and tree species diversity. Annual Review of
Ecology and Systematics 18:431-451.
Denslow, J. S. and A. E. Gomez Diaz. 1990. Seed rain to tree-fall gaps in a Neotropical
rain forest. Canadian Journal of Forest Research 20:642-648.
Denslow, J. S. and G. S. Hartshorn. 1994. Tree-fall gap environments and forest dynamic
processes. Pages 120-127 in L. A. McDade, K. S. Bawa, H. A. Hespenheide, and
G. S. Hartshorn, editors. La Selva: Ecology and natural history of a neotropical
rainforest. University of Chicago Press, Chicago, Illinois.
DeSteven, D. and F. E. Putz. 1984. Impact of mammals on early recruitment of a tropical
tree, Dipteryxpanamensis, in Panama. Oikos 43:207-216.
Dinerstein, E. 1991. Seed dispersal by greater one-homed rhinoceros (Rhinoceros
unicornis) and the flora of Rhinoceros latrines. Mammalia 55:355-362.
Dinerstein, E. and C. M. Wemmer. 1988. Fruits Rhinoceros eat: dispersal of Trewia
nudiflora (Euphorbiaceae) in lowland Nepal. Ecology 69:1768-1774.
Dirzo, R. and C. A. Dominguez. 1986. Seed shadows, seed predation and the advantages
of dispersal. Pages 237-249 in A. Estrada and T. H. Fleming, editors. Frugivores
and seed dispersal. Dr. W. Junk Publishers, Dordrecht, Netherlands.
Docters van Leeuwen, W. M. 1954. On the biology of some Javanese Loranthaceae and
the role birds play in their life-historie. Beauforda 4:103-207.
Duncan, R. S. 1997. Seed dispersal in a degraded agricultural landscape: the first steps
toward reforestation. M.A. Thesis. University of Florida, Gainesville, FL.
Ellner, S. and A. Shmida. 1981. Why are adaptations for long-range seed dispersal rare in
desert plants? Oecologia 51:133-144.
Emmons, L. H. 1990. Neotropical rainforest mammals. University of Chicago Press,
Chicago, Illinois.

158
Endler, J. A. and M. Théry. 1996. Interacting effects of lek placement, display behavior,
ambient light, and color patterns in three neotropical forest-dwelling birds.
American Naturalist 148:421-452.
Englund, R. 1993. Fruit removal in Viburnum opulus: copious seed predation and sporadic
massive seed dispersal in a temperate shrub. Oikos 67:503-510.
Estrada, A. and R. Coates-Estrada. 1991. Howler monkeys (Alouatta palliatd), dung
beetles (Scarabaeidae) and seed dispersal: ecological interactions in the tropical rain
forest of Los Tuxtlas, Mexico. Journal of Tropical Ecology 7:459-474.
Estrada, A., G. Halffter, R. Coates-Estrada, and D. A. Meritt Jr. 1993. Dung beetles
attracted to mammalian herbivore (.Alouatta palliatd) and omnivore (Nasua narica)
dung in the tropical rain forest of Los Tuxtlas, Mexico. Journal of Tropical Ecology
9:45-54.
Fenner, M. 1985. Seed ecology. Chapman and Hall, London, UK.
Fischer, E. A. 1997. The role of plumes in Eriotheca pentaphylla (Bombacaceae) seed
survival in south-eastern Brazil. Journal of Tropical Ecology 13:133-138.
Fleming, T. H. 1988. The short-tailed fruit bat: a study in plant-animal interactions.
University of Chicago Press, Chicago, Illinois.
Fogden, M. 1993. An annotated checklist of the birds of Monteverde and Peñas Blancas.
Green Mountain Publications, Monteverde, Costa Rica.
Forget, P.-M. 1990. Seed-dispersal of Vouacapoua americana (Caesalpiniaceae) by
caviomorph rodents in French Guiana. Journal of Tropical Ecology 6:459-468.
Forget, P.-M. 1991. Scatterhoarding of Astrocaryum paramaca by Proechimys in French
Guiana: comparison with Myoprocta exilis. Tropical Ecology 32:155-167.
Forget, P.-M. 1993. Post-dispersal predation and scatterhoarding of Dipteryx panamensis
(Papilionaceae) seeds by rodents in Panama. Oecologia 94:255-261.
Forget, P.-M. 1996. Removal of seeds of Car apa procera (Meliaceae) by rodents and their
fate in rainforest in French Guiana. Journal of Tropical Ecology 12:751-761.
Forget, P.-M. 1997. Effect of microhabitat on seed fate and seedling performance in two
rodent-dispersed tree species in rain forest in French Guiana. Journal of Ecology
85:693-703.
Forget, P.-M. and T. Milleron. 1991. Evidence for secondary dispersal by rodents in
Panama. Oecologia 87:596-599.
Foster, S. A. and C. H. Janson. 1985. The relationship between seed size and
establishment conditions in tropical woody plants. Ecology 66:773-780.
Fowler, N. L. 1988. What is a safe site? Neighbor, litter, germination date, and patch
effects. Ecology 69:947-961.
Fragoso, J. M. V. 1997. Tapir-generated seed shadows: scale-dependent patchiness in the
Amazon rain forest. Journal of Ecology 85:519-529.

159
Franco-Pizaña, J. G., T. E. Fulbright, D. T. Gardiner, and A. R. Tipton. 1996. Shrub
emergence and seedling growth in microenvironments created by Prosopis
glandulosa. Journal of Vegetation Science 7:257-264.
Fulbright, T. E., J. O. Kuti, and A. R. Tipton. 1995. Effects of nurse-plant canopy
temperatures on shrub seed germination and seedling growth. Acta Oecologia
16:621-632.
Garwood, N. C. 1996. Functional morphology of tropical tree seedlings. Pages 59-129 in
M. D. Swaine, editor. Ecology of tropical forest tree seedlings.
UNESCO/Parthenon, Paris/Camforth.
Gentry, A. H. 1982. Patterns of neotropical plant species diversity. Evolutionary Biology
15:1-84.
Gentry, A. H. 1990. Floristic similarities and differences between southern Central
America and upper and central Amazonia. Pages 141-159 in A. H. Gentry, editor.
Four neotropical forests. Yale University Press, New Haven, Connecticut.
Gentry, A H. 1993. A field guide to the families and genera of woody plants of northwest
South America (Colombia, Ecuador, Peru). Conservation International,
Washington, D.C.
Geritz, S. A. H., T. J. de Jong, and P. G. L. Klinkhamer. 1984. The efficacy of dispersal
in relation to safe site area and seed production. Oecologia 62:219-221.
Gibson, J. P. and N. T. Wheelwright. 1995. Genetic structure in a population of a tropical
tree Ocotea teñera (Lauraceae): influence of avian seed dispersal. Oecologia
103:49-54.
Gibson, W. 1993. Selective advantages to hemi-parasitic annuals, genus Melampyrum, of
a seed-dispersal mutualism involving ants: n. Seed-predator avoidance. Oikos
67:345-350.
Gladstone, D. E. 1979. Description of a seed-shadow of a wind-dispersed tropical tree.
Brenesia 16:81-86.
Gray, A. N. and T. A. Spies. 1996. Gap size, within-gap position and canopy structure
effects of conifer seedling establishment Journal of Ecology 84:635-645.
Green, D. S. 1983. The efficacy of dispersal in relation to safe site density. Oecologia
56:356-358.
Greene, D. F. and E. A. Johnson. 1989. A model of wind dispersal of winged or plumed
seeds. Ecology 70:339-347.
Greene, D. F. and E. A. Johnson. 1996. Wind dispersal of seeds from a forest into a
clearing. Ecology 77:595-609.
Greene, D. F. and E. A. Johnson. 1997. Secondary dispersal of tree seeds on snow.
Journal of Ecology 85:329-340.
Grubb, P. J. 1977. The maintenance of species-richness in plant communities: the
importance of the regeneration niche. Biological Reviews 52:107-145.

160
Guevara, S. and J. Laborde. 1993. Monitoring seed dispersal at isolated standing trees in
tropical pastures: consequences for local species availability. Vegetado
107/108:319-338.
Guindon, C. F. 1996. The importance of forest fragments to the maintenance of regional
biodiversity in Costa Rica. Pages 168-186 in J. Schelhas and R. Greenberg,
editors. Forest patches in tropical landscapes. Island Press, Washington, D.C.
Guix, J. C. and X. Ruiz. 1997. Weevil larvae dispersal by guans in southeastern Brazil.
Biotropica 29:522-525.
Haber, W. A. 1991. Lista provisional de las plantas de Monteverde, Costa Rica. Brenesia
34:63-120.
Haber, W. A., W. Zuchowski, and E. Bello. 1996. An introduction to cloud forest trees:
Monteverde, Costa Rica. La Nación, San José, Costa Rica.
Hall, G. 1987. Seed dispersal by birds of prey. The Zimbabwe Science News 21:12.
Hallwachs, W. 1986. Agoutis (Dasyprocta punctata): inheritors of guapinol (Hymenaea
courbaril: Leguminosae). Pages 285-304 in A. Estrada and T. H. Fleming, editors.
Frugivores and seed dispersal. Dr. W. Junk Publishers, Dordrecht, The
Netherlands.
Hamilton, W. D. and R. M. May. 1977. Dispersal in stable habitats. Nature 269:578-581.
Hammond, D. S. 1995. Post-dispersal seed and seedling mortality of tropical dry forest
trees after shifting agriculture, Chiapas, Mexico. Journal of Tropical Ecology
11:295-313.
Hammond, D. S. and V. K. Brown. 1995. Seed size of woody plants in relation to
disturbance, dispersal, soil type in wet neotropical forests. Ecology 76:2544-2561.
Hamrick, J. L. and M. D. Loveless. 1986. The influence of seed dispersal mechanisms on
the genetic structure of plant populations. Pages 211-223 in A. Estrada and T. H.
Fleming, editors. Frugivores and seed dispersal. Dr. W. Junk Publishers,
Dordrecht, Netherlands.
Hamrick, J. L., D. A. Murawski, and J. D. Nason. 1993. The influence of seed dispersal
mechanisms on the genetic structure of tropical tree populations. Vegetado
107/108:281-297.
Hamrick, J. L. and J. D. Nason. 1996. Consequences of dispersal in plants. Pages 203-
236 in O. E. Rhodes, R. K. Chesser, and M. H. Smith, editors. Population
dynamics in ecological space and time. University of Chicago Press, Chicago,
Illinois.
Hanzawa, F. M., A. J. Beattie, and D. C. Culver. 1988. Directed dispersal: demographic
analysis of an ant-seed mutualism. American Naturalist 131:1-13.
Harper, J. L. 1977. Population Biology of Plants. Academic Press, London, UK.

161
Hartshorn, G. S. 1978. Tree falls and tropical forest dynamics. Pages 617-638 in P. B.
Tomlinson and M. H. Zimmerman, editors. Tropical trees as living systems.
Cambridge University Press, Cambridge, UK.
Hartshorn, G. S. 1983. Plants. Pages 118-157 in D. H. Janzen, editor. Costa Rican
natural history. University of Chicago Press, Chicago.
Heithaus, E. R. 1981. Seed predation by rodents on three ant-dispersed plants. Ecology
62:136-145.
Herrera, C. M. 1984a. Seed dispersal and fitness determinants in wild rose: Combined
effects of hawthorn, birds, mice, and browsing ungulates. Oecologia 63:386-393.
Herrera, C. M. 1984b. A study of avian frugivores, bird-dispersed plants, and their
interaction in Mediterranean scrublands. Ecological Monographs 54:1-23.
Herrera, C. M. 1985. Determinants of plant-animal coevolution: the case of mutualistic
dispersal of seeds by vertebrates. Oikos 44:132-141.
Herrera, C. M. 1986. Vertebrate-dispersed plants: why they don't behave the way they
should. Pages 5-18 in A. Estrada and T. H. Fleming, editors. Frugivores and seed
dispersal. Dr. W. Junk Publishers, Dordrecht, Netherlands.
Herrera, C. M. and P. Jordano. 1981. Prunus mahaleb and birds: the high-efficiency seed
dispersal system of a temperate fruiting tree. Ecological Monographs 51:203-218.
Herrera, C. M., P. Jordano, L. Lopez-Soria, and J. A. Amat. 1994. Recruitment of a
mast-fruiting, bird-dispersed tree: bridging frugivore activity and seedling
establishment. Ecological Monographs 64:315-344.
Heymann, E. W. 1995. Sleeping habits of tamarins, Saguinus mystax and Saguinus
fuscicollis (Mammalia; Primates; Callitrichidae), in north-east Peru. Journal of
Zoology 237:211-226.
Hofgaard, A. 1993. 50 years of change in a Swedish boreal old-growth Picea abies forest.
Journal of Vegetation Science 4:773-782.
Holl, K. D„ and M. E. Lulow. 1997. Effect of species habitat and distance from edge on
post-dispersal seed predation in a tropical rainforest. Biotropica 29:459-468.
Holm, S. 1979. A simple sequentially rejective multiple test procedure. Scandinavian
Journal of Statistics 6:65-70.
Holthuijzen, A. M. A., T. L. Sharik, and J. D. Fraser. 1987. Dispersal of eastern red
cedar (Juniperus virginiana) into pastures: an overview. Canadian Journal of
Botany 65:1092-1095.
Hoppes, W. G. 1987. Pre- and post-foraging movements of frugivorous birds in an
eastern deciduous forest woodland, USA. Oikos 49:281-290.
Hoppes, W. G. 1988. Seedfall pattern of several species of bird-dispersed plants in an
Illinois woodland. Ecology 69:320-329.

162
Horvitz, C. C. 1981. Analysis of how ant behaviors affect germination in a tropical
myrmecochore Calathea microcephala (P. & E.) Koemicke (Marantaceae): microsite
selection and aril removal by neotropical ants, Odontomachus, Pachycondyla, and
Solenopsis (Formicidae). Oecologia 51:47-52.
Horvitz, C. C. and D. W. Schemske. 1986a. Ant-nest soil and seedling growth in a
neotropical ant-dispersed herb. Oecologia 70:318-320.
Horvitz, C. C. and D. W. Schemske. 1986b. Seed dispersal and environmental
heterogeneity in a neotropical herb: a model of population and patch dynamics.
Pages 169-186 in A. Estrada and T. H. Fleming, editors. Frugivores and seed
dispersal. Dr. W. Junk Publishers, Dordrecht, Netherlands.
Houle, G. 1995. Seed dispersal and seedling recruitment: the missing link(s). Écoscience
2:238-244.
Howe, H. F. 1984. Constraints on the evolution of mutualisms. American Naturalist
123:764-777.
Howe, H. F. 1985. Gomphothere fruits: a critique. American Naturalist 125:853-865.
Howe, H. F. 1986. Seed dispersal by fruit-eating birds and mammals. Pages 123-189 in
D. R. Murray, editor. Seed dispersal. Academic Press, Sydney, Australia.
Howe, H. F. 1989. Scatter- and clump-dispersal and seedling demography: hypothesis and
implications. Oecologia 79:417-426.
Howe, H. F. 1990a. Seed dispersal by birds and mammals: implications for seedling
demography. Pages 191-218 in K. S. Bawa and M. Hadley, editors. Reproductive
ecology of tropical forest plants. UNESCO/Parthenon Publishing Group,
Paris/Camforth/Park Ridge.
Howe, H. F. 1990b. Survival and growth of juvenile Virola surinamensis in Panama:
effects of herbivory and canopy closure. Journal of Tropical Ecology 6:259-280.
Howe, H. F. 1993a. Aspects of variation in a neotropical seed dispersal system. Vegetatio
107/108:149-162.
Howe, H. F. 1993b. Specialized and generalized dispersal systems: where does 'the
paradigm' stand? Vegetatio 107/108:3-13.
Howe, H. F. and D. DeSteven. 1979. Fruit production, migrant bird visitation, and seed
dispersal of Guarea glabra in Panama. Oecologia 39:185-196.
Howe, H. F. and G. F. Estabrook. 1977. On intraspecific competition for avian dispersers
in tropical trees. American Naturalist 111:817-832.
Howe, H. F. and W. M. Richter. 1982. Effects of seed size on seedling size in Virola
surinamensis; a within and between tree analysis. Oecologia 53:347-351.
Howe, H. F., E. W. Schupp, and L. C. Westley. 1985. Early consequences of seed
dispersal for a neotropical tree (Virola surinamensis). Ecology 66:781-791.

163
Howe, H. F. and J. Smallwood. 1982. Ecology of seed dispersal. Annual Review of
Ecology and Systematics 13:201-228.
Howe, H. F. and G. A. Vande Kerckhove. 1980. Nutmeg dispersal by tropical birds.
Science 210:925-926.
Howe, H. F. and G. A. Vande Kerckhove. 1981. Removal of wild nutmeg (Virola
surinamensis) crops by birds. Ecology 62:1093-1106.
Hubbell, S. P. 1980. Seed predation and the coexistence of tree species in tropical forests.
Oikos 35:214-229.
Hulme, P. E. 1993. Post-dispersal seed predation by small mammals. Symposium of the
Zoological Society of London 65:269-287.
Hulme, P. E. 1996. Natural regeneration of yew (Taxus baccata L.): microsite, seed or
herbivore limitation. Journal of Ecology 84:853-861.
Hulme, P. E. 1997. Post-dispersal seed predation and the establishment of vertebrate
dispersed plants in Mediterranean scrublands. Oecologia 111:91-98.
Hunter, J. R. 1989. Seed dispersal and germination of Enterolobium cyclocarpum (Jacq.)
Griseb. Leguminosae: Mimosoideae): are megafauna necessary? Journal of
Biogeography 16:369-378.
Hutchinson, T. F. and J. L. Vankat. 1997. Invasibility and effects of amur honeysuckle in
southwestern Ohio forests. Conservation Biology 11:1117-1124.
Izhaki, L, P. B. Walton, and U. N. Safriel. 1991. Seed shadows generated by frugivorous
birds in an eastern Mediterranean scrub. Journal of Ecology 79:575-590.
Jackson, P. S. W., Q. C. B. Cronk, and J. A. N. Parnell. 1988. Notes on the regeneration
of two rare Mauritian endemic trees. Tropical Ecology 29:98-106.
James, F. C. and C. E. McCulloch. 1990. Multivariate analysis in ecology and
systematics: panacea or pandora's box? Annual Review of Ecology and Systematics
21:129-166.
Janzen, D. H. 1970. Herbivores and the number of tree species in tropical forests.
American Naturalist 104:501-528.
Janzen, D. H. 1980. When is it coevolution? Evolution 34:611-612.
Janzen, D. H. 1981. The defenses of legumes against herbivores. Pages 951-977 in R. M.
Polhill and P. H. Raven, editors. Advances in Legume Systematics. Royal
Botanical Garden, Kew, U.K.
Janzen, D. H. 1982a. Seeds in tapir dung in Santa Rosa National Park, Costa Rica.
Brenesia 19/20:129-135.
Janzen, D. H. 1982b. Simulation of Andira fruit pulp removal by bats reduces seed
predation by Cleogonus weevils. Brenesia 19/20:165-170.

164
Janzen, D. H. 1983a. Dispersal of seeds by vertebrate guts. Pages 232-262 in D. J.
Futuyma and M. Slatkin, editors. Coevolution. Sinauer Associates, Sunderland,
Massachusetts.
Janzen, D. H. 1983b. Food webs: who eats what, why, how, and with what consequences
in a tropical forest? Pages 167-182 in F. B. Golley, editor. Tropical rain forest
ecosystems. Elsevier, Amsterdam.
Janzen, D. H. 1983c. Seed and pollen dispersal by animals: convergence in the ecology of
contamination and sloppy harvest Biological Journal of the Linnean Society
20:103-113.
Janzen, D. H. 1983d. Tapirus bairdii (Danto, Danta, Baird’s Tapir). Pages 496-497 in D.
H. Janzen, editor. Costa Rican Natural History. University of Chicago Press,
Chicago.
Janzen, D. H. 1986. Mice, big mammals, and seeds: it matters who defecates what where.
Pages 251-271 in A. Estrada and T. H. Fleming, editors. Frugivores and seed
dispersal. Dr. W. Junk Publishers, Dordrecht, Netherlands.
Janzen, D. H. and P. S. Martin. 1982. Neotropical anachronisms: the fruits the
gomphotheres ate. Science 215:19-27.
Johnson, C. N. 1996. Interactions between mammals and ectomycorrhizal fungi. Trends in
Ecology and Evolution 11:503-507.
Johnson, W. C., C. S. Adkisson, T. R. Crow, and M. D. Dixon. 1997. Nut caching by
blue jays (Cyanocitta cristata L.): implications for tree demography. American
Midland Naturalist 138:357-370.
Johnson, W. C. and T. Webb. 1989. The role of blue jays (Cyanocitta cristata L.) in the
post-glacial dispersal of fagaceous trees in eastern North America. Journal of
Biogeography 16:561-571.
Jordano, P. 1992. Fruits and frugivory. Pages 105-156 in M. Fenner, editor. Seeds: the
ecology of regeneration in plant communities. CAB International, Wallingford,
UK.
Jordano, P. 1995. Spatial and temporal variation in the avian-frugivore assemblage of
Prunus mahaleb: patterns and consequences. Oikos 71:479-491.
Jordano, P. and C. M. Herrera. 1995. Shuffling the offspring: uncoupling and spatial
discordance of multiple stages in vertebrate seed dispersal. Écoscience 2:230-237.
Joshi, A. R., J. L. D. Smith, and F. J. Cuthbert. 1995. Influence of food distribution and
predation pressure on spacing behavior in palm civets. Journal of Mammalogy
76:1205-1212.
Julien-Laferriere, D. 1993. Radio-tracking observations on ranging and foraging patterns
by kinkajous {Potos flavus) in French Guiana. Journal of Tropical Ecology 9:19-
32.

165
Julliot, C. 1996. Seed dispersal by red howling monkeys (Alouatta seniculus) in the
tropical rain forest of French Guiana. International Journal of Primatology 17:239-
258.
Julliot, C. 1997. Impact of seed dispersal by red howler monkeys Alouatta seniculus on the
seedling population in the understorey of tropical rain forest. Journal of Ecology
85:431-440.
Kamil, A. C. and J. E. Jones. 1997. The seed storing corvid Clark's nutcracker learns
geometric relationships among landmarks. Nature 390:276-279.
Kaspari, M. 1993. Removal of seeds from Neotropical frugivore droppings. Oecologia
95:81-88.
Kauffman, S., D. B. McKey, M. Hossaert-McKey, and C. C. Horvitz. 1991. Adaptations
for a two-phase seed dispersal system involving vertebrates and ants in a
hemiepiphytic fig (Ficus microcarpa: Moraceae). American Journal of Botany
78:971-977.
Kelly, C. K. and A. Purvis. 1993. Seed size and establishment conditions in tropical trees.
Oecologia 94:356-360.
Kiltie, R. A. 1981. Distribution of palm fruits on a rainforest floor: why white-lipped
peccaries forage near objects. Biotropica 13:141-145.
Kitajima, K. 1996. Ecophysiology of tropical tree seedlings. Pages 559-596 in S. S.
Mulkey, R. L. Chazdon, and A. P. Smith, editors. Tropical forest plant
ecophysiology. Chapman and Hall, New York.
Kitajima, K. and C. K. Augspurger. 1989. Seed and seedling ecology of a monocarpic
tropical tree, Tachigalia versicolor. Ecology 70:1102-1114.
Kitchings, J. T. and D. J. Levey. 1981. Habitat patterns in a small mammal community.
Journal of Mammalogy 62:814-820.
Krefting, L. W. and E. I. Roe. 1949. The role of some birds and mammals in seed
germination. Ecological Monographs 19:270-286.
Krijger, C. L., M. Opdam, M. Théry, and F. Bongers. 1997. Courtship behaviour of
manakins and seed bank composition in a French Guianan rain forest. Journal of
Tropical Ecology 13:631-636.
Lack, A. J. 1991. Dead logs as substrate for rain forest trees in Dominica. Journal of
Tropical Ecology 7:401-405.
Laman, T. G. 1995. Ficus stupenda germination and seedling establishment in a Bornean
rain forest canopy. Ecology 76:2617-2626.
Laman, T. G. 1996a. Ficus seed shadows in a Bornean rain forest. Oecologia 107:347-
355.
Laman, T. G. 1996b. Specialization for canopy position by hemiepiphytic Ficus species in
a Bornean rain forest. Journal of Tropical Ecology 12:789-803.

166
Lamont, B. B., E. T. F. Witkowski, and N. J. Enright. 1993. Post-fire litter microsites:
safe for seeds, unsafe for seedlings. Ecology 74:501-512.
Langtimm, C. A. 1992. Specialization for vertical habitats within a cloud forest community
of mice. Dissertation. University of Florida, Gainesville, Florida.
Lanner, R. M. 1996. Made for each other: a symbiosis of birds and pines. Oxford
University Press, New York.
Larson, D. and H. F. Howe. 1987. Dispersal and destruction of Virola surinamensis seeds
by agoutis: appearance and reality. Journal of Mammalogy 68:859-860.
Larson, D. L. 1996. Seed dispersal by specialist versus generalist foragers: the plant's
perspective. Oikos 76:113-120.
Lawton, R. and V. Dryer. 1980. The vegetation of the Monteverde Cloud Forest Preserve.
Brenesia 18:101-116.
Lawton, R. O. 1990. Canopy gaps and light penetration into a wind-exposed tropical lower
montane rain forest Canadian Journal of Forest Research 20:659-667.
Lawton, R. O. and F. E. Putz. 1988. Natural disturbance and gap-phase regeneration in a
wind-exposed tropical cloud forest. Ecology 69:764-777.
Lemmon, P. E. 1957. A new instrument for measuring forest overstory density. Journal of
Forestry 55:667-668.
Levey, D. J. 1986. Methods of seed processing by birds and seed deposition patterns.
Pages 147-158 in A. Estrada and T. H. Fleming, editors. Frugivores and seed
dispersal. Dr. W. Junk Publishers, Dordrecht, Netherlands.
Levey, D. J. 1987. Seed size and fruit-handling techniques of avian frugivores. American
Naturalist 129:471-485.
Levey, D. J. 1988a. Spatial and temporal variation in Costa Rican fruit and fruit-eating bird
abundance. Ecological Monographs 58:251-269.
Levey, D. J. 1988b. Tropical wet forest treefall gaps and distributions of understory birds
and plants. Ecology 69:1076-1089.
Levey, D. J. 1990. Habitat-dependent fruiting behaviour of an understorey tree, Miconia
centrodesma, and tropical treefall gaps as keystone habitats for frugivores in Costa
Rica. Journal of Tropical Ecology 6:409-420.
Levey, D. J. and M. M. Byrne. 1993. Complex ant-plant interactions: rain forest ants as
secondary dispersers and post-dispersal seed predators. Ecology 74:1802-1819.
Levey, D. J., T. C. Moermond, and J. S. Denslow. 1994. Frugivory: an overview. Pages
282-294 in L. A. McDade, K. S. Bawa, H. A. Hespenheide, and G. S. Hartshorn,
editors. La Selva: ecology and natural history of a neotropical rainforest
University of Chicago Press, Chicago, Illinois.

167
Levey, D. J. and F. G. Stiles. 1992. Evolutionary precursors of long-distance migration:
resource availability and movement patterns in neotropical landbirds. American
Naturalist 140:447-476.
Levey, D. J. and F. G. Stiles. 1994. Birds: ecology, behavior, and taxonomic affinities.
Pages 217-228 in L. A. McDade, K. S. Bawa, H. A. Hespenheide, and G. S.
Hartshorn, editors. La Selva: ecology and natural history of a neotropical
rainforest. University of Chicago Press, Chicago, Illinois.
Levin, D. A. and H. W. Kerster. 1974. Gene flow in seed plants. Evolutionary Biology
7:139-220.
Li, M., M. Lieberman, and D. Lieberman. 1996. Seedling demography in undisturbed
tropical wet forest in Costa Rica. Pages 285-314 in M. D. Swaine, editor. Ecology
of tropical forest tree seedlings. UNESCO/Parthenon, Paris/Camforth.
Lieberman, D. 1996. Demography of tropical tree seedlings: a review. Pages 131-137 in
M. D. Swaine, editor. Ecology of tropical forest tree seedlings.
UNESCO/Parthenon, Paris/Camforth.
Lieberman, D., J. B. Hall, M. D. Swaine, and M. Lieberman. 1979. Seed dispersal by
baboons in the Shai Hills, Ghana. Ecology 60:65-75. -
Livingston, R. B. 1972. Influence of birds, stones and soil on the establishment of pasture
juniper, Juniperus communis, and red cedar, J. virginiana in New England
pastures. Ecology 53:1141-1147.
Loiselle, B. A. 1990. Seeds in droppings of tropical fruit-eating birds: importance of
considering seed composition. Oecologia 82:494-500.
Loiselle, B. A., E. Ribbens, and O. Vargas. 1996. Spatial and temporal variation of seed
rain in a tropical lowland wet forest. Biotropica 28:82-95.
Lovelock, C. E„ D. Kyllo, and K. Winter. 1996. Growth responses to vesicular-
arbuscular mycorrhizae and elevated C02 in seedlings of a tropical tree,
Beilschmiedia péndula. Functional Ecology 10:662-667.
Malo, J. E. and F. Suarez. 1997. Dispersal mechanism and transcontinental naturalization
proneness among Mediterranean herbaceous species. Journal of Biogeography
24:391-394.
Martinez del Rio, C., M. Hourdequin, A. Silva, and R. Medel. 1995. The influence of
cactus size and previous infection on bird deposition of mistletoe seeds. Australian
Journal of Ecology 20:571-576.
Martinez del Rio, C., A. Silva, R. Medel, and M. Hourdequin. 1996. Seed dispersers as
disease vectors: bird transmission of mistletoe seeds to plants. Ecology 77:912-
921.
Martínez-Ramos, M. and E. Alvarez-Buylla. 1986. Seed dispersal, gap dynamics and tree
recruitment: the case of Cecropia obtusifolia at Los Tuxtlas, Mexico. Pages 333-
346 in A. Estrada and T. H. Fleming, editors. Frugivores and seed dispersal. Dr.
W. Junk Publishers, Dordrecht, Netherlands.

168
Martínez-Ramos, M. and E. R. Alvarez-Buylla. 1995. Seed dispersal and patch dynamics
in tropical rain forests: a demographic approach. Écoscience 2:223-229.
Martínez-Ramos, M., E. R. Alvarez-Buylla, J. Sarukhan, and D. Piñero. 1988. Treefall
age determination and gap dynamics in a tropical forest. Journal of Ecology
76:700-716.
Martínez-Ramos, M. and A. Soto-Castro. 1993. Seed rain and advanced regeneration in a
tropical rain forest. Vegetado 107/108:299-318.
Masaki, T., Y. Kominami, and T. Nakashizuka. 1994. Spatial and seasonal patterns of
seed dissemination of Comus controversa in a temperate forest. Ecology 75:1903-
1910.
Matlack, G. R. 1989. Secondary dispersal of seed across snow in Betula lenta, a gap-
colonizing tree species. Journal of Ecology 77:853-869.
Mazer, S. J. and N. T. Wheelwright. 1993. Fruit size and shape: allometry at different
taxonomic levels in bird-dispersed plants. Evolutionary Ecology 7:556-575.
McCanny, S. J. 1985. Alternatives in parent-offspring relationships in plants. Oikos
45:148-149.
McClanahan, T. R. and R. W. Wolfe. 1993. Accelerating forest succession in a
fragmented landscape: the role of birds and perches. Conservation Biology 7:279-
288.
McDonnell, M. J. 1986. Old Field vegetation height and the dispersal pattern of bird-
disseminated woody plants. Bulletin of the Torrey Botanical Club 113:6-11.
McDonnell, M. J. and E. W. Stiles. 1983. The structural complexity of old field vegetation
and the recruitment of bird-dispersed plant species. Oecologia 56:109-116.
McKey, D. 1975. The ecology of coevolved seed dispersal systems. Pages 159-191 in L.
E. Gilbert and P. H. Raven, editors. Coevolution of Plants and Animals.
University of Texas Press, Austin, Texas, USA.
Merg, K. F. 1994. Utility of seed dispersal in escape from seed predators: an experiment in
a lowland rainforest of New Guinea. M.S. Thesis. University of Florida,
Gainesville.
Meyer, J.-Y. and J. Florence. 1996. Tahiti's native flora endangered by the invasion of
Miconia calvescens DC. (Melastomataceae). Journal of Biogeography 23:775-781.
Moegenburg, S. M. 1994. Sabal palmetto seed-animal interactions. M.S. Thesis. Florida,
Gainesville.
Moegenburg, S. M. 1996. Sabal palmetto seed size: causes of variation, choices of
predators, and consequences for seedlings. Oecologia 106:539-543.
Moermond, T. C. and J. S. Denslow. 1983. Fruit choice in neotropical birds: effects of
fruit type and accessibility on selectivity. Journal of Animal Ecology 52:407-420.

169
Moermond, T. C. and J. S. Denslow. 1985. Neotropical avian frugivores: patterns of
behavior, morphology, and nutrition, with consequences for fruit selection.
Ornithological Monographs 36:865-897.
Moermond, T. C., J. S. Denslow, D. J. Levey, and E. Santana C. 1986. The influence of
morphology on fruit choice in neotropical birds. Pages 137-146 in A. Estrada and
T. H. Fleming, editors. Frugivores and seed dispersal. Dr. W. Junk Publishers,
Dordrecht, Netherlands.
Molofsky, J. and C. K. Augspurger. 1992. The effect of leaf litter on early seedling
establishment in a tropical forest. Ecology 73:68-77.
Monteiro, R. F., R. P. Martins, and K. Yamamoto. 1992. Host specificity and seed
dispersal of Psittaccmthus robustus (Loranthaceae) in south-east Brazil. Journal of
Tropical Ecology 8:307-314.
Morton, E. S. 1973. On the evolutionary advantages and disadvantages of fruit eating in
tropical birds. American Naturalist 107:8-22.
Mulkey, S. S., R. L. Chazdon, and A. P. Smith, editors. 1996. Tropical forest plant
ecophysiology. Chapman and Hall, New York, New York, USA.
Murray, K. G. 1986. Consequences of seed dispersal for gap-dependent plants:
relationships between seed shadows, germination requirements, and forest
dynamics. Pages 187-198 in A. Estrada and T. H. Fleming, editors. Frugivores
and seed dispersal. Dr. W. Junk Publishers, Dordrecht, The Netherlands.
Murray, K. G. 1988. Avian seed dispersal of three neotropical gap-dependent plants.
Ecological Monographs 58:171-198.
Myster, R. W. and S. T. A. Pickett. 1992. Effects of palatability and dispersal mode on
spatial patterns of trees in oldfields. Bulletin of the Torrey Botanical Club 119:145-
151.
Myster, R. W. and S. T. A. Pickett. 1993. Effects of litter, distance, density and
vegetation patch type on postdispersal tree seed predation in old fields. Oikos
66:381-388.
Nabhan, G. 1997. Why chilies are hot. Natural History June:24-29.
Nadkami, N. M„ T. J. Matelson, and W. A. Haber. 1995. Structural characteristics and
floristic composition of a Neotropical cloud forest, Monteverde, Costa Rica.
Journal of Tropical Ecology 11:481-495.
Nadkami, N. M. and N. T. Wheelwright, editors, in press. Monteverde: the ecology and
conservation of a tropical cloud forest. Oxford University Press, New York.
Ne'eman, G. and I. Izhaki. 1996. Colonization in an abandoned East-Mediterranean
vineyard. Journal of Vegetation Science 7:465-472.
Nepstad, D. C., C. Uhl, C. A. Pereira, and J. M. C. da Silva. 1996. A comparative study
of tree establishment in abandoned pasture and mature forest of eastern Amazonia.
Oikos 76:25-39.

170
Nogales, M., F. M. Medina, and A. Valido. 1996. Indirect seed dispersal by the feral cats
Felis cams in island ecosystems (Canary Islands). Ecography 19:3-6.
Nogales, M., A. Valido, and F. M. Medina. 1995. Frugivory of Plocama péndula
(Rubiaceae) by the rabbit (Oryctolagus cuniculus) in xerophytic zones of Tenerife
(Canary Islands). Acta Oecologica 16:585-591.
O'Dowd, D. J. and M. E. Hay. 1980. Mutualism between harvester ants and a desert
ephemeral: seed escape from rodents. Ecology 61:531-540.
Ohkawara, K., M. Ohara, and S. Higashi. 1997. The evolution of ant-dispersal in a
spring-ephemeral Corydalis ambigua (Papaveraceae): timing of seed-fall and effects
of ants and ground beetles. Ecography 20:217-223.
Okubo, A. and S. A. Levin. 1989. A theoretical framework for data analysis of wind
dispersal of seeds and pollen. Ecology 70:329-338.
Orians, G. H. 1969. The number of bird species in some tropical forests. Ecology 50:783-
801.
Oswald, B. P. and L. F. Neuenschwander. 1993. Microsite variability and safe site
description for western larch germination and establishment Bulletin of the Torrey
Botanical Club 120:148-156.
Owens, M. K., R. B. Wallace, and S. R. Archer. 1995. Landscape and microsite
influences on shrub recruitment in a disturbed semi-arid Quercus-Juniperus
woodland. Oikos 74:493-502.
Peres, C. A. and C. Baider. 1997. Seed dispersal, spatial distribution and population
structure of brazilnut trees (Bertholletia excelsa) in southeastern Amazonia. Journal
of Tropical Ecology 13:595-616.
Peres, C. A., L. C. Schiesari, and C. L. Dias-Leme. 1997. Vertebrate predation of Brazil-
nuts (Bertholletia excelsa, Lecythidaceae), an agouti-dispersed Amazonian seed
crop: a test of the escape hypothesis. Journal of Tropical Ecology 13:69-79.
Pignozzi, G. 1992. Frugivory and seed dispersal by the European badger in a
Mediterranean habitat Journal of Mammalogy 73:630-639.
Popma, J„ F. Bongers, M. Martinez-Ramos, and E. Veneklaas. 1988. Pioneer species
distribution in treefall gaps in Neotropical rain forest; a gap definition and its
consequences. Journal of Tropical Ecology 4:77-88.
Portnoy, S. and M. F. Willson. 1993. Seed dispersal curves: behavior of the tail of the
distribution. Evolutionary Ecology 7:25-44.
Powell, G. V. N. and R. Bjork. 1995. Implications of intratropical migration on reserve
design: a case study using Pharomachrus mocinno. Conservation Biology 9:354-
362.
Price, M. V. and S. H. Jenkins. 1986. Rodents as seed consumers and dispersers. Pages
191-235 in D. R. Murray, editor. Seed dispersal. Academic Press, Sydney,
Australia.

171
Pyke, D. A. and J. N. Thompson. 1986. Statistical analysis of survival and removal rate
experiments. Ecology 67:240-245.
Redbo-Torstensson, P. and A. Telenius. 1995. Primary and secondary dispersal by wind
and water in Spergualria salina. Ecography 18:230-237.
Reid, N. 1989. Dispersal of mistletoes by honeyeaters and flowerpeckers: components of
seed dispersal quality. Ecology 70:137-145.
Reid, N. 1991. Coevolution of mistletoes and frugivorous birds? Australian Journal of
Ecology 16:457-469.
Restrepo, C. 1987. Aspectos ecológicos de la diseminación de cinco especies de
muérdagos por aves. Humboltia 1:65-116.
Rice, B. L. and M. Westoby. 1986. Evidence against the hypothesis that ant-dispersed
seeds reach nutrient-enriched microsites. Ecology 67:1270-1274.
Ridley, H. N. 1930. The dispersal of plants throughout the world. Reeve, Ashford, UK.
Riley, C. M. and K. G. Smith. 1992. Sexual dimorphism and foraging behavior of
Emerald Toucanets Aulacorhynchus prasinus in Costa Rica. Omis Scandinavica
23:459-466.
Roberts, J. T. and E. R. Heithaus. 1986. Ants rearrange the vertebrate-generated seed
shadow of a neotropical fig tree. Ecology 67:1946-1051.
Robinson, G. R. and S. N. Handel. 1993. Forest restoration on a closed landfill: rapid
addition of new species by bird dispersal. Conservation Biology 7:271-278.
Runkle, J. R. 1990. Gap dynamics in an Ohio Acer-Fagus forest and speculations on the
geography of disturbance. Canadian Journal of Forest Research 20:632-641.
Santana C., E. and B. G. Milligan. 1984. Behavior of toucanets, bellbirds, and quetzals
feeding on lauraceous fruits. Biotropica 16:152-154.
Sargent, S. 1995. Seed fate in a tropical mistletoe: the importance of host twig size.
Functional Ecology 9:197-204.
SAS Institute. 1988. SAS/STAT user’s guide. SAS Institute, Inc., Cary, North Carolina.
SAS Institute. 1989. JMP user's guide, version 2. SAS Institute, Inc., Cary, North
Carolina.
Schemske, D. W. and N. V. L. Brokaw. 1981. Treefalls and the distribution of understory
birds in a tropical forest. Ecology 62:938-945.
Schupp, E. W. 1988a. Factors affecting post-dispersal seed survival in a tropical forest.
Oecologia 76:525-530.
Schupp, E. W. 1988b. Seed and early seedling predation in the forest understory and in
treefall gaps. Oikos 51:71-78.

172
Schupp, E. W. 1993. Quantity, quality and the effectiveness of seed dispersal by animals.
Vegetado 107/108:15-29.
Schupp, E. W. 1995. Seed-seedling conflicts, habitat choice, and patterns of plant
recruitment. American Journal of Botany 82:399-409.
Schupp, E. W. and E. J. Frost. 1989. Differential predation of Welfia georgii seeds in
treefall gaps and the forest understory. Biotropica 21:200-203.
Schupp, E. W. and M. Fuentes. 1995. Spatial pattern of seed dispersal and the unification
of plant population ecology. Écoscience 2:267-275.
Schupp, E. W., J. M. Gómez, J. E. Jiménez, and M. Fuentes. 1997a. Dispersal of
Juniperus occidentals (western juniper) seeds by frugivorous mammals on Juniper
Mountain, southeast Oregon. Great Basin Naturalist 57:74-78.
Schupp, E. W., H. J. Heaton, and J. M. Gómez. 1997b. Lagomorphs and the dispersal of
seeds into communities dominated by exotic annual weeds. Great Basin Naturalist
57:253-258.
Schupp, E. W., H. F. Howe, C. K. Augspurger, and D. J. Levey. 1989. Arrival and
survival in tropical treefall gaps. Ecology 70:562-564.
Scowcroft, P. G. 1992. Role of decaying logs and other organic seedbeds in natural
regeneration of Hawaiian forest species on abandoned montane pasture. USDA
Forest Service General Technical Report PSW-129:67-73.
Shaw, R. G. and T. Mitchell-Olds. 1993. ANOVA for unbalanced data: an overview.
Ecology 74:1638-1645.
Shepherd, V. E. and C. A. Chapman, in press. Dung beetles as secondary seed dispersers:
impact of seed predation and germination. Journal of Tropical Ecology 14:xx-xx.
Skeate, S. T. 1987. Interactions between birds and fruits in a northern Florida hammock
community. Ecology 68:297-309.
Skutch, A. F. 1967. Life histories of Central American highland birds. Publications of the
Nuttall Ornithological Club 7:1-213.
Smith, A. J. 1975. Invasion and ecesis of bird-disseminated woody plants in a temperate
forest sere. Ecology 56:19-34.
Smith, A. P. 1984. Postdispersal parent-offspring conflict in plants: antecedent and
hypothesis from the Andes. American Naturalist 123:354-370.
Smith, B. H., P. D. Forman, and A. E. Boyd. 1989. Spatial patterns of seed dispersal and
predation of two myrmecochorous forest herbs. Ecology 70:1649-1656.
Smythe, N. 1978. The natural history of the Central American agouti (Dasyprocta
punctata). Smithsonian Contributions to Zoology 257:1-52.
Smythe, N. 1989. Seed survival in the palm Astrocaryum standleyanum: evidence for
dependence upon its seed dispersers. Biotropica 21:50-56.

173
Sneddon, I. A. 1991. Latrine use by the European rabbit (Oryctolagus cuniculus). Journal
of Mammalogy 72:769-775.
Snow, B. K. 1961. Notes on the behavior of three Cotingidae. Auk 78:150-161.
Snow, B. K. 1970. A field study of the Bearded Bellbird in Trinidad. Ibis 112:299-329.
Snow, B. K. 1973a. Notes on the behavior of the White Bellbird. Auk 90:743-751.
Snow, B. K. 1977. Territorial behavior and courtship of the male Three-wattled Bellbird.
Auk 94:623-645.
Snow, B. K. 1979. The oilbirds of Los Tayos. Wilson Bulletin 91:457-461.
Snow, D. W. 1962a. A field study of the black and white manakin, Manacus manacus, in
Trinidad. Zoológica 47:65-104.
Snow, D. W. 1962b. A field study of the golden-headed manakin, Pipra erythrocephala, in
Trinidad, W.I. Zoológica 47:183-198.
Snow, D. W. 1965. A possible selective factor in the evolution of fruiting seasons in
tropical forest. Oikos 15:274-281.
Snow, D. W. 1971. Evolutionary aspects of fruit-eating by birds. Ibis 113:194-202.
Snow, D. W. 1973b. Distribution, ecology and evolution of the bellbirds (Procnias,
Cotingidae). Bulletin of the British Museum (Natural History) 25:369-391.
Snow, D. W. 1976. The web of adaptation. Quadrangle, New York.
Snow, D. W. 1981. Tropical frugivorous birds and their food plants: a world survey.
Biotropica 13:1-14.
Snow, D. W. 1982. The Cotingas: bellbirds, umbrellabirds and other species. British
Museum and Cornell University Press, Ithaca, New York.
Sork, V. L. 1983. Distribution of pignut hickory (Carya glabra) along a forest to edge
transect, and factors affecting seedling establishment Bulletin of the Torrey
Botanical Club 110:491-506.
Stiles, E. W. 1982. Expansions of mockingbird and multiflora rose in the northeastern
United States and Canada. American Birds 36:358-364.
Stiles, E. W. 1989. Fruits, seeds, and dispersal agents. Pages 87-122 in W. G.
Abrahamson, editor. Plant-animal interactions. McGraw-Hill, New York, New
York.
Stiles, F. G. 1985. On the role of birds in the dynamics of neotropical forests. Pages 49-59
in A. W. Diamond and T. E. Lovejoy, editors. Conservation of tropical forest
birds. International Council for Bird Preservation, Cambridge, UK.
Stiles, F. G. and L. Rosselli. 1993. Consumption of fruits of the Melastomataceae by
birds: how diffuse is coevolution. Vegetatio 107/108:57-73.

174
Stiles, F. G. and A. F. Skutch. 1989. A guide to the birds of Costa Rica. Cornell
University Press, Ithaca, New York, USA.
Sun, C., T. C. Moermond, and T. J. Givnish. 1997. Nutritional determinants of diet in
three turacos in a tropical montane forest. Auk 114:200-211.
Swaine, M. D., editor. 1996. The Ecology of Tropical Forest Seedlings.
UNESCO/Parthenon, Paris/Camforth.
Swaine, M. D. and T. C. Whitmore. 1988. On the definition of ecological species groups
in tropical rain forests. Vegetatio 75:81-86.
Tanner, E. V. J. 1982. Species diversity and reproductive mechanisms in Jamaican trees.
Biological Journal of the Linnean Society 18:263-278.
Terborgh, J., E. Losos, M. P. Riley, and M. B. Riley. 1993. Predation by vertebrates and
invertebrates on the seeds of five canopy tree species of an Amazonian forest.
Vegetatio 107/108:375-386.
Tester, M„ D. C. Patton, N. Reid, and R. T. Lange. 1987. Seed dispersal by birds and
densities of shrubs under trees in arid south Australia. Transactions of the Royal
Society of South Australia 111:1-5.
Théry, M. and D. Larpin. 1993. Seed dispersal and vegetation dynamics at a cock-of-the-
rock's lek in the tropical forest of French Guiana. Journal of Tropical Ecology
9:109-116.
Thomas, D. W., D. Cloutier, M. Provencher, and C. Houle. 1988. The shape of bird- and
bat-generated seed shadows around a tropical fruiting tree. Biotropica 20:347-348.
Thompson, J. N. 1980. Treefalls and colonization patterns of temperate forest herbs.
American Midland Naturalist 104:176-184.
Thompson, J. N. and M. F. Willson. 1978. Disturbance and the dispersal of fleshy fruits.
Science 200:1161-1163.
Tomback, D. F. 1982. Dispersal of whitebark pine seeds by Clark's nutcracker: a
mutualism hypothesis. Journal of Ecology 51:451-467.
Tomback, D. F. and Y. B. Linhart. 1990. The evolution of bird-dispersed pines.
Evolutionary Ecology 4:185-219.
Trainer, J. M. and T. C. Will. 1984. Avian methods of feeding on Bursera simaruba
(Burseraceae) fruits in Panama. Auk 101:193-195.
Traveset, A. and M. F. Willson. 1997. Effect of birds and bears on seed germination of
fleshy-fruited plants in temperate rainforests of southeast Alaska. Oikos 80:89-95.
Trexler, J. C. and J. Travis. 1993. Nontraditional regression analyses. Ecology 74:1629-
1637.
Turin, C. E. G., E. A. Williamson, M. E. Rogers, and M. Fernandez. 1991. A case study
of a plant-animal relationship: Cola lizae and lowland gorillas in the Lopé Reserve,
Gabon. Journal of Tropical Ecology 7:181-199.

175
Valiente-Banuet, A., F. Vite, and J. A. Zavala-Hurtado. 1991. Interaction between the
cactus Neobuxbaumia tetezo and the nurse shrub Mimosa lusiana. Journal of
Vegetation Science 2:11-14.
Van der Meer, P. J. and F. Bongers. 1996. Patterns of tree-fall and branch-fall in a tropical
forest in French Guiana. Journal of Ecology 84:19-29.
van der Pijl, L. 1972. Principles of dispersal in higher plants. Springer-Verlag, Berlin.
Vander Wall, S. B. 1990. Food hoarding in animals. University of Chicago Press,
Chicago, Illinois.
Vander Wall, S. B. 1992. The role of animals in dispersing a "wind-dispersed" pine.
Ecology 73:614-621.
Vander Wall, S. B. 1997. Dispersal of singleleaf piñón pine (Pinus monophylla) by seed¬
caching rodents. Journal of Mammalogy 78:181-191.
Vander Wall, S. B. and R. P. Baida. 1977. Coadaptations of the Clark's nutcracker and
the piñón pine for efficient seed harvest and dispersal. Ecological Monographs
47:89-111.
Vásquez, R. A. 1996. Patch utilization by three species of Chilean rodents differing in
body size and mode of locomotion. Ecology 77:2343-2351.
Veenendaal, E. M., M. D. Swaine, V. K. Agyeman, D. Blay, I. K. Abebrese, and C. E.
Mullins. 1995. Differences in plant and soil water relations in and around a forest
gap in West Africa during the dry season may influence seedling establishment and
survival. Journal of Ecology 83:83-90.
Venable, D. L. and J. S. Brown. 1993. The population-dynamic functions of seed
dispersal. Vegetado 107/108:31-55.
Vieira, I. C. G„ C. Uhl, and N. D. 1994. The role of the shrub Cordia multisipcata Cham,
as a 'succession facilitator' in an abandoned pasture, Paragominas, Amazonia.
Vegetado 115:91-99.
Wahaj, S. A., D. J. Levey, A. K. Sanders, and M. L. Cipollini. in press. Control of gut
retention time by secondary metabolites in ripe Solanum fruits. Ecology 79:xx-xx.
Walsberg, G. E. 1975. Digestive adaptations of Phainopepla nitens associated with the
eating of mistletoe berries. Condor 77:169-174.
Weiblen, G. D. and J. D. Thomson. 1995. Seed dispersal in Erythronium grandiflorum
(Liliaceae). Oecologia 102:211-219.
Wheelwright, N. T. 1983. Fruits and the ecology of Resplendent Quetzals. Auk 100:286-
301.
Wheelwright, N. T. 1985a. Competition for dispersers, and the timing of flowering and
fruiting in a guild of tropical trees. Oikos 44:465-477.
Wheelwright, N. T. 1985b. Fruit size, gape width, and the diets of fruit-eating birds.
Ecology 66:808-818.

176
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.
Wheelwright, N. T. 1988. Four constraints on coevolution between fruit-eating birds and
fruiting plants: a tropical case study. Proceedings of the International Ornithological
Congress XIX:827-845.
Wheelwright, N. T. 1991. How long do fruit-eating birds stay in the plants where they
feed? Biotropica 23:29-40.
Wheelwright, N. T., W. A. Haber, K. G. Murray, and C. A. Guindon. 1984. Tropical
fruit-eating birds and their food plants: a survey of a Costa Rican lower montane
forest. Biotropica 16:173-192.
Wheelwright, N. T. and G. H. Orians. 1982. Seed dispersal by animals: contrasts with
pollen dispersal, problems of terminology, and constraints on coevolution.
American Naturalist 119:402-413.
Whelan, C. J., M. L. Dilger, D. Robson, N. Hallyn, and S. Dilger. 1994. Effects of
olfactory cues on artificial-nest experiments. Auk 111:945-952.
Whitmore, T. C. 1996. A review of some aspects of tropical rain forest seedling ecology
with suggestions for further enquiry. Pages 3-39 in M. D. Swaine, editor. Ecology
of tropical forest tree seedlings. UNESCO/Parthenon, Paris/Camforth.
Wilkinson, D. M. 1997a. Plant colonization: are wind dispersed seeds really dispersed by
birds at larger spatial and temporal scales? Journal of Biogeography 24:61-65.
Wilkinson, D. M. 1997b. The role of seed dispersal in the evolution of mycorrhizae. Oikos
78:394-396.
Willson, M. F. 1992. The ecology of seed dispersal. Pages 61-85 in M. Fenner, editor.
Seeds: the ecology of regeneration in plant communities. CAB International,
Wallingford, UK.
Willson, M. F. 1993. Dispersal mode, seed shadows, and colonization patterns. Vegetado
107/108:261-280.
Willson, M. F. and F. H. J. Crome. 1989. Patterns of seed rain at the edge of a tropical
Queensland rain forest Journal of Tropical Ecology 5:301-308.
Willson, M. F„ A. K. Irvine, and N. G. Walsh. 1989. Vertebrate dispersal syndromes in
some Australian and New Zealand plant communities, with geographic
comparisons. Biotropica 21:133-147.
Willson, M. F., E. A. Porter, and R. Condit. 1982. Avian frugivore activity in relation to
forest light gaps. Caribbean Journal of Science 18:1-6.
Witmer, M. C. and A. S. Cheke. 1991. The dodo and the tambalacoque tree: an obligate
mutualism reconsidered. Oikos 61:133-137.

177
Wright, S. J., M. E. Gompper, and B. DeLeon. 1994. Are large predators keystone
species in Neotropical forests? The evidence from Barro Colorado Island. Oikos
71:279-294.
Young, B. and D. McDonald, in press. Lessons from 25 years of bird research at
Monteverde in N. M. Nadkami and N. T. Wheelwright, editors. Monteverde: the
ecology and conservation of a tropical cloud forest. Oxford University Press, New
York.
Young, T. P. and S. P. Hubbell. 1991. Crown asymmetry, treefalls, and repeat
disturbance of broad-leaved forest gaps. Ecology 72:1464-1471.

BIOGRAPHICAL SKETCH
Born in Poughkeepsie, New York, I have always enjoyed being outside. Two of
my earliest memories are of climbing a spruce tree behind the house and rolling down the
hill in the front yard into the leaves below. In the same year I was bom, my parents bought
some land in New Hampshire, where the whole family spent two to three months every
summer. We lived in a large tent, cooked on an open wood fire, and kept food in a
propane refrigerator. Water was supplied by several natural springs (one of which doubled
as a 'fridge for the first couple years). We feasted on wild strawberries, blueberries,
raspberries, and blackberries, as well as apples from an abandoned orchard. These
summers on the 270 acres of Thompson Hill near Hillsboro were the foundation of my
interest in biology. I began by identifying ferns and trees, later catching snakes, frogs, and
salamanders, and finally, watching birds when I was given binoculars. My maternal
grandparents owned land adjacent to a lake nearby (within sight of the hill), where my
grandfather had been coming every summer since he was a small boy. The lake added
swimming, canoeing, and fishing to the activities. When I turned four, my family moved
to Wilmington, Delaware, where I eventually graduated from Brandywine High School.
My main activities in high school were bird-watching and listening to Firesign Theatre
records with my friend Jeff Gordon (now a guide for Victor Emanuel Nature Tours),
playing soccer on the school team, and playing street hockey.
I attended Earlham College where my most important mentors were ornithologist
Bill Buskirk and botanist Brent Smith. The highlights included a semester in Kenya on a
biology/sociology program, and a three week tropical ecology field trip to Costa Rica.
Both trips were led by Bill Buskirk. I became frustrated with the competitiveness of
college soccer and switched to ultimate frisbee, which is much more fun, a better workout,
178

179
and less injury-prone. After graduating from Earlham, I worked one summer in North
Dakota as a field assistant for Cornell University graduate student Carola Haas (now at
Virginia Tech). Then I moved on to a surprisingly quick and efficient master's degree at
the University of Missouri-Columbia under the direction of John Faaborg. One day in the
library at UMC, I read an article by Doug Levey, and decided I should write him to ask
about the possibility of continuing graduate work on some aspect of seed dispersal by
tropical birds. At the time I did not know that Doug had also graduated from Earlham. In
my last semester at UMC, I participated in an Organization for Tropical Studies course.
While at Missouri, I met Wendy Gibbons, another biology graduate student, and
we were married several years later. I began conducting field work in Costa Rica, and
enjoyed Monteverde so much that I convinced Wendy to move down with me so I could
work year-round instead of doing research only during the northern summers. Wendy
worked at the Monteverde Institute editing a newsletter. We lived in Costa Rica long
enough to distinguish a cold day and a hot one. Our first child, Malia, was bom in Costa
Rica in 1994, and, to complete our clutch, Jack was bom in Gainesville in 1996. Now we
know the real meaning of the phrase "sleeps like a baby."
My family and I will be moving to northwestern Illinois where I am taking a job
with the Illinois Natural History Survey as an avian ecologist. I will be based at the
Savanna Army Depot, which is being decommissioned and restored to prairie and oak
savanna. Future work in the tropics is also a possibility.

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.
Douglas Y. Levey, Qr&ir
Associate Professor'of Zoology
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.
Karen A. Bjomcml
Professor of Zoology
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.
(V
Colin A. Chapman
Assistant Professor of Zoology
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.
GL-5^ Holling
Arthur R. Marshall, Jr.,
Ecological Sciences
ofessor of
I certify that I have read this study and that in myopinion it conforms to acceptable
standards of scholarly presentation and is fully ade^uate/íhscope andjjuality, as a
dissertation for the degree of Doctor of Philos^
E. Putz
essor of Botany
This dissertation was submitted to the Graduate Faculty of the Department of
Zoology in the College of Liberal Arts and Sciences and to the Graduate School and was
accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.
May 1998
Dean, Graduate School

LO
1780
199¿?
,mu
UNIVERSITY OF FLORIDA
3 1262 08555 1223




PAGE 1

6((' ',63(56$/ $1' 3267',63(56$/ 6((' )$7( 2) )285 75(( 63(&,(6 ,1 $ 1(27523,&$/ &/28' )25(67 %\ '$1,(/ :(11< $ ',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 P\ JUDQGIDWKHU 'RQDOG %DNHU ZKR OHG PH RQ KLNHV LQ WKH ZRRGV RI 1HZ +DPSVKLUH DQG WR P\ NLGV 0DOLD DQG -DFN ZLWK ZKRP FDQ EDUHO\ NHHS XS

PAGE 3

$&.12:/('*0(176 )LUVW DQG IRUHPRVW WKDQN P\ DGYLVRU 'RXJ /HYH\ IRU KLV DGYLFH HQFRXUDJHPHQW SDWLHQFH DQG KHOS ZLWK DOO DVSHFWV RI WKLV SURMHFW +LV LQSXW ZDV LQVWUXPHQWDO LQ LWV VXFFHVV DQG ZLOO DOZD\V EH JUDWHIXO IRU KLV KHOS +H IRXQG D IHZ VHHGV HYHQ WKRXJK WKH VXQELWWHP HOXGHG KLP ,I GLVFRYHU D QHZ VSHFLHV ZLOO QDPH LW DIWHU KLP LQ DSSUHFLDWLRQ DOVR WKDQN P\ FRPPLWWHH .DUHQ %MRPGDO &ROLQ &KDSPDQ -DFN 3XW] DQG %X]] +ROOLQJ DOVR )UDQN %RQDFFRUVR DQG 3HWH )HLQVLQJHU DW DQ HDUOLHU VWDJHf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f IHHO VR ZHOFRPH WKHUH 6SHFLDO WKDQNV JR WR %HQLWR &DUORV DQG 7RP£V *XLQGRQ IRU VKDULQJ WKHLU NQRZOHGJH RI LQ

PAGE 4

WKH DUHD DQG WR *UHJ 0XUUD\ 0DXULFLR *DUFLD 5RGULJR 6RODQR %LOO +DEHU :LOORZ =XFKRZVNL )UDQN -R\FH $ODQ 3RXQGV DQG $ODQ DQG .DUHQ 0DVWHUV IRU SDWLHQWO\ DQVZHULQJ UD\ HQGOHVV TXHVWLRQV )LQDOO\ WKDQNV JR WR -XG\ 3RH DQG %UXFH 3DFN IRU UHQWLQJ XV WKHLU EHDXWLIXO RFWDJRQDO KRXVH RQ WKH FOLII HGJH DQG WR /H\Q 5RFNZHOO IRU HPHUJHQF\ UHSDLUV &UXFLDO ILQDQFLDO VXSSRUW DW WKH EHJLQQLQJ RI WKLV SURMHFW ZDV UHFHLYHG IURP WKH 2UJDQL]DWLRQ IRU 7URSLFDO 6WXGLHV D SLORW VWXG\ JUDQW 3HZ &KDULWDEOH 7UXVWf DQG D 7URSLFDO )HOORZVKLS -HVVH 6PLWK 1R\HV )RXQGDWLRQf 7KDQNV JR WR /XFLQGD 0F'DGH DQG 6KDXQ %HQQHWW IRU IDFLOLWDWLQJ 276 VXSSRUW 2WKHU HDUO\ IXQGLQJ IURP 6LJPD ;L *UDQWVLQ $LG RI 5HVHDUFK DQG $PHULFDQ 0XVHXP RI 1DWXUDO +LVWRU\ )UDQN 0 &KDSPDQ 0HPRULDO )XQGf LV JUDWHIXOO\ DFNQRZOHGJHG 7KH EXON RI WKH ILHOGZRUN ZDV VXSSRUWHG E\ D 'LVVHUWDWLRQ ,PSURYHPHQW *UDQW IURP WKH 1DWLRQDO 6FLHQFH )RXQGDWLRQ '(% f ,QGLUHFW ILQDQFLDO VXSSRUW ZDV DOVR FRQWULEXWHG E\ WKH *LEERQV DQG :HQQ\ IDPLOLHV DP DOVR JUDWHIXO IRU ORJLVWLF VXSSRUW IURP WKH 2UJDQL]DWLRQ IRU 7URSLFDO 6WXGLHV 0RQWHYHUGH ,QVWLWXWH HPDLO DFFHVVf DQG WKH 7URSLFDO 6FLHQFH &HQWHU ODE VSDFHf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

PAGE 5

%LJHORZ 6FRW 'XQFDQ 5RQ (GZDUGV 6XVDQ 0RHJHQEXUJ -D\ 2n6XOOLYDQ *UHJ 3U\RU 5DIDHO 6DPXGLR /HQQ\ 6DQWHVWHEDQ 0DUN 6SULW]HU 0DUNXV 7HOONDPS DQG 6SHFLDO WKDQNV JR WR 6XVDQ 0RHJHQEXUJ IRU EHLQJ WKH LGHDOO\ FKHHUIXO RIILFHPDWH DQG WR -DVRQ (YHUW IRU EHLQJ WKH LGHDOO\ LQYLVLEOH RIILFHPDWH Y

PAGE 6

7$%/( 2) &217(176 $&.12:/('*0(176 L $%675$&7 YLLL &+$37(56 *(1(5$/ ,1752'8&7,21 6((' ',63(56$/ 6((' 35('$7,21 $1' 6(('/,1* 5(&58,70(17 2) 2&27($ (1'5(6,$1$ /$85$&($(f ,QWURGXFWLRQ 6WXG\ 6LWH 6WXG\ 6SHFLHV 0HWKRGV 5HVXOWV 'LVFXVVLRQ &RQFOXVLRQ 6((' ',63(56$/ 2) $ +,*+48$/,7< )58,7 %< 63(&,$/,=(' )58*,925(6 +,*+48$/,7< ',63(56$/" ,QWURGXFWLRQ 6WXG\ 6LWH 6WXG\ 6SHFLHV 0HWKRGV 5HVXOWV 'LVFXVVLRQ 7:267$*( ',63(56$/ 2) 7:2 63(&,(6 2) *8$5($ 0(/,$&($(f ,QWURGXFWLRQ 6WXG\ 6LWH 6WXG\ 6SHFLHV 0HWKRGV 5HVXOWV 'LVFXVVLRQ YL

PAGE 7

$'9$17$*(6 2) 6((' ',63(56$/ $ 5((9$/8$7,21 2) ',5(&7(' ',63(56$/ ,QWURGXFWLRQ $GYDQWDJHV RI 'LVSHUVDO 7KH 'LIIXVH 0XWXDOLVP 3DUDGLJP &ODVVLF ([DPSOHV RI 'LUHFWHG 'LVSHUVDO 2WKHU ([DPSOHV RI 3RVVLEOH 'LUHFWHG 'LVSHUVDO &RQFOXVLRQ *RLQJ WKH 'LVWDQFH DQG %H\RQG $33(1',; 2&27($(1'5(6,$1$ 6,7( /2&$7,216 5()(5(1&(6 %,2*5$3+,&$/ 6.(7&+ YQ

PAGE 8

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n GLVSHUVDO IDWH LV LPSRUWDQW LQ GHWHUPLQLQJ SDWWHUQV RI SODQW UHFUXLWPHQW VWXGLHG GLVSHUVDO SDWWHUQV DQG SRVWGLVSHUVDO VHHG IDWHV RI IRXU WUHH VSHFLHV LQ 0RQWHYHUGH &RVWD 5LFD 7KH IRXU VSHFLHV VWXGLHG LQFOXGHG WZR VSHFLHV RI /DXUDFHDH 2FRWHD HQGUHVLDQD DQG %HLOVFKPLHGLD SQGXODf DQG WZR 0HOLDFHDH *XDUHD JODEUD DQG NXQWKLDQDf GHWHUPLQHG ORFDWLRQV RI GLVSHUVHG VHHGV E\ IROORZLQJ ELUGV XQWLO WKH\ UHJXUJLWDWHG RU GHIHFDWHG VHHGV DQG E\ V\VWHPDWLFDOO\ VHDUFKLQJ WKH VWXG\ VLWH IRU UHFHQWO\ GLVSHUVHG VHHGV )RU HDFK RI WKH IRXU VSHFLHV DSSUR[LPDWHO\ b RI WKH VHHGV GLVSHUVHG E\ ELUGV ZHUH GHSRVLWHG ZLWKLQ P RI WKH SDUHQW WUHHV 2FRWHD HQGUHVLDQD VKRZHG D ELPRGDO SDWWHUQ LQ ZKLFK EHOOELUGV 3URFQLDV WULFDUXQFXODWDf GLVSHUVHG PDQ\ VHHGV XQGHU VRQJ SHUFKHV DQG WKXV SURGXFHG D VHFRQG SHDN LQ VHHG UDLQ LQ DGGLWLRQ WR WKH SHDN QHDU SDUHQW WUHHV 0DUNHG VHHGV ZHUH XVHG WR GHWHUPLQH LI SRVWGLVSHUVDO UHPRYDO UHVXOWHG LQ VHHG SUHGDWLRQ RU VHFRQGDU\ GLVSHUVDO 5HPRYDO UDWHV ZHUH KLJK IRU 2FRWHD b UHPRYHGf DQG *XDUHD bf DQG ORZ IRU YLLL

PAGE 9

%HLOVFKPLHGLD bf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

PAGE 10

&+$37(5 *(1(5$/ ,1752'8&7,21 :LWK SODQWV WKHUH LV D YDVW GHVWUXFWLRQ RI VHHGV 6HHGOLQJV DOVR DUH GHVWUR\HG LQ YDVW QXPEHUV E\ YDULRXV HQHPLHV &KDUOHV 'DUZLQ f 3ODQWV DUH QRW FKDULWDEOH EHLQJV :LOOLDP %HDO f 6HHG GLVSHUVDO E\ DQLPDOV LV ZLGHVSUHDG DQG UHFRJQL]HG DV LPSRUWDQW E\ YLUWXH RI WKH ODUJH QXPEHU RI SODQW VSHFLHV ZLWK IUXLWV SUHVXPDEO\ DGDSWHG IRU DQLPDO FRQVXPSWLRQ ,Q WURSLFDO IRUHVWV LQ SDUWLFXODU b RI WKH WUHHV DQG VKUXEV DUH WKRXJKW WR EH GLVSHUVHG E\ DQLPDOV 6WLOHV +RZH :LOOVRQ HW DO -RUGDQR f 7KLV VHHG GLVSHUVDO LQWHUDFWLRQ LV FKDUDFWHUL]HG DV D GLIIXVH PXWXDOLVP EHFDXVH PRVW SODQWV KDYH PXOWLSOH GLVSHUVDO DJHQWV DQG PRVW VHHG GLVSHUVLQJ DQLPDOV GLVSHUVH VHYHUDO WR PDQ\ GLIIHUHQW SODQW VSHFLHV 7KXV IUXLWIUXJLYRUH FRHYROXWLRQ JHQHUDOO\ RFFXUV DPRQJ JURXSV RI SODQWV DQG JURXSV RI GLVSHUVHUV UDWKHU WKDQ DW WKH VSHFLHVVSHFLILF OHYHO VHHQ LQ PDQ\ SROOLQDWLRQ V\VWHPV :KHHOZULJKW DQG 2ULDQV f 7KH DGYDQWDJH RI WKLV PXWXDOLVP WR DQLPDOV LV D QXWULWLYH UHZDUG ZKHUHDV WKH DGYDQWDJHV WR SODQWV DUH HVFDSH IURP GHQVLW\ GHSHQGHQW PRUWDOLW\ QHDU WKH SDUHQW SODQW DQG DUULYDO LQ VXLWDEOH VLWHV IRU HVWDEOLVKPHQW $UULYDO LQ VXLWDEOH ORFDWLRQV FDQ EH HLWKHU D UDQGRP SURFHVV LQ ZKLFK ZLGHVSUHDG GLVVHPLQDWLRQ RI VHHGV ZRXOG LQFUHDVH WKH FKDQFH RI FRORQL]LQJ D JRRG VLWH FRORQL]DWLRQ K\SRWKHVLVf RU D QRQUDQGRP SURFHVV LQ ZKLFK VHHGV DUH GLUHFWHG WR VXLWDEOH VLWHV E\ DWWUDFWLQJ FHUWDLQ GLVSHUVDO DJHQWV GLUHFWHG GLVSHUVDO K\SRWKHVLVf (VFDSH DQG FRORQL]DWLRQ DUH WKRXJKW WR EH WKH PDLQ DGYDQWDJHV RI VHHG GLVSHUVDO IRU PRVW SODQWV +RZH f DOWKRXJK IHZ VWXGLHV KDYH H[DPLQHG WKHVH K\SRWKHVHV LQ GHWDLO

PAGE 11

2QH RI WKH PDLQ IDFWRUV OLPLWLQJ RXU XQGHUVWDQGLQJ RI WKH DGYDQWDJHV RI GLVSHUVDO LV WKDW PRVW SUHYLRXV VWXGLHV KDYH H[DPLQHG DVSHFWV RI GLVSHUVDO ZLWKRXW FRQVLGHULQJ VXEVHTXHQW VWDJHV WKDW PD\ DOVR LQIOXHQFH UHFUXLWPHQW 'LVSHUVDO LV SDUW RI WKH SODQW UHFUXLWPHQW SURFHVV ZKLFK LV FRPSRVHG RI VHYHUDO VWDJHV DQG PRVW SUHYLRXV VWXGLHV KDYH EHHQ VWDJHVSHFLILF =RRORJLVWV KDYH IRFXVHG RQ IUXLW UHPRYDO IRUDJLQJ EHKDYLRU DQG JXW WUHDWPHQW RI VHHGV ZKLOH ERWDQLVWV KDYH VWXGLHG JHUPLQDWLRQ DQG VHHGOLQJ JURZWK +RZH Ef 6HHG SUHGDWLRQ KDV DWWUDFWHG WKH DWWHQWLRQ RI ERWK ]RRORJLVWV DQG ERWDQLVWV EXW IHZ VWXGLHV KDYH LQWHJUDWHG DOO WKH VWDJHV RI UHFUXLWPHQW LQ RQH VWXG\ EXW VHH +HUUHUD HW DO f 2XU XQGHUVWDQGLQJ RI WKH LPSRUWDQFH RI GLVSHUVDO E\ DQLPDOV LV OLPLWHG EHFDXVH ZH NQRZ VR OLWWOH DERXW ZKHUH DQLPDOV WDNH VHHGV DQG ZKDW KDSSHQV WR WKHP DIWHU GLVSHUVDO )XUWKHUPRUH PDQ\ VWXGLHV DVVXPH WKDW ZKHQ VHHGV DUH GLVSHUVHG WKH\ ZLOO HLWKHU OLYH RU GLH LQ WKDW SODFH 5HFHQW VWXGLHV KRZHYHU LQGLFDWH WKDW PDQ\ VHHGV H[SHULHQFH D VHFRQG VWDJH RI GLVSHUVDO E\ DQWV URGHQWV GXQJ EHHWOHV DQG RWKHU DQLPDOV (VWUDGD DQG &RDWHV(VWUDGD /HYH\ DQG %\PH )UDJRVR f 7KXV FRQFOXVLRQV EDVHG RQ WKH SDWWHUQ RI LQLWLDO GLVSHUVDO PD\ EH PLVOHDGLQJ 7KH JRDO RI WKLV VWXG\ ZDV WR H[DPLQH WKH VWDJHV RI UHFUXLWPHQW LQFOXGLQJ VHHG GLVSHUVDO VHHG SUHGDWLRQ VHFRQGDU\ GLVSHUVDO JHUPLQDWLRQ DQG VHHGOLQJ HVWDEOLVKPHQW WR OLQN SDWWHUQV RI GLVSHUVDO ZLWK SDWWHUQV RI UHFUXLWPHQW 7KLV VWXG\ WRRN SODFH LQ 0RQWHYHUGH &RVWD 5LFD ZKLFK KDV EHHQ WKH VLWH RI VRPH RI WKH NH\ VWXGLHV LQ WKH VHHG GLVSHUVDO OLWHUDWXUH HJ :KHHOZULJKW DQG 2ULDQV 0XUUD\ 1DGNDPL DQG :KHHOZULJKW LQ SUHVVf ,Q &KDSWHUV DQG SUHVHQW VWXGLHV RI WZR WUHH VSHFLHV LQ WKH SODQW IDPLO\ /DXUDFHDH 2FRWHD HQGUHVLDQD DQG %HLOVFKPLHGLD SQGXOD /DXUDFHRXV VSHFLHV DUH DQ LPSRUWDQW FRPSRQHQW RI 1HRWURSLFDO PRQWDQH IRUHVWV LQ WHUPV RI VSHFLHV ULFKQHVV *HQWU\ +DEHU HW DO f DQG WKH YDULHW\ RI ELUGV WKDW DUH DW OHDVW SDUWLDOO\ GHSHQGHQW RQ WKHLU IUXLWV :KHHOZULJKW HW DO f ,Q &KDSWHU SUHVHQW D VWXG\ RI WZR VSHFLHV RI 0HOLDFHDH *XDUHD JODEUD DQG NXQWKLDQD 7KH PDLQ REMHFWLYH IRU DOO IRXU VSHFLHV ZDV WR ILQG WKH ORFDWLRQV ZKHUH VHHGV DUH GLVSHUVHG DQG IROORZ WKH IDWH RI WKRVH

PAGE 12

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b DUH GLVSHUVHG WR JDSV E\ PDOH EHOOELUGV 3URFQLDV WULFDUXQFXODWDf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f SUHFOXGHG IROORZLQJ WKH ELUGV ORQJ HQRXJK WR ILQG WKH VHHGV 2WKHUZLVH WKH PHWKRGV ZHUH VLPLODU WR WKRVH XVHG IRU 2FRWHD 7KH GLVWULEXWLRQ RI GLVSHUVHG VHHGV ZDV VLPLODU WR WKDW RI 2FRWHD H[FHSW WKDW IHZ VHHGV ZHUH GLVSHUVHG WR JDSV 6HHG SUHGDWLRQ ZDV YHU\ ORZ DQG DJDLQ VHFRQGDU\ GLVSHUVDO ZDV QRW REVHUYHG +DELWDW FKDUDFWHULVWLFV ZHUH QRW XVHIXO LQ SUHGLFWLQJ VHHGOLQJ

PAGE 13

VXUYLYDO 7KH KLJKHVW SUREDELOLW\ RI VHHGOLQJ UHFUXLWPHQW ZDV P IURP WKH HGJH RI SDUHQW WUHH FURZQV GLVFXVV WKH UHVXOWV LQ WHUPV RI GLVSHUVDO TXDOLW\ DQG VXJJHVW WKDW IRU WKLV VSHFLHV GLVSHUVDO ZLWKLQ P RI WKH FURZQ LV UHODWLYHO\ KLJKTXDOLW\ GLVSHUVDO DQG DSSUR[LPDWHO\ b RI WKH VHHGV UHFHLYH VXFK WUHDWPHQW ,Q &KDSWHU SUHVHQW GDWD RQ WKH GLVSHUVDO V\VWHPV RI WZR VSHFLHV RI *XDUHD 0HOLDFHDHf WKDW SURGXFHG IUXLW DW D WLPH ZKHQ 2FRWHD DQG %HLOVFKPLHGLD GLG QRW $V LQ WKH %HLOVFKPLHGLD VWXG\ GLVSHUVHG VHHGV ZHUH ORFDWHG E\ V\VWHPDWLF JURXQG VHDUFKHV 7KH GLVWULEXWLRQ RI GLVSHUVHG VHHGV ZDV VLPLODU WR WKH SUHYLRXV WZR VSHFLHV EXW b RI WKH VHHGV KDG D VHFRQG VWDJH RI GLVSHUVDO SUHVXPDEO\ E\ DJRXWLV 'DV\SURFWD SXQFWDWDf $JRXWLV EXULHG PDQ\ RI WKH VHHGV WKH\ UHPRYHG LQ VKDOORZ VXUIDFH FDFKHV 7KXV VHFRQGDU\ GLVSHUVDO UHVXOWHG LQ D UHDUUDQJHPHQW RI WKH LQLWLDO VHHG VKDGRZ ,Q &KDSWHU IXUWKHU H[SORUH WKH JHQHUDO LGHD RI GLVSHUVDO GLUHFWHG GLVSURSRUWLRQDWHO\ WR VLWHV VXLWDEOH IRU VXUYLYDO 7KH LPSOLFDWLRQ RI GLUHFWHG GLVSHUVDO LV D GLVSURSRUWLRQDWH HIIHFW RQ SODQW UHFUXLWPHQW EXW H[DPSOHV RI GLUHFWHG GLVSHUVDO DUH WKRXJKW WR EH UDUH DQG XQXVXDO +RZH f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

PAGE 14

XQGHUVWDQGLQJ RI VHHG GLVSHUVDO DQG VXEVHTXHQW SRVWGLVSHUVDO IDWHV PD\ EH XVHIXO LQ FRQVHUYDWLRQ DQG UHVWRUDWLRQ HFRORJ\

PAGE 15

&+$37(5 6((' ',63(56$/ 6((' 35('$7,21 $1' 6(('/,1* 5(&58,70(17 2) $ 1(27523,&$/ 0217$1( 75(( 1R RQH KDV VXFFHHGHG LQ PHDVXULQJ DFWXDO SDWWHUQV RI VHHG GLVSHUVDO SURGXFHG E\ GLIIHUHQW ELUG VSHFLHV IRU DQ\ ELUGGLVSHUVHG SODQW VSHFLHV (YHQ LI ZH FRXOG GHWHUPLQH ZKHUH ELUGV GURSSHG DOO VHHGV LW ZRXOG VWLOO EH GLIILFXOW WR UDQN GLIIHUHQW ELUG VSHFLHV DFFRUGLQJ WR GLVSHUVDO TXDOLW\ EHFDXVH VR OLWWOH LV NQRZQ DERXW VHHGOLQJ DQG VDSOLQJ PLFURKDELWDW UHTXLUHPHQWV 1DWKDQLHO7 :KHHOZULJKW f ,QWURGXFWLRQ 6HHG GLVSHUVDO GHWHUPLQHV WKH VSDWLDO DUUDQJHPHQW DQG SK\VLFDO HQYLURQPHQW RI VHHGV IURP ZKLFK WKH QH[W FRKRUW RI VHHGOLQJV LV VHOHFWHG 7KH OLQNV EHWZHHQ VHHG GLVSHUVDO DQG VHHGOLQJ UHFUXLWPHQW KRZHYHU DUH SRRUO\ XQGHUVWRRG 'LVSHUVDO LV WKH ILUVW RI D VHULHV RI HYHQWV RQH RU PRUH RI ZKLFK PD\ EH LPSRUWDQW LQ OLPLWLQJ UHFUXLWPHQW )RU YHUWHEUDWHGLVSHUVHG SODQWV WKHVH VWDJHV LQFOXGH IUXLW UHPRYDO VHHG GLVVHPLQDWLRQ SRVWn GLVSHUVDO VHHG SUHGDWLRQ SRWHQWLDOO\ VHFRQGDU\ GLVSHUVDO JHUPLQDWLRQ DQG VHHGOLQJ HVWDEOLVKPHQW 3UHYLRXV VWXGLHV KDYH JHQHUDOO\ IRFXVHG RQ RQO\ RQH RU D IHZ RI WKHVH VWDJHV +HUUHUD HW DO 6FKXSS DQG )XHQWHV f 'HVSLWH PDQ\ VWXGLHV RQ IUXLW FRQVXPSWLRQ DQG VHHGKDQGOLQJ E\ ELUGV DQG PDPPDOV VHHG VKDGRZV VSDWLDO SDWWHUQV RI GLVSHUVHG VHHGVf JHQHUDWHG E\ DQLPDOV DUH SRRUO\ NQRZQ :LOOVRQ f :KLOH FRQFHQWUDWLRQV RI VHHGV XQGHU RU QHDU IUXLWLQJ WUHHV KDYH EHHQ QRWHG DQG LQGHHG DUH H[SHFWHG -DQ]HQ +RZH :LOOVRQ f WKH SDWWHUQ RI VHHG GLVWULEXWLRQ IDUWKHU IURP IUXLWLQJ WUHHV LV OLNHO\ WR EH WKH PRVW LPSRUWDQW SDUW RI WKH VHHG VKDGRZ IRU SODQW ILWQHVV DQG SRSXODWLRQ HFRORJ\ 3RUWQR\ DQG :LOOVRQ 6FKXSS DQG )XHQWHV f ,Q SDUWLFXODU WKH WDLO RI WKH GLVWULEXWLRQ LV RIWHQ KHWHURJHQHRXV

PAGE 16

DV D UHVXOW RI GLVSHUVHU EHKDYLRU 0F'RQQHOO DQG 6WLOHV +RSSHV &KDYH] 5DPLUH] DQG 6ODFN -XOOLRW f )UXLWHDWLQJ YHUWHEUDWHV GLIIHU LQ IRUDJLQJ EHKDYLRU &UX] 6DQWDQD & DQG 0LOOLJDQ 7UDLQHU DQG :LOO 0RHUPRQG DQG 'HQVORZ f IUXLW UHPRYDO UDWHV +RZH DQG 9DQGH .HUFNKRYH %URQVWHLQ DQG +RIIPDQ (QJOXQG f VHHGKDQGOLQJ WHFKQLTXHV -DQ]HQ D /HYH\ &RUOHWW DQG /XFDV 6WLOHV DQG 5RVVHOOL f DQG HIIHFWV RQ VHHG JHUPLQDWLRQ .UHIWLQJ DQG 5RH &RPSWRQ HW DO 7UDYHVHW DQG :LOOVRQ f EXW IHZ VWXGLHV KDYH FRPSDUHG VHHG VKDGRZV JHQHUDWHG E\ GLIIHUHQW GLVSHUVHUV EXW VHH 7KRPDV HW DO &KDYH]5DPLUH] DQG 6ODFN f RU WKH FRQVHTXHQFHV RI GLVSHUVDO SDWWHUQ RQ UHFUXLWPHQW EXW VHH +RZH D +HUUHUD HW DO 6FKXSS 6FKXSS DQG )XHQWHV f $IWHU GLVSHUVDO VHHG SUHGDWLRQ DQG VHHGOLQJ PRUWDOLW\ DUH RIWHQ H[WHQVLYH IRU WURSLFDO WUHHV 'H6WHYHQ DQG 3XW] +RZH HW DO &KDSPDQ D +DPPRQG &LQWUD DQG +RPD 3HUHV HW DO f DQG PD\ EH LPSRUWDQW LQ LQIOXHQFLQJ WKH VSDWLDO SDWWHUQ RI UHFUXLWPHQW &RQQHOO -DQ]HQ +DUSHU +XEEHOO &ODUN DQG &ODUN %HFNHU HW DO 0F&DQQ\ +RZH f ,Q DGGLWLRQ VWXGLHV RQ IRUHVW G\QDPLFV KDYH VXJJHVWHG WKDW FDQRS\ JDSV DUH RIWHQ FUXFLDO UHFUXLWPHQW VLWHV IRU WUHH VHHGOLQJV +DUWVKRUQ 'HQVORZ 6ZDLQH DQG :KLWPRUH 6ZDLQH f )HZ VWXGLHV KRZHYHU KDYH H[DPLQHG VHHG SUHGDWLRQ DV D OLQN EHWZHHQ VHHG GLVSHUVDO DQG VHHGOLQJ UHFUXLWPHQW ,Q IDFW DVLGH IURP UHVHDUFK RQ 9LUROD LQ 3DQDPD +RZH DQG 9DQGH .HUFNKRYH +RZH HW DO +RZH E Df 3KLOO\UHD LQ 6SDLQ +HUUHUD HW DO -RUGDQR DQG +HUUHUD f DQG )LFXV LQ %RUQHR /DPDQ D Ef LW LV GLIILFXOW WR ILQG VWXGLHV WKDW FRQVLGHU WKH YDULRXV VWDJHV OHDGLQJ WR UHFUXLWPHQW IRU DQ\ VSHFLHV RI IOHVK\IUXLWHG YHUWHEUDWHGLVSHUVHG SODQW )XUWKHUPRUH VRPH SODQW VSHFLHV W\SLFDOO\ FRQVLGHUHG WR EH ELUG RU PDPPDOGLVSHUVHG PD\ KDYH D VHFRQG RU HYHQ WKLUGf VWDJH RI GLVSHUVDO E\ DQ HQWLUHO\ GLIIHUHQW GLVSHUVDO YHFWRU 5REHUWV DQG +HLWKDXV &OLIIRUG DQG 0RQWHLWK )RUJHW DQG 0LOOHURQ (VWUDGD HW DO

PAGE 17

/HYH\ DQG %\UQH 1RJDOHV HW DO f 7R XQGHUVWDQG WKH UHODWLYH LPSRUWDQFH RI VHHG GLVSHUVHUV DQG VHHG SUHGDWRUV LQ IRUHVW G\QDPLFV LW LV QHFHVVDU\ WR VWXG\ HDFK VWDJH OHDGLQJ WR UHFUXLWPHQW DQG LGHDOO\ WR UHSURGXFWLYH DJHf 7KH PDLQ IDFWRU WKDW OLPLWV RXU XQGHUVWDQGLQJ RI WKH OLQN EHWZHHQ VHHG GLVSHUVHUV DQG VHHGOLQJ UHFUXLWPHQW LV WKH GLIILFXOW\ RI ILQGLQJ GLVSHUVHG VHHGV +DUSHU -DQ]HQ D 6FKXSS DQG )XHQWHV f 7KXV PRVW VWXGLHV RQ YHUWHEUDWHGLVSHUVHG SODQWV KDYH DVVHVVHG WKH LPSRUWDQFH IRU SODQW ILWQHVV RI GLIIHUHQW GLVSHUVHUV E\ WKH DPRXQW RI IUXLW LQ WKH GLHW DQG JXW WUHDWPHQW RI VHHGV DQG KDYH QRW FRQVLGHUHG SRVWGLVSHUVDO IDWH RI VHHGV UHYLHZHG E\ +RZH 6WLOHV -RUGDQR :LOOVRQ /HYH\ HW DO f 6LPLODUO\ VWXGLHV RQ VHHG SUHGDWLRQ DQG VHHGOLQJ UHFUXLWPHQW KDYH UHOLHG RQ H[SHULPHQWDOO\ GLVSHUVHG VHHGV XVXDOO\ ZLWKRXW GDWD RQ WKH DFWXDO SDWWHUQV RI VHHG UDLQ 3ULFH DQG -HQNLQV &UDZOH\ +XOPH 6FKXSS f 7KH REMHFWLYH RI WKLV VWXG\ ZDV WR OLQN SDWWHUQV RI VHHG GLVSHUVDO DQG VHHG SUHGDWLRQ ZLWK WKRVH RI VHHGOLQJ UHFUXLWPHQW E\ GHWHUPLQLQJ WKH ORFDWLRQV RI QDWXUDOO\ GLVSHUVHG VHHGV DQG WKHQ IROORZLQJ WKH IDWH RI WKRVH VHHGV DIWHU GLVSHUVDO VWXGLHG D FRPPRQ WUHH FRWHD HQGUHVLDQD 0H] /DXUDFHDHf LQ ROG JURZWK PRQWDQH IRUHVW LQ &RVWD 5LFD WR DQVZHU WKH IROORZLQJ TXHVWLRQV f 'R WKH ILYH PDLQ IUXLW FRQVXPHUV RI 2 HQGUHVLDQD JHQHUDWH GLIIHUHQW SDWWHUQV RI VHHG UDLQ DQG VXEVHTXHQWO\ LQIOXHQFH JHUPLQDWLRQ DQG WKH SDWWHUQ RI VHHGOLQJ UHFUXLWPHQW" f :KDW SURSRUWLRQ RI WKH GLVSHUVHG VHHGV DUH VXEVHTXHQWO\ NLOOHG E\ VHHG SUHGDWRUV DQG ZKDW DQLPDOV DUH WKH PRVW LPSRUWDQW VHHG SUHGDWRUV" f 'R VFDWWHUKRDUGLQJ URGHQWV SURYLGH D VHFRQG VWDJH RI GLVSHUVDO" f :KDW SURSRUWLRQ RI WKH GLVSHUVHG VHHGV JHUPLQDWH DQG HVWDEOLVK DV VHHGOLQJV XQGHU GLIIHUHQW PLFURKDELWDW FRQGLWLRQV" 7R DQVZHU WKHVH TXHVWLRQV XVHG D FRPELQDWLRQ RI QDWXUDO DQG PDQLSXODWLYH H[SHULPHQWV WR DVVHVV WKH W\SHV RI VLWHV WR ZKLFK VHHGV DUH GLVSHUVHG DQG KRZ WKH KDELWDW FKDUDFWHULVWLFV DQG VSDWLDO DUUDQJHPHQW RI WKRVH VLWHV UHODWLYH WR WKH SDUHQW WUHHV LQIOXHQFH SRVWGLVSHUVDO VXUYLYDO

PAGE 18

$ VHFRQGDU\ REMHFWLYH ZDV WR DVVHVV WKUHH K\SRWKHVL]HG DGYDQWDJHV RI GLVSHUVDO f HVFDSH IURP KLJK PRUWDOLW\ FDXVHG E\ GLVWDQFH RU GHQVLW\GHSHQGHQW IDFWRUV QHDU FRQVSHFLILFV HVFDSH K\SRWKHVLVf f FRORQL]DWLRQ RI UDUH XQSUHGLFWDEOH HSKHPHUDO VLWHV VXFK DV WUHHIDOO JDSV FRORQL]DWLRQ K\SRWKHVLVf DQG f GLUHFWHG GLVSHUVDO WR SDUWLFXODU PLFURKDELWDWV VXLWDEOH IRU VXUYLYDO GLUHFWHG GLVSHUVDO K\SRWKHVLVf +RZH DQG 6PDOOZRRG +RZH :LOOVRQ f $ IRXUWK DGYDQWDJH RI GLVSHUVDO JHQH IORZ /HYLQ DQG .HUVWHU +DPULFN DQG 1DVRQ f ZDV EH\RQG WKH VFRSH RI WKLV VWXG\ 7KHVH K\SRWKHVHV DUH QRW PXWXDOO\ H[FOXVLYH DQG FDQ EH DVVHVVHG RQO\ LI GLVSHUVDO VLWHV DQG SRVWn GLVSHUVDO IDWHV DUH NQRZQ (YLGHQFH IRU WKH HVFDSH K\SRWKHVLV VKRXOG LQFOXGH KLJKHU UDWHV RI VHHG SUHGDWLRQ DQGRU VHHGOLQJ PRUWDOLW\ FORVHU WR WKH SDUHQW WUHH RU ZKHUH VHHGV DQG VHHGOLQJV RFFXU LQ GHQVH FRQFHQWUDWLRQV 7KH GLIIHUHQFH EHWZHHQ FRORQL]DWLRQ DQG GLUHFWHG GLVSHUVDO LV ZKHWKHU VHHGV DUULYH DQG VXUYLYH LQ VSHFLILF KDELWDWV PRUH RIWHQ WKDQ H[SHFWHG E\ FKDQFH +RZH DQG 6PDOOZRRG +RZH 0XUUD\ 6FKXSS HW DO :LOOVRQ f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f 6WXG\ 6LWH 7KLV VWXG\ ZDV FRQGXFWHG IURP 0D\ WR -XO\ LQ WKH 0RQWHYHUGH &ORXG )RUHVW 3UHVHUYH rf1 r:f LQ WKH &RUGLOOHUD GH 7LODUDQ QRUWKHUQ &RVWD 5LFD 7KLV KD SUHVHUYH LV DGMDFHQW WR RWKHU SURWHFWHG ODQGV HQFRPSDVVLQJ RQH RI WKH ODUJHVW DUHDV RI UHODWLYHO\ XQEURNHQ IRUHVW LQ &RVWD 5LFD 7KH VWXG\ DUHD FRQWDLQV WKH IXOO FRPSOHPHQW RI ELUGV DQG PDPPDOV LQFOXGLQJ WRS SUHGDWRUV WKDW KDYH KLVWRULFDOO\ RFFXUUHG

PAGE 19

LQ WKH DUHD
PAGE 20

0RVW ZRRG\ SODQWV DW 0RQWHYHUGH DUH DQLPDOGLVSHUVHG DV LV W\SLFDO RI 1HRWURSLFDO IRUHVWV LQ JHQHUDO DQG WURSLFDO PRQWDQH IRUHVWV LQ SDUWLFXODU *HQWU\ +RZH DQG 6PDOOZRRG 7DQQHU 6WLOHV /HYH\ DQG 6WLOHV f %LUG GLVSHUVDO SUHGRPLQDWHV b RI WKH XQGHUVWRU\ WUHHV DQG VKUXEV DQG b RI WKH FDQRS\ WUHHV KDYH IUXLW PRUSKRORJ\ VXJJHVWLQJ GLVSHUVDO E\ ELUGV 6WLOHV f $W OHDVW UHVLGHQW DQG PLJUDQW ELUG VSHFLHV IHHG RQ WKH IUXLWV RI RYHU VSHFLHV RI WUHHV DQG VKUXEV :KHHOZULJKW HW DO f 0RVW RI WKHVH ELUGV UHJXUJLWDWH RU GHIHFDWH VHHGV LQWDFW :KHHOZULJKW HW DO 0XUUD\ f 7KH DYHUDJH DQQXDO UDLQIDOO DW P RQ WKH 3DFLILF VORSH DERXW NP IURP WKH VWXG\ VLWH LV DSSUR[LPDWHO\ PP ZLWK PRVW RI WKH SUHFLSLWDWLRQ RFFXUULQJ EHWZHHQ 0D\ DQG 1RYHPEHU $FWXDO UDLQIDOO LQ WKH VWXG\ VLWH ZDV SUREDEO\ JUHDWHU WKDQ PP EXW WKH VHDVRQDO SDWWHUQ ZDV VLPLODU 1DGNDPL DQG :KHHOZULJKW LQ SUHVVf 5DQJH JDXJHV XQGHUHVWLPDWH WKH DPRXQW RI SUHFLSLWDWLRQ IURP PLVW DQG FORXG LQWHUFHSWLRQ ZKLFK FRQWULEXWH XS WR b RI WKH SUHFLSLWDWLRQ LQ VRPH 1HRWURSLFDO PRQWDQH IRUHVWV &DYHOLHU f 7HPSHUDWXUHV UHFRUGHG DW WKH VWXG\ VLWH GXULQJ WKLV SURMHFW UDQJHG IURP r WR r& 6WXG\ 6SHFLHV 7KH /DXUDFHDH LV DQ LPSRUWDQW IDPLO\ LQ 1HRWURSLFDO IRUHVWV LQ WHUPV RI VSHFLHV ULFKQHVV D IRRG UHVRXUFH IRU ELUGV DQG HFRQRPLF YDOXH :KHHOZULJKW D %XUJHU DQG YDQ GHU :HUII *HQWU\ 0DUWLQH]5DPRV DQG 6RWR&DVWUR *XLQGRQ f 0HPEHUV RI WKH /DXUDFHDH DUH DOVR LPSRUWDQW GLHWDU\ FRPSRQHQWV IRU IUXJLYRURXV ELUGV LQ $IULFD 6RXWKHDVW $VLD DQG $XVWUDOLD &URPH 6QRZ 6XQ HW DO f 2FRWHD HQGUHVLDQD 0H] OLVWHG DV 2 DXVWLQLL LQ :KHHOZULJKW HW DO :KHHOZULJKW D f LV D FRPPRQ FDQRS\ WUHH LQ PRQWDQH IRUHVWV LQ FHQWUDO DQG QRUWKZHVWHUQ &RVWD 5LFD IURP P %XUJHU DQG YDQ GHU :HUII f ,Q WKH 0RQWHYHUGH DUHD LW RFFXUV IURP P DORQJ WKH FRQWLQHQWDO GLYLGH 7ZHQW\ WR

PAGE 21

RWKHU VSHFLHV RI /DXUDFHDH RFFXU LQ WKH VDPH DUHD :KHHOZULJKW D +DEHU f 2FRWHD HQGUHVLDQD EHJLQV IORZHULQJ LQ $XJXVW WKH PLGUDLQ\ VHDVRQ DQG LV SROOLQDWHG E\ VPDOO IOLHV DQG RWKHU LQVHFWV )UXLWV ULSHQ WKH IROORZLQJ 0D\ DQG -XQH WKH HDUO\ UDLQ\ VHDVRQ 5LSH IUXLWV [ PP Jf KDYH EOXHEODFN VNLQ OLSLGULFK SXOS DQG DUH KHOG LQ D VKDOORZ UHGGLVK UHFHSWDFOH W\SLFDO RI WKH /DXUDFHDH JHQHUD 2FRWHD DQG 1HFWDQGUD :KHHOZULJKW HW DO %XUJHU DQG YDQ GHU :HUII f 0RVW RI WKH YROXPH RI WKH IUXLW LV D VLQJOH VHHG [ PP J IUHVK ZHLJKWf FRPSRVHG RI D VPDOO HPEU\R DQG WZR ODUJH FRW\OHGRQV VXUURXQGHG E\ D WKLQ PPf VHHG FRDW $OWKRXJK VPDOO FRPSDUHG WR VRPH RWKHU /DXUDFHDH VSHFLHV 2 HQGUHVLDQD IUXLWV DQG VHHGV DUH DPRQJ WKH ODUJHVW DW 0RQWHYHUGH :KHHOZULJKW HW DO f 7KLUW\HLJKW 2 HQGUHVLDQD WUHHV ZHUH LQ WKH VWXG\ DUHD PRVW GDWD LQ WKLV VWXG\ DUH IURP WUHHV LQ WKH FRUH KD VWXG\ VLWH )LJ f )UXLWV DQG VHHGV IRU VRPH H[SHULPHQWV ZHUH FROOHFWHG IURP WUHHV RQ WKH SHULSKHU\ RI WKH PDLQ VWXG\ DUHD /DUJH IUXLW FURSV s IUXLWVWUHHf ZHUH SURGXFHG LQ DQG EXW QRW LQ (DFK 2 HQGUHVLDQD WUHH ZDV DQ DYHUDJH RI P s f IURP WKH WKUHH FORVHVW FRQVSHFLILF DGXOWV 7KH PRVW FRPPRQ DYLDQ YLVLWRUV WR IUXLWLQJ /DXUDFHDH LQFOXGH WKH ODUJHVW DQG PRVW IUXJLYRURXV VSHFLHV DW 0RQWHYHUGH :KHHOZULJKW HW DO f 2FRWHD HQGUHVLDQD IUXLWV DUH HDWHQ SULPDULO\ E\ ILYH VSHFLHV RI ELUGV HPHUDOG WRXFDQHW 5DPSKDVWLGDH $XODFRUK\QFKXV SUDVLQXVf UHVSOHQGHQW TXHW]DO 7URJRQLGDH 3KDURPDFKUXV PRFLQQRf WKUHHZDWWOHG EHOOELUG &RWLQJLGDH 3URFQLDV WULFDUXQFXODWDf PRXQWDLQ URELQ 7XUGLGDH 7XUGXV SOHEHMXVf DQG EODFN JXDQ &UDFLGDH &KDPDHSHWHV XQLFRORUf DOO RI ZKLFK EUHHG LQ WKH VWXG\ VLWH GXULQJ WKH IUXLWLQJ VHDVRQ 7KH ILUVW IRXU VSHFLHV VZDOORZ IUXLWV LQWDFW DQG UHJXUJLWDWH VHHGV LQ YLDEOH FRQGLWLRQ 6HHGV LQ WKLV VL]H UDQJH DUH JHQHUDOO\ UHJXUJLWDWHG PLQXWHV DIWHU LQJHVWLRQ :KHHOZULJKW f *XDQ GHIHFDWH VHHGV LQ YLDEOH FRQGLWLRQ :KHHOZULJKW f 7KHVH ELUG VSHFLHV HVSHFLDOO\ TXHW]DOVf W\SLFDOO\ UHPDLQ LQ D IUXLWLQJ WUHH DIWHU HDWLQJ VHYHUDO IUXLWV DQG RIWHQ UHJXUJLWDWH VHHGV XQGHU WKH VDPH WUHH RU QHDUE\ :KHHOZULJKW f 6HHG SURFHVVLQJ WLPHV IRU JXDQV DUH XQNQRZQ EXW WKH\

PAGE 22

)LJXUH 0DS RI WKH VWXG\ VLWH VKRZLQJ ORFDWLRQV RI IUXLWLQJ 2FRWHD HQGUHVLDQD WUHHV GDUN FLUFOHV ZLWK QXPEHUVf DQG EHOOELUG VRQJ SHUFKHV ;f 7KLFN OLQHV LQGLFDWH VWUHDPV DQG WKLQ OLQHV DUH WUDLOV *ULG OLQHV DUH P DSDUW 1RUWK LV DW WKH WRS 'DWD RQ VHHG ORFDWLRQV IRU HDFK RI WKH WUHHV QXPEHUHG RQ WKLV PDS DUH SUHVHQWHG LQ WKH $SSHQGL[

PAGE 24

JHQHUDOO\ OHDYH D IUXLWLQJ WUHH EHIRUH GHIHFDWLQJ WKH VHHGV IURP WKDW IRUDJLQJ ERXW :HQQ\ SHUVRQDO REVHUYDWLRQf 6HYHUDO RWKHU ELUG VSHFLHV :KHHOZULJKW HW DO f DV ZHOO DV VSLGHU PRQNH\V $WHOHV JHRIIUR\Lf RFFDVLRQDOO\ HDW 2 HQGUHVLDQD IUXLWV DQG SUREDEO\ GLVSHUVH YLDEOH VHHGV HJ &KDSPDQ Df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f 6LWHV ZHUH FODVVLILHG DV QRQGLVSHUVHG LI GLUHFWO\ XQGHU WKH FURZQ RI D IUXLWLQJ 2 HQGUHVLDQD WUHH RU DV GLVSHUVHG LI QRW XQGHU VXFK D WUHH (YHQ WKRXJK DOO GLVSHUVHG DQG QRQGLVSHUVHG VHHGV LQ WKLV VWXG\ KDG EHHQ UHJXUJLWDWHG RU GHIHFDWHG E\ ELUGV DQG WKHUHIRUH GLVSHUVHG VHQVX -DQ]HQ Df LW LV JHQHUDOO\ EHOLHYHG WKDW VHHGV GHSRVLWHG XQGHU WKH SDUHQW WUHHV KDYH YHU\ OLWWOH FKDQFH RI VXUYLYDO HJ -DQ]HQ +RZH HW DO f 7KHUHIRUH WKH WHUP GLVSHUVHG LQ WKLV SDSHU DOZD\V UHIHUV WR VHHGV WKDW DUH QRW GHSRVLWHG GLUHFWO\ XQGHU WKH FURZQ RI D IUXLWLQJ FRQVSHFLILF DQG QRQ GLVSHUVHG UHIHUV WR VHHGV WKDW DUH UHJXUJLWDWHG RU GHIHFDWHG E\ ELUGV GLUHFWO\ XQGHU WKH FURZQ RI D IUXLWLQJ WUHH )RU FRQYHQLHQFH ZLOO UHIHU WR VLWHV RI GLVSHUVHG VHHGV DV GLVSHUVHG VLWHV DQG RI QRQGLVSHUVHG VHHGV DV QRQGLVSHUVHG VLWHV HYHQ WKRXJK WKH VLWHV WKHPVHOYHV ZHUH QRW FDSDEOH RI EHLQJ GLVSHUVHG

PAGE 25

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f /HDI OLWWHU ZDV WKH QXPEHU RI OHDYHV SLHUFHG E\ D PHWDO VWDNH WKUXVW LQWR WKH VRLO DW WKH VLWH 9HJHWDWLRQ GHQVLW\ ZDV WKH QXPEHU RI VWHPV ZLWKLQ D FP UDGLXV RI WKH VLWH 7KH GLVWDQFHV WR WKH QHDUHVW ZRRG\ VWHP P LQ KHLJKW WUHH FP '%+ GLDPHWHU DW EUHDVW KHLJKWf WUXQN RI IUXLWLQJ 2FRWHD WUHH DQG IDOOHQ ORJ ZHUH PHDVXUHG ZLWK D ILEHUJODVV PHDVXULQJ WDSH 7KHVH YDULDEOHV ZHUH VHOHFWHG EDVHG RQ WKHLU GHPRQVWUDWHG LPSRUWDQFH LQ SUHYLRXV VWXGLHV 6HHG VL]H PD\ LQIOXHQFH WKH SUREDELOLW\ RI VHHG SUHGDWLRQ 3ULFH DQG -HQNLQV +XOPH f RU VHHGOLQJ VL]H +RZH DQG

PAGE 26

5LFKWHU f &DQRS\ FRYHU OLJKW DYDLODELOLW\f LV NQRZQ WR EH DQ LPSRUWDQW IDFWRU IRU JHUPLQDWLRQ DQG WURSLFDO VHHGOLQJ JURZWK +RZH HW DO 0XONH\ HW DO 6ZDLQH f 9HJHWDWLRQ GHQVLW\ DQG GLVWDQFH WR REMHFWV PD\ LQIOXHQFH URGHQW DFWLYLW\ DQG VHHG SUHGDWLRQ 6P\WKH .LOWLH .LWFKLQJV DQG /HYH\ f )LQDOO\ OHDI OLWWHU PD\ LQIOXHQFH VHHG SUHGDWLRQ RU JHUPLQDWLRQ 6FKXSS D 0RORIVN\ DQG $XJVSXUJHU 0\VWHU DQG 3LFNHWW f 3RVWGLVSHUVDO 6HHG 3UHGDWLRQ $W HDFK GLVSHUVHG QRQGLVSHUVHG DQG UDQGRP VLWH D PDUNHG VHHG ZDV XVHG WR DVVHVV UDWH RI VHHG SUHGDWLRQ WR LGHQWLI\ VHHG SUHGDWRUV DQG WR GHWHUPLQH LI VHFRQGDU\ GLVSHUVDO RFFXUUHG )RU WKLV WUHDWPHQW XVHG VHHGV WKDW ZHUH UHJXUJLWDWHG E\ ELUGV DQG FROOHFWHG XQGHU IUXLWLQJ WUHHV DGMDFHQW WR WKH VWXG\ VLWH 6HHGV ZHUH PDUNHG E\ JOXLQJ FP RI XQZD[HG GHQWDO IORVV WR WKH VHHG DQG W\LQJ DERXW FP RI IODJJLQJ WDSH WR WKH GLVWDO HQG RI WKH IORVV %HFDXVH WKH JOXH KHOG EHVW RQ VHHGV ZLWK D GU\ VHHG FRDW VHHGV ZHUH WDNHQ LQVLGH D ODE URRP DQG DOORZHG WR GU\ IRU K EHIRUH JOXLQJ (DFK PDUNHG VHHG ZDV SODFHG DW D VLWH WKH QH[W PRUQLQJ 7R GHWHUPLQH LI SUHVHQFH RI GHQWDO IORVV DQG IODJJLQJ WDSH LQIOXHQFHG VHHG UHPRYDO FRQGXFWHG D SLORW VWXG\ LQ 0D\ )LIW\ PDUNHG VHHGV DQG XQPDUNHG VHHGV ZHUH SODFHG VLQJO\ DW UDQGRP ORFDWLRQV DQG FHQVXVHG RQH ZHHN ODWHU 5HPRYDO RI PDUNHG DQG XQPDUNHG VHHGV ZDV QRW VLJQLILFDQWO\ GLIIHUHQW DQG VHHGV UHPRYHG UHVSHFWLYHO\ ; GI 3 f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

PAGE 27

2OIDFWRU\ FXHV OHIW E\ PDUNLQJ DQG KDQGOLQJ WKH VHHGV DUH DQRWKHU SRVVLEOH FRQIRXQGLQJ IDFWRU EXW WKH IUHTXHQW UDLQV SUREDEO\ GLPLQLVKHG WKHVH :KHODQ HW DO f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f 7KH GLVWDQFH IURP WKH GLVSHUVDO VLWH WR WKH SUHGDWLRQ VLWH ZDV PHDVXUHG DQG HDFK VLWH RI D UHPRYHG VHHG ZDV FODVVLILHG LQ WKH IROORZLQJ FDWHJRULHV LQ EXUURZ LQ XQGHU RU QHDU D IDOOHQ ORJ FP GLDPHWHUf DW WKH EDVH RI D ODUJH WUHH LQ GHQVH YHJHWDWLRQ GHILQHG DV DW OHDVW b FRYHU RI SODQWV OHVV WKDQ FP WDOO ZLWKLQ FP UDGLXV RI VLWH DVVHVVHG YLVXDOO\f XQGHU RU EHVLGH D FOXPS RI IDOOHQ EUDQFKHV RU RQ WKH OHDI OLWWHU ,I D PDUNHG VHHG ZDV UHPRYHG EXW QRW HDWHQ LW ZDV OHIW LQ WKH QHZ ORFDWLRQ DQG LQFOXGHG LQ VXEVHTXHQW FHQVXVHV 'LVWDQFH HIIHFW 7ZR RWKHU H[SHULPHQWV RQH LQ DQG RQH LQ f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

PAGE 28

DV RWKHU VWXGLHV HJ +RZH HW DO f LQGLFDWHG WKDW WKH GLVWDQFH HIIHFW FDQ EH GHWHFWHG EH\RQG UD IURP FRQVSHFLILF DGXOW ,Q WKLV FDVH WKH FRPSDVV GLUHFWLRQV ZHUH UDQGRPO\ VHOHFWHG DOWKRXJK VRPH GLUHFWLRQV ZHUH GLVFDUGHG LI WKH P WUHDWPHQW ZDV QRW DW OHDVW P IURP DOO IUXLWLQJ FRQVSHFLILFV 7KLV H[SHULPHQW ZDV UXQ WZLFH RQFH LQ WKH HDUO\ IUXLWLQJ VHDVRQ ODWH 0D\f DQG RQFH WZR ZHHNV ODWHU GXULQJ WKH SHDN IUXLWLQJ VHDVRQ PLG-XQHf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f QR URGHQWV H[FORVXUHV PDGH RI FPA JDOYDQL]HG ZLUH PHVK P WDOO f VPDOO URGHQWV RQO\ H[FORVXUHV PDGH RI FKLFNHQ ZLUH PHVK ZLWK FP KROHV RU P WDOO DQG f DOO URGHQWV FRQWURO SORWV ZLWK RQO\ ZRRGHQ VWDNHV PDUNLQJ WKH FRPHUV 7KH ERWWRP HGJHV RI WKH ZLUH H[FORVXUHV ZHUH EXULHG FP EHORZ JURXQG DQG KHOG ZLWK WZR RU WKUHH FP PHWDO VWDNHV RQ HDFK VLGH 7KH FRPHUV ZHUH VXSSRUWHG E\ P OHQJWKV RI PP WKLFN PHWDO VWDNHV 7KH WRSV ZHUH RSHQ WR DOORZ QRUPDO DFFXPXODWLRQ RI IDOOHQ OHDYHV (DFK VHW ZDV ORFDWHG ZKHUH FRQYHQLHQW DYRLGLQJ WUHHV DQG IDOOHQ ORJVf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

PAGE 29

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f RU UHPRYHG WKH HQWLUH VKRRW KHUELYRUHVf 6RPH VHHGV WKDW DSSHDUHG WR KDYH JHUPLQDWHG ZHUH NLOOHG E\ EHHWOH ODUYDH +HLOLSXV VS &XUFXOLRQLGDHf GHYHORSLQJ LQ WKH VHHG ,QVHFWNLOOHG VHHGV IUHTXHQWO\ GHYHORSHG D URRW EXW QHYHU KDG D VKRRW FP WDOO 6HHGOLQJV NLOOHG E\ IXQJDO SDWKRJHQV ZHUH FKDUDFWHUL]HG E\ D ZLOWHG DQG GLVFRORUHG VKRRW DERYH DERXW FP $XJVSXUJHU f 3K\VLFDO GDPDJH FRQVLVWHG RI WUDPSOLQJ E\ SHFFDULHV DQG RWKHU ODUJH PDPPDOV RU GDPDJH IURP IDOOLQJ WUHHV EUDQFKHV RU ODUJH OHDYHV &ODUN DQG &ODUN f 8QNQRZQ FDXVHV RI PRUWDOLW\ LQFOXGHG FDVHV WKDW ILW PRUH WKDQ RQH FDWHJRU\ ZKHUH WKH VHTXHQFH RI HYHQWV FRXOG QRW EH GHWHUPLQHG DQG FDVHV WKDW GLG QRW FOHDUO\ IDOO LQWR DQ\ FDWHJRU\ $ VHHG ZDV FRQVLGHUHG DOLYH LI WKH VHHG UHPDLQHG ILUP HYHQ LI WKH VKRRW KDG EHHQ HDWHQ RU RWKHUZLVH GDPDJHG 6XFK VHHGV UHVSURXWHG UHSHDWHGO\ :HQQ\ SHUVRQDO REVHUYDWLRQf 6HHGV IURP WKDW KDG QRW JHUPLQDWHG RU UHVSURXWHG DIWHU RQH \HDU ZHUH FXW RSHQ DQG FODVVLILHG DV LQVHFWNLOOHG LI ILOOHG ZLWK IUDVV RU YLDEOH LI WKH HPEU\R DQG FRW\OHGRQV ZHUH QRW GLVFRORUHG RU PHDO\ LQ DSSHDUDQFH *HUPLQDWLRQ WULDOV 7R GHWHUPLQH WKH HIIHFW RQ HQGUHVLDQD JHUPLQDWLRQ RI LQJHVWLRQ E\ ELUGV DQG VHHG EXULDO ZKLFK LQLWLDOO\ WKRXJKW ZRXOG DSSO\ WR VHHGV WDNHQ E\ PDPPDOVf WULDOV ZHUH FRQGXFWHG LQ D JUHHQKRXVH FRQVWUXFWHG RI Q\ORQ ZLQGRZ VFUHHQ RYHU D [ P ZRRGHQ IUDPH RQ D FRQFUHWH IRXQGDWLRQ VHW FP LQWR WKH JURXQG 7KH URRI ZDV FRUUXJDWHG SODVWLF 6HHGV ZHUH SODQWHG LQGLYLGXDOO\ LQ FDUGERDUG PLON FDUWRQV DQG FXW SODVWLF ERWWOHV RI YDULRXV VL]HV $OO FRQWDLQHUV ZHUH ZDVKHG ZLWK KRW VRDS\ ZDWHU DQG GULHG

PAGE 30

LQ WKH VXQ EHIRUH XVH 6RLO ZDV FROOHFWHG IURP D QHDUE\ VHFRQGDU\ IRUHVW DQG PL[HG ZLWK VDQG 7KUHH VHHG WUHDWPHQWV ZHUH FRPSDUHG f QDWXUDOO\ UHJXUJLWDWHG VHHGV FROOHFWHG XQGHU IUXLWLQJ WUHHV RU DORQJ WKH WUDLO WR WKH VWXG\ VLWH f VHHGV UHPRYHG E\ KDQG IURP ULSH IUXLWV FROOHFWHG XQGHU IUXLWLQJ WUHHV DQG f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f ZHUH PHDVXUHG DQG PDSSHG LQ SDLUHG [ P SORWV )RU HDFK RI WUHHV RQH SORW ZDV ORFDWHG QHDU WKH WUHH ZLWK DSSUR[LPDWHO\ KDOI RI WKH SORW GLUHFWO\ XQGHU WKH FURZQ 7KH VHFRQG SORW ZDV ORFDWHG WR P IURP WKH HGJH RI WKH ILUVW SORW DQG DW OHDVW P IURP WKH FURZQ HGJH 1R SORWV ZHUH ORFDWHG LQ UHFHQW \Uf FDQRS\ JDSV RU DORQJ VWUHDPV 6WDWLVWLFDO $QDO\VHV 7KH LQIOXHQFH RI WKH PLFURKDELWDW YDULDEOHV RQ UHPRYDO RI PDUNHG VHHGV DW GD\ RQO\f DQG ZHHNV DQG RQ JHUPLQDWLRQ VHHGOLQJ HVWDEOLVKPHQW DQG \HDU VXUYLYDO RI FDJHG VHHGV ZHUH H[DPLQHG ZLWK PXOWLSOH ORJLVWLF UHJUHVVLRQ XVLQJ 6 $6 -03 6$6 ,QVWLWXWH f )RU HDFK RI WKH VHYHQ ELQRPLDO OLYH RU GHDGf UHVSRQVH YDULDEOHV PHQWLRQHG DERYH WKH PRGHO ZDV UXQ ILUVW ZLWK DOO SUHGLFWRU YDULDEOHV DQG WKHQ ZLWK DQG ZLWKRXW HDFK SUHGLFWRU YDULDEOH WR GHWHUPLQH LI WKH GHOHWLRQ RI D JLYHQ YDULDEOH KDG D VLJQLILFDQW HIIHFW RQ WKH DPRXQW RI YDULDWLRQ H[SODLQHG 7KLV GHOHWLRQ SURFHGXUH ZDV UHSHDWHG XQWLO RQO\

PAGE 31

VLJQLILFDQW SUHGLFWRUV ZHUH UHWDLQHG 7UH[OHU DQG 7UDYLV f 0RGHOV ZHUH VHOHFWHG PDQXDOO\ WR DYRLG WKH SUREOHPV RI DXWRPDWLF VWHSZLVH SURFHGXUHV -DPHV DQG 0F&XOORFK f 7KH SUHGLFWRU YDULDEOHV LQFOXGHG WKUHH PHDVXUHV RI VHHG VL]H OHQJWK ZLGWK DQG PDVVf HLJKW PLFURKDELWDW FKDUDFWHULVWLFV OHDI OLWWHU FDQRS\ FRYHU QXPEHU RI VWHPV DQG GLVWDQFHV WR QHDUHVW FDJHG VHHG KHUEDFHRXV VWHP ZRRG\ VWHP FP WUHH SDUHQW WUHH WUXQN DQG IDOOHQ ORJf GDWH RI GLVSHUVDO -XOLDQ GDWHf DQG WUHH QXPEHU 2QO\ VLQJOH WHUPV ZHUH LQFOXGHG LQ WKH PRGHOV DV ODFNRIILW WHVWV LQGLFDWHG WKDW WKH VLQJOH WHUP PRGHOV ZHUH DGHTXDWH DQG WKXV LQWHUDFWLRQ WHUPV ZHUH QRW UHTXLUHG 6$6 ,QVWLWXWH f 7KH UHODWLRQVKLSV RI WKH PLFURKDELWDW YDULDEOHV ZLWKLQ DQG DPRQJ WKH GLVSHUVHG QRQGLVSHUVHG DQG UDQGRP VLWHV ZHUH H[DPLQHG ZLWK SULQFLSDO FRPSRQHQWV DQDO\VLV 352& )$&725f IURP WKH 6$6 VWDWLVWLFDO SDFNDJH 6$6 ,QVWLWXWH f 7KH DYHUDJH ORDGLQJV IRU WKH WKUHH W\SHV RI VLWHV ZHUH FRPSDUHG ZLWK RQHZD\ DQDO\VLV RI YDULDQFH IURP 6XSHU $QRYD $EDFXV &RQFHSWV f 7\SH +, VXPV RI VTXDUHV ZHUH XVHG WR FRPSHQVDWH IRU WKH XQHTXDO VDPSOH VL]HV 6KDZ DQG 0LWFKHOO2OGV f 5HPRYDO UDWHV ZHUH FRPSDUHG DPRQJ WUHDWPHQWV ZLWK VXUYLYDO DQDO\VLV DQG *HKDQ:LOFR[RQ WHVWV IURP 6$6 -03 6$6 ,QVWLWXWH f 7KH *HKDQ:LOFR[RQ WHVW SODFHV JUHDWHU ZHLJKW RQ HDUO\ HYHQWV WKDQ ODWHU HYHQWV DQG ZDV GHHPHG DSSURSULDWH IRU WKLV VWXG\ EHFDXVH PRVW VHHGV ZHUH UHPRYHG GXULQJ WKH ILUVW ZHHN 3\NH DQG 7KRPSVRQ f 3DUDPHWULF WHVWV ZHUH XVHG XQOHVV WKH GDWD YLRODWHG WKH DVVXPSWLRQV RI QRUPDOLW\ DQG HTXDO YDULDQFH LQ ZKLFK FDVH QRQSDUDPHWULF SURFHGXUHV ZHUH XVHG 'DWD LQ WKH IRUP RI SURSRUWLRQV ZHUH DUFVLQ VTXDUHURRW WUDQVIRUPHG EHIRUH DQDO\VLV :KHUH PXOWLSOH FRPSDULVRQV RQ D GDWD VHW ZHUH LQYROYHG WKH DOSKD YDOXH ZDV DGMXVWHG DFFRUGLQJ WR WKH QXPEHU RI FRPSDULVRQV SODQQHG %RQIHUURQL WHFKQLTXH +ROP f 'DWD IURP WKH WZR \HDUV ZHUH YHU\ VLPLODU DQG ZHUH FRPELQHG IRU DQDO\VHV LQ ZKLFK VDPSOH VL]HV ZRXOG KDYH EHHQ ORZ RWKHUZLVH 7KURXJKRXW WKLV SDSHU PHDQ YDOXHV DUH IROORZHG E\ s 6'

PAGE 32

5HVXOWV 6HHG 'LVSHUVDO ,Q VHHGV UHJXUJLWDWHG RU GHIHFDWHG E\ ELUGV ZHUH IRXQG GLVSHUVHG DQG QRQGLVSHUVHG ZKLOH LQ GLVSHUVHG DQG QRQGLVSHUVHG VHHGV ZHUH IRXQG 1LQHW\ILYH DQG UDQGRP VLWHV ZHUH HVWDEOLVKHG LQ DQG UHVSHFWLYHO\ 7ZHQW\IRXU SHUFHQW RI WKH VHHGV DQG b RI WKH VHHGV ZHUH IRXQG E\ IROORZLQJ RU REVHUYLQJ ELUGV 1 f DQG WKH UHVW 1 f E\ VHDUFKLQJ WKH JURXQG 7KH GLVSHUVHG VHHGV ZHUH PRVW FRPPRQ ZLWKLQ P RI WKH FURZQ HGJH b bf EXW VRPH VHHGV ZHUH DV IDU DV P DZD\ ,Q UDUH FDVHV bf GLVSHUVHG VHHGV ZHUH ZLWKLQ P RI DQ HQGUHVLDQD WUXQN )LJ f 'HVSLWH WKHLU FORVH SUR[LPLW\ WR WKH SDUHQW WKH\ ZHUH FODVVLILHG DV GLVSHUVHG EHFDXVH WKH\ ZHUH QRW GLUHFWO\ XQGHU WKH SDUHQW FURZQ DV D UHVXOW RI DV\PPHWULFDO FURZQ JHRPHWU\ RU D WLOWHG WUXQN RU ERWKf 5DQGRP VLWHV ZHUH PRUH HYHQO\ GLVWULEXWHG WKDQ GLVSHUVHG RU QRQGLVSHUVHG VLWHV EXW VKRZHG D SHDN DW P LQ ERWK \HDUV )LJ f :LWK GLVSHUVHG DQG QRQGLVSHUVHG VLWHV FRPELQHG WKH VHHG GLVWULEXWLRQ JHQHUDWHG E\ ELUGV LQ ERWK \HDUV LV EHVW GHVFULEHG E\ D ORJDULWKPLF IXQFWLRQ RI GLVWDQFH IURP D IUXLWLQJ FRQVSHFLILF )LJ f 0RVW VHHGV ODQGHG ZLWKLQ P RI WKH IUXLWLQJ WUHHV EXW WKH WDLO RI WKH VHHG GLVWULEXWLRQ FXUYH EH\RQG P DFFRXQWHG IRU b RI WKH VLWHV LQ DQG b LQ 6HHG UDLQ ZDV KLJKO\ YDULDEOH ZKHQ DYHUDJHG DPRQJ WUHHV DQG \HDUV )LJ f ,Q SDUWLFXODU QRWH WKDW VHHG GLVWULEXWLRQV VKRZHG VHFRQGDU\ SHDNV DW DQG P LQ ERWK \HDUV )LJ f 7KHVH SHDNV FRUUHVSRQGHG WR KDELWXDO VRQJ SHUFKHV XVHG E\ EHOOELUGV VHH EHORZf ZKLFK ZHUH ORFDWHG P IURP IUXLWLQJ WUHHV LQ WKH \HDUV WKLV VWXG\ ZDV FRQGXFWHG )LJ f 0RUH LQIRUPDWLRQ RQ WKH ORFDWLRQV RI VHHGV LV SUHVHQWHG LQ WKH $SSHQGL[ 0LFURKDELWDW FKDUDFWHULVWLFV 7KH ORFDWLRQV RI GLVSHUVHG DQG QRQGLVSHUVHG VHHGV DQG VHHGV DW UDQGRP VLWHV GLIIHUHG VLJQLILFDQWO\ LQ ERWK \HDUV ZLWK UHJDUG WR FDQRS\ FRYHU GLVWDQFH WR SDUHQW DQG QXPEHU RI VWHPV 7DEOH f 'LVSHUVHG VHHG VLWHV KDG ORZHU DYHUDJH FDQRS\ FRYHU WKDQ QRQGLVSHUVHG RU UDQGRP VLWHV LQ ERWK \HDUV SRVWKRF 61.

PAGE 33

&2 ,' 8 2 '& 8O P ',67$1&( )520 3$5(17 Pf )LJXUH 7KH GLVWULEXWLRQ RI VHHGV QDWXUDOO\ UHJXUJLWDWHG RU GHIHFDWHG E\ ELUGV GLUHFWO\ XQGHU WKH FURZQ RI D IUXLWLQJ WUHH QRQGLVSHUVHGf DQG DZD\ IURP IUXLWLQJ WUHHV GLVSHUVHGf DQG WKH GLVWULEXWLRQ RI UDQGRP VLWHV DW GLIIHUHQW GLVWDQFH FDWHJRULHV IURP WUXQNV RI IUXLWLQJ WUHHV LQ DERYHf DQG EHORZf 9DOXHV RQ WKH [D[LV UHSUHVHQW WKH PD[LPXP GLVWDQFH IRU HDFK FDWHJRU\ Pf 7KH VHHG VKDGRZ LV WUXQFDWHG DW P EHFDXVH WKH DEXQGDQFH RI 2FRWHD HQGUHVLDQD WUHHV LQ WKH VWXG\ VLWH PDNH VLWHV YHU\ UDUH

PAGE 34

8XL FF FF 84 WR 4 882 ',67$1&( )520 &52:1 ('*( Pf )LJXUH 7KH DYHUDJH QXPEHU 6'f RI QDWXUDOO\ UHJXUJLWDWHG RU GHIHFDWHG VHHGV SHU WUHH 1 f IRU ERWK \HDUV FRPELQHG DW GLIIHUHQW GLVWDQFHV IURP WKH FORVHVW IUXLWLQJ WUHH 'LVWDQFH FODVV LQFOXGHV DOO QRQGLVSHUVHG VHHGV GLUHFWO\ EHORZ SDUHQWDO FURZQV 7KH RWKHU GLVWDQFH FDWHJRULHV P HWFf DUH PHDVXUHG IURP WKH HGJH RI WKH FURZQ RI WKH FORVHVW IUXLWLQJ 2FRWHD HQGUHVLDQD WUHH 7KH YDOXHV IRU HDFK GLVWDQFH FDWHJRU\ UHSUHVHQWV WKH QXPEHU RI VHHGV DW D SDUWLFXODU GLVWDQFH DYHUDJHG DPRQJ DOO WUHHV IRU WKH HQWLUH VWXG\ VLWH

PAGE 35

7DEOH 6XPPDU\ RI RQHZD\ DQDO\VLV RI YDULDQFH WHVWV IRU HDFK KDELWDW YDULDEOH FRPSDUHG DPRQJ VLWHV RI GLVSHUVHG 'f QRQGLVSHUVHG 1f VHHGV DQG UDQGRPO\ ORFDWHG VLWHV 5f LQ DQG $129$ WHVWV KDG D %RQIHUURQLDGMXVWHG DOSKD YDOXH 0HDQ YDOXHV DUH VKRZQ IRU HDFK WUHDWPHQW IRU HDFK YDULDEOH 6WDQGDUG GHYLDWLRQV DUH LQ SDUHQWKHVHV EHORZ HDFK PHDQ /HDI OLWWHU ZDV WKH QXPEHU RI OHDYHV SLHUFHG E\ D PHWDO VWDNH WKUXVW LQWR WKH VRLO DW WKH VLWH &DQRS\ FRYHU ZDV HVWLPDWHG ZLWK D VSKHULFDO GHQVLRPHWHU 9HJHWDWLRQ GHQVLW\ ZDV HVWLPDWHG DV WKH QXPEHU RI VWHPV ZLWKLQ D FP UDGLXV RI WKH VLWH 7KH ODVW IRXU YDULDEOHV DUH GLVWDQFHV WR FORVHVW VWHP FORVHVW ZRRG\ VWHP P KHLJKWf FORVHVW WUHH FP '%+f DQG FORVHVW ORJ FPf :LWKLQ HDFK \HDU DQG YDULDEOH IRU ZKLFK D VLJQLILFDQW GLIIHUHQFH DPRQJ WUHDWPHQWV ZDV GHWHFWHG LQGLFDWHG ZLWK r DIWHU WKH ) YDOXHf UHVXOWV RI SRVWKRF WHVWV 6WXGHQW1HZPDQ.XHOV WHVWVf DUH LQGLFDWHG ZLWK D OHWWHU DIWHU WKH PHDQ 0HDQV IROORZHG E\ GLIIHUHQW OHWWHUV DUH VLJQLILFDQWO\ GLIIHUHQW 3 f 9DULDEOH 1 5 A 1 5 A OHDI OLWWHU f f f f f f f FDQRS\ FRYHU bf D f E f E f r D f E f E f r YHJHWDWLRQ GHQVLW\ D f E f D f r D f E f E f r FORVHVW VWHP FPf f f f D f E f DE f r ZRRG\ VWHP FPf f f f f f f FORVHVW WUHH Pf f f f f f f FORVHVW ORJ Pf f f f f f f r3

PAGE 36

WHVWV ) V f ZKLOH WKH QRQGLVSHUVHG DQG UDQGRP VLWHV ZHUH VLPLODU ^3 f ,Q GLVSHUVHG DQG UDQGRP VLWHV KDG VLPLODUO\ 3 f KLJK QXPEHUV RI VWHPV FRPSDUHG WR QRQGLVSHUVHG VLWHV ^3 f ,Q GLVSHUVHG VLWHV KDG PRUH VWHPV WKDQ ERWK WKH RWKHU VLWHV 3 f ZKLOH UDQGRP DQG QRQGLVSHUVHG VLWHV KDG VLPLODU QXPEHUV RI VWHPV 3 f 7KH RQO\ RWKHU YDULDEOH WKDW GLIIHUHG DPRQJ WKH WUHDWPHQWV ZDV GLVWDQFH WR FORVHVW VWHP LQ 'LVSHUVHG VLWHV ZHUH FORVHU WR VWHPV WKDQ QRQGLVSHUVHG VLWHV 3 f ZKLOH UDQGRP VLWHV GLG QRW GLIIHU IURP WKH RWKHU WZR WUHDWPHQWV 3 f ,Q D SULQFLSDO FRPSRQHQW DQDO\VLV RI WKH GDWD WKH ILUVW WZR FRPSRQHQWV H[SODLQHG b RI WRWDO YDULDQFH ZKLOH IRU WKH ILUVW WZR FRPSRQHQWV DFFRXQWHG IRU b RI WRWDO YDULDQFH 7DEOH f 7KH UHODWLYHO\ ORZ DPRXQW RI YDULDQFH H[SODLQHG LV GXH LQ SDUW WR WKH LQFOXVLRQ RI UDQGRP VLWHV -DFNVRQ f ZKLFK DSSHDUHG DV D VSKHULFDO FORXG RI SRLQWV FHQWHUHG RQ WKH RULJLQ )LJ f 5HPRYDO RQ UDQGRP VLWHV KRZHYHU LQFUHDVHV WKH WRWDO YDULDQFH H[SODLQHG E\ WKH ILUVW WZR FRPSRQHQWV E\ RQO\ b 7KH SDWWHUQ ZDV VLPLODU LQ ERWK \HDUV WKH ILUVW PXOWLYDULDWH D[LV 3& f ZDV FKDUDFWHUL]HG E\ KLJK QHJDWLYH ORDGLQJ RI FDQRS\ FRYHU DQG KLJK SRVLWLYH ORDGLQJV RI OHDI OLWWHU GHSWK DQG FORVHVW FP '%+ WUHH 7DEOH f 7KH VHFRQG D[LV 3& f ZDV FKDUDFWHUL]HG E\ SRVLWLYH ORDGLQJ RI GLVWDQFH WR FORVHVW VWHP LQ ERWK \HDUV DQG QHJDWLYH ORDGLQJ RI YHJHWDWLRQ GHQVLW\ LQ EXW SRVLWLYH ORDGLQJ LQ 7DEOH f 7KH DYHUDJH ORDGLQJV IRU 3& GLIIHUHG VLJQLILFDQWO\ DPRQJ WKH GLVSHUVHG QRQGLVSHUVHG DQG UDQGRP VLWHV LQ ) 3 f DQG ) 3 f (DFK W\SH RI VLWH GLIIHUHG IURP WKH RWKHUV )LVKHUnV /6' WHVWV DOO 3nV f /RDGLQJV IRU 3& GLIIHUHG RQO\ EHWZHHQ GLVSHUVHG DQG QRQGLVSHUVHG VLWHV LQ A 3 f 7KLV DQDO\VLV VKRZV WKDW VLWH FKDUDFWHULVWLFV RI GLVSHUVHG DQG QRQGLVSHUVHG VHHGV ZHUH QRQUDQGRP :KHQ WKH WKUHH WUHDWPHQWV GLVSHUVHG QRQGLVSHUVHG DQG UDQGRPf DUH LGHQWLILHG RQ D SORW RI WKH ILUVW WZR SULQFLSDO FRPSRQHQWV WKH\ VKRZHG EURDG RYHUODS H[FHSW IRU VRPH

PAGE 37

3& 3& Â’ ',63(56(' f 121',63(56(' 5$1'20 3& Â’ ',63(56(' f 121',63(56(' 5$1'20 3& )LJXUH 3ORW RI VLWHV ZLWK GLVSHUVHG RSHQ VTXDUHVf QRQGLVSHUVHG VROLG FLUFOHVf DQG UDQGRPO\ ORFDWHG VHHGV FURVVHVf IURP DERYHf DQG EHORZf RQ WKH ILUVW WZR D[HV GHWHUPLQHG E\ SULQFLSDO FRPSRQHQWV DQDO\VLV RI KDELWDW YDULDEOHV $[LV LV SULPDULO\ D IXQFWLRQ RI FDQRS\ FRYHU OHDI OLWWHU DQG FORVHVW FP '%+ WUHH DQG $[LV LV SULPDULO\ D IXQFWLRQ RI YHJHWDWLRQ GHQVLW\ DQG FORVHVW VWHP VHH 7DEOH f

PAGE 38

7DEOH /RDGLQJV HLJHQYHFWRUVf RI HDFK KDELWDW YDULDEOH RQ WKH ILUVW WZR FRPSRQHQWV GHWHUPLQHG E\ SULQFLSDO FRPSRQHQW DQDO\VLV RI WKH FRUUHODWLRQ PDWUL[ 7KH SURSRUWLRQ RI WRWDO YDULDQFH H[SODLQHG E\ HDFK FRPSRQHQW LV OLVWHG LQ WKH ODVW URZ 3& 3& 3& 3& OHDI 8WWHU FDQRS\ FRYHU VWHPV FORVHVW VWHP ZRRG\ VWHP FP WUHH FORVHVW ORJ YDULDWLRQ b b b b

PAGE 39

GLVSHUVHG VLWHV WKDW KDYH KLJKHU YDOXHV IRU WKH ILUVW FRPSRQHQW 7KHVH FRUUHVSRQG WR VLWHV ZLWK ORZ FDQRS\ FRYHU UHODWLYHO\ IDU IURP WKH SDUHQW WUHHV )LJ f 6HHGV LQ WKH VLWHV ZLWK YDOXHV LQ DQG LQ IRU 3& ZHUH DOO GLVSHUVHG E\ EHOOELUGV LQWR ODUJH JDSV VXUURXQGLQJ VRQJ SHUFKHV $GGLWLRQDOO\ WZR RI WKH VHHGV ZHUH GLVSHUVHG E\ %ODFN *XDQV RQWR ORJV LQ DQRWKHU VPDOOHU JDS %HOOELUGGLVSHUVHG VHHGV VHHGV GLVSHUVHG E\ RWKHU VSHFLHV DQG UDQGRP VLWHV KDG VLJQLILFDQWO\ GLIIHUHQW 3& ORDGLQJV LQ r 3 f DQG ) 3 f %HOOELUG VLWHV KDG KLJKHU IDFWRU ORDGLQJV WKDQ RWKHU VSHFLHVn VLWHV DQG UDQGRP VLWHV 3nV f EXW VLWHV RI WKH RWKHU VSHFLHV DQG UDQGRP VLWHV GLG QRW GLIIHU 3 f &RPELQLQJ WKH VHHGV IURP ERWK \HDUV DQG XVLQJ RQO\ VHHGV IRU ZKLFK WKH GLVSHUVHU ZDV NQRZQ WKH DYHUDJH FDQRS\ FRYHU DW VLWHV RI VHHGV GLVSHUVHG E\ EHOOELUGV ZDV VLJQLILFDQWO\ ORZHU WKDQ RI VLWHV RI VHHGV GLVSHUVHG E\ DOO WKH RWKHU VSHFLHV b DQG b UHVSHFWLYHO\ .UXVNDO:DOOLV WHVW ; GI 3 )LJ f 6LPLODUO\ WKH DYHUDJH GLVWDQFH IURP WKH FORVHVW SDUHQW WUHH RI EHOOELUG VLWHV ZDV JUHDWHU WKDQ DQ\ RI WKH RWKHU IRXU VSHFLHV .UXVNDO:DOOLV ; GI 3 f EXW WKH GLIILFXOW\ LQ IROORZLQJ WKH ELUGV HVSHFLDOO\ URELQV ZKLFK WHQGHG WR IO\ DERYH WKH FDQRS\f ELDVHV WKH UHVXOWV LQ IDYRU RI EHOOELUGV 1HYHUWKHOHVV WKH GDWD VKRZ WKDW EHOOELUGV WHQG QRW WR GURS PDQ\ VHHGV QHDU SDUHQW WUHHV )LJ f &RQVLGHULQJ WKDW DSSUR[LPDWHO\ b RI WKH VWXG\ DUHD ZDV LQ JDSV VHH 6WXG\ 6LWHf RYHUDOO VHHG DUULYDO LQ JDSV bf RFFXUUHG PRUH RIWHQ WKDQ H[SHFWHG E\ FKDQFH ; GI 3 f ZKHUHDV FORVH WR WKH H[SHFWHG QXPEHU RI UDQGRP VLWHV bf ZDV LQ JDSV ; GI 3 VHH )LJ f )XUWKHUPRUH EHOOELUGV GLVSHUVHG VLJQLILFDQWO\ PRUH VHHGV WR JDSV WKDQ GLG WKH RWKHU VSHFLHV ; GI 3 f 1RWH WKDW EHOOELUGV GLVSHUVHG VHHGV WR RQO\ RI JDSV LQ WKH VWXG\ DUHD

PAGE 40

&$123< &29(5 bf Q %(//%,5' *8$1 48(7=$/ 52%,1 728&$1 (7 ',63(56(5 63(&,(6 )LJXUH 0HDQ s 6'f FDQRS\ FRYHU bf RI VLWHV RI VHHG GLVSHUVHG E\ EHOOELUGV JXDQV TXHW]DOV URELQV DQG WRXFDQHWV IRU ERWK \HDUV FRPELQHG 6DPSOH VL]H IRU HDFK VSHFLHV VKRZQ DW WKH ERWWRP RI HDFK EDU

PAGE 41

180%(5 2) 6(('6 Â’ 728&$1(7 Â’ 52%,1 Â’ 48(7=$/ *8$1 %(//%,5' ',67$1&( )520 &52:1 ('*( Pf )LJXUH 7KH QXPEHU RI VHHGV GLVSHUVHG DW GLIIHUHQW GLVWDQFHV IURP SDUHQWDO WUXQNV E\ EHOOELUGV JXDQV TXHW]DOV URELQV DQG WRXFDQHWV IRU ERWK \HDUV FRPELQHG 1 f

PAGE 42

3RVWGLVSHUVDO 6HHG 3UHGDWLRQ 0RUH WKDQ b RI PDUNHG VHHGV ZHUH UHPRYHG ZLWKLQ RQH ZHHN LQ ERWK \HDUV )LJ f ,Q VHHGV DW GLVSHUVHG QRQGLVSHUVHG DQG UDQGRP VLWHV ZHUH UHPRYHG DW VLPLODU UDWHV ZLWK GLVSHUVHG VHHGV VKRZLQJ D QRQVLJQLILFDQW WUHQG IRU VORZHU UHPRYDO WKDQ VHHGV DW QRQGLVSHUVHG RU UDQGRP VLWHV :LOFR[RQ ; GI 3 f ,Q GLVSHUVHG VHHGV ZHUH UHPRYHG PRUH VORZO\ WKDQ WKRVH DW QRQGLVSHUVHG DQG UDQGRP VLWHV :LOFR[RQ ; GI 3 f 8OWLPDWHO\ RQO\ RI PDUNHG VHHGV RQH GLVSHUVHG DQG RQH UDQGRP VLWH ERWK IURP DQG ERWK A P IURP FRQVSHFLILF WUHHVf LQ WKH HQWLUH VWXG\ VXUYLYHG WR -XQH DQG JHUPLQDWHGf IRU DQ RYHUDOO SUHGDWLRQ UDWH RI b IRXQG b RI PDUNHG VHHGV DIWHU UHPRYDO LQ DQG b LQ ,Q DOO FDVHV WKH VHHG ZDV NLOOHG DQG LQ PRVW FDVHV bf HQWLUHO\ FRQVXPHG )HZ VHHGV ZHUH HDWHQ bf EXW QRW PRYHG 7KH JHQHUDO SDWWHUQ ZDV WKDW VHHGV ZHUH WDNHQ D VKRUW GLVWDQFH UDQJH P 1 f WR D VLWH ZLWK FRYHU SUHVXPDEO\ ZKHUH WKH VHHG FRXOG EH HDWHQ LQ VDIHW\ 2Q DYHUDJH UHPRYHG VHHGV QRW FRXQWLQJ VHHGV HDWHQ EXW QRW PRYHGf ZHUH WDNHQ P s 1 f LQ DQG P s 1 f LQ $ERXW b RI WKRVH UHPRYHG ZHUH IRXQG RQ WKH OHDI OLWWHU EXW VRPH ZHUH IRXQG LQ VPDOO EXUURZV FP GLDPHWHUf LQ RU XQGHU IDOOHQ ORJV RU LQ FUHYLFHV DW WKH EDVHV RI ODUJH WUHHV )LJ f 0RVW VHHGV WDNHQ WR EXUURZV WUHHV DQG ORJV ZHUH FOHDUO\ WDNHQ E\ VPDOO DQLPDOV SUREDEO\ URGHQWV EHFDXVH WKH VPDOO GLDPHWHU RI WKH RSHQLQJ ZRXOG H[FOXGH ODUJHU VSHFLHV VXFK DV DJRXWLV 'DV\SURFWD SXQFWDWDf IRXQG QR HYLGHQFH RI VFDWWHUKRDUGLQJ RU VHFRQGDU\ GLVSHUVDO 6LJQLILFDQW SUHGLFWRUV RI VXUYLYDO RI PDUNHG VHHGV WR ZHHNV LQ LQFOXGHG FDQRS\ FRYHU OHDI OLWWHU GHSWK DQG GLVSHUVDO GDWH 7DEOH f 6HHGV LQ VLWHV ZLWK PRUH RSHQ FDQRS\ FRYHU GHHSHU OHDI OLWWHU DQG ODWHU GLVSHUVDO GDWH ZHUH PRUH OLNHO\ WR VXUYLYH ,Q G VXUYLYDO ZDV VLJQLILFDQWO\ SUHGLFWHG E\ JUHDWHU VHHG PDVV HDUOLHU GLVSHUVDO GDWH DQG WKH SDUWLFXODU SDUHQW WUHH $W ZN VHHGV IDUWKHU IURP WKH SDUHQW WUHH GLVSHUVHG

PAGE 43

6(('6 5(0$,1,1* bf 6(('6 5(0$,1,1* bf )LJXUH 5HPRYDO RI PDUNHG VHHGV IURP VLWHV RI GLVSHUVHG RSHQ VTXDUHVf QRQ GLVSHUVHG VROLG VTXDUHVf DQG UDQGRPO\ ORFDWHG FURVVHVf VHHGV LQ DERYHf DQG EHORZf 6DPSOH VL]HV DUH VKRZQ LQ SDUHQWKHVHV

PAGE 44

3(5&(17$*( 2) 6,7(6 )LJXUH 3HUFHQWDJH RI PDUNHG VHHGV IRXQG LQ GLIIHUHQW ORFDWLRQV DIWHU UHPRYDO E\ DQLPDOV 6HH WH[W 0HWKRGV SRVWGLVSHUVDO VHHG SUHGDWLRQf IRU GHILQLWLRQV RI FDWHJRULHV

PAGE 45

7DEOH 5HVXOWV RI ORJLVWLF UHJUHVVLRQV RI SRVWGLVSHUVDO VXUYLYDO RI PDUNHG VHHGV DJDLQVW KDELWDW YDULDEOHV 2QO\ VLJQLILFDQW HIIHFWV DUH OLVWHG 5HVSRQVH U RJ OLNHOLKRRG ;f SUHGLFWRUV ZN VXUYLYDO rrr OHDI OLWWHUrrr FDQRS\ FRYHUr GLVSHUVDO GDWHrrr GD\ VXUYLYDO rrr VHHG PDVVrr GLVSHUVDO GDWHrrr WUHH QXPEHUr ZN VXUYLYDO rrr GLVWDQFH WR SDUHQWrrr GLVSHUVDO GDWHrrr WUHH QXPEHUr r3 rr3 rrr3 DVLJQ LQ IURQW RI HDFK SUHGLFWRU YDULDEOH LQGLFDWHV SRVLWLYH RU QHJDWLYH FRUUHODWLRQ ZLWK WKH UHVSRQVH YDULDEOH

PAGE 46

ODWHU LQ WKH VHDVRQ DQG DW FHUWDLQ SDUHQW WUHHV ZHUH PRUH OLNHO\ WR VXUYLYH /RJLVWLF UHJUHVVLRQ PRGHOV RI VXUYLYDO SDVW ZN ZHUH QRW VLJQLILFDQW EHFDXVH VR IHZ VHHGV VXUYLYHG WKDW ORQJ 'LVWDQFH HIIHFW 6HHG UHPRYDO IURP WUDQVHFWV UDGLDWLQJ IURP IUXLWLQJ 2 HQGUHVLDQD WUHHV ZDV VORZHU IRU VHHGV IDUWKHU IURP WKH SDUHQW WUHHV LQ ERWK DQG 7KH VHHG UHPRYDO H[SHULPHQW ZKLFK UDQ IRU ZHHNV VKRZHG WKDW WKH HIIHFW RI GLVWDQFH RQ VHHG UHPRYDO GHFOLQHG RYHU WLPH DQG WKDW VHHG SUHGDWLRQ HYHQWXDOO\ DSSURDFKHG b DW DOO GLVWDQFHV )LJ Df 2QO\ VHHG DW Pf VXUYLYHG ZN 5HPRYDO RI VHHGV ZDV VLJQLILFDQWO\ IDVWHU IRU VHHGV ZLWKLQ P RI SDUHQWV WKDQ IRU VHHGV SODFHG RU P DZD\ :LOFR[RQ ; GI 3 f 7KH H[SHULPHQW )LJ Ef VKRZHG WKDW VHHG UHPRYDO ZDV VLJQLILFDQWO\ IDVWHU IRU VHHGV H[SHULPHQWDOO\ GLVSHUVHG GXULQJ WKH SHDN RI IUXLWLQJ WKDQ EHIRUH WKH IUXLWLQJ SHDN EXW RQO\ IRU VHHGV P XQGHU WKH FURZQf IURP SDUHQWV ; GI 3 f DQG QRW IRU VHHGV P DZD\ ; GI 3 f 5HPRYDO UDWHV GLG QRW GLIIHU EHWZHHQ VHHGV SODFHG DQG P IURP SDUHQWV DW HLWKHU WLPH LQ WKH IUXLWLQJ VHDVRQ HDUO\ ; GI 3 PLGGOH ; GI 3 f 7KH VDPSOH RI VHHGV ZDV FHQVXVHG DW GXVN f DQG GDZQ f IRU WZR FRQVHFXWLYH GD\V )LYH VHHGV KDG EHHQ UHPRYHG E\ WKH ILUVW GXVN FHQVXV WZR RI ZKLFK FOHDUO\ ZHUH WDNHQ E\ VPDOO URGHQWV LQWR VPDOO KROHV LQ IDOOHQ ORJV %\ WKH QH[W GDZQ PRUH VHHGV KDG EHHQ UHPRYHG 2QH VHHG ZDV WDNHQ GXULQJ WKDW GD\ DQG IRXU PRUH E\ WKH QH[W GDZQ 7KXV PRVW VHHG UHPRYDO bf WRRN SODFH DW QLJKW ; GI 3 f SUREDEO\ E\ VPDOO URGHQWV ([FORVXUHV 5HPRYDO UDWHV GLIIHUHG VLJQLILFDQWO\ DPRQJ WKH WKUHH WUHDWPHQWV VHHGV DFFHVVLEOH WR DOO URGHQWV ZHUH WDNHQ IDVWHU WKDQ WKH RWKHU WZR WUHDWPHQWV :LOFR[RQ ; GI 3 f $IWHU G WKUHH VHHGV KDG EHHQ UHPRYHG IURP WKH VPDOO PHVK H[FORVXUHV ZKLFK ZHUH GHVLJQHG WR H[FOXGH DOO URGHQWV )LJ f 7ZR RI WKHVH H[FORVXUHV KDG EXUURZV LQVLGH WKDW ZHUH SUREDEO\ GXJ GXULQJ WKH WKUHH PRQWKV EHWZHHQ

PAGE 47

'$<6 )LJXUH $f 5HPRYDO RI PDUNHG VHHGV H[SHULPHQWDOO\ GLVSHUVHG DQG P IURP WUXQNV RI WKH IUXLWLQJ 2FRWHD HQGUHVLDQD WUHHV VWDUWLQJ LQ -XQH 1 DW HDFK GLVWDQFH %f 5HPRYDO RI PDUNHG VHHGV DW DQG P IURP SDUHQW WUHHV LQ WKH HDUO\ ODWH 0D\f DQG DW WKH SHDN PLG-XQHf RI IUXLWLQJ VHDVRQ 1 IRU HDFK FDWHJRU\ $OO VHHGV DW DQG P ZHUH XQGHU WKH SDUHQWDO FURZQV

PAGE 48

6(('6 5(0$,1,1* bf )LJXUH 3HUFHQWDJH RI VHHGV UHPDLQLQJ LQ WKUHH W\SHV RI H[FORVXUHV DIWHU GD\V )RU HDFK WUHDWPHQW 1 7KH nQR URGHQWn H[FORVXUH ZDV GHVLJQHG WR H[FOXGH DOO WHUUHVWULDO DQLPDOV nVPDOO URGHQWn H[FORVXUHV H[FOXGHG ODUJH DQLPDOV EXW DOORZHG DFFHVV WR VPDOO DQLPDOV WKURXJK FP PHVK DQG nDOO URGHQWn SORWV ZHUH FRQWUROV ZLWK QR H[FORVXUH

PAGE 49

FRQVWUXFWLRQ RI H[FORVXUHV DQG WKH EHJLQQLQJ RI WKH H[SHULPHQW 7KH WKLUG H[FORVXUH ZDV KLW E\ D IDOOHQ WUHH WKXV FUHDWLQJ DFFHVV ([FHSW IRU WKRVH WKUHH VHHGV ZHUH QRW UHPRYHG IURP QR URGHQW H[FORVXUHV 6HYHUDO RWKHU H[FORVXUHV ZHUH GDPDJHG E\ IDOOLQJ EUDQFKHV RU EUHDFKHG E\ SHFFDULHV IRUDJLQJ IRU ,QJD SRGV EXW WKHVH LQFLGHQWV HLWKHU RFFXUUHG DIWHU WKH VHHG KDG EHHQ UHPRYHG RU GLG QRW OHDG WR VHHG UHPRYDO $IWHU G DOO WKH VHHGV IURP WKH FRQWURO SORWV KDG EHHQ UHPRYHG ZKLOH b RI WKH VHHGV IURP WKH VPDOO URGHQWV RQO\ H[FORVXUHV KDG EHHQ UHPRYHG )LJ f $IWHU PR b RI WKH VHHGV IURP WKH VPDOO URGHQWV RQO\ H[FORVXUH KDG EHHQ UHPRYHG 7KXV DOWKRXJK ODUJH PDPPDOV PD\ WDNH VRPH VHHGV VPDOO PDPPDOV SUHVXPDEO\ URGHQWVf ZLOO HYHQWXDOO\ ILQG DQG HDW PRVW VHHGV *HUPLQDWLRQ DQG 6HHGOLQJ 6XUYLYDO 2FRWHD HQGUHVLDQD VHHGV EHJDQ JHUPLQDWLQJ DERXW ZHHNV DIWHU GLVSHUVDO 7KH JHUPLQDWLRQ UDWH RI FDJHG VHHGV UDQJHG IURP WR b LQ ERWK \HDUV DQG GLG QRW GLIIHU DPRQJ WKH VHHGV DW GLVSHUVHG QRQGLVSHUVHG DQG UDQGRP VLWHV )LJ f 2I WKH VHHGV WKDW JHUPLQDWHG LQ DQG b DQG b UHVSHFWLYHO\ VXUYLYHG RQH \HDU DV VHHGOLQJV 2YHUDOO b DQG b RI VHHGV LQ DQG UHVSHFWLYHO\ JHUPLQDWHG DQG VXUYLYHG RQ \HDU )RU WKH FDJHG VHHGV DQQXDO VXUYLYDO IRU WKH ILUVW WKUHH \HDUV DYHUDJHG b 7DEOH f 6HHGV DW IRXU GLVSHUVHG DQG ILYH UDQGRP VLWHV IURP VXUYLYHG XQWLO -XQH IRU DQ RYHUDOO \HDU VXUYLYDO UDWH RI b )LJ f :LWKLQ HDFK VWDJH JHUPLQDWLRQ \U \U DQG \U VXUYLYDOf WKH WKUHH ORFDWLRQ WUHDWPHQWV KDG VLPLODU VXUYLYDO UDWHV H[FHSW IRU RQH\HDU VXUYLYDO LQ ) 3 f 6HHGV DW GLVSHUVHG VLWHV KDG VLJQLILFDQWO\ KLJKHU RQH\HDU VXUYLYDO WKDQ WKRVH DW QRQ GLVSHUVHG DQG UDQGRP VLWHV 3 f 5DQGRP DQG QRQGLVSHUVHG VLWHV KDG VLPLODU RQH \HDU VXUYLYDO 3 f *HUPLQDWLRQ VXFFHVV RI VHHGV GHIHFDWHG E\ JXDQV GLG QRW GLIIHU IURP WKDW RI VHHGV UHJXUJLWDWHG E\ WKH RWKHU IRXU VSHFLHV ; GI 3 f 1RQH RI WKH YDULDEOHV LQ WKH ORJLVWLF UHJUHVVLRQ PRGHOV SUHGLFWHG JHUPLQDWLRQ VXFFHVV LQ HLWKHU \HDU 6LJQLILFDQW SUHGLFWRUV RI \HDU VXUYLYDO LQFOXGHG ORZHU FDQRS\

PAGE 50

3523257,21 6859,9,1* + Â’ 'LVSHUVHG 1RQGLVSHUVHG Â’ 5DQGRP JHUP \ U \U 6,7(6 \U JHUP \U 6,7(6 )LJXUH *HUPLQDWLRQ VXFFHVV DQG DQQXDO VHHGOLQJ VXUYLYDO RI GLVSHUVHG DQG QRQ GLVSHUVHG VHHGV DQG VHHGV DW UDQGRPO\ ORFDWHG VLWHV PHDQ 6' SHU WUHHf LQ OHIWf DQG ULJKWf :LWKLQ HDFK VWDJH ZKHUH D GLIIHUHQFH DPRQJ VLWHV ZDV GHWHFWHG EDUV ZLWK GLIIHUHQW OHWWHUV DERYH DUH VLJQLILFDQWO\ GLIIHUHQW 3 %RQIHUURQLDGMXVWHG DOSKD f

PAGE 51

7DEOH 5HVXOWV RI ORJLVWLF UHJUHVVLRQV RI SRVWGLVSHUVDO VXUYLYDO RI VHHGOLQJV SUHGLFWHG E\ KDELWDW YDULDEOHV 7KHVH VHHGV ZHUH SURWHFWHG IURP VHHG SUHGDWRUV ZLWK FDJHV XQWLO JHUPLQDWLRQ 2QO\ VLJQLILFDQW HIIHFWV DUH OLVWHG 1R YDULDEOHV VLJQLILFDQWO\ SUHGLFWHG JHUPLQDWLRQ VXFFHVV QRW VKRZQf 5HVSRQVH U RJ OLNHOLKRRG r SUHGLFWRUV \U VXUYLYDO rrr FDQRS\ FRYHUr GLVSHUVDO GDWHr \U VXUYLYDO rrr FDQRS\ FRYHUr GLVSHUVDO GDWHrr WUHH QXPEHUr r3 rr3 rrr! ALJQ LQ IURQW RI HDFK SUHGLFWRU YDULDEOH LQGLFDWHV SRVLWLYH RU QHJDWLYH FRUUHODWLRQ ZLWK WKH UHVSRQVH YDULDEOH U

PAGE 52

FRYHU DQG ODWHU GLVSHUVDO GDWH LQ ERWK DQG 7DEOH f $GGLWLRQDOO\ LQ \U VXUYLYDO ZDV KHWHURJHQHRXV ZLWK UHVSHFW WR ZKLFK SDUWLFXODU IUXLWLQJ WUHH ZDV FORVHVW DOWKRXJK QR FOHDUFXW SDWWHUQ ZDV GLVFHUQLEOH LQ WHUPV RI FURS VL]H WUHH VL]H RU SUR[LPLW\ WR RWKHU IUXLWLQJ WUHHV 7KH FDXVHV RI PRUWDOLW\ RI FDJHG VHHGV DQG VHHGOLQJV GXULQJ WKH ILUVW \HDU GLIIHUHG DPRQJ GLVSHUVHG QRQGLVSHUVHG DQG UDQGRP VLWHV LQ ERWK ; GI 3 f DQG ; GI 3 )LJ f 0DQ\ VHHGOLQJV ZHUH NLOOHG ZKHQ VHHG SUHGDWRUV SUHVXPDEO\ URGHQWVf UHPRYHG WKH VHHG DQG VHYHUHG WKH FRQQHFWLRQ EHWZHHQ WKH VKRRW DQG URRW 0DPPDOLDQ KHUELYRUHV DOVR NLOOHG VHHGOLQJV E\ HQWLUHO\ UHPRYLQJ WKH OHDYHV DQG VKRRW 0RUWDOLW\ E\ WKH FRPELQDWLRQ RI PDPPDOLDQ VHHG SUHGDWRUV DQG KHUELYRUHV ZDV VLJQLILFDQWO\ GLIIHUHQW DPRQJ WKH WKUHH ORFDWLRQ WUHDWPHQWV LQ 3 3 f EXW QRW ) 3 f ,Q PDPPDOV NLOOHG PRUH VHHGV DQG VHHGOLQJV DW GLVSHUVHG WKDQ DW UDQGRP VLWHV 3 f EXW PRUWDOLW\ FDXVHG E\ PDPPDOV DW QRQGLVSHUVHG VLWHV GLG QRW GLIIHU IURP WKDW DW HLWKHU GLVSHUVHG RU UDQGRP VLWHV 3 )LJ f 0RUWDOLW\ FDXVHG E\ EHHWOH ODUYDH +HLOLSXV VS DQG DW OHDVW RQH RWKHU XQLGHQWLILHG VSHFLHVf GLIIHUHG DPRQJ WKH WKUHH WUHDWPHQWV LQ ERWK \HDUV ) 3 r 3 f ,Q ERWK \HDUV PRUH QRQGLVSHUVHG WKDQ GLVSHUVHG VHHGV ZHUH NLOOHG E\ EHHWOHV 3 )LJ f 0RUWDOLW\ FDXVHG E\ IXQJDO SDWKRJHQV DQG SK\VLFDO GDPDJH IDOOLQJ EUDQFKHV WUDPSOLQJf GLG QRW GLIIHU DPRQJ WUHDWPHQWV RU \HDUV )RU WKH WKUHH WUHDWPHQWV FRPELQHG OHYHOV RI PRUWDOLW\ E\ HDFK FDXVH ZHUH VLPLODU EHWZHHQ WKH WZR \HDUV H[FHSW IRU LQVHFW SUHGDWLRQ ZKLFK ZDV VLJQLILFDQWO\ KLJKHU LQ WKDQ 6WXGHQWnV WWHVW W GI 3 f DQG PRUWDOLW\ E\ PDPPDOV ZKLFK ZDV VLJQLILFDQWO\ KLJKHU LQ WKDQ LQ W GI 3 f $YHUDJH FDQRS\ FRYHU IRU VHHGOLQJV WKDW VXUYLYHG RQH \HDU DW GLVSHUVHG VLWHV b s 1 f ZDV VLJQLILFDQWO\ ORZHU WKDQ IRU VHHGOLQJV WKDW GLHG b s 1 f ZKHUHDV FDQRS\ FRYHU GLG QRW GLIIHU ZLWK RQH\HDU VXUYLYDO IRU QRQGLVSHUVHG RU

PAGE 53

R /8 R K && 2 D 2 RF D E PDPPDOV LQVHFWV 2 'LVSHUVHG 1 f 1RQGLVSHUVHG 1 f Â’ 5DQGRP 1 f IXQJDO SK\VLFDO D L U E PDPPDOV LQVHFWV Â’ 'LVSHUVHG 1 f 1RQGLVSHUVHG 1 f Â’ 5DQGRP 1 f )LJXUH $YHUDJH 6'f SURSRUWLRQ SHU WUHH RI FDJHG GLVSHUVHG DQG QRQGLVSHUVHG VHHGV DQG UDQGRPO\ ORFDWHG VHHGV NLOOHG E\ GLIIHUHQW FDXVHV GXULQJ WKH ILUVW \HDU IRU VHHGV IURP DERYHf DQG EHORZf &DJHV ZHUH UHPRYHG DIWHU JHUPLQDWLRQ DQG VKRRW GHYHORSPHQW 6DPSOH VL]HV RI VHHGV RU VHHGOLQJV NLOOHG VKRZQ LQ SDUHQWKHVHV :LWKLQ HDFK FDXVH RI PRUWDOLW\ ZKHUH D GLIIHUHQFH DPRQJ VLWHV ZDV GHWHFWHG EDUV ZLWK GLIIHUHQW OHWWHUV DERYH DUH VLJQLILFDQWO\ GLIIHUHQW %RQIHUURQLDGMXVWHG DOSKD f 6HH WH[W 0HWKRGV *HUPLQDWLRQ DQG VHHGOLQJ VXUYLYDOf IRU GHILQLWLRQV RI W\SHV RI PRUWDOLW\

PAGE 54

UDQGRP VLWHV ZD\ $129$ ) 3 f 'HQVHU FDQRS\ FRYHU ZDV DOVR D VLJQLILFDQW SUHGLFWRU RI KLJKHU PRUWDOLW\ E\ IXQJDO SDWKRJHQV IORJLVWLF UHJUHVVLRQ 3 f 6HHGOLQJV WKDW VXUYLYHG RQH \HDU ZHUH VLJQLILFDQWO\ WDOOHU DW VLWHV ZLWK OHVV GHQVH FDQRS\ FRYHU OLQHDU UHJUHVVLRQ U 3 )LJ f 6HHGOLQJV IURP VHHGV GLVSHUVHG E\ EHOOELUGV ZHUH VLJQLILFDQWO\ WDOOHU WKDQ WKRVH DW RWKHU VSHFLHVn VLWHV WWHVW 1 3 f *HUPLQDWLRQ WULDOV 5HJXUJLWDWHG VHHGV DQG VHHGV FOHDQHG RI SXOS E\ KDQG KDG KLJKHU JHUPLQDWLRQ UDWHV WKDQ WKH VHHGV LQ LQWDFW IUXLWV ; GI 3 f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f DQG VHHPHG YLDEOH EXW WKH RWKHU HLJKW VHHGV URWWHG $IWHU PR QHLWKHU RI WKH DSSDUHQWO\ JHUPLQDWLQJ VHHGV KDG DQ\ VLJQ RI URRW GHYHORSPHQW QRU KDG WKH FRW\OHGRQV EHJXQ WR VHSDUDWH 6L[ RI EXULHG VHHGV KDG ZKLWLVK PROG RQ WKH VHHG FRDW 7KXV IHZ EXULHG VHHGV DSSHDU WR HVWDEOLVK DV VHHGOLQJV 6HHGOLQJ DQG VDSOLQJ SORWV 6HHGOLQJV DQG VDSOLQJV ZHUH HTXDOO\ FRPPRQ LQ [ P SORWV XQGHU s LQGLYLGXDOVf DQG DZD\ s f IURP WKH SDUHQW WUHHV SDLUHG WWHVW W 3 f ,QGLYLGXDOV LQ SORWV DZD\ IURP WKH SDUHQW WUHHV KRZHYHU ZHUH WDOOHU s FPf WKDQ WKRVH QHDU WKH SDUHQW WUHHV s 0DQQ:KLWQH\ 8 WHVW 3 f 5HFUXLWPHQW %DVHG RQ WKH H[SHULPHQW ZLWK PDUNHG VHHGV WKH SUREDELOLW\ RI UHFUXLWPHQW LV GHWHUPLQHG PRVWO\ E\ SRVWGLVSHUVDO VHHG SUHGDWRUV $VVXPLQJ WKH H[SHULPHQW UHIOHFWV WKH

PAGE 55

6(('/,1* +(,*+7 FPf )LJXUH 6HHGOLQJ KHLJKW 1 f SORWWHG DV D IXQFWLRQ RI SHUFHQW FDQRS\ FRYHU IRU ERWK \HDUV FRPELQHG KHLJKW [ U 3 f 9DOXHV IRU FDQRS\ FRYHU EHORZ DSSUR[LPDWHO\ b DUH DVVRFLDWHG ZLWK FDQRS\ JDSV

PAGE 56

DEVROXWH OHYHO RI VHHG SUHGDWLRQ DQG DOO VHHGV WKDW HVFDSH SUHGDWLRQ JHUPLQDWH DQG VXUYLYH WKHQ UHFUXLWPHQW ZDV b LQ DQG b LQ 2Q WKH RWKHU KDQG LI VHHG UHPRYDO DW ZN bf UHIOHFWV WKH XOWLPDWH SDWWHUQ RI VXUYLYDO HJ /LHEHUPDQ f WKHQ LW LV SRVVLEOH WR FDOFXODWH WKH LQIOXHQFH RI VWDJHVSHFLILF PRUWDOLW\ SDWWHUQV RQ WKH UHODWLYH SUREDELOLW\ RI UHFUXLWPHQW 7DEOH )LJ f 7KH LQLWLDO SDWWHUQ RI VHHG UDLQ VKRZV D VWURQJ SHDN QHDU WKH SDUHQW WUHHV DQG VHFRQGDU\ SHDNV EH\RQG P WKDW FRUUHVSRQG ZLWK EHOOELUG VRQJ SHUFKHV )LJ Df 7KH EHOOELUG SHUFKHV HQFRPSDVV D UDQJH RI FDQRS\ FRYHU YDOXHV IURP IRUHVWJDS HGJH bf WR JDS FHQWHUV bf DQG WKXV VHHG UDLQ LV VSUHDG DFURVV D ZLGHU UDQJH RI FDQRS\ FRQGLWLRQV ZLWK LQFUHDVLQJ GLVWDQFH IURP WKH SDUHQW WUHHV 7ZRZHHN VHHG VXUYLYDO LV KLJKHU IRU VHHGV P IURP IUXLWLQJ WUHHV EXW UHODWLYHO\ FRQVLVWHQW DFURVV WKH UDQJH RI FDQRS\ FRYHU )LJ Ef *HUPLQDWLRQ LV KLJK IRU DOO VHHGV UHJDUGOHVV RI GLVWDQFH IURP SDUHQW RU FDQRS\ FRYHU )LJ Ff 2QH\HDU VHHGOLQJ VXUYLYDO LV KLJKHVW IRU VLWHV P IURP SDUHQWV DQG b FDQRS\ FRYHU )LJ Gf 7KH FXPXODWLYH SUREDELOLW\ RI UHFUXLWPHQW WKH SURGXFW RI VHHG UDLQ ZN VHHG VXUYLYDO JHUPLQDWLRQ DQG RQH \HDU VXUYLYDOf VKRZV D SHDN QHDU WKH SDUHQW WUHHV DQG VHFRQGDU\ SHDNV FRUUHVSRQGLQJ WR WKH EHOOELUG SHUFKHV )LJ Hf 1RWH WKDW WKH SUREDELOLW\ RI D VHHG GLVSHUVHG E\ ELUGV VXUYLYLQJ RQH \HDU LV b DW DQ\ ORFDWLRQ 'LVFXVVLRQ 7KH UHVXOWV LOOXVWUDWH IRXU PDLQ SRLQWV f EHOOELUGV GLVSHUVHG VHHGV LQ D GLIIHUHQW SDWWHUQ WKDQ WKH RWKHU VSHFLHV f SRVWGLVSHUVDO VHHG SUHGDWLRQ LV KLJKHVW QHDU WKH SDUHQWV EXW LV DOVR KLJK HYHU\ZKHUH f WKH SDWWHUQ RI UHFUXLWPHQW LV ELPRGDO ZLWK D SHDN QHDU WKH SDUHQW WUHH LQ FORVHG FDQRS\ IRUHVW DQG DQRWKHU LQ JDSV FRUUHVSRQGLQJ WR EHOOELUG SHUFKHV DQG f DOO WKUHH DGYDQWDJHV RI GLVSHUVDO HVFDSH FRORQL]DWLRQ DQG GLUHFWHG GLVSHUVDOf DUH VXSSRUWHG 7KHVH SRLQWV ZLOO EH GLVFXVVHG LQ GHWDLO EHORZ

PAGE 57

7DEOH 6XPPDU\ RI SRVWGLVSHUVDO VWDJHV OHDGLQJ WR VHHGOLQJ UHFUXLWPHQW 9DOXHV IRU VWDJHVSHFLILF SUREDELOLWLHV UHSUHVHQW WKH DYHUDJH SURSRUWLRQ RI LQGLYLGXDOV SHU WUHH WKDW HQWHUHG DQG VXUYLYHG HDFK VWDJH &XPXODWLYH SUREDELOLWLHV DUH WKH SURGXFW RI WKH FXUUHQW DQG SUHYLRXV VWDJHVSHFLILF SUREDELOLWLHV 'DWD IRU DQG \U VXUYLYDO RI WKH FRKRUW DUH XQDYDLODEOH 6WDJH 6WDJHVSHFLILF SUREDELOLW\ &XPXODWLYH SUREDELOLW\ ZN VHHG JHUPLQDWLRQ s \U VHHGOLQJ s \U VHHGOLQJ f§ f§ \U VHHGOLQJ f§ f§

PAGE 58

)LJXUH $YHUDJH VHHG UDLQ Df ZN VHHG VXUYLYDO Ef JHUPLQDWLRQ Ff \U VHHGOLQJ VXUYLYDO Gf DQG FXPXODWLYH SUREDELOLW\ RI UHFUXLWPHQW Hf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f LV FDOFXODWHG DV WKH SURGXFW D [ E [ F [ Gf IRU HDFK GLVWDQFHFRYHU FDWHJRU\ )RU FODULW\ VWDQGDUG GHYLDWLRQV DUH QRW VKRZQ

PAGE 59

6((' 5$,1 )LJXUH 6((' 5$,1 bf

PAGE 60

:. 6((' 6859,9$/ )LJXUH f§FRQWLQXHG 6((' 6859,9$/ bf

PAGE 61

*(50,1$7,21 )LJXUH FRQWLQXHG *(50,1$7,21 bf

PAGE 62

<5 6(('/,1* 6859,9$/ )LJXUH f§FRQWLQXHG <5 6859,9$/ bf

PAGE 63

5(&58,70(17 )LJXUH FRQWLQXHG 3Uf 5(&58,70(17

PAGE 64

6HHG 'LVSHUVDO 0RVW 2 HQGUHVLDQD VHHGV KDQGOHG E\ ELUGV ODQG XQGHU RU MXVW EH\RQG WKH FURZQV RI SDUHQW WUHHV 7KLV GLVWULEXWLRQ RI VHHGV LV W\SLFDO IRU YHUWHEUDWHGLVSHUVHG SODQWV 'LU]R DQG 'RPLQJXH] +RSSHV :LOOVRQ /DPDQ Df 7KH DEXQGDQFH RI 2 HQGUHVLDQD DGXOWV LQ WKH VWXG\ DUHD PDNHV GLVSHUVDO EH\RQG P IURP DQ\ FRQVSHFLILF WUHH YLUWXDOO\ LPSRVVLEOH &RQVLGHULQJ WKH UHVWULFWHG HOHYDWLRQDO UDQJH RI WKLV VSHFLHV ORQJHU GLVWDQFH GLVSHUVDO PD\ OHDG WR DUULYDO LQ KDELWDWV XQVXLWDEOH IRU HVWDEOLVKPHQW :KHHOZULJKW f 7KXV ZKLOH UHJXUJLWDWLQJ VHHGV MXVW RXWVLGH WKH FURZQ RI D IUXLWLQJ WUHH PD\ VHHP OLNH SRRU TXDOLW\ GLVSHUVDO 0F.H\ f WKH VKHDU DEXQGDQFH RI VHHGV LQ WKDW UHJLRQ PD\ OHDG WR D VPDOO SHDN LQ UHFUXLWPHQW QHDU WKH SDUHQW WUHH )LJ Hf DV SUHGLFWHG E\ +XEEHOO f DQG &RQGLW HW DO f 7KH ORQJ WDLO RI WKH VHHG GLVWULEXWLRQ KRZHYHU LV DSSDUHQWO\ LPSRUWDQW IRU UHFUXLWPHQW EHFDXVH VHHGV GLVSHUVHG IDUWKHU IURP WKH SDUHQW WUHHV KDYH D JUHDWHU FKDQFH RI DUULYLQJ LQ D KDELWDW ZKHUH WKH SUREDELOLW\ RI VHHGOLQJ VXUYLYDO LV KLJKHU LH D VDIH VLWHf DV ZLOO EH VKRZQ EHORZ 2I WKH ILYH VSHFLHV RI ELUGV WKDW GLVSHUVHG PRVW RI WKH VHHGV LQ WKLV VWXG\ DOO EXW RQH GHSRVLWHG PRVW VHHGV LQ FORVHG FDQRS\ IRUHVW 2QO\ PDOH EHOOELUGV IUHTXHQWO\ GLVSHUVHG VHHGV WR WUHHIDOO JDSV ,Q P\ VWXG\ VLWH EHOOELUG VRQJ SHUFKHV ZHUH LQ GHDG 6DSLXP ROLJRQHXURQ WUHHV ERUGHULQJ ODUJH WUHHIDOO JDSV 6XFK VLWHV DUH W\SLFDO IRU EHOOELUGV 6QRZ f UHSRUWHG WKDW EHOOELUG VRQJ SHUFKHV DW ORZHU HOHYDWLRQV LQ WKH 0RQWHYHUGH DUHD ZKHUH WKH IRUHVW LV IUDJPHQWHGf DUH XVXDOO\ RQ GHDG EUDQFKHV LQ WDOO WUHHV RQ WKH IRUHVW HGJH ,Q RWKHU 1HRWURSLFDO IRUHVWV WKUHH RWKHU VSHFLHV RI EHOOELUGV 3URFQLDV VSSf H[KLELW VLPLODU EHKDYLRU LQ KDELWXDO XVH RI WDOO H[SRVHG VRQJ SHUFKHV RIWHQ RQ GHDG EUDQFKHV 6QRZ Df 6QDJV XVHG E\ EHOOELUGV LQ P\ VLWH KDG VHYHUDO EUDQFKHV DQG WKH ELUGV PDGH XVH RI PDQ\ GLIIHUHQW EUDQFKHV ZLWKLQ HDFK WUHH VXFK WKDW VHHGV XQGHU WKHP ZHUH VFDWWHUHG RYHU DSSUR[LPDWHO\ PA LQFOXGLQJ D JUDGLHQW RI VLWH FRQGLWLRQV IURP JDS WR IRUHVW XQGHUVWRU\ 0DOH EHOOELUGV DOVR GLVSHUVHG VHHGV LQ IRUHVW DV GR WKH RWKHU VSHFLHV EXW WKH\ W\SLFDOO\ VSHQG b RI WKH GD\ LQ WKH YLFLQLW\ RI WKH VRQJ

PAGE 65

SHUFK 6QRZ f DQG SUREDEO\ GLVSHUVH D VLPLODU SHUFHQWDJH RI VHHGV WKH\ SURFHVV XQGHU VRQJ SHUFKHV )HPDOH EHOOELUGV ZHUH UDUHO\ VHHQ DW WKH VRQJ SHUFKHV DQG WKHQ RQO\ IRU D IHZ PLQXWHV 7KH LQIOXHQFH RI EHOOELUGV RQ WKH SDWWHUQ RI VHHG IDOO LV VKRZQ E\ WKH VOLJKW LQFUHDVH LQ WKH QXPEHU RI VHHGV P IURP WKH SDUHQW WUHHV )LJ f %HOOELUG SHUFKHV LQ P\ VLWH KDSSHQHG WR EH IDU Pf IURP DQ\ IUXLWLQJ 2 HQGUHVLDQD WUHHV 7KH RWKHU IRXU ELUG VSHFLHV WKDW GLVSHUVH 2 HQGUHVLDQD VHHGV RFFDVLRQDOO\ SHUFK RQ WKH HGJHV RI JDSV :HQQ\ SHUVRQDO REVHUYDWLRQf DOWKRXJK ZKHWKHU WKH\ GR VR PRUH RIWHQ WKDQ H[SHFWHG EDVHG RQ SHUFK DYDLODELOLW\ LV XQNQRZQ %HFDXVH ELUGV GR QRW FDQQRW"f UHJXUJLWDWH LQ IOLJKW DQG JXDQV DSSDUHQWO\ GR QRW GHIHFDWH LQ IOLJKW SHUVRQDO REVHUYDWLRQf WKH PRVW OLNHO\ ZD\ DQ 2 HQGUHVLDQD VHHG FDQ DUULYH LQ D JDS LV YLD PDOH EHOOELUGV QHYHU REVHUYHG WKH RWKHU IRXU VSHFLHV RQ EHOOELUG SHUFKHV 7KXV 2 HQGUHVLDQD KDV WZR W\SHV RI GLVSHUVHUV WKDW GHSRVLW VHHGV LQ WZR GLIIHUHQW EXW SUHGLFWDEOH SDWWHUQV PDOH EHOOELUGV GHSRVLWLQJ ODUJH QXPEHUV bf RI WKH VHHGV WKH\ GLVSHUVH LQ DQG QHDU JDSV DQG WKH RWKHU IRXU VSHFLHV GHSRVLWLQJ VHHGV LQ IRUHVW PRVW bf ZLWKLQ P RI SDUHQW WUHHV DQG IHZHU PRUH WKDQ P DZD\ $GGLWLRQDOO\ RQ VHYHUDO RFFDVLRQV IROORZHG TXHW]DOV DQG JXDQV IURP RQH IUXLWLQJ 2 HQGUHVLDQD WR DQRWKHU ZKHUH WKH\ GHSRVLWHG VHHGV EHIRUH HDWLQJ IUXLWV LQ WKH VHFRQG WUHH $OWKRXJK KDG QR PHWKRG RI LGHQWLI\LQJ WKH VRXUFH WUHH RI WKH VHHGV LW LV OLNHO\ WKDW GLVSHUVDO RI VHHGV IURP RQH WUHH WR DQRWKHU FRQVSHFLILF RFFXUV UHJXODUO\ :KHHOZULJKW *LEVRQ DQG :KHHOZULJKW f 7KH SDWWHUQV RI VHHG GLVWULEXWLRQ GHVFULEHG KHUH PD\ EH VLPLODU IRU RWKHU WUHHV LQ FORXG IRUHVWV ZKHUH WKHVH GLVSHUVHU VSHFLHV RFFXU GXULQJ WKH EUHHGLQJ VHDVRQ EXW PD\ EH YHU\ GLIIHUHQW GXULQJ WKH UHVW RI WKH \HDU 7KH ILYH VSHFLHV RI ELUGV WKDW GLVSHUVH 2 HQGUHVLDQD DUH WKH PRVW LPSRUWDQW IUXLW FRQVXPHUV RI FDQRS\ WUHHV LQ WKH DUHD :KHHOZULJKW HW DO f $W OHDVW RWKHU VSHFLHV RI WUHHV DUH GLVSHUVHG E\ WKHVH ELUGV DQG SUREDEO\ KDYH VLPLODU SDWWHUQV RI VHHGIDOO DV 2 HQGUHVLDQD ZLWK D SHDN QHDU WKH SDUHQW WUHH DQG D VHFRQG SHDN DW EHOOELUG SHUFKHV +RZHYHU DOO ILYH GLVSHUVHUV H[FHSW SHUKDSV WRXFDQHWVf

PAGE 66

XQGHUWDNH HOHYDWLRQDO VKLIWV WKDW FRUUHVSRQG ZLWK IUXLW DYDLODELOLW\ :KHHOZULJKW /HYH\ DQG 6WLOHV 3RZHOO DQG %MRUN *XLQGRQ f DQG OHDYH WKH VWXG\ DUHD LQ ODWH -XO\ RU HDUO\ $XJXVW ,Q WKH\ OHIW LQ -XQH ZKHQ 2 HQGUHVLDQD IDLOHG WR IUXLWf 0DOH EHOOELUGV PD\ QRW SURYLGH VXFK D SURQRXQFHG SHDN LQ VHHG GLVSHUVDO WR VRQJ SHUFKHV DIWHU WKH EUHHGLQJ VHDVRQ EHFDXVH WKH\ VHHP WR XVH PDQ\ GLIIHUHQW SHUFKHV LQ WKH QRQ EUHHGLQJ VHDVRQ DQG WHQG WR XVH VXEFDQRS\ SHUFKHV WKDW DUH QRW H[SRVHG 6WLOHV DQG 6NXWFK f 7KXV VHHG IDOO IRU WKH FRPPXQLW\ DV D ZKROH LV OLNHO\ WHPSRUDOO\ DV ZHOO DV VSDWLDOO\ KHWHURJHQHRXV 7KH LPSOLFDWLRQV RI VHHG GLVSHUVDO WR KDELWXDO SHUFKHV GHVHUYHV IXUWKHU VWXG\ 'R KDELWXDO SHUFKHV UHSUHVHQW IRFL RI SODQW UHFUXLWPHQW RU RI VHHG SUHGDWLRQ" 6HHG GLVSHUVDO WR SHUFKHV KDV EHHQ IUHTXHQWO\ VWXGLHG LQ VXFFHVVLRQDO ODQGVFDSHV /LYLQJVWRQ 0F'RQQHOO +ROWKXLM]HQ HW DO *XHYDUD DQG /DERUGH 0F&ODQDKDQ DQG :ROIH 5RELQVRQ DQG +DQGHO 9LHLUD HW DO 1HSVWDG HW DO 'XQFDQ f ,Q WKHVH V\VWHPV GLUHFWHG GLVSHUVDO PD\ SOD\ D NH\ UROH LQ WKH UHVWRUDWLRQ RI DEDQGRQHG SDVWXUHV ORJJHG IRUHVWV DQG RWKHU GHJUDGHG ODQGV ,W LV SRVVLEOH WKDW GLUHFWHG GLVSHUVDO LV PRUH FRPPRQ LQ LQWDFW IRUHVWV WKDQ SUHYLRXVO\ H[SHFWHG )RU H[DPSOH PDQDNLQV 3LSULGDHf DQG FRFNVRIWKHURFN 5XSLFROD &RWLQJLGDHf FKRRVH OHN SHUFKHV WKDW DUH PRUH VXQOLW WKDQ DYHUDJH XQGHUVWRU\ SHUFKHV (QGOHU DQG 7KU\ f DQG VXFK VLWHV PD\ SURYLGH JURZWK DGYDQWDJHV IRU VHHGOLQJV 6QRZ Ef QRWHV WKDW RWKHU VSHFLHV RI EHOOELUGV 3URFQLDVf SUHIHUHQWLDOO\ VHOHFW SHUFKHV RQ GHDG WUHHV RU EUDQFKHV RU LQ VSDUVHO\YHJHWDWHG WUHHV )RU WKHVH VSHFLHV FRPSHWLWLRQ DPRQJ PDOHV IRU IHPDOHV GULYHV WKHP WR EH DV FRQVSLFXRXV DV SRVVLEOH 7KH IRUWXLWRXV RXWFRPH RI WKLV EHKDYLRU PD\ EH D GLVSURSRUWLRQDWH HIIHFW RQ SODQW UHFUXLWPHQW LQ WKH YLFLQLW\ RI WKHLU GLVSOD\ DUHDV 7KU\ DQG /DUSLQ f :KHWKHU KDELWXDO SHUFKHV UHSUHVHQW IRFL RI VHHGOLQJ UHFUXLWPHQW 0F'RQQHOO DQG 6WLOHV 0F'RQQHOO f RU OHDG WR GHQVLW\GHSHQGHQW VHHG DQG VHHGOLQJ PRUWDOLW\ :KHHOZULJKW f QHHGV WR EH H[DPLQHG LQ PRUH GHWDLO

PAGE 67

6HHG 3UHGDWLRQ 5HPRYDO RI PDUNHG VHHGV DIWHU GLVSHUVDO LQ WKLV VWXG\ DOZD\V UHVXOWHG LQ SUHGDWLRQ $OWKRXJK VHHGV ZHUH WDNHQ LQWR EXUURZV DQG ORJV QHYHU IRXQG DQ\ WUHDWPHQW RI VHHGV LQGLFDWLYH RI VFDWWHUKRDUGLQJ VXFK DV EXULDO LQ WKH VRLO 6P\WKH +DOOZDFKV )RUJHW f RU XQGHU SLOHV RI OHDI OLWWHU )RUJHW f ,Q DGGLWLRQ WKH IDLOXUH RI EXULHG VHHGV WR JHUPLQDWH DQG VXUYLYH LQ WKH JUHHQKRXVH H[SHULPHQW VXJJHVWV WKDW VFDWWHUKRDUGLQJ ZRXOG QRW EH DGYDQWDJHRXV IRU 2 HQGUHVLDQD UHFUXLWPHQW 2QO\ WZR VSHFLHV RI PDPPDOV NQRZQ WR VFDWWHUKRDUG VHHGV RFFXU LQ WKH VWXG\ VLWH VSLQ\ SRFNHW PLFH +HWHURP\V GHVPDUHVWLDQXVf DQG DJRXWLV 'DV\SURFWDSXQFWDWDf %RWK VSHFLHV DUH PXFK OHVV FRPPRQ KHUH WKDQ DW ORZHU HOHYDWLRQV :HQQ\ SHUVRQDO REVHUYDWLRQ .* 0XUUD\ SHUVRQDO FRPPXQLFDWLRQf $OWKRXJK VRPH VSHFLHV RI VTXLUUHOV 6FLXUXVf DQG GHHU PLFH 3HURP\VFXVf DUH NQRZQ WR FDFKH VHHGV LQ RWKHU UHJLRQV RI WKH ZRUOG 9DQGHU :DOO f LQIRUPDWLRQ RQ WURSLFDO VSHFLHV LQ WKLV VLWH 6 GHSSHL DQG 3 PH[LFDQXVf LV ODFNLQJ (PPRQV f 7KH PDLQ VHHG SUHGDWRUV LQ WKLV VWXG\ ZHUH SUREDEO\ URGHQWV SDUWLFXODUO\ WKH GHHU PRXVH 3HURP\VFXV PH[LFDQXV E\ IDU WKH PRVW FRPPRQ WHUUHVWULDO URGHQW LQ WKH VWXG\ DUHD $QGHUVRQ /DQJWLPP f %DVHG RQ WKH H[FORVXUH H[SHULPHQW KRZHYHU LW LV SRVVLEOH WKDW VSHFLHV ODUJHU WKDQ PLFH LH DJRXWL SDFD SHFFDU\ ZRRGTXDLOf DUH DOVR LPSRUWDQW VHHG SUHGDWRUV %HFDXVH UHPRYDO RI VHHGV IURP WKH RSHQ FRQWURO SORWV ZDV DERXW WZLFH WKDW RI WKH VPDOO URGHQWV RQO\ SORWV DJRXWLV DQG RWKHU ODUJH PDPPDOV PD\ KDYH EHHQ UHVSRQVLEOH IRU DV PXFK DV b RI 2 HQGUHVLDQD VHHG SUHGDWLRQ 1HYHUWKHOHVV WKH H[FORVXUHV DOVR VKRZHG WKDW LQ WKH DEVHQFH RI DJRXWLV VPDOO URGHQWV ZLOO ILQG DQG HDW PRVW RI WKH VHHGV DOEHLW RYHU D ORQJHU WLPH SHULRG $GGLWLRQDOO\ WKH ILQGLQJ WKDW PRVW PDUNHG VHHGV IURP WKH UHPRYDO H[SHULPHQW ZHUH WDNHQ DW QLJKW DQG UHPRYDO RI VHHGV DOZD\V UHVXOWHG LQ SUHGDWLRQ UDWKHU WKDQ VFDWWHUKRDUGLQJ VXJJHVW WKDW DJRXWLV ZKLFK DUH PRVWO\ GLXUQDO DQG NQRZQ VFDWWHUKRDUGHUV 6P\WKH +DOOZDFKV f PD\ QRW EH WDNLQJ

PAGE 68

PDQ\ VHHGV DOWKRXJK IHZ GDWD H[LVW RQ WKH H[WHQW WR ZKLFK DJRXWLV HDW EXW GR QRW VFDWWHUKRDUG FHUWDLQ VSHFLHV HJ 6P\WKH f 7KH KLJK GHJUHH RI SRVWGLVSHUVDO VHHG SUHGDWLRQ UHSRUWHG KHUH LV QRW XQXVXDO +DUSHU &DYHUV f 5HFHQW UHYLHZV KDYH IRXQG b ORVVHV RI VHHGV WR SUHGDWRUV LQ RI VWXGLHV &UDZOH\ +XOPH f (VWLPDWHV RI VHHG SUHGDWLRQ UDWHV IRU 1HRWURSLFDO WUHHV UDQJH IURP b IRU WKH FDQRS\ WUHH %URVLPXP DOLFDVWUXP %XUNH\ f b IRU WKH FDQRS\ HPHUJHQW WUHH 'LSWHU\[SDQDPHQVLV 'H6WHYHQ DQG 3XW] f b IRU 'LSWHU\[ PLFUDQWKD &LQWUD DQG +RPD f b IRU WKH VXEFDQRS\ WUHH )DUDPHD RFFLGHQWDOV 6FKXSS Ef b IRU WKH VXEFDQRS\ SDOP :HOILD JHRUJLL 6FKXSS DQG )URVW f b IRU WKH FDQRS\ SDOP $VWURFDU\XP PXUXPXUX &LQWUD DQG +RPD f DQG b IRU WKH FDQRS\ WUHH 9LUROD QRELOLV +RZH HW DO f 7KH VWXG\ E\ %XUNH\ f ZDV VKRUWWHUP GD\Vf DQG PD\ XQGHUHVWLPDWH WRWDO SUHGDWLRQ $OVR LQ WKH VWXG\ E\ 'H6WHYHQ DQG 3XW] f D VLWH ZLWK RQO\ b SUHGDWLRQ KDG ORZ SRSXODWLRQV RI VHHG SUHGDWRUV DV D UHVXOW RI KXQWLQJ 3UHYLRXV VWXGLHV RQ VHHG SUHGDWLRQ RI /DXUDFHDH IRXQG b SUHGDWLRQ :KHHOZULJKW &KDSPDQ D +ROO DQG /XORZ f 7KH KLJK OHYHO RI VHHG SUHGDWLRQ LQ P\ VWXG\ VLWH FRXOG EH H[SODLQHG E\ GHQVH YHJHWDWLRQ DQG FORXG\ FRQGLWLRQV SURYLGLQJ VPDOO URGHQWV SURWHFWLRQ IURP SUHGDWRUV %RZHUV DQG 'RROH\ 9£VTXH] f +LJK SUHGDWLRQ LV SUREDEO\ QRW D FRQVHTXHQFH RI KLJK URGHQW SRSXODWLRQV GXH WR ODFN RI SUHGDWRUV EHFDXVH SRWHQWLDO SUHGDWRUV RI PLFH VXFK DV RZOV IRUHVWIDOFRQV DQG FDWV DUH UHODWLYHO\ FRPPRQ :HQQ\ SHUVRQDO REVHUYDWLRQf 'XULQJ WKH GD\ OLJKW OHYHOV LQ WKH VWXG\ DUHD DUH RIWHQ UHGXFHG DV D UHVXOW RI FORXG\ ZHDWKHU &DYHOLHU &KD]GRQ HW DO f DV ZHOO DV GHQVH YHJHWDWLRQ ,QGHHG RQ D IHZ RFFDVLRQV PLFH ZHUH VHHQ GXULQJ WKH GD\ HVSHFLDOO\ 3HURP\VFXV DQG 6FRWLQRP\Vf DQG VRPH PDUNHG VHHGV ZHUH UHPRYHG E\ VPDOO URGHQWV GXULQJ WKH GD\ VHH 3RVWGLVSHUVDO VHHG SUHGDWLRQ GLVWDQFH HIIHFWf $OVR ORFDO QDWXUDOLVWV UHSRUW VHHLQJ PRUH PLFH RQ PLVW\ RU UDLQ\ GD\V DQG QLJKWV WKDQ GXULQJ FOHDU ZHDWKHU 7 *XLQGRQ SHUVRQDO FRPPXQLFDWLRQf

PAGE 69

7KXV WKH KLJK OHYHOV RI SUHGDWLRQ IRU 2 HQGUHVLDQD FRXOG EH D UHVXOW RI ORQJHU DFWLYLW\ SDWWHUQV E\ VPDOO URGHQWV DQG PD\ EH W\SLFDO IRU PDQ\ ODUJHVHHGHG WUHH VSHFLHV LQ FORXG IRUHVWV 6HHGOLQJ 6XUYLYDO DQG 5HFUXLWPHQW 0RVW RI WKH VHHGV SURWHFWHG IURP URGHQWV JHUPLQDWHG +LJK JHUPLQDWLRQ UDWHV IRU /DXUDFHDH DUH DSSDUHQWO\ W\SLFDO :KHHOZULJKW Ef 6RPH VHHGV WKDW JHUPLQDWHG ZHUH HYHQWXDOO\ NLOOHG E\ EHHWOH ODUYDH +HLOLSXV VS &XUFXOLRQLGDHf GHYHORSLQJ LQ WKH FRW\OHGRQV $IWHU WKH FDJHV ZHUH UHPRYHG IURP WKH JURZLQJ VHHGOLQJV WKH VHHGV DV ZHOO DV WKH VHHGOLQJV ZHUH VXVFHSWLEOH WR SUHGDWLRQ E\ PDPPDOV ,W LV GLIILFXOW WR GHWHUPLQH WKH UROH RI HDFK VSHFLHV LQYROYHG EXW WKH VXLWH RI VSHFLHV WKDW NLOO VHHGOLQJV LV SUREDEO\ ODUJHU WKDQ WKH VXLWH RI VHHG SUHGDWRUV DQG FRXOG LQFOXGH SDFDV $JRXWL SDFDf EURFNHW GHHU 0DPPD DPHULFDQDf WDSLU 7DSLUXV EDLUGLLf DQG SHFFDULHV 7D\DVVX WDMDFXf LQ DGGLWLRQ WR DJRXWLV DQG VPDOOHU URGHQWV ,QVHFWV ZHUH PRVW OLNHO\ WR NLOO QRQGLVSHUVHG VHHGV RU VHHGOLQJV ZKLOH PDPPDOV NLOOHG DOO W\SHV RI VHHGV GLVSHUVHG QRQGLVSHUVHG DQG UDQGRPf 7KLV ILQGLQJ LV FRQVLVWHQW ZLWK +RZH Df DQG 7HUERUJK HW DO f ZKR IRXQG WKDW LQVHFWFDXVHG PRUWDOLW\ ZDV GLVWDQFHGHSHQGHQW EXW PRUWDOLW\ E\ PDPPDOV ZDV QRW ,Q 1HZ *XLQHD 0HUJ f IRXQG WKH RSSRVLWH SDWWHUQ LQVHFWV WHQGHG WR NLOO GLVSHUVHG VHHGV DQG PDPPDOV NLOOHG QRQ GLVSHUVHG VHHGV ,Q FRQWUDVW WR LQVHFWV DQG PDPPDOV IXQJDO SDWKRJHQV NLOOHG SURSRUWLRQDWHO\ PRUH 2 HQGUHVLDQD VHHGOLQJV LQ FORVHGFDQRS\ IRUHVW WKDQ LQ JDSV $XJVSXUJHU f DOVR VKRZHG IXQJDO SDWKRJHQV RI VHHGOLQJV ZHUH OHVV SUHYDOHQW LQ JDSV WKDQ IRUHVW XQGHUVWRU\ IRU QLQH VSHFLHV RI WURSLFDO WUHHV 7KH JUHDWHU KHLJKW RI VHHGOLQJV DQG VDSOLQJV LQ SORWV P IURP WKH SDUHQW WUHHV WKDQ LQ SORWV EHQHDWK WKH FURZQV VXJJHVWV WKDW VHHGOLQJ PRUWDOLW\ ZDV KLJKHU FORVHU WR WKH SDUHQW WUHHV 6HHGOLQJ DJHV DUH QRW NQRZQ EXW WKLV GLIIHUHQFH VXJJHVWV WKDW ORQJHUWHUP VXUYLYDO VHYHUDO \HDUV DIWHU GLVSHUVDO LV KLJKHU IRU VHHGOLQJV P IURP WKH FURZQ RI

PAGE 70

SDUHQW WUHHV WKDQ IRU VHHGOLQJV EHQHDWK FURZQV 7KXV VHHG GLVSHUVDO DZD\ IURP WKH FURZQ RI D FRQVSHFLILF DSSHDUV WR LQFUHDVH WKH FKDQFH RI UHFUXLWPHQW RYHU WKH ORQJ WHUP &ODUN DQG &ODUN /L HW DO f 7KH RYHUDOO SDWWHUQ RI UHFUXLWPHQW ZLWK UHVSHFW WR GLVWDQFH IURP SDUHQW WUHHV DQG FDQRS\ FRYHU ZDV ELPRGDO 7KLV SDWWHUQ ZDV FDXVHG E\ WKH FRPELQDWLRQ RI GLVWDQFH GHSHQGHQW VHHG SUHGDWLRQ )LJ Ef DQG WKH LQIOXHQFH RI FDQRS\ FRYHU RQ VHHGOLQJ VXUYLYDO )LJ Gf GHVSLWH WKH IDFW WKDW PRVW VHHGV ODQGHG FORVH WR WKH SDUHQW WUHHV LQ FORVHGFDQRS\ IRUHVW )LJ Df 7KLV SDWWHUQ LV DQ H[DPSOH RI VSDWLDO GLVFRUGDQFH FDXVHG E\ WKH ODFN RI FRQJUXLW\ DPRQJ WKH VWDJHV OHDGLQJ WR UHFUXLWPHQW +HUUHUD HW DO -RUGDQR f 7KH RFFXUUHQFH RI VXFK GLVFRUGDQFH HPSKDVL]HV WKH LPSRUWDQFH RI VWDJH VSHFLILF VXUYLYDO SDWWHUQV RQ SDWWHUQV RI UHFUXLWPHQW DQG WKH QHHG IRU GDWD RQ WKH VHTXHQWLDO VWDJHV RI SODQW UHSURGXFWLRQ UDWKHU WKDQ RQ RQO\ RQH RU D IHZ VWDJHV %HOOELUGV DUH FOHDUO\ DQ LPSRUWDQW SDUW RI WKH GLVSHUVDO V\VWHP RI 2 HQGUHVLDQD DV RYHU KDOI bf RI WKH VHHGV WKH\ GLVSHUVHG ODQGHG LQ D ]RQH RI KLJKHU UHFUXLWPHQW 6\QWKHVLV $GYDQWDJHV RI 'LVSHUVDO 'HVSLWH QHDUO\ FRPSOHWH SUHGDWLRQ DOO WKUHH DGYDQWDJHV RI GLVSHUVDO UHFHLYH VRPH VXSSRUW IRU 2FRWHD HQGUHVLDQD :LWK UHVSHFW WR WKH HVFDSH K\SRWKHVLV PRUWDOLW\ IRU ERWK VHHGV DQG VHHGOLQJV KDG D FRPSRQHQW RI GHQVLW\ DQGRU GLVWDQFH GHSHQGHQFH 3RVWn GLVSHUVDO VHHG SUHGDWLRQ ZDV KLJKHU IRU VHHGV GLUHFWO\ XQGHU WKH SDUHQWDO FURZQV WKDQ IRU VHHGV GLVSHUVHG DZD\ IURP WKH FURZQV )LJ f 6HHG SUHGDWLRQ ZDV KLJKHU IRU VHHGV FORVHU WR WKH WUHHV )LJ f $OVR RI VHHGV LQLWLDOO\ SURWHFWHG IURP PDPPDOLDQ VHHG SUHGDWRUV PRUH GLVSHUVHG WKDQ QRQGLVSHUVHG VHHGV HVWDEOLVKHG DV VHHGOLQJV DQG VXUYLYHG RQH \HDU )LJ f 2YHUDOO VHHGV GLVSHUVHG DZD\ IURP WKH SDUHQW WUHHV KDG D JUHDWHU SUREDELOLW\ RI VXUYLYDO DQG UHFUXLWPHQW 7KHVH UHVXOWV VXSSRUW WKH JHQHUDO SUHGLFWLRQV RI WKH -DQ]HQ&RQQHOO PRGHO RI KLJKHU VHHG DQG VHHGOLQJ PRUWDOLW\ QHDU WKH SDUHQW WUHH &RQQHOO -DQ]HQ f +RZHYHU UHPRYDO RI H[SHULPHQWDOO\ GLVSHUVHG VHHGV VKRZHG WKDW

PAGE 71

DOWKRXJK GLVWDQFH LV LPSRUWDQW LWV HIIHFW GHFUHDVHG RYHU WLPH EHFDXVH UHPRYDO DSSURDFKHG b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f DVVXPLQJ WKH SDWWHUQ RI VHHG SUHGDWLRQ DW ZN )LJ Ef 7KLV SHDN LV WKH RQH SUHGLFWHG E\ -DQ]HQ f DQG LW RFFXUV ZKHUH +XEEHOO f VXJJHVWHG LW ZRXOG DV D UHVXOW RI KLJK VHHG UDLQ FORVH WR WKH SDUHQW WUHHV DQG LQFRPSOHWH GHQVLW\GHSHQGHQW PRUWDOLW\ VHH DOVR &RQGLW HW DO f 7KXV IDFWRUV LQ DGGLWLRQ WR HVFDSH IURP VHHG SUHGDWLRQ PD\ EH LPSRUWDQW LQ UHFUXLWPHQW 'XULQJ D WUHHnV OLIHWLPH UHFUXLWPHQW PD\ EH HSLVRGLF ZKLFK LI WUXH FRXOG H[SODLQ ZK\ 2 HQGUHVLDQD VHHGOLQJV DUH IDLUO\ FRPPRQ GHVSLWH WKH KLJK OHYHOV RI VHHG SUHGDWLRQ REVHUYHG LQ WKH WZR \HDUV RI WKLV VWXG\ 3HUKDSV VWRFKDVWLF HYHQWV FRPELQH WR UHVXOW LQ ORZ SRSXODWLRQV RI VHHG SUHGDWRUV DQG DQ LQFUHDVH LQ VHHGOLQJ UHFUXLWPHQW RQO\ D IHZ WLPHV GXULQJ D WUHHnV UHSURGXFWLYH \HDUV /RQJWHUP VWXGLHV DUH QHFHVVDU\ WR WHVW +XEEHOOnV f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

PAGE 72

DFFRUGLQJO\ D VHHGOLQJ EDQN GHYHORSV LQ WKH XQGHUVWRU\ 0\ UHVXOWV DUH SUREDEO\ UHSUHVHQWDWLYH RI PDQ\ VKDGHWROHUDQW FDQRS\ WUHHV LQ WURSLFDO ZHW IRUHVWV +RZH D /LHEHUPDQ :KLWPRUH f %HFDXVH PRVW 2 HQGUHVLDQD VHHGV DUH GLVSHUVHG WR VLWHV ZLWKLQ FORVHG FDQRS\ IRUHVW PRVW VXUYLYLQJ VHHGOLQJV DUH SDUW RI WKH XQGHUVWRU\ VHHGOLQJ EDQN 'XULQJ WKUHH \HDUV RI LQWHQVLYH ILHOG ZRUN LQ WKLV VWXG\ VLWH RQO\ IRXQG WZR VXEDGXOW 2 HQGUHVLDQD P WDOOf DQG ERWK ZHUH LQ RU QHDU ROG \Uf WUHHIDOO JDSV 7KH GLUHFWHG GLVSHUVDO K\SRWKHVLV LV DOVR VXSSRUWHG 6SHFLHV WKDW KDYH GLUHFWHG GLVSHUVDO JDLQ DQ DGYDQWDJH LQ DUULYDO DW VSHFLILF VLWHV DVVRFLDWHG ZLWK D KLJKHU SUREDELOLW\ RI VXUYLYDO 7KH IHZ H[DPSOHV RI GLUHFWHG GLVSHUVDO LQ ELUGGLVSHUVHG SODQWV DUH IRU SDUDVLWLF HSLSK\WHV PLVWOHWRHVf WKDW FDQ HVWDEOLVK RQO\ RQ EUDQFKHV 'DYLGDU 5HLG 6DUJHQW /DUVRQ f )RU 2 HQGUHVLDQD PDOH EHOOELUGV SURYLGH GLUHFWHG GLVSHUVDO WR JDSV ZKHUH VHHGV KDYH D VOLJKWO\ ORZHU SUREDELOLW\ RI HDUO\ SUHGDWLRQ E\ PDPPDOV 7DEOH f DQG D VOLJKWO\ KLJKHU FKDQFH RI UHFUXLWPHQW WKDQ LQ IRUHVW XQGHUVWRU\ )LJ Hf 2QH PXVW WDNH LQWR DFFRXQW WKH IDFW WKDW PRVW VHHGV ODQG D IHZ PHWHUV EH\RQG WKH FURZQV RI WKH SDUHQW WUHHV DQG WKHUHIRUH WKLV ]RQH FRQWULEXWHV WKH PRVW LQGLYLGXDOV WR WKH VHHGOLQJVDSOLQJ VWDJH 6HHGV WKDW ODQG LQ D JDS HGJH ]RQH VHQVX 3RSPD HW DO f LQ WKLV FDVH RIWHQ DVVRFLDWHG ZLWK EHOOELUG VRQJ SHUFKHV JDLQ DQ DGYDQWDJH LQ VXUYLYDO DQG JURZWK 7KXV DOWKRXJK EHOOELUG SHUFKHV DUH QRW WKH RQO\ SODFHV VHHGOLQJV FDQ VXUYLYH D VHHG GLVSHUVHG E\ D EHOOELUG WR D JDS KDV D VOLJKWO\ KLJKHU FKDQFH RI JHUPLQDWLQJ DQG VXUYLYLQJ WKDQ RQH LQ WKH IRUHVW XQGHUVWRU\ $GGLWLRQDOO\ EHFDXVH JDSV LQ 0RQWHYHUGH WHQG WR H[SDQG ZKHQ WUHHV RQ WKH HGJH RI WKH JDS IDOO /DZWRQ DQG 3XW] f VHHGV GLVSHUVHG E\ EHOOELUGV WKDW JHUPLQDWH DQG VXUYLYH DUH PRUH OLNHO\ WR HQFRXQWHU D IDYRUDEOH JURZWK HQYLURQPHQW LH KLJKHU OLJKW OHYHOVf LQ WKH IXWXUH ZKHUHDV WKH SURVSHFWV IRU VHHGOLQJV LQ WKH XQGHUVWRU\ DUH XQSUHGLFWDEOH 7KHVH UHVXOWV DUH FRQVLVWHQW ZLWK RWKHU VWXGLHV WKDW VKRZ VKDGH WROHUDQW VSHFLHV WR EH FDSDEOH RI UHFUXLWPHQW XQGHU D ZLGHU UDQJH RI FRQGLWLRQV WKDQ VKDGHLQWROHUDQW SLRQHHU

PAGE 73

VSHFLHV &DQKDP 'HQVORZ DQG +DUWVKRUQ /LHEHUPDQ f ,Q PRQWDQH IRUHVWV VKDGHWROHUDQW VSHFLHV PD\ EH EHWWHU DEOH WR UHFUXLW LQ JDSV WKDQ LQ ORZODQG IRUHVWV EHFDXVH FORXG\ FRQGLWLRQV PRGHUDWH WKH WHPSHUDWXUH IOXFWXDWLRQV &DYHOLHU f 7KXV VHHGOLQJV RI VKDGHWROHUDQW VSHFLHV LQ PRQWDQH IRUHVWV PD\ JDLQ DQ DGYDQWDJH RI KLJKHU OLJKW OHYHOV LQ JDSV ZLWKRXW WKH LQFUHDVHG ULVN RI GHVLFFDWLRQ IRXQG LQ RWKHU UDLQ IRUHVWV %D]]D] DQG 3LFNHWW 9HHQHQGDDO HW DO :KLWPRUH f ,Q DGGLWLRQ VKDGH LQWROHUDQW SLRQHHU VSHFLHV DUH OHVV FRPPRQ DQG UHTXLUH ODUJHU JDSV ZLWK LQFUHDVLQJ HOHYDWLRQ LQ VRPH IRUHVWV :KLWPRUH f VXJJHVWLQJ WKDW WKH VHHGOLQJV RI QRQSLRQHHUV DUH PRUH FRPPRQ LQ PRQWDQH IRUHVW JDSV &RQFOXVLRQ 7KH RYHUDOO FRQFOXVLRQ RI WKLV VWXG\ LV WKDW WKH SDWWHUQ RI UHFUXLWPHQW RI 2 HQGUHVLDQD GHSHQGV RQ WKH FRPELQHG HIIHFWV RI VHHG GLVSHUVHUV VHHG SUHGDWRUV DQG VHHGOLQJ PRUWDOLW\ 6HOHFWLRQ RQ SODQW WUDLWV RFFXUV GXULQJ HDFK VWDJH DQG VHOHFWLRQ GXULQJ VHTXHQWLDO VWDJHV PD\ EH RSSRVHG :KHHOZULJKW DQG 2ULDQV +HUUHUD :KHHOZULJKW +HUUHUD HW DO f )RU H[DPSOH KLJK VHHG SUHGDWLRQ RYHUDOO DQG WKH VOLJKW SUHIHUHQFH IRU ODUJHU VHHGV E\ VHHG SUHGDWRUV PD\ VHOHFW IRU VPDOOHU VHHGV RU ODUJHU VHHG FURSV ZKLOH VHHG GLVSHUVHUV PD\ SUHIHU ODUJHU IUXLWV ZKLFK KDYH ODUJHU VHHGV EXW RFFXU LQ VPDOOHU IUXLW FURSV EXW VHH +RZH DQG 9DQGH .HUFNKRYH :KHHOZULJKW 0D]HU DQG :KHHOZULJKW f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f )XUWKHU VWXGLHV WKDW FRPSDUH

PAGE 74

GLVSHUVDO SDWWHUQV DQG WKH VXEVHTXHQW VWDJHV OHDGLQJ WR UHFUXLWPHQW DW GLIIHUHQW VLWHV DV ZHOO DV RYHU ORQJHU WLPH SHULRGV DUH HVSHFLDOO\ QHHGHG

PAGE 75

&+$37(5 6((' ',63(56$/ 2) $ +,*+48$/,7< )58,7 %< 63(&,$/,=(' )58*,925(6 +,*+48$/,7< ',63(56$/" ,W LV ZHOO NQRZQ WKDW PRVW /DXUDFHDH VHHGV DUH IUHH RI SUHGDWLRQ 2VFDU & &DVWUR f ,QWURGXFWLRQ ,Q D VHPLQDO SDSHU 0F.H\ f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fV IUDPHZRUN ZDV EDVHG RQ WKUHH LGHDV )LUVW WKH JHQHUDO VLPLODULW\ RI IUXLW FKDUDFWHULVWLFV RI SODQWV GLVSHUVHG E\ GLIIHUHQW W\SHV RI GLVSHUVDO DJHQWV LH GLVSHUVDO V\QGURPHV 5LGOH\ YDQ GHU 3LMO f 6HFRQG REVHUYDWLRQV WKDW IUXLW ZDV DQ LPSRUWDQW GLHWDU\ FRPSRQHQW IRU PDQ\ VSHFLHV EXW IHZ VSHFLHV ZHUH WRWDOO\ GHSHQGHQW RQ IUXLW 2ULDQV 0RUWRQ f 7KLUG 6QRZnV f LGHDV DERXW SRVVLEOH FRHYROXWLRQ EHWZHHQ SODQWV ZLWK ODUJHVHHGHG QXWULWLRXV IUXLWV DQG KLJKO\ IUXJLYRURXV ELUGV WKDW GLVSHUVHG WKH VHHGV 7KHVH LGHDV ZHUH IXUWKHU H[SDQGHG E\ +RZH DQG (VWDEURRN f WR

PAGE 76

LQFRUSRUDWH FURS VL]H SKHQRORJ\ DQG YLVLWDWLRQ UDWHV $OWKRXJK FRPSRQHQWV RI WKH VSHFLDOLVWJHQHUDOLVW IUDPHZRUN KDYH EHHQ FRQILUPHG GDWD WR IXOO\ WHVW LW DUH VWLOO ODFNLQJ +RZH E 6FKXSS DQG )XHQWHV f +RZH Ef VXJJHVWHG WKDW WKH PDLQ UHDVRQ WKH IUDPHZRUN UHPDLQV XQWHVWHG LV EHFDXVH SODQW HFRORJLVWV DQG ]RRORJLVWV VWXG\ GLIIHUHQW DVSHFWV RI WKH PXOWLSOH VWDJHV RI WKH SODQW UHFUXLWPHQW SURFHVV 4XDQWLI\LQJ GLVSHUVDO TXDOLW\ LV NH\ WR WHVWLQJ WKH PRGHO \HW WR H[DPLQH GLVSHUVDO TXDOLW\ LW LV QHFHVVDU\ WR VWXG\ ERWK WKH GLVSHUVDO SDWWHUQ DQG SRVWGLVSHUVDO IDWH RI VHHGV -DQ]HQ F +RZH E -RUGDQR DQG +HUUHUD 6FKXSS DQG )XHQWHV f $QRWKHU UHDVRQ WKDW WKH VSHFLDOLVWJHQHUDOLVW IUDPHZRUN KDV QRW EHHQ DGHTXDWHO\ WHVWHG LV EHFDXVH VSHFLHVVSHFLILF FRHYROXWLRQ EHWZHHQ SODQWV DQG GLVSHUVHUV LV QRZ FRQVLGHUHG XQOLNHO\ GLIIXVH FRHYROXWLRQ VHQVX -DQ]HQ f EHWZHHQ JURXSV RI SODQWV DQG JURXSV RI GLVSHUVHUV LV WKRXJKW WR EH PRUH W\SLFDO :KHHOZULJKW DQG 2ULDQV -DQ]HQ F +RZH +HUUHUD /HYH\ HW DO f 7KH GLIIXVH PXWXDOLVP SDUDGLJP LV EDVHG RQ IRXU PDLQ SRLQWV )LUVW EHFDXVH SODQWV RIIHU D UHZDUG WKH IUXLW SXOSf WR GLVSHUVHUV IRU IUXLW UHPRYDO EXW QR UHZDUG IRU VHHG GLVSHUVDO WR DQ DSSURSULDWH VLWH SODQWV KDYH OLWWOH FRQWURO RYHU ZKDW KDSSHQV WR VHHGV LQJHVWHG E\ SRWHQWLDO GLVSHUVHUV :KHHOZULJKW DQG 2ULDQV f 6HFRQG PRVW ODUJH KLJKO\ IUXJLYRURXV DQLPDOV HDW D ZLGH UDQJH RI IUXLWV LQFOXGLQJ ERWK KLJKTXDOLW\ DQG ORZTXDOLW\ VSHFLHV :KHHOZULJKW :KHHOZULJKW HW DO f 6LPLODUO\ VPDOOERGLHG IUXJLYRUHV FDQ EH KLJKO\ VHOHFWLYH DPRQJ WKH DYDLODEOH IUXLWV DQG VSHFLHV RI DSSUR[LPDWHO\ WKH VDPH ERG\ VL]H PD\ KDQGOH IUXLWV DQG VHHGV YHU\ GLIIHUHQWO\ 0RHUPRQG DQG 'HQVORZ 0RHUPRQG HW DO /HYH\ f 7KLUG PRVW SODQW VSHFLHV ZLWK IOHVK\ IUXLWV DUH HDWHQ DQG SUHVXPDEO\ GLVSHUVHGf E\ VHYHUDO WR PDQ\ VSHFLHV RI DQLPDOV RI GLIIHULQJ ERG\ VL]H VHHG KDQGOLQJ WHFKQLTXHV DQG PRYHPHQW SDWWHUQV :KHHOZULJKW HW DO %URQVWHLQ DQG +RIIPDQ -RUGDQR f )RXUWK SODQWV SUREDEO\ HYROYH DW GLIIHUHQW UDWHV WKDQ WKH DQLPDOV WKDW GLVSHUVH WKHP +HUUHUD f 7KHUHIRUH WKH RSSRUWXQLWLHV IRU GLVSHUVDO UHODWHG SODQW DQG DQLPDO WUDLWV WR FRHYROYH DW D VSHFLHVVSHFLILF OHYHO DUH PLQLPDO DQG QHLWKHU

PAGE 77

ERG\ VL]H QRU DPRXQW RI IUXLW LQ WKH GLHW FDQ EH XVHG WR PDNH DQ\ FRQVLVWHQW SUHGLFWLRQ DERXW WKH TXDOLW\ RI GLVSHUVDO SURYLGHG E\ WKDW VSHFLHV +HUUHUD E +RZH /HYH\ f $OWKRXJK PDQ\ VWXGLHV KDYH DGGUHVVHG GLVSHUVDO TXDOLW\ LQ WHUPV RI JXW WUHDWPHQW RI VHHGV .UHIWLQJ DQG 5RH 7UDYHVHW DQG :LOOVRQ :DKDM HW DO LQ SUHVVf IHZ KDYH H[DPLQHG GLVSHUVDO TXDOLW\ LQ WHUPV RI VXLWDELOLW\ IRU JURZWK DQG VXUYLYDO RI VLWHV ZKHUH VHHGV DUH GLVSHUVHG 6FKXSS f ,W LV UHDVRQDEOH WR H[SHFW VHHGV IURP DQ LQGLYLGXDO SODQW WR EH GLVSHUVHG WR D YDULHW\ RI VLWHV DQG LW LV ZHOO NQRZQ WKDW DOO SRWHQWLDO GLVSHUVDO VLWHV DUH QRW HTXDOO\ VXLWDEOH IRU VHHGOLQJ HVWDEOLVKPHQW RU JURZWK WR PDWXULW\ *UXEE +DUSHU 0XUUD\ %D]]D] 6FKXSS f 7KXV GLVSHUVDO TXDOLW\ PD\ YDU\ ZLWKLQ DQG DPRQJ FRQVSHFLILF WUHHV DV ZHOO DV DPRQJ GLVSHUVHUV 0DQ\ PRGHOV RI GLVSHUVDO DUH EDVHG RQ ZLQGGLVSHUVHG VSHFLHV *UHHQ *HULW] HW DO *UHHQH DQG -RKQVRQ 2NXER DQG /HYLQ $QGHUVHQ f EHFDXVH VR OLWWOH LV NQRZQ DERXW ZKHUH DQLPDOV GLVSHUVH VHHGV -DQ]HQ D :LOOVRQ /DPDQ Df 8QWLO UHFHQWO\ WKH GLIILFXOW\ LQ ILQGLQJ VHHGV GLVSHUVHG E\ DQLPDOV KDV OLPLWHG LQYHVWLJDWLRQV RI GLVSHUVDO TXDOLW\ LQ WHUPV RI WKH VXLWDELOLW\ RI GLVSHUVDO VLWHV IRU UHFUXLWPHQWf WR SDUDVLWLF PLVWOHWRHV ZKLFK KDYH VSHFLILF DQG HDVLO\ TXDQWLILDEOH VDIH VLWHV 'DYLGDU 5HLG 6DUJHQW f DQG WR DQWGLVSHUVHG VSHFLHV WKDW DUH GLVSHUVHG RYHU UHODWLYHO\ VPDOO VFDOHV +RUYLW] DQG 6FKHPVNH E +DQ]DZD HW DO f 'LVSHUVDO TXDOLW\ KDV EHHQ GHILQHG DV WKH SUREDELOLW\ WKDW D GLVSHUVHG VHHG ZLOO VXUYLYH WR UHSURGXFWLYH DJH 6FKXSS f 7KH SURGXFW RI GLVSHUVDO TXDOLW\ DQG GLVSHUVDO TXDQWLW\ QXPEHU RI VHHGV GLVSHUVHGf HTXDOV GLVSHUVHU HIIHFWLYHQHVV ZKLFK LV GHILQHG DV WKH SURSRUWLRQ RI VHHGOLQJV 5HLG f RU LGHDOO\ DGXOW SODQWV 6FKXSS f LQ D SRSXODWLRQ UHVXOWLQJ IURP DFWLYLWLHV RI D SDUWLFXODU GLVSHUVDO DJHQW VHH DOVR %XVWDPDQWH DQG &DQDOV / f 0RVW SUHYLRXV VWXGLHV KDYH H[SHULPHQWDOO\ H[DPLQHG VHHG DQG VHHGOLQJ VXUYLYDO ZLWKRXW FRQVLGHULQJ WKH DFWXDO SDWWHUQ RI VHHG GHSRVLWLRQ JHQHUDWHG E\ GLVSHUVHUV 7KHUHIRUH GLVSHUVDO TXDOLW\ FDQQRW EH DVVHVVHG +RZH E +HUUHUD HW DO 6FKXSS

PAGE 78

DQG )XHQWHV f $OWKRXJK GLVSHUVDO TXDQWLW\ KDV EHHQ VWXGLHG PRUH LQWHQVLYHO\ WKDQ GLVSHUVDO TXDOLW\ SUHOLPLQDU\ HVWLPDWHV IRU 9LUROD LQGLFDWH WKDW TXDOLW\ LV PRUH VWURQJO\ FRUUHODWHG ZLWK GLVSHUVHU HIIHFWLYHQHVV WKDQ LV TXDQWLW\ 6FKXSS f 2QH RI WKH SULPH H[DPSOHV RI D VSHFLDOL]HG GLVSHUVDO V\VWHP DQG LQGHHG DQ LQWHJUDO SDUW RI WKH GHYHORSPHQW RI VHHG GLVSHUVDO WKHRU\ LV WKH SODQW IDPLO\ /DXUDFHDH 6QRZ 0F.H\ :KHHOZULJKW DQG 2ULDQV f 0RVW ODXUDFHRXV WUHHV SURGXFH ODUJH RQHVHHGHG OLSLGULFK IUXLWV DQG DUH GLVSHUVHG E\ ODUJH KLJKO\ IUXJLYRURXV ELUGV VXFK DV EHOOELUGV &RWLQJLGDHf WURJRQV 7URJRQLGDHf DQG WRXFDQV 5DPSKDVWLGDHf WKURXJKRXW WKH 1HRWURSLFV 6QRZ :KHHOZULJKW $YLOD + HW DO f 6SHFLHV RI /DXUDFHDH DUH DOVR LPSRUWDQW IRU IUXJLYRURXV ELUGV LQ WKH 3DOHRWURSLFV &URPH 6XQ HW DO f ,Q WKLV VWXG\ H[DPLQHG VHHG GLVSHUVDO DQG VHHGOLQJ HVWDEOLVKPHQW RI D 1HRWURSLFDO /DXUDFHDH %HLOVFKPLHGLD SQGXOD KHUHDIWHU %HLOVFKPLHGLDf GLVSHUVHG E\ IRXU VSHFLHV RI KLJKO\ IUXJLYRURXV ELUGV %HLOVFKPLHGLD LV RQH RI WKH ODUJHVW ELUGGLVSHUVHG VHHGV WKURXJKRXW PXFK RI LWV JHRJUDSKLF UDQJH ,Q WKH PRQWDQH IRUHVWV RI QRUWKZHVWHUQ &RVWD 5LFD LW LV WKH ODUJHVW ELUGGLVSHUVHG VHHG UDQJH J PHDQ s 6' s J :KHHOZULJKW HW DO %XUJHU DQG YDQ GHU :HUII +DEHU HW DO f %HLOVFKPLHGLD SURGXFHV IUXLWV WKDW FOHDUO\ ILW WKH SDWWHUQ RI KLJKTXDOLW\ IUXLWV 7KH\ DUH ODUJH KDYH ORZ SXOSVHHG UDWLRV DQG DUH KLJK LQ OLSLGV UHODWLYH WR RWKHU VSHFLHV RI IUXLWV 6QRZ :KHHOZULJKW HW DO 0RHUPRQG DQG 'HQVORZ f )UXLWV DUH SURGXFHG SULRU WR DQG GXULQJ WKH EUHHGLQJ VHDVRQV RI WKH PDLQ GLVSHUVHUV DW D WLPH RI \HDU ZKHQ WKH QXPEHU RI WUHH VSHFLHV LQ IUXLW LV LQWHUPHGLDWH +DEHU HW DO f :LWKLQ WKH VWXG\ VLWH %HLOVFKPLHGLD IUXLW DYDLODELOLW\ RYHUODSV ZLWK IHZ RWKHU /DXUDFHDH VSHFLHV :KHHOZULJKW Df 7KH IRXU PDLQ VHHG GLVSHUVHUV DUH ODUJH GHSRVLW LQWDFW VHHGV DQG DUH UHOLDEOH FRQVXPHUV LQ WKH VHQVH WKDW DOO IRXU VSHFLHV DUH UHODWLYHO\ FRPPRQ DQG UHJXODUO\ HDW %HLOVFKPLHGLD IUXLWV :KHHOZULJKW :KHHOZULJKW HW DO E *XLQGRQ f 1DWXUDOLVW JXLGHV LQ WKH 0RQWHYHUGH &ORXG )RUHVW 3UHVHUYH QRWH WKDW IUXLWLQJ %HLOVFKPLHGLD WUHHV DUH SUHGLFWDEOH VSRWV WR ILQG ODUJH IUXJLYRURXV ELUGV ,Q \HDUV RU DUHDV

PAGE 79

ZKHQ WKH WUHHV GR QRW IUXLW TXHW]DOV GHOD\ EUHHGLQJ RU PRYH HOVHZKHUH 7 *XLQGRQ 5 *XLQGRQ $ 9LOOHJDV SHUVRQDO FRPPXQLFDWLRQVf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f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rn1 rn:f LQ WKH &RUGLOOHUD GH 7LODUDQ &RVWD 5LFD 7KLV KD SUHVHUYH LV DGPLQLVWHUHG E\ WKH 7URSLFDO 6FLHQFH &HQWHU RI 6DQ -RVH &RVWD 5LFD

PAGE 80

7KH VWXG\ DUHD ZDV LQ UHODWLYHO\ XQGLVWXUEHG ORZHU PRQWDQH UDLQ IRUHVW +DUWVKRUQ f DORQJ WKH FRQWLQHQWDO GLYLGH DW P HOHYDWLRQ $ KD DUHD P IURP WKH EHJLQQLQJ RI WKH 9DOOH\ 7UDLO 6HQGHUR (O 9DOOHf ZDV PDSSHG DQG PDUNHG LQWR [ P TXDGUDWV ZLWK 39& WXELQJ DW HYHU\ JULG SRLQW 7KH VLWHnV YHJHWDWLRQ LV FODVVLILHG DV OHHZDUG FORXG IRUHVW E\ /DZWRQ DQG 'U\HU f 7KH FDQRS\ LV P WDOO DQG GRPLQDWHG E\ VHYHUDO VSHFLHV RI /DXUDFHDH 6DSLXP ROLJRQHXURQ (XSKRELDFHDHf )LFXV FUDVVLXVFXOD 0RUDFHDHf ,QJD VSS /HJXPLQRVDHf DQG 3RXWHULD YLULGLV 6DSRWDFHDHf 7KH XQGHUVWRU\ LV GRPLQDWHG E\ 5XELDFHDH $FDQWKDFHDH *HVQHULDFHDH +HOLFRQLDFHDH DQG $UDFDFHDH 7KH YHJHWDWLRQ RI WKH DUHD LV GHVFULEHG LQ PRUH GHWDLO E\ /DZWRQ DQG 'U\HU f DQG 1DGNDPL HW DO f 2WKHU FKDUDFWHULVWLFV RI WKH 0RQWHYHUGH DUHD DUH GHVFULEHG E\ 1DGNDPL DQG :KHHOZULJKW LQ SUHVVf 7KH DYHUDJH DQQXDO UDLQIDOO DW P RQ WKH 3DFLILF VORSH DERXW NP IURP WKH VWXG\ VLWH LV DSSUR[LPDWHO\ PP ZLWK PRVW RI WKH SUHFLSLWDWLRQ RFFXUULQJ EHWZHHQ 0D\ DQG 1RYHPEHU $FWXDO UDLQIDOO LQ WKH VWXG\ VLWH ZDV SUREDEO\ JUHDWHU WKDQ PP GXH WR WKH KLJKHU HOHYDWLRQ RI WKH VWXG\ VLWH UHODWLYH WR WKH UDLQ JDXJHf EXW WKH VHDVRQDO SDWWHUQ ZDV VLPLODU 1DGNDPL DQG :KHHOZULJKW LQ SUHVVf 5DQJH JDXJHV XQGHUHVWLPDWH WKH DPRXQW RI SUHFLSLWDWLRQ IURP PLVW DQG FORXG LQWHUFHSWLRQ ZKLFK FRQWULEXWH XS WR b RI WKH SUHFLSLWDWLRQ LQ VRPH 1HRWURSLFDO PRQWDQH IRUHVWV &DYHOLHU f 7HPSHUDWXUHV UHFRUGHG DW WKH VWXG\ VLWH GXULQJ WKLV SURMHFW UDQJHG IURP r WR r& 6WXG\ 6SHFLHV %HLOVFKPLHGLD SQGXOD >6Zf +HPVOH\@ LV D FRPPRQ FDQRS\ WUHH LQ &RVWD 5LFDQ PRQWDQH IRUHVWV IURP P %XUJHU DQG YDQ GHU :HUII f ,Q WKH 0RQWHYHUGH DUHD LW RFFXUV IURP P +DEHU HW DO f %HLOVFKPLHGLD EHJLQV IORZHULQJ LQ WKH ODWH GU\ VHDVRQ 0DUFKf DQG LV SROOLQDWHG E\ VPDOO IOLHV DQG RWKHU LQVHFWV 5LSH IUXLWV DUH DYDLODEOH IURP PLG-DQXDU\ WKURXJK ODWH $SULO )UXLWV KDYH EODFN VNLQ DQG OLSLGULFK SXOS :KHHOZULJKW HW DO %XUJHU DQG YDQ GHU :HUII f 0RVW RI WKH YROXPH RI

PAGE 81

WKH IUXLW LV D VLQJOH VHHG 4HQJWK s PP ZLGWK PP PDVV s J PHDQ s 6' 1 f WKDW LV FRPSRVHG RI D VPDOO HPEU\R DQG WZR ODUJH FRW\OHGRQV IRU DGGLWLRQDO PHDVXUHPHQWV VHH 0D]HU DQG :KHHOZULJKW f 6HHG VL]H YDULHV IURP WR J &RPSDUHG WR RWKHU JHQHUD RI /DXUDFHDH %HLOVFKPLHGLD KDV D UHODWLYHO\ WKLFN PPf HQGRFDUS DQG WKH SXOS LV PRUH WLJKWO\ DWWDFKHG WR WKH VHHG 6HYHQ %HLOVFKPLHGLD WUHHV ZHUH LQ WKH KD VWXG\ VLWH 'DWD IRU WZR WUHHV WKDW KDG DGMDFHQW FURZQV ZHUH SRROHG IRU DQDO\VHV )UXLWV DQG VHHGV IRU VRPH H[SHULPHQWV ZHUH FROOHFWHG IURP WUHHV RQ WKH SHULSKHU\ RI WKH PDLQ VWXG\ DUHD /DUJH IUXLW FURSV s IUXLWVWUHHf ZHUH SURGXFHG E\ WKH VHYHQ WUHHV LQ EXW QRW LQ RU VHH DOVR :KHHOZULJKW f %HLOVFKPLHGLD IUXLWV DUH HDWHQ SULPDULO\ E\ IRXU VSHFLHV RI ELUGV HPHUDOG WRXFDQHW 5DPSKDVWLGDH $XODFRUK\QFKXV SUDVLQXVf UHVSOHQGHQW TXHW]DO 7URJRQLGDH 3KDURPDFKUXV PRFLUPRf WKUHHZDWWOHG EHOOELUG &RWLQJLGDH 3URFQLDV WULFDUXQFXODWDf DQG EODFN JXDQ &UDFLGDH &KDPDHSHWHV XQLFRORUf 7KH ILUVW WKUHH VSHFLHV W\SLFDOO\ UHPDLQ LQ D IUXLWLQJ WUHH DIWHU HDWLQJ VHYHUDO IUXLWV DQG RIWHQ UHJXUJLWDWH YLDEOH VHHGV PLQ ODWHU XQGHU WKH VDPH WUHH RU QHDUE\ :KHHOZULJKW f *XDQV GHIHFDWH VHHGV LQ YLDEOH FRQGLWLRQ DQG WKH\ JHQHUDOO\ OHDYH D IUXLWLQJ WUHH EHIRUH GHIHFDWLQJ WKH VHHGV IURP WKDW IRUDJLQJ ERXW $OWKRXJK VHHG UHWHQWLRQ WLPHV E\ JXDQV ZHUH QRW UHFRUGHG LQ WKLV VWXG\ *XL[ DQG 5XL] f UHSRUWHG D PHGLDQ UHWHQWLRQ WLPH RI KU IRU D ODUJHU FUDFLG ^3HQHORSH REVFXUDf $OO IRXU VSHFLHV FDQ EH FRQVLGHUHG IUXLW VSHFLDOLVWV LQ WKH VHQVH WKDW WKH\ GHSHQG RQ IUXLW IRU PRVW LI QRW DOO RI WKHLU GLHWDU\ UHTXLUHPHQWV DW OHDVW DW VRPH WLPHV RI \HDU :KHHOZULJKW $YLOD + HW DO f +RZHYHU TXHW]DOV DQG WRXFDQHWV DOVR HDW ODUJH LQVHFWV DQG VPDOO YHUWHEUDWHV 6NXWFK :KHHOZULJKW 5LOH\ DQG 6PLWK $YLOD + HW DO f ZKLOH JXDQV DOVR HDW OHDYHV +DEHU HW DO f %HOOELUGV DSSDUHQWO\ HDW RQO\ IUXLWV DOWKRXJK GLHWV RI IHPDOH EHOOELUGV DUH SRRUO\ NQRZQ 6QRZ f 'URSSHG RU IDOOHQ IUXLWV DUH HDWHQ E\ DJRXWLV ^'DV\SURFWD SXQFWDWDf DQG SRVVLEO\ SDFDV ^$JRXWL SDFDf DQG RWKHU URGHQWV $JRXWLV FKHZ RII SXOS DQG VRPHWLPHV WKH

PAGE 82

HQGRFDUS DQG OHDYH WKH VHHGV XQGHU WKH WUHHV EXW GR QRW HDW RU EXU\ VHHGV 7KXV WKHVH VSHFLHV SUREDEO\ GR QRW SURYLGH VLJQLILFDQW GLVSHUVDO IRU %HLOVFKPLHGLD 0HWKRGV 6HHG 'LVSHUVDO 6HHGV ZHUH ORFDWHG E\ V\VWHPDWLFDOO\ VHDUFKLQJ WKH JURXQG IRU IUHVKO\ UHJXUJLWDWHG RU GHIHFDWHG VHHGV IURP ODWH -DQXDU\ WKURXJK PLG$SULO 7KH JURXQG VHDUFKHV VWDUWHG DW WKH EDVH RI D IUXLWLQJ WUHH DQG SURFHHGHG DORQJ P ZLGH WUDQVHFWV GHOLQHDWHG E\ WKH 39& PDUNHUVf WR P IURP HDFK %HLOVFKPLHGLD WUXQN $GGLWLRQDO VLWHV DZD\ IURP WKH IRFDO WUHHV ZHUH DOVR VHDUFKHG VRPH ZHUH UDQGRPO\ VHOHFWHG DQG RWKHUV ZHUH FKRVHQ GXH WR ELUG DFWLYLW\ LQ WKH DUHD ,W ZDV LPSRVVLEOH WR VHDUFK WKH HQWLUH VLWH ZLWK HTXDO LQWHQVLW\ EXW DQ HIIRUW ZDV PDGH WR FRYHU WKH HQWLUH VLWH DW OHDVW RQFH HYHU\ WKUHH ZHHNV VR WKDW RYHU WKH FRXUVH RI WKH WKUHHPRQWK IUXLWLQJ VHDVRQ HDFK [ P SORW ZDV FKHFNHG DW OHDVW IRXU WLPHV 'HIHFDWHG VHHGV ZHUH DVVXPHG WR EH GLVSHUVHG E\ EODFN JXDQV EHFDXVH JXDQV DUH WKH RQO\ DYLDQ GLVSHUVHU LQ WKH DUHD WKDW GHIHFDWHV ODUJH VHHGV 6LWHV ZHUH FODVVLILHG DV QRQ GLVSHUVHG LI GLUHFWO\ XQGHU WKH FURZQ RI D IUXLWLQJ %HLOVFKPLHGLD WUHH RU DV GLVSHUVHG LI QRW XQGHU VXFK D WUHH $OO WKH GLVSHUVHG DQG QRQGLVSHUVHG VHHGV DV ZHOO DV WKH VHHGV SODFHG DW UDQGRP VLWHV VHH EHORZf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

PAGE 83

*HUPLQDWLRQ DQG 6HHGOLQJ 6XUYLYDO 7KH RULJLQDO GLVSHUVHG DQG QRQGLVSHUVHG VHHGV DQG DOO VHHGV SODFHG DW UDQGRPO\ ORFDWHG VLWHV ZHUH SURWHFWHG E\ D [ [ FP FDJH PDGH RI FP JDOYDQL]HG ZLUH PHVK KHOG LQ SODFH E\ WZR FP PHWDO VWDNHV &DJHG VHHGV ZHUH XVHG WR GHWHUPLQH JHUPLQDWLRQ UDWHV DQG LQVHFW SUHGDWLRQ UDWHV LQ WKH DEVHQFH RI PDPPDOLDQ VHHG SUHGDWRUV (DFK VLWH ZDV FKHFNHG ZHHNO\ IRU DW OHDVW ZHHNV DQG DW PR DIWHU GLVSHUVDO KHUHDIWHU \U VXUYLYDOf *HUPLQDWLRQ ZDV GHILQHG DV WKH VSOLWWLQJ RI WKH VHHG FRDW DQG VSUHDGLQJ RI WKH FRW\OHGRQV 7\SLFDOO\ WKH UDGLFOH KDG HPHUJHG E\ WKH FHQVXV DIWHU JHUPLQDWLRQ DQG D ZHHN ODWHU WKH VWHP ZDV YLVLEOH $V HDFK VHHG JHUPLQDWHG DQG WKH VKRRW EHJDQ WR JURZ WKH FDJH ZDV UHPRYHG WR DOORZ QRUPDO VHHGOLQJ JURZWK 7KH VHHGOLQJ ORFDWLRQ ZDV PDUNHG ZLWK RQH RI WKH VWDNHV IURP WKH FDJH &DXVHV RI VHHGOLQJ PRUWDOLW\ ZHUH FODVVLILHG DV PDPPDO LQVHFW IXQJDO SDWKRJHQ SK\VLFDO RU XQNQRZQ 0DPPDOV HLWKHU DWH WKH VHHG DQG OHIW WKH GDPDJHG VKRRW EHKLQG VHHG SUHGDWRUVf RU UHPRYHG WKH HQWLUH VKRRW KHUELYRUHVf 6RPH VHHGV WKDW DSSHDUHG WR KDYH JHUPLQDWHG ZHUH HYHQWXDOO\ NLOOHG E\ EHHWOHV LQVLGH WKH FRW\OHGRQV ,QVHFWNLOOHG VHHGV IUHTXHQWO\ GHYHORSHG D URRW EXW QHYHU KDG D VKRRW FP WDOO 6HHGOLQJV NLOOHG E\ IXQJDO SDWKRJHQV ZHUH FKDUDFWHUL]HG E\ D ZLOWHG DQG GLVFRORUHG VKRRW $XJVSXUJHU f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f )RU WKLV WUHDWPHQW XVHG UHJXUJLWDWHG VHHGV FROOHFWHG XQGHU IUXLWLQJ WUHHV DGMDFHQW WR WKH VWXG\ VLWH

PAGE 84

6HHGV ZHUH PDUNHG E\ JOXLQJ FP RI XQZD[HG GHQWDO IORVV WR WKH VHHG DQG W\LQJ DERXW FP RI IODJJLQJ WDSH WR WKH GLVWDO HQG RI WKH IORVV %HFDXVH WKH JOXH KHOG EHVW RQ VHHGV ZLWK D GU\ VHHG FRDW VHHGV ZHUH WDNHQ LQVLGH DQG DOORZHG WR GU\ IRU K EHIRUH JOXLQJ (DFK PDUNHG VHHG ZDV SODFHG DW D VLWH WKH QH[W PRUQLQJ $OO PDUNHG VHHGV ZHUH FHQVXVHG RQ GD\V DQG DQG RQFH HDFK ZHHN r DIWHUZDUGV XQWLO ZHHN ,I D PDUNHG VHHG ZDV UHPRYHG WKH VXUURXQGLQJ DUHD ZDV VHDUFKHG WR ILQG WKH IODJJLQJ WDSHGHQWDO IORVV DVVHPEO\ 7KH HQG RI WKH IORVV ZKHUH WKH VHHG KDG EHHQ DWWDFKHG ZDV H[DPLQHG WR GHWHUPLQH WKH IDWH SUHVHQW RU DEVHQWf RI WKH VHHG ,Q WKH IHZ FDVHV ZKHUH VHHGV ZHUH SDUWLDOO\ FRQVXPHG WHHWK RU ELOO PDUNV ZHUH H[DPLQHG WR LGHQWLI\ WKH FRQVXPHU 6HHGOLQJ DQG 6DSOLQJ 3ORWV 7R GHWHUPLQH LI %HLOVFKPLHGLD VHHGOLQJV DQG VDSOLQJV DUH PRUH OLNHO\ WR UHFUXLW XQGHU RU DZD\ IURP FRQVSHFLILFV VHHGOLQJV DQG VDSOLQJV XS WR P KHLJKWf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f /HDI OLWWHU GHSWK ZDV PHDVXUHG DV WKH QXPEHU RI OHDYHV SLHUFHG E\ D

PAGE 85

PHWDO VWDNH WKUXVW LQWR WKH VRLO DW WKH VLWH 9HJHWDWLRQ GHQVLW\ ZDV WKH QXPEHU RI VWHPV ZLWKLQ D FP UDGLXV RI WKH VLWH 7KH GLVWDQFHV WR WKH QHDUHVW ZRRG\ VWHP WUHH FP '%+ GLDPHWHU DW EUHDVW KHLJKWf FURZQ HGJH RI IUXLWLQJ %HLOVFKPLHGLD WUHH DQG IDOOHQ ORJ ZHUH PHDVXUHG ZLWK D ILEHUJODVV PHDVXULQJ WDSH 6WDWLVWLFDO $QDO\VHV 'DWD ZHUH DQDO\]HG ZLWK WHVWV IURP 6$6 -03 6$6 ,QVWLWXWH f 3DUDPHWULF WHVWV ZHUH XVHG XQOHVV GDWD YLRODWHG WKH DVVXPSWLRQV RI QRUPDOLW\ DQG HTXDO YDULDQFH LQ ZKLFK FDVH QRQSDUDPHWULF SURFHGXUHV ZHUH XVHG 6XUYLYDO SURSRUWLRQV ZHUH DUFVLQ VTXDUH URRW WUDQVIRUPHG EHIRUH DQDO\VLV 7KURXJKRXW WKLV SDSHU PHDQ YDOXHV DUH IROORZHG E\ s 6' 5HVXOWV 2YHU WKH WKUHHPRQWK IUXLWLQJ VHDVRQ UHJXUJLWDWHG DQG GHIHFDWHG VHHGV ZHUH IRXQG 2QH KXQGUHG WZHQW\QLQH bf RI WKHVH ZHUH GHSRVLWHG EH\RQG WKH FURZQV RI IUXLWLQJ WUHHV GLVSHUVHGf ZKLOH bf ZHUH GLUHFWO\ XQGHU WKH WUHHV QRQGLVSHUVHGf 2I WKH GLVSHUVHG VHHGV b ODQGHG ZLWKLQ P RI FURZQ HGJH EXW VRPH VHHGV ZHUH IRXQG XS WR P DZD\ )LJ f 'LVSHUVDO GLVWDQFH IRU GLVSHUVHG VHHGVf ZDV QRW FRUUHODWHG ZLWK VHHG PDVV 6SHDUPDQnV 5KR 3 f OHQJWK 5KR 3 f RU ZLGWK 5KR 3 f 6HHGV GHIHFDWHG SUHVXPDEO\ E\ JXDQVf DYHUDJHG OLJKWHU :LOFR[RQ ;A 3 f DQG VKRUWHU ;A 3 f EXW QRW QDUURZHU ;A 3 f WKDQ VHHGV UHJXUJLWDWHG E\ WKH RWKHU WKUHH VSHFLHV 7DEOH f 'HIHFDWHG VHHGV ZHUH GLVSHUVHG VLJQLILFDQWO\ IDUWKHU s Pf IURP WKH SDUHQWDO FURZQV WKDQ UHJXUJLWDWHG VHHGV s P :LOFR[RQ ;A 3 f 9HU\ IHZ PDUNHG VHHGV bf ZHUH HDWHQ RU UHPRYHG DIWHU GLVSHUVDO RU SODFHPHQW DW UDQGRP VLWHVf DQG UHPRYDO UDWHV GLG QRW GLIIHU DPRQJ GLVSHUVHG QRQGLVSHUVHG RU

PAGE 86

XL L8 '& '& /8 D FR 2 /8 8L LQ ',67$1&( )520 &52:1 ('*( Pf )LJXUH 7KH DYHUDJH QXPEHU RI UHJXUJLWDWHG RU GHIHFDWHG %HLOVFKPLHGLD VHHGV 6'f DW P LQWHUYDOV IURP FURZQ HGJHV RI VL[ IUXLWLQJ %HLOVFKPLHGLD WUHHV 7KH ILUVW FDWHJRU\ GLVWDQFH f LQFOXGHV DOO WKH QRQGLVSHUVHG VHHGV ZKLOH WKH UHPDLQLQJ FDWHJRULHV LQFOXGH RQO\ GLVSHUVHG VHHGV

PAGE 87

7DEOH $YHUDJH s 6'f PDVV OHQJWK DQG ZLGWK RI %HLOVFKPLHGLD VHHGV GHIHFDWHG SUHVXPDEO\ E\ EODFN JXDQVf RU UHJXUJLWDWHG E\ TXHW]DOV WRXFDQHWV DQG EHOOELUGV 6L]HV RI UHJXUJLWDWHG DQG GHIHFDWHG VHHGV ZHUH FRPSDUHG ZLWK :LOFR[RQ UDQN VXP WHVW 0HDVXUHPHQWV IRU DOO VHHGV LQFOXGLQJ VHHGV SODFHG DW UDQGRP VLWHV 1 f DUH VKRZQ IRU FRPSDULVRQ 'HIHFDWHG 1 5HJXUJLWDWHG $OO 6HHGV 0DVV Jf /HQJWK PPf :LGWK PPf f f f frr f frr f fQV f rr 3 QV 3

PAGE 88

UDQGRP VLWHV VXUYLYDO DQDO\VLV :LOFR[RQ ; GI 3 )LJ f 7KH PRVW FRQVSLFXRXV SRVWGLVSHUVDO VHHG SUHGDWRUV ZHUH WZR VSHFLHV RI EHHWOHV RQH RI ZKLFK 1LWLGXODHf EXULHG WKH VHHGV DQG IHG XSRQ WKH URWWLQJ FRW\OHGRQDU\ UHVHUYHV 7KH VHFRQG EHHWOH VSHFLHV &XUFXOLRQLGDHf FRQVXPHG WKH VHHG E\ WXQQHOLQJ WKURXJK WKH HQGRFDUS WR UHDFK WKH VHHG 1RQH RI WKH VHHGV EXULHG E\ EHHWOHV 1 bf SURGXFHG VHHGOLQJV DOWKRXJK D IHZ JHUPLQDWHGf DQG WKXV WKH EHHWOHV OLNHO\ GR QRW DFW DV VHFRQGDU\ GLVSHUVHUV 6PDOO URGHQWV UHPRYHG HLJKW VHHGV bf DQG WRRN WKHP LQWR EXUURZV DW OHDVW FP GHHS 6XFK D GHSWK ZRXOG SUHFOXGH VHHGOLQJ HVWDEOLVKPHQW 7KHUHIRUH URGHQWV SUREDEO\ DUH QRW VHFRQGDU\ GLVSHUVHUV RI %HLOVFKPLHGLD %ODFNEUHDVWHG ZRRGTXDLO 2GRQWRSKRUXV OHXFRODHPXVf SHFNHG DSDUW VHYHQ VHHGV EDVHG RQ RQH GLUHFW REVHUYDWLRQ DQG ELOO PDUNV OHIW RQ SLHFHV RI VHHGVf $Q DGGLWLRQDO PDUNHG VHHGV ZHUH UHPRYHG EXW ZHUH QRW IRXQG 1HLWKHU ZLGWK OHQJWK RU PDVV ZHUH VLJQLILFDQW HIIHFWV LQ ORJLVWLF UHJUHVVLRQ PRGHOV SUHGLFWLQJ ZN VHHG VXUYLYDO 7DEOH f 6HHGV DW GLVSHUVHG DQG UDQGRPO\ ORFDWHG VLWHV ZHUH PRUH OLNHO\ WR VXUYLYH ZN WKDQ VHHGV DW QRQGLVSHUVHG VLWHV :DOG ; 3 f DQG VLPLODUO\ VHHG VXUYLYDO ZDV SRVLWLYHO\ FRUUHODWHG ZLWK GLVSHUVDO GLVWDQFH :DOG ; 3 7DEOH f 6HHG VXUYLYDO ZDV DOVR SRVLWLYHO\ FRUUHODWHG ZLWK DPRXQW RI OHDI OLWWHU :DOG ;" 3 f YHJHWDWLRQ GHQVLW\ :DOG ; 3 f DQG GDWH RI GLVSHUVDO :DOG ;A 3 f 9LUWXDOO\ DOO FDJHG VHHGV LQLWLDWHG JHUPLQDWLRQ bf DQG WKH PDMRULW\ bf KDG HVWDEOLVKHG VHHGOLQJV E\ ODWH -XO\ WKUHH WR ILYH PRQWKV DIWHU GLVSHUVDO 7KH DYHUDJH SURSRUWLRQ SHU WUHH RI VHHGV WKDW JHUPLQDWHG HVWDEOLVKHG VHHGOLQJV RU VXUYLYHG RQH \HDU GLG QRW GLIIHU DPRQJ WKH WKUHH WUHDWPHQWV .UXVNDO:DOOLV WHVWV 3nV )LJ f 2YHUDOO KRZHYHU WKH QXPEHU RI QRQGLVSHUVHG VHHGV WKDW HVWDEOLVKHG VHHGOLQJV ZDV JUHDWHU WKDQ H[SHFWHG LI DOO WKUHH WUHDWPHQWV KDG HTXDO VXUYLYDO ;A GI 3 f 6HHG PDVV ZDV QRW FRUUHODWHG ZLWK WLPH WR JHUPLQDWLRQ DIWHU GLVSHUVDO U 3 f EXW VKRRWV IURP ODUJHU VHHGV JUHZ IDVWHU DIWHU JHUPLQDWLRQ U 3 f

PAGE 89

3523257,21 5(0$,1,1* )LJXUH 3RVWGLVSHUVDO SUHGDWLRQ RI VHHGV IURP GLVSHUVHG 1 f QRQGLVSHUVHG 1 f DQG UDQGRP 1 f ORFDWLRQV 5HPRYDO UDWHV GLG QRW GLIIHU DPRQJ WKH WKUHH WUHDWPHQWV

PAGE 90

7DEOH 5HVXOWV RI ORJLVWLF UHJUHVVLRQV RI SRVWGLVSHUVDO VXUYLYDO RI PDUNHG VHHGV DIWHU ZHHNV DQG RQH\HDU VXUYLYDO RI VHHGV LQLWLDOO\ FDJHG DJDLQVW KDELWDW YDULDEOHV 2QO\ VLJQLILFDQW HIIHFWV DUH OLVWHG 1RQVLJQLILFDQW YDULDEOHV LQFOXGHG VHHG PDVV VHHG OHQJWK VHHG ZLGWK FDQRS\ FRYHU GLVWDQFH WR ORJ GLVWDQFH WR ZRRG\ VWHP DQG GLVWDQFH WR FP WUHH 5HVSRQVH 5 ORJ OLNHOLKRRG [f SUHGLFWRUVD ZN VHHG VXUYLYDO rrr OHDI OLWWHUr GLVSHUVDO GDWHrrr YHJHWDWLRQ GHQVLW\rrr GLVWDQFH IURP SDUHQWr W\SH 8 5fr \U VHHGOLQJ VXUYLYDO rrr PR VHHGOLQJ KHLJKWrr GLVSHUVDO GDWHrrr r3A rr3A rrr3 DVLJQ LQ IURQW RI HDFK SUHGLFWRU LQGLFDWHV SRVLWLYH RU QHJDWLYH FRUUHODWLRQ ZLWK WKH UHVSRQVH

PAGE 91

3523257,21 6859,9,1* *(50,1$7,21 (67$%/,6+0(17 <5 6859,9$/ )LJXUH 3URSRUWLRQ RI VHHGV JHUPLQDWLQJ HVWDEOLVKLQJ VHHGOLQJV PR DIWHU GLVSHUVDO DQG VXUYLYLQJ RQH \HDU DW GLVSHUVHG 1 f QRQGLVSHUVHG 1 f DQG UDQGRP 1 f ORFDWLRQV 7KH DYHUDJH SURSRUWLRQ VXUYLYLQJ HDFK VWDJH GLG QRW GLIIHU DPRQJ WKH WKUHH WUHDWPHQWV .UXVNDO:DOOLV WHVWV 3nV f

PAGE 92

6HHGV WKDW HVWDEOLVKHG VHHGOLQJV E\ -XO\ DYHUDJHG VLJQLILFDQWO\ KHDYLHU s J 1 f WKDQ VHHGV WKDW GLG QRW HVWDEOLVK VHHGOLQJV s J 1 W 3 f EXW VHHGOLQJV WKDW VXUYLYHG RQH \HDU s J 1 f ZHUH QRW IURP ODUJHU VHHGV WKDQ WKRVH VHHGOLQJV WKDW GLG QRW VXUYLYH s J 1 W 3 f 2I WKH VHHGOLQJV WKDW VXUYLYHG RQH \HDU VHHGOLQJ KHLJKW ZDV SRVLWLYHO\ FRUUHODWHG ZLWK LQLWLDO VHHG PDVV 1 U 3 f 7KH SUHGRPLQDQW VRXUFH RI VHHGOLQJ PRUWDOLW\ ZDV IXQJDO SDWKRJHQV DOWKRXJK PDQ\ VHHGV UHVSURXWHG VHYHUDO WLPHV HYHQ DIWHU IXQJDO DWWDFN )XQJDO SDWKRJHQV NLOOHG PRUH VHHGOLQJV GLUHFWO\ EHQHDWK WKH SDUHQWDO FURZQV QRQGLVSHUVHGf WKDQ VHHGOLQJV DW GLVSHUVHG RU UDQGRP ORFDWLRQV ;A GI 3 f $V IRU WKH PDUNHG VHHGV VRPH FDJHG VHHGV bf ZHUH EXULHG RU HDWHQ E\ EHHWOHV +HUELYRU\ E\ LQVHFWV DQG PDPPDOV ZDV SUREDEO\ WKH SULPDU\ FDXVH RI VHHGOLQJ PRUWDOLW\ DIWHU VHHGOLQJ HVWDEOLVKPHQW EXW WKH HIIHFW RI KHUELYRU\ ZDV GLIILFXOW WR TXDQWLI\ EHFDXVH WKH VHHGOLQJV ZHUH H[DPLQHG WRR LQIUHTXHQWO\ 6HHGOLQJ VXUYLYDO DW RQH \HDU ZDV SUHGLFWHG E\ RQO\ WZR YDULDEOHV LQ WKH ORJLVWLF UHJUHVVLRQ DQDO\VLV 7DEOH f ,Q FRQWUDVW WR VHHG VXUYLYDO DW ZHHNV RQH\HDU VHHGOLQJ VXUYLYDO ZDV QHJDWLYHO\ FRUUHODWHG ZLWK GLVSHUVDO GDWH ;A 3 f 6HHGOLQJ VXUYLYDO ZDV SRVLWLYHO\ FRUUHODWHG ZLWK VHHGOLQJ KHLJKW DW PR ;A 3 7DEOH f 7KH DEXQGDQFH RI VHHGOLQJV DQG VDSOLQJV XS WR P KHLJKWf ZDV KLJKHU LQ SORWV FORVH WR WKH IUXLWLQJ WUHHV SDLUHG LWHVW GI 3 f EXW WKH PHGLDQ KHLJKW RI LQGLYLGXDOV ZDV JUHDWHU LQ WKH SORWV P DZD\ IURP WKH WUHHV :LOFR[RQ 5DQN 6XPV ; GI 3 )LJ f 0RVW VHHGOLQJV bf LQ WKH SORWV FORVH WR IUXLWLQJ WUHHV ZHUH OHVV WKDQ FP LQ KHLJKW ZKLOH PRVW LQGLYLGXDOV bf LQ WKH SORWV IDU IURP DGXOW WUHHV ZHUH RYHU FP DQG UDQJHG XS WR P %HFDXVH WKH %HLOVFKPLHGLD WUHHV LQ WKH VWXG\ VLWH GLG QRW SURGXFH IUXLWV LQ WKH \RXQJHVW LQGLYLGXDOV ZHUH DW OHDVW \U ROG

PAGE 93

&/26( )$5 &/26( )$5 )LJXUH $YHUDJH QXPEHU DQG KHLJKW RI VHHGOLQJV DQG VDSOLQJV XS WR P WDOO LQ SDLUHG [ P SORWV 1 SDLUVf XQGHU DQG P DZD\ IURP %HLOVFKPLHGLD FURZQV $VWHULVNV LQGLFDWH VLJQLILFDQW GLIIHUHQFHV rr3 r3 f

PAGE 94

)LJXUH $YHUDJH 6'f SURSRUWLRQ RI VHHGV VXUYLYLQJ SRVWGLVSHUVDO VHHG SUHGDWRUV Df JHUPLQDWLQJ Ef HVWDEOLVKLQJ VHHGOLQJV Ff DQG VXUYLYLQJ RQH \HDU Gf DV IXQFWLRQV RI GLVWDQFH IURP WKH FURZQ HGJH 6HHGV GHSRVLWHG GLUHFWO\ EHQHDWK WKH SDUHQWDO FURZQV QRQ GLVSHUVHGf DUH LQFOXGHG LQ GLVWDQFH 7KH SUREDELOLW\ RI UHFUXLWPHQW Hf ZDV FDOFXODWHG DV WKH SURGXFW RI HDFK RI WKH IRXU SUHYLRXV VWDJHV IRU HDFK WUHH 'LIIHUHQW OHWWHUV DERYH EDUV LQGLFDWH VLJQLILFDQWO\ GLIIHUHQW PHDQV ^3 f EDVHG RQ $129$ DQG )LVKHUnV /6' RQ DUFVLQWUDQVIRUPHG GDWD 3DQHOV ZLWK QR OHWWHUV KDG QR VLJQLILFDQW GLIIHUHQFHV

PAGE 95

3523257,21 6859,9,1* 3523257,21 6859,9,1* 6((' 6859,9$/ *(50,1$7,21 % (67$%/,6+0(17 <($5 6859,9$/ 5(&58,70(17 ',67$1&( Pf

PAGE 96

7KH SUREDELOLW\ RI UHFUXLWPHQW SHU VHHG DV D IXQFWLRQ RI GLVWDQFH IURP SDUHQW WUHHV ZDV FDOFXODWHG DV WKH SURGXFW RI WKH SUREDELOLWLHV RI VXUYLYLQJ VHHG SUHGDWLRQ JHUPLQDWLRQ VHHGOLQJ HVWDEOLVKPHQW DQG \U VHHGOLQJ VXUYLYDO 6HHG SUHGDWLRQ DQG JHUPLQDWLRQ ZHUH QRW DV LPSRUWDQW LQ OLPLWLQJ UHFUXLWPHQW DV ZHUH VHHGOLQJ HVWDEOLVKPHQW DQG \U VXUYLYDO )LJ f (VWDEOLVKPHQW DQG \U VXUYLYDO ZHUH KLJKHU IRU VHHGV XQGHU WKH FURZQ DQG XS WR P IURP WKH FURZQ WKDQ IRU VHHGV GLVSHUVHG P IURP WKH FURZQ )LJ f 7KH RYHUDOO SUREDELOLW\ RI UHFUXLWPHQW YDULHG ZLWK GLVWDQFH 2QHZD\ $129$ RQ DUFVLQ WUDQVIRUPHG GDWD ) 3 f 5HFUXLWPHQW SUREDELOLW\ ZDV KLJKHU P IURP WKH FURZQ WKDQ ZLWKLQ P )LVKHUnV /6' 3nV f RU EH\RQG P 3nV )LJ (f 5HFUXLWPHQW ZDV KLJKHU LQ WKH ]RQH P WKDQ P IURP DGXOWV EXW WKH GLIIHUHQFH ZDV PDUJLQDOO\ VLJQLILFDQW 3 f 1RWH WKDW WKHVH SUREDELOLWLHV GR QRW LQFOXGH WKH QXPEHU RI VHHGV SHU GLVWDQFH LQWHUYDO )LJ f EHFDXVH WKDW LV D FRPSRQHQW RI GLVSHUVDO TXDQWLW\ QRW TXDOLW\ 'LVFXVVLRQ 7KH REVHUYHG VSDWLDO SDWWHUQ RI GLVSHUVHG %HLOVFKPLHGLD VHHGV LQ UHODWLRQ WR SDUHQW WUHHV ZDV VLPLODU WR WKDW RI RWKHU YHUWHEUDWHGLVSHUVHG WUHHV +RZH D /DPDQ Df 7KH YDVW PDMRULW\ RI WKH VHHGV GLVVHPLQDWHG E\ ELUGV ODQGHG XQGHU RU ZLWKLQ P RI WKH FURZQV RI SDUHQW WUHHV 2QH RI WKH IRXU GLVSHUVHU VSHFLHV EODFN JXDQ GLVSHUVHG VHHGV RYHU JUHDWHU GLVWDQFHV WKDQ WKH RWKHU GLVSHUVHU VSHFLHV ,I GLVWDQFH DORQH ZHUH WKH PRVW LPSRUWDQW IDFWRU LQ GHWHUPLQLQJ WKH SUREDELOLW\ RI UHFUXLWPHQW LH KLJK TXDOLW\ GLVSHUVDOf WKHQ JXDQV PLJKW SURYLGH KLJKHU TXDOLW\ GLVSHUVDO WKDQ WKH RWKHU VSHFLHV *XDQV KRZHYHU SDVV VHHGV WKURXJK WKH GLJHVWLYH WUDFW DQG GLVSHUVHG VPDOOHU VHHGV WKDQ WKH RWKHU VSHFLHV DQG VPDOO VHHGV ZHUH OHVV OLNHO\ WR HVWDEOLVK DV VHHGOLQJV WKDQ ZHUH ODUJH VHHGV 7KXV ODUJH VHHGV DUH PRUH OLNHO\ WR ODQG QHDU WKH SDUHQW WUHHV ZKHUH WKH\ IDFH WKH SRVVLELOLW\ RI GHQVLW\GHSHQGHQW PRUWDOLW\ IURP EHHWOHV DQG IXQJDO SDWKRJHQV DQG FRPSHWLWLRQ IURP VLEOLQJV :KHHOZULJKW Ef DOVR QRWHG WKDW TXHW]DOV EHOOELUGV DQG WRXFDQHWV DUH KLJKO\

PAGE 97

VHOHFWLYH ZLWK UHJDUG WR IUXLW VL]H DQG WHQG WR GURS ODUJH IUXLWV DQG FRQVXPH VPDOOHU IUXLWV 7KHVH GURSSHG IUXLWV DUH QRW GLVSHUVHG DQ\ DSSUHFLDEOH GLVWDQFH E\ WHUUHVWULDO DQLPDOV DQG WKH VHHGV DUH QRW HDWHQ YHU\ RIWHQ E\ YHUWHEUDWHV %RWK JHUPLQDWLRQ DQG VHHGOLQJ VXUYLYDO DUH UHODWLYHO\ KLJK $ VHHG KDV DERXW D b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nV Ef VXPPDU\ RI WKH FKDUDFWHULVWLFV RI VSHFLDOL]HG GLVSHUVDO V\VWHPV WKH RQO\ SRVWGLVSHUVDO FRPSRQHQW OLVWHG WKDW FRXOG EH FRQVWUXHG DV GLVSHUVDO TXDOLW\ LV VHHG GLVSHUVDO DZD\ IURP SDUHQWV LVf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f 'LVWDQFH IURP SDUHQW WUHHV LV WKH RQO\ YDULDEOH WKDW VHHPV LPSRUWDQW IRU RQH\HDU VXUYLYDO RI %HLOVFKPLHGLD 6HHG VL]H DQG WKH PHDVXUHG KDELWDW YDULDEOHV KDG PLQLPDO HIIHFWV DW EHVW 7KH GDWD RQ RQH\HDU VXUYLYDO SUHVHQWHG DERYH FRPELQHG ZLWK WKH GLVWULEXWLRQ RI ROGHU VHHGOLQJV DQG VDSOLQJV VXJJHVW WKDW GLVSHUVDO D VKRUW GLVWDQFH DZD\ IURP WKH SDUHQW WUHH OHDGV WR D KLJKHU SUREDELOLW\ RI VXUYLYDO WKDQ HLWKHU ORQJHU GLVSHUVDO RU GHSRVLWLRQ GLUHFWO\ EHORZ WKH IUXLWLQJ WUHH 'HQVLW\GHSHQGHQW VXUYLYDO

PAGE 98

LV VKRZQ E\ WKH KLJKHU VXUYLYDO IURP VHHG SUHGDWRUV DQG IXQJDO SDWKRJHQV ZLWK GLVWDQFH DQG VXJJHVWHG E\ WKH WDOOHU VDSOLQJV IDUWKHU IURP DGXOW WUHHV 7KH UHVXOWLQJ SUREDELOLW\ RI UHFUXLWPHQW RI RQH\HDU VHHGOLQJV UHDFKHV D PRGH P IURP WKH SDUHQW WUHHV $SSUR[LPDWHO\ b RI WKH VHHGV UHFHLYH KLJK TXDOLW\ GLVSHUVDO E\ EHLQJ GHSRVLWHG LQ WKLV ]RQH :KHWKHU RU QRW VHHGOLQJV LQ WKLV ]RQH UHWDLQ WKH KLJKHVW SUREDELOLW\ UHFUXLWPHQW WR UHSURGXFWLYH DJH WR VDWLVI\ 6FKXSSnV GHILQLWLRQ RI GLVSHUVDO TXDOLW\f DZDLWV ORQJHUWHUP VWXGLHV 8QIRUWXQDWHO\ ORQJ WHUP VWXGLHV RQ UHFUXLWPHQW RI ORQJOLYHG WUHHV DUH ORJLVWLFDOO\ GLIILFXOW (YHQ ZLWK VRPH FRPSRQHQW RI KLJKTXDOLW\ GLVSHUVDO WKH PDMRULW\ RI VHHGV UHFHLYHV QR RU ORZ TXDOLW\ GLVSHUVDO 6HHG PDVV RI %HLOVFKPLHGLD UDQJHV IURP WR J DQG WKH ODUJHVW VHHGV DUH IUHTXHQWO\ GURSSHG E\ WKH FXUUHQW GLVSHUVHUV :KHHOZULJKW Ef 7KHVH DSSDUHQWO\ WRRODUJH VHHGV VXJJHVW HLWKHU DQ DOWHUQDWLYH GLVSHUVDO VWUDWHJ\ RU D PLVVLQJ GLVSHUVHU $OWKRXJK WKH FRQFHSW RI GHWHUPLQLQJ H[WLQFW GLVSHUVHUV LV IUDXJKW ZLWK DVVXPSWLRQV -DQ]HQ DQG 0DUWLQ +RZH +XQWHU :LWPHU DQG &KHNH f :KHHOZULJKW Ef VSHFXODWHG WKDW EDUHQHFNHG XPEUHOODELUGV &HSKDORSWHUXV JODEULFROOLVf PD\ EH PLVVLQJ IURP WKH GLVSHUVHU DVVHPEODJH RI /DXUDFHDH LQ WKH 0RQWHYHUGH FRQWLQHQWDO GLYLGH DUHD VHH DOVR )RJGHQ f 8PEUHOODELUGV ZHUH REVHUYHG LQ WKH VWXG\ VLWH RQ VHYHUDO RFFDVLRQV EXW DUH QRW NQRZQ WR EUHHG WKHUH FXUUHQWO\ 8PEUHOODELUGV FRXOG HDVLO\ VZDOORZ WKH ODUJHVW %HLOVFKPLHGLD IUXLWV 7KH DGGLWLRQ RI GLVSHUVDO E\ XPEUHOODELUGV ZRXOG FRQVLGHUDEO\ DOWHU WKH UHVXOWV SUHVHQWHG LQ WKLV VWXG\ 7DSLUV 7DSLUXV EDLUGLLf PD\ DOVR EH D PLVVLQJ GLVSHUVHU 7DSLU SRSXODWLRQV LQ &RVWD 5LFD KDYH EHHQ JUHDWO\ UHGXFHG E\ KXQWLQJ -DQ]HQ Gf 7KH\ GR RFFXU UDUHO\ LQ WKH VWXG\ VLWH EXW PRVW REVHUYDWLRQV ZHUH LQ WKH ZHW VHDVRQ ZKHQ %HLOVFKPLHGLD ZDV QRW IUXLWLQJ $OWKRXJK WDSLU WUDFNV ZHUH QRW REVHUYHG XQGHU IUXLWLQJ %HLOVFKPLHGLD WUHHV WKH VWUXFWXUH RI WKH IUXLW DQG VHHG LV VLPLODU WR RWKHU VSHFLHV HDWHQ DQG GLVSHUVHG E\ WDSLUV -DQ]HQ D %RGPHU )UDJRVR f

PAGE 99

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f LQ FRQWUDVW WR RWKHU ODUJHVHHGHG VSHFLHV &ODUN DQG &ODUN +RZH HW DO &UDZOH\ EXW VHH &KDSPDQ DQG &KDSPDQ f ,W LV JHQHUDOO\ EHOLHYHG WKDW /DXUDFHDH VHHGV KDYH VHFRQGDU\ FRPSRXQGV WKDW SURWHFW WKHP IURP VHHG SUHGDWRUV &DVWUR f HLWKHU E\ PDNLQJ WKH VHHGV WR[LF RU XQSDODWDEOH RU E\ OLPLWLQJ WKH QXPEHU RI VHHGV DQ LQGLYLGXDO VHHG SUHGDWRU FDQ SK\VLRORJLFDOO\ WROHUDWH GXULQJ D JLYHQ GD\ $OWKRXJK WKH FKHPLVWU\ RI %HLOVFKPLHGLD VHHGV KDV QRW EHHQ H[DPLQHG WKH UHVLQRXV IUDJUDQFH RI FXW VHHGV DQG WKH ORZ SUHGDWLRQ UDWHV VXJJHVW WKDW WKH VHHGV DUH FKHPLFDOO\ GHIHQGHG 7KHVH GHIHQVHV DUH OLNHO\ DQ LPSRUWDQW SDUW RI WKH WUDGHRIIV LQYROYHG LQ WKH %HLOVFKPLHGLD GLVSHUVDO VWUDWHJ\ /DUJH VHHG VL]H LV WKRXJKW WR IRVWHU UHFUXLWPHQW LQ WKH VKDGHG XQGHUVWRU\ )RVWHU DQG -DQVRQ EXW VHH .HOO\ DQG 3XUYLV +DPPRQG DQG %URZQ f %HFDXVH VXLWDEOH VLWHV IRU %HLOVFKPLHGLD JHUPLQDWLRQ DQG VKRUWWHUP VXUYLYDO DUH SUHGLFWHG RQO\ E\ D UHODWLYHO\ VKRUW GLVWDQFH IURP SDUHQW WUHHV UDWKHU WKDQ VSHFLILF KDELWDW FKDUDFWHULVWLFV WKH ODUJH VHHGV PD\ HQDEOH UHFUXLWPHQW LQ WKH VKDGHG XQGHUVWRU\ GHVSLWH KLJK FRQVSHFLILF GHQVLW\ LI WKH\ DUH FKHPLFDOO\ GHIHQGHG DQG FDQ DYRLG GHQVLW\ GHSHQGHQW VHHG SUHGDWRUV DQG IXQJDO SDWKRJHQV HJ .LWDMLPD f $OWKRXJK VHHG SUHGDWLRQ E\ EHHWOHV DQG URGHQWV DQG VHHGOLQJ PRUWDOLW\ E\ IXQJDO SDWKRJHQV VKRZHG VRPH LQGLFDWLRQ RI GHQVLW\ GHSHQGHQFH WKH RYHUDOO OHYHOV RI PRUWDOLW\

PAGE 100

ZHUH QRW KLJK HQRXJK WR SUHFOXGH RQH\HDU VXUYLYDO RI PDQ\ VHHGOLQJV XQGHU DQG QHDU WKH SDUHQW WUHHV 7KH GDWD IRU %HLOVFKPLHGLD VXJJHVW QR DGYDQWDJH RI GLVSHUVDO RYHU ORQJ GLVWDQFHV $OWKRXJK WKHRUHWLFDOO\ GLVSHUVDO LV DOZD\V DGYDQWDJHRXV IRU JHQH IORZ DQG WR LQFUHDVH WKH FKDQFH RI RFFXS\LQJ YDFDQW VLWHV +DPLOWRQ DQG 0D\ f WKH SUREDELOLW\ RI %HLOVFKPLHGLD UHFUXLWPHQW GURSV FRQVLGHUDEO\ P IURP WKH SDUHQW WUHHV DQG IHZ VHHGV ZHUH REVHUYHG GLVSHUVHG EH\RQG P 3HUKDSV %HLOVFKPLHGLD UHFUXLWPHQW LV OLPLWHG E\ HGDSKLF IDFWRUV RU RWKHU KDELWDW UHTXLUHPHQWV QRW PHDVXUHG LQ WKLV VWXG\ 2QH IDFWRU LPSRUWDQW IRU PDQ\ WURSLFDO VSHFLHV LV PXWXDOLVWLF DVVRFLDWLRQ ZLWK P\FRUUKL]DO IXQJL %HLOVFKPLHGLD VHHGOLQJV LQFUHDVHG UHODWLYH JURZWK UDWHV DIWHU H[SHULPHQWDO LQRFXODWLRQ /RYHORFN HW DO f 6KRUWGLVWDQFH GLVSHUVDO PD\ LQFUHDVH WKH FKDQFH RI HQFRXQWHULQJ P\FRUUKL]DO IXQJL VLPLODU WR WKRVH DVVRFLDWHG ZLWK WKH SDUHQW WUHHV DOWKRXJK P\FRUUKL]DH DUH QRW NQRZQ WR EH KLJKO\ VSHFLHV VSHFLILF :LONLQVRQ Ef $QRWKHU H[SODQDWLRQ IRU WKH DGYDQWDJH RI VKRUW RYHU ORQJ GLVWDQFH GLVSHUVDO LV WKDW FRQGLWLRQV IRU JURZWK DQG HVWDEOLVKPHQW EHFRPH PRUH XQSUHGLFWDEOH RU DW OHDVW SUHGLFWDEO\ QRW EHWWHU WKDQ FORVH WR WKH SDUHQW ZLWK LQFUHDVLQJ GLVWDQFH 7KH ODFN DGDSWDWLRQV IRU ORQJGLVWDQW GLVSHUVDO LQ PDQ\ GHVHUW SODQWV PD\ EH H[SODLQHG E\ WKH KDUVK HQYLURQPHQW (OOQHU DQG 6KPLGD f )RU WKHVH SODQWV LQYHVWPHQW LQ GLVSHUVDO VWUXFWXUHV LV ZDVWHG HQHUJ\ LI WKH FKDQFHV RI VXUYLYDO DUH ORZ HYHU\ZKHUH DQG GHWHUPLQHG ODUJHO\ E\ VWRFKDVWLF SURFHVVHV 3HUKDSV WKH ODUJH LQYHVWPHQW SHU VHHG LQ %HLOVFKPLHGLD OLPLWV GLVSHUVDO WR DUHDV QHDU DQG WKXV LQ VLPLODU KDELWDWV WR WKH SDUHQW WUHHV EXW IDYRUV SHUVLVWHQW VKDGHWROHUDQW VHHGOLQJV $ VLPLODU WUDGHn RII EHWZHHQ GLVSHUVDO DQG VHHG VXUYLYDO KDV EHHQ QRWHG IRU VRPH $IULFDQ WUHHV &KDSPDQ DQG &KDSPDQ f 'LVSHUVDO TXDOLW\ IRU WURSLFDO WUHHV KDV UHFHLYHG OLWWOH DWWHQWLRQ IURP HFRORJLVWV GHVSLWH LQVLJKWV LW PD\ SURYLGH DERXW IRUHVW G\QDPLFV +RZH E 6FKXSS f QRW WR PHQWLRQ WKH WKHRUHWLFDO LPSRUWDQFH LW LV DVVXPHG WR KDYH 0F.H\ 9HQDEOH DQG %URZQ f 7KH UHVXOWV IURP WKLV VWXG\ VXJJHVW WKDW GLVSHUVDO TXDOLW\ LV PRUH D FKDUDFWHULVWLF RI

PAGE 101

WKH SODQWV WKDQ RI WKH GLVSHUVHUV SHU VH $OWKRXJK GLVSHUVHUV DUH LPSRUWDQW PDQ\ DUH LQWHUFKDQJHDEOH IRU D JLYHQ SODQW VSHFLHV RU WUHDW GLIIHUHQW SODQW VSHFLHV GLIIHUHQWO\ :KHHOZULJKW DQG 2ULDQV /HYH\ f %HFDXVH GLVSHUVDO TXDOLW\ LV WKH PRVW GLIILFXOW DVSHFW RI GLVSHUVDO WR TXDQWLI\ LW PD\ EH PRVW SURILWDEOH WR VWXG\ SODQW VSHFLHV WKDW IDOO DW WKH HQGSRLQWV RI WKH IUXLW TXDOLW\ FRQWLQXXP QRWHG E\ 0F.H\ f 6SHFLILFDOO\ WKHVH LQFOXGH VPDOO IUXLWV ZLWK PDQ\ VPDOO VHHGV DQG SXOS ORZ LQ OLSLGV HJ 6RODQDFHDH 0HODVWRPDWDFHDHf RQ RQH HQG RI WKH VSHFWUXP DQG ODUJH VLQJOHVHHGHG QXWULHQWULFK IUXLWV RU DULOVf VXFK DV /DXUDFHDH 3DOPDH RU 0\ULVWDFHDH RQ WKH RWKHU 0XFK RI WKH DWWHQWLRQ E\ UHVHDUFKHUV KDV EHHQ RQ VSHFLHV WKDW DUH XQXVXDO DQG QRW UHSUHVHQWDWLYH RI PRVW SODQWV HJ )LFXV &HFURSLD PLVWOHWRHVf )LFXV LQFOXGHV PDQ\ VSHFLHV RI KHPLHSLSK\WHV VWUDQJOHU ILJVf DQG PLVWOHWRHV DUH REOLJDWH VWHP SDUDVLWHV 7KH IUXLWOLNH VWUXFWXUHV RI ERWK )LFXV V\QFRQLDf DQG &HFURSLD VSDGLFHVf DUH UHODWLYHO\ ODUJH ZLWK PDQ\ VPDOO VHHGV DQG DUH HDWHQ SLHFHV UDWKHU WKDQ VZDOORZHG ZKROH E\ PDQ\ SRWHQWLDO GLVSHUVHUV LQFOXGLQJ ELUGV HJ /HYH\ f )DPLOLHV RU JHQHUD ZLWK ZLGH JHRJUDSKLF GLVWULEXWLRQV HLWKHU SDQWURSLFDO HJ /DXUDFHDH 3DOPDHf RU FRVPRSROLWDQ HJ 6RODULXP -XQLSHUXV 3UXQXVf DUH OLNHO\ WR OHDG WR IUXLWIXO FRPSDUDWLYH DQDO\VHV WKDW DGGUHVV WKH XQGHUO\LQJ HYROXWLRQDU\ IDFWRUV LQYROYHG LQ VHHG GLVSHUVDO V\VWHPV

PAGE 102

&+$37(5 7:267$*( ',63(56$/ 2) 7:2 63(&,(6 2) *8$5($ 0(/,$&($(f ,W LV ZHOO HVWDEOLVKHG WKDW DQLPDOVEXU\ QXWV DQG LW LV ZLGHO\ DFFHSWHG WKDW WKHVH DQLPDOV DUH WKH SULPDU\ GLVSHUVHUV RI PDQ\ WUHHV DQG VKUXEV +RZHYHU H[DFWO\ ZKDW KDSSHQV WR WKH QXWV DIWHU WKH\ KDYH EHHQ EXULHG LV SRRUO\ XQGHUVWRRG 6WHSKHQ % 9DQGHU :DOO f ,QWURGXFWLRQ 6HHG GLVSHUVDO V\VWHPV LQYROYLQJ WZR VWDJHV ZLWK GLIIHUHQW GLVSHUVDO DJHQWV KDYH EHHQ UHSRUWHG LQ D YDULHW\ RI WURSLFDO SODQWV 7\SLFDOO\ WKH ILUVW VWDJH RI GLVSHUVDO LV E\ ELUGV RU PDPPDOV WKDW GHSRVLW LQWDFW VHHGV RQ WKH JURXQG DQG WKH VHFRQG VWDJH LV HLWKHU E\ DQWV WKDW GLVFDUG VHHGV LQ UHIXVH SLOHV RU QHVWV 5REHUWV DQG +HLWKDXV .DVSDUL /HYH\ DQG %\UQH f RU FDYLRPRUSK URGHQWV 'DV\SURFWLGDHf WKDW VFDWWHUKRDUG VHHGV DQG IDLO WR UHWULHYH VRPH RI WKHP )RUJHW DQG 0LOOHURQ )RUJHW )UDJRVR f 'XQJ EHHWOHV DOVR DFW DV VHFRQGDU\ VHHG GLVSHUVHUV ZKHQ WKH\ FDUU\ RII DQG EXU\ GXQJ FRQWDLQLQJ VHHGV (VWUDGD DQG &RDWHV(VWUDGD 6KHSKHUG DQG &KDSPDQ LQ SUHVVf DQG VLPLODUO\ FDUQLYRUHV PD\ RFFDVLRQDOO\ GLVSHUVH VHHGV LQJHVWHG E\ WKHLU SUH\ VSHFLHV +DOO 'HDQ DQG 0LOWRQ 1RJDOHV HW DO f ,Q DGGLWLRQ WR WKHVH H[DPSOHV 9DQGHU:DOO f UHSRUWHG WZRVWDJH GLVSHUVDO RI D WHPSHUDWH SLQH LQYROYLQJ ZLQG DQG VPDOO URGHQWV DQG &OLIIRUG DQG 0RQWHLWK f GHVFULEHG WKUHHVWDJH GLVSHUVDO RI DQ $XVWUDOLDQ VKUXE LQYROYLQJ ELUGV HPXVf DQWV DQG H[SORVLYH GLVSHUVDO ([DPSOHV RI WZR VWDJH DELRWLF GLVSHUVDO KDYH DOVR EHHQ GHVFULEHG 5HGER7RUVWHQVVRQ DQG 7HOHQLXV )LVFKHU *UHHQH DQG -RKQVRQ f )LQDOO\ SRVWGLVSHUVDO VHHG GRUPDQF\ FRXOG EH YLHZHG DV D VHFRQG VWDJH RI GLVSHUVDO WKURXJK WLPH UDWKHU WKDQ VSDFH HJ )HQQHU 0XUUD\ f

PAGE 103

7KH NH\ FRPSRQHQW RI DOO WKHVH V\VWHPV LV WKDW VHFRQGDU\ GLVSHUVHUV UHDUUDQJH WKH SDWWHUQ RI VHHG GLVWULEXWLRQ VHHG VKDGRZf SURGXFHG E\ WKH ILUVW VWDJH RI GLVSHUVDO DQG LQ GRLQJ VR SODFH VRPH VHHGV LQ D GLIIHUHQW PLFURKDELWDW WKDW PD\ SURYLGH EHWWHU FKDQFHV RI VHHG VXUYLYDO )RU H[DPSOH GLVSHUVDO E\ DQWV LV RIWHQ WR QXWULHQWULFK VLWHV %HDWWLH DQG &XOYHU +RUYLW] DQG 6FKHPVNH D EXW VHH 5LFH DQG :HVWRE\ +DQ]DZD HW DO %RQG DQG 6WRFN f ZKLOH VHHG GLVSHUVDO E\ VFDWWHUKRDUGLQJ URGHQWV GHFUHDVHV WKH SUREDELOLW\ RI VHHG SUHGDWLRQ E\ RWKHU VSHFLHV RI VHHG SUHGDWRUV HYHQ WKRXJK PDQ\ LI QRW PRVW VHHGV DUH UHWULHYHG DQG HDWHQ E\ WKH VFDWWHUKRDUGHU +DOOZDFKV 6P\WKH f 7KXV VHFRQGDU\ GLVSHUVDO PD\ KDYH D JUHDW LPSDFW RQ SODQW UHFUXLWPHQW DQG PD\ FRXQWHUDFW RU DFFHQWXDWH WKH SDWWHUQV JHQHUDWHG GXULQJ WKH ILUVW VWDJH RI GLVSHUVDO HJ +HUUHUD HW DO f ,Q PDQ\ 1HRWURSLFDO IRUHVWV WKH PRVW LPSRUWDQW VFDWWHUKRDUGLQJ URGHQWV DUH DJRXWLV 'DV\SURFWD VSSf DQG DFRXFKLV 0\RSURFWDf 7KH\ EXU\ VHHGV FP LQ WKH VRLO UDWKHU WKDQ VLPSO\ FDFKLQJ VHHGV XQGHU OHDI OLWWHU DV GRHV 3URHFKLP\V )RUJHW f DQG VHHGV XQGHU OHDI OLWWHU DUH PRUH DFFHVVLEOH WR RWKHU VHHG SUHGDWRUV VXFK DV SHFFDULHV DQG LQVHFWV WKDQ DUH EXULHG VHHGV 6P\WKH )RUJHW f 0DQ\ RWKHU VPDOO URGHQWV FDFKH VHHGV LQ GHHS EXUURZV ODUGHUKRDUGLQJf RU RQ EUDQFKHV (PPRQV 9DQGHU :DOO f ZKLFK DUH SUREDEO\ QRW VXLWDEOH VLWHV IRU VHHGOLQJ HVWDEOLVKPHQW $OWKRXJK PDQ\ VSHFLHV RI SODQWV DQG VHHGHDWLQJ DQLPDOV DUH LQYROYHG LQ WKLV LQWHUDFWLRQ WKH HIIHFWV RI VFDWWHUKRDUGLQJ KDYH EHHQ VWXGLHG LQ RQO\ D IHZ VSHFLHV +DOOZDFKV 6P\WKH )RUJHW 3HUHV HW DO f DQG WKXV WKH LQIOXHQFH RI VHFRQGDU\ GLVSHUVHUV LQ UHDUUDQJLQJ WKH LQLWLDO VHHG VKDGRZ LV SRRUO\ NQRZQ ,Q DGGLWLRQ WKH LPSRUWDQFH RI VFDWWHUKRDUGLQJ DV VHFRQGDU\ GLVSHUVDO IRU VRPH SODQWV LV LQ GLVSXWH /DUVRQ DQG +RZH )RUJHW DQG 0LOOHURQ f 7KH SXUSRVH RI WKLV VWXG\ ZDV WR IROORZ WKH SRVWGLVSHUVDO IDWH IRU QDWXUDOO\ GLVSHUVHG VHHGV RI WZR VSHFLHV RI FRQJHQHULF WUHHV *XDUHD JODEUD DQG NXQWKLDQDf 7KH VSHFLILF REMHFWLYHV ZHUH WR GHWHUPLQH WKH SDWWHUQ RI GLVSHUVDO JHQHUDWHG E\ ELUGV DQG

PAGE 104

FRPSDUH LW WR WKH SDWWHUQ DIWHU VHHG SUHGDWLRQ DQG VHFRQGDU\ GLVSHUVDO WR GHWHUPLQH LI VHFRQGDU\ GLVSHUVDO UHVXOWHG LQ D VKLIW LQ PLFURKDELWDW DQG RYHUDOO GLVSHUVDO GLVWDQFH 3RVWn GLVSHUVDO IDWHV RI WKH WZR VSHFLHV ZHUH FRPSDUHG WR DVVHVV WKH LQIOXHQFH RI VHHG VL]H RQ UHPRYDO UDWHV ([SHULPHQWV LQ D VLPSOH JUHHQKRXVH ZHUH DOVR FRQGXFWHG ZLWK JODEUD WR H[DPLQH WKH HIIHFWV RI VHHG EXULDO RQ JHUPLQDWLRQ DQG VHHGOLQJ HVWDEOLVKPHQW 6WXG\ 6LWH 7KLV VWXG\ ZDV FRQGXFWHG 0D\ WR $XJXVW LQ WKH 0RQWHYHUGH &ORXG )RUHVW 3UHVHUYH rn1 rn:f LQ WKH &RUGLOOHUD GH 7LODUDQ LQ QRUWKHUQ &RVWD 5LFD 7KLV SUHVHUYH LV DGPLQLVWHUHG E\ WKH 7URSLFDO 6FLHQFH &HQWHU RI 6DQ -RVH &RVWD 5LFD 7KH DYHUDJH DQQXDO UDLQIDOO LV DERXW PP ZLWK PRVW RFFXUULQJ EHWZHHQ 0D\ DQG 1RYHPEHU $GGLWLRQDO SUHFLSLWDWLRQ IURP PLVW DQG FORXG LQWHUFHSWLRQ LV SUREDEO\ VXEVWDQWLDO HJ &DYHOLHU DQG *ROGVWHLQ /DZWRQ f EXW KDV QRW EHHQ ZHOO TXDQWLILHG DW 0RQWHYHUGH 7KH VWXG\ DUHD LV LQ UHODWLYHO\ XQGLVWXUEHG ORZHU PRQWDQH UDLQ IRUHVW +DUWVKRUQ f DORQJ WKH FRQWLQHQWDO GLYLGH DW P HOHYDWLRQ $ KD DUHD P IURP WKH EHJLQQLQJ RI WKH 9DOOH\ 7UDLO 6HQGHUR (O 9DOOHf ZDV PDSSHG DQG PDUNHG LQWR P [ P TXDGUDWV ZLWK WKH JULG SRLQWV PDUNHG ZLWK 39& SLSHV 7KH YHJHWDWLRQ RI WKH DUHD LV GHVFULEHG E\ /DZWRQ DQG 'U\HU f DQG 1DGNDPL DQG :KHHOZULJKW LQ SUHVVf 6WXG\ 6SHFLHV 7KH JHQXV *XDUHD LV ZLGHVSUHDG LQ WKH 1HRWURSLFV DQG D IHZ VSHFLHV RFFXU LQ $IULFD *HQWU\ f 7KH WD[RQRP\ DQG UHODWLRQVKLSV RI &RVWD 5LFDQ *XDUHD VSHFLHV DUH QRW ZHOO NQRZQ +DEHU HW DO f *XDUHD JODEUD 9DKOf LV DQ XQGHUVWRU\ WUHH UDQJLQJ P LQ KHLJKW ZLWK FURZQ UDGLL DSSUR[LPDWHO\ P ZKLOH NXQWKLDQD $GU -XVVf LV D WUHH RI WKH FDQRS\ DQG XSSHU XQGHUVWRU\ P LQ KHLJKW ZLWK DSSUR[LPDWHO\ P FURZQ UDGLL %RWK VSHFLHV SURGXFH GHKLVFHQW FDSVXOHV ZLWK WR XVXDOO\ f DULOODWH VHHGV SHU IUXLW )UHVK VHHG PDVV DYHUDJHG J s f IRU NXQWKLDQD DQG s f

PAGE 105

IRU JODEUD 1 IRU HDFK VSHFLHVf 7KH VHHGV KDYH D GLVWLQFWLYH VHHG FRDW DQG LUUHJXODU VKDSHV $ORQJ WKH ORQJ D[LV RI WKH VHHG RQH VLGH LV VPRRWK ZLWK D OLJKWHUFRORUHG GRW QHDU WKH FHQWHU ZKLOH WKH RWKHU VLGH KDV D SLWWHG VXUIDFH IURP ZKLFK ILEHUV HPHUJH DQG FRQYHUJH LQWR DQ DWWDFKPHQW WKDW KROGV WKH DULOODWH VHHG WR WKH FDSVXOH 7KH UHG WR RUDQJH DULO FRYHUV DOO RI WKH VPRRWK VLGH DQG JHWV SURJUHVVLYHO\ WKLQQHU RQ WKH RWKHU VLGH FORVHU WR WKH DWWDFKPHQW 7KH VHHG FRDW RI JODEUD LV PP s f WKLFN ZKLOH WKDW RI NXQWKLDQD LV PXFK WKLFNHU s PP 1 IRU HDFK VSHFLHVf %DVHG RQ KU RI ZDWFKHV DW IUXLWLQJ WUHHV DQG FDVXDO REVHUYDWLRQV WKH PDLQ FRQVXPHUV RI NXQWKLDQD DQG JODEUD DULOV DUH ELUGV SULPDWHV DQG VTXLUUHOV 7KH DQLPDOV PRVW IUHTXHQWO\ VHHQ IHHGLQJ LQ WKHVH WUHHV ZHUH WRXFDQHWV $XODFRUK\QFKXV SUDVLQXVf DQG JXDQV &KDPDHSHWHV XQLFRORUf ,Q DGGLWLRQ JODEUD LV GLVSHUVHG E\ URELQV 7XUGXV SOHEHMXVf DQG ERWK WUHH VSHFLHV DUH RFFDVLRQDOO\ REVHUYDWLRQVf GLVSHUVHG E\ PDQWOHG KRZOHU $ORXDWWD SDOOLDWDf DQG ZKLWHIURQWHG FDSXFKLQ &HEXV FDSXFLQXVf PRQNH\V DQG UDUHO\ REVHUYDWLRQf E\ VSLGHU PRQNH\V $WHOHV JHRIIUR\Lf 6TXLUUHOV 6FLXUXV GHSSHLf FKHZ WKH DULOV EXW GURS VHHGV EHQHDWK WKH SDUHQW WUHHV 'HWDLOHG IRUDJLQJ REVHUYDWLRQV ZHUH QRW XQGHUWDNHQ EXW DGGLWLRQDO GLVSHUVDO DJHQWV DUH H[SHFWHG HVSHFLDOO\ IRU WKH VPDOOHUVHHGHG JODEUD HJ +RZH DQG 'H6WHYHQ f $W ORZHU HOHYDWLRQV RQ WKH 3DFLILF VORSH 0DVNHG 7LW\UD 7LW\UD VHPLIDVFLDWDf *ROGHQROLYH :RRGSHFNHUV 3LFXOXV UXELJLQRVXVf 2OLYHVWULSHG )O\FDWFKHUV 0LRQHFWHV ROLYDFHXVf DQG %ODFNIDFHG 6ROLWDLUHV 0\DGHVWHV PHODQRSVf LQ DGGLWLRQ WR JXDQV WRXFDQHWV DQG URELQV DUH NQRZQ WR FRQVXPH DULOV RI *XDUHD VSHFLHV :KHHOZULJKW HW DO f )LYH IUXLWLQJ NXQWKLDQD DQG HLJKW IUXLWLQJ JODEUD RFFXUUHG LQ WKH VWXG\ VLWH 0HWKRGV 'LVSHUVHG VHHGV ZHUH ORFDWHG LQ -XQH $XJXVW E\ V\VWHPDWLFDOO\ VHDUFKLQJ WKH JURXQG IRU IUHVKO\ UHJXUJLWDWHG GURSSHG RU GHIHFDWHG VHHGV 7KH VHDUFKHV VWDUWHG DW WKH

PAGE 106

EDVH RI D IUXLWLQJ WUHH DQG SURFHHGHG DORQJ P ZLGH WUDQVHFWV GHOLQHDWHG E\ WKH 39& PDUNHUVf P IURP WKH WUXQN ,W ZDV LPSRVVLEOH WR VHDUFK WKH HQWLUH VLWH ZLWK HTXDO LQWHQVLW\ EXW DQ HIIRUW ZDV PDGH WR FRYHU WKH HQWLUH VLWH DW OHDVW RQFH HYHU\ WZR ZHHNV VR WKDW RYHU WKH FRXUVH RI WKH ZN IUXLWLQJ VHDVRQ HDFK [ P SORW ZDV FKHFNHG DW OHDVW IRXU WLPHV 6HHGV ZLWK D GDPDJHG VHHG FRDW E\ JQDZLQJ RI VTXLUUHOV RU RWKHU DQLPDOVf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f DV ZHOO DV ZLWK RWKHU VSHFLHV LQ WKLV VDPH VWXG\ VLWH &KDSWHU f (DFK PDUNHG VHHG ZDV UHWXUQHG WR LWV RULJLQDO ORFDWLRQ WKH QH[W PRUQLQJ 7KH GLVWDQFH IURP HDFK GLVSHUVHG VHHG WR WKH WUXQN RI WKH FORVHVW IUXLWLQJ FRQVSHFLILF KHUHDIWHU UHIHUUHG WR DV WKH SDUHQW WUHHf ZDV PHDVXUHG ZLWK D WDSH PHDVXUH LI OHVV WKDQ Pf RU HVWLPDWHG IURP WKH VWXG\ VLWH PDS LI PRUH WKDQ P 6HHGV ZHUH FODVVLILHG DV QRQGLVSHUVHG LI GLUHFWO\ EHORZ WKH SDUHQWDO FURZQ DQG DV GLVSHUVHG LI EH\RQG WKH FURZQ EXW WKH GLVWDQFH IURP WKH HGJH RI WKH FURZQ ZDV QRW PHDVXUHG )RU DOO VLWHV PHDVXUHG PLFURKDELWDW YDULDEOHV WKRXJKW PLJKW LQIOXHQFH VHHG UHPRYDO &DQRS\ FRYHU ZDV HVWLPDWHG ZLWK D VSKHULFDO GHQVLRPHWHU /HPPRQ f /HDI OLWWHU GHSWK ZDV HVWLPDWHG DV WKH QXPEHU RI OHDYHV SLHUFHG E\ D PHWDO VWDNH WKUXVW LQWR WKH VRLO DW WKH VLWH 9HJHWDWLRQ GHQVLW\ ZDV HVWLPDWHG DV WKH QXPEHU RI VWHPV ZLWKLQ D FP UDGLXV RI WKH VLWH 7KH GLVWDQFHV WR WKH QHDUHVW WUHH WUXQN RYHU FP '%+ GLDPHWHU DW EUHDVW KHLJKWf DQG IDOOHQ ORJ RYHU FP GLDPHWHU ZHUH PHDVXUHG ZLWK D ILEHUJODVV PHDVXULQJ WDSH 7KH VDPH YDULDEOHV ZHUH PHDVXUHG DW WKH VLWHV RI EXULHG VHHGV DIWHU VFDWWHUKRDUGLQJ ,Q DGGLWLRQ WKH GLVWDQFH IURP WKH LQLWLDO GLVSHUVDO ORFDWLRQ WR WKH VHFRQGDU\ GLVSHUVDO ORFDWLRQ ZDV

PAGE 107

PHDVXUHG 7KHVH YDULDEOHV ZHUH VHOHFWHG EDVHG RQ WKHLU LPSRUWDQFH LQ SUHYLRXV VWXGLHV &DQRS\ FRYHU YHJHWDWLRQ GHQVLW\ DQG GLVWDQFH WR REMHFWV PD\ LQIOXHQFH URGHQW IRUDJLQJ SDWWHUQV .LOWLH .LWFKLQJV DQG /HYH\ %RZHUV DQG 'RROH\ 9£VTXH] f /HDI OLWWHU PD\ LQIOXHQFH VHHG SUHGDWLRQ RU JHUPLQDWLRQ 0RORIVN\ DQG $XJVSXUJHU 0\VWHU DQG 3LFNHWW &LQWUD Df $OO VLWHV ZHUH FHQVXVHG RQ GD\V DQG RQFH HDFK ZHHN DIWHUZDUGV IRU D WRWDO RI ZHHNV ,I D PDUNHG VHHG ZDV UHPRYHG WKH VXUURXQGLQJ DUHD ZDV VHDUFKHG XQWLO WKH IODJJLQJ WDSHGHQWDO IORVV DVVHPEO\ ZDV IRXQG 7KH HQG RI WKH IORVV ZKHUH WKH VHHG ZDV DWWDFKHG ZDV H[DPLQHG WR GHWHUPLQH WKH IDWH RI WKH VHHG ,I D VHHG ZDV UHPRYHG IURP WKH IORVV DQG SLHFHV RI WKH VHHG FRDW UHPDLQHG WKH VHHG ZDV FODVVLILHG DV HDWHQ ,I D VHHG ZDV EXULHG LQ WKH VRLO QRW MXVW XQGHU OHDI OLWWHUf LW ZDV FODVVLILHG DV FDFKHG E\ DJRXWLV 'DV\SURFWD SXQFWDWDf 6HHG SUHGDWLRQ E\ SHFFDULHV 7D\DVVX WDMDFXf ZDV GHWHUPLQHG RQ WKH EDVLV RI WUDFNV DQG RYHUWXUQHG OHDI OLWWHU W\SLFDO RI SHFFDU\ IRUDJLQJ SDFNV .LOWLH f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f ,Q HDFK RI SODVWLF FRQWDLQHUV DSSUR[LPDWHO\ ; ; FPf ZLWK GUDLQ KROHV IRXU VHHGV ZHUH SODQWHG WZR ZHUH EXULHG FP GHHS DQG WZR ZHUH SODFHG RQ WKH VRLO VXUIDFH IRU D WRWDO VDPSOH RI EXULHG VHHGV DQG VXUIDFH VHHGV 7KH FRQWDLQHUV ZHUH ZDWHUHG IUHTXHQWO\ WR NHHS WKH VRLO PRLVW DV LW ZRXOG EH LQ WKH IRUHVW DW WKDW WLPH RI \HDUf 7KH\ ZHUH FKHFNHG DW OHDVW RQFH HDFK ZHHN IRU VHHGOLQJ HPHUJHQFH DQG JURZWK

PAGE 108

6WDWLVWLFDO WHVWV IURP 6$6 -03 6$6 ,QVWLWXWH f ZHUH XVHG 'LVSHUVDO GLVWDQFHV ZHUH QRW QRUPDOO\ GLVWULEXWHG HYHQ DIWHU DSSURSULDWH WUDQVIRUPDWLRQV VR LQ PRVW FDVHV QRQSDUDPHWULF WHVWV ZHUH XVHG 5HPRYDO UDWHV RI WKH WZR VSHFLHV ZHUH FRPSDUHG ZLWK VXUYLYDO DQDO\VLV DQG WKH :LOFR[RQ WHVW ZKLFK SODFHV JUHDWHU HPSKDVLV RQ HDUOLHU HYHQWV 3\NH DQG 7KRPSVRQ f )RU WKH JUHHQKRXVH JHUPLQDWLRQ H[SHULPHQW WKH HIIHFWV RI VHHG VL]H DQG EXULDO WUHDWPHQW RQ JHUPLQDWLRQ VXFFHVV ZHUH DQDO\]HG ZLWK ORJLVWLF UHJUHVVLRQ PRGHOV 7UH[OHU DQG 7UDYLV f 0LFURKDELWDW FKDUDFWHULVWLFV RI LQLWLDO GLVSHUVDO VLWHV ZHUH FRPSDUHG ZLWK WKRVH DW FDFKH VLWHV ZLWK SDLUHG WHVWV ZLWK DQ DOSKD YDOXH DGMXVWHG IRU PXOWLSOH FRPSDULVRQV +ROP f 7KURXJKRXW WKLV SDSHU PHDQ YDOXHV DUH IROORZHG E\ 6' 5HVXOWV 'LVSHUVDO EY %LUGV 7KH ILUVW VWDJH RI GLVSHUVDO LV ODUJHO\ E\ ELUGV DQG RFFDVLRQDOO\ E\ PDPPDOV 'XULQJ HDUO\ -XQH WR PLG$XJXVW NXQWKLDQD DQG JODEUD VHHGV ZHUH IRXQG 7ZHQW\ bf NXQWKLDQD DQG bf JODEUD VHHGV ZHUH GHSRVLWHG XQGHU WKH FURZQV )LJ f )RU VHHGV GLVSHUVHG EH\RQG WKH FURZQV NXQWKLDQD VHHGV ZHUH GLVSHUVHG IDUWKHU IURP WKH WUXQN PHGLDQ P PHDQ s Pf WKDQ JODEUD VHHGV PHGLDQ P PHDQ s P :LOFR[RQ UDQN VXP WHVW ; GI 3 f $SSUR[LPDWHO\ b RI f RI JODEUD VHHGV KDG VLJQV RI SUHn GLVSHUVDO LQVHFW LQIHVWDWLRQ ZKLOH b RI f RI WKH NXQWKLDQD VHHGV ZHUH XQXVXDOO\ VPDOO DQGRU OLJKW DQG ODWHU GHWHUPLQHG WR EH KROORZ DQG LQYLDEOH +ROORZ VHHGV RI JODEUD ZHUH QHYHU IRXQG DQG RQO\ NXQWKLDQD VHHGV KDG REYLRXV VLJQV RI SUHn GLVSHUVDO LQVHFW LQIHVWDWLRQ 7DEOH f

PAGE 109

&2 4 /8 8O &2 FF P = &2 4 +L 8O &2 FF //, P ',67$1&( )520 7581. Pf )LJXUH 1XPEHU RI VHHGV UHJXUJLWDWHG RU GHIHFDWHG E\ ELUGV DQG PDPPDOV EDUVf DQG WKH QXPEHU RI VHHGV EXULHG E\ URGHQWV OLQHVf DW P GLVWDQFH LQWHUYDOV IURP IUXLWLQJ HLJKW *XDUHD JODEUD DQG ILYH KPWKLDQD WUHHV 'LVWDQFH FODVV LQFOXGHV DOO VHHGV GLUHFWO\ EHORZ WKH FURZQV ZKLOH WKH RWKHU GLVWDQFH FODVVHV DUH PHDVXUHG IURP WKH WUXQNV &URZQ UDGLL ZHUH DSSUR[LPDWHO\ P IRU JODEUD DQG P IRU KPWKLDQD

PAGE 110

7DEOH 3RVWGLVSHUVDO IDWH RI *XDUHD VHHGV DIWHU ZHHNV 6HHGV UHJXUJLWDWHG RU GHIHFDWHG E\ ELUGV DQG PDPPDOV ZHUH PDUNHG ZLWK D GHQWDO IORVV DQG IODJJLQJ WDSH DVVHPEO\ VHH 0HWKRGVf DQG VHDUFKHG IRU LI UHPRYHG 6HHGV ZHUH FRQVLGHUHG HDWHQ E\ PDPPDOV LI UHPRYHG IURP WKH IORVV DQG SLHFHV RI WKH VHHG FRDW UHPDLQHG HDWHQ E\ LQVHFWV LI HQWU\ RU H[LW KROHV ZHUH YLVLEOH FDFKHG LI IRXQG EXULHG LQ WKH VRLO DQG LQYLDEOH LI DSSHDULQJ PHDO\ NXQWKLDQD JODEUD )DWH 1 bf 1 bf UHPRYHG HDWHQ f f UHPRYHG FDFKHG HDWHQ f f UHPRYHG FDFKHG f f UHPRYHG PLVVLQJ f f QRW UHPRYHG YLDEOH f f QRW UHPRYHG LQVHFWV f f QRW UHPRYHG LQYLDEOH f f 7RWDO VDPSOH

PAGE 111

3RVWGLVSHUVDO 6HHG )DWH 6HHG 3UHGDWLRQ DQG 6FDWWHUKRDUGLQJ 7KH PRVW VWULNLQJ GLIIHUHQFH LQ WKH SRVWGLVSHUVDO IDWH EHWZHHQ WKH WZR WUHH VSHFLHV ZDV WKDW SHFFDULHV UDSLGO\ FRQVXPHG PDQ\ NXQWKLDQD VHHGV WKDW IHOO XQGHU RU ZHUH GLVSHUVHG QHDU WKH SDUHQW WUHHV ZKHUHDV WKH\ GLG QRW FRQVXPH JODEUD 7KH FRPELQDWLRQ RI VHHG SUHGDWLRQ E\ SHFFDULHV DQG URGHQWV DQG VFDWWHUKRDUGLQJ E\ DJRXWLV OHIW QR YLDEOH XQFDFKHG NXQWKLDQD VHHGV RI WKH RULJLQDO VDPSOH RI DIWHU ZHHNV $JRXWLV EXULHG bf VHHGV VLQJO\ FP GHHS 2QO\ bf RI WKHVH EXULHG VHHGV KDG QRW EHHQ GXJ XS DQG HDWHQ DIWHU ZHHNV 7DEOH f 7KUHH FDFKHG VHHGV ZHUH GXJ XS DQG UHEXULHG EXW DOO WKUHH ZHUH ODWHU HDWHQ ,Q WKH GHVFULSWLRQV RI FDFKH VLWH PLFURKDELWDWV EHORZ RQO\ WKH FKDUDFWHULVWLFV RI WKH ILUVW FDFKH DUH LQFOXGHG ,Q FRQWUDVW WR NXQWKLDQD b RI f RI JODEUD VHHGV ZHUH IRXQG EXULHG DQG RI WKRVH bf UHPDLQHG EXULHG DIWHU ZHHNV ,Q DGGLWLRQ bf DSSDUHQWO\ YLDEOH VHHGV ZHUH QRW HDWHQ RU FDFKHG 7DEOH f ,QVHFWV NLOOHG PDQ\ VHHGV bf EHQHDWK DQG QHDU SDUHQW WUHHV FRPSDUHG WR NXQWKLDQD ZKHUH UHPRYDO E\ SHFFDULHV VXSHUVHGHG LQVHFW VHHG SUHGDWLRQ 7KH RYHUDOO UHPRYDO UDWHV VHHGV HDWHQ SOXV VHHGV FDFKHGf ZHUH VLPLODU IRU WKH WZR VSHFLHV )LJ f EXW NXQWKLDQD VHHGV ZHUH PRUH RIWHQ HDWHQ WKDQ FDFKHG VXUYLYDO DQDO\VLV :LOFR[RQ GI 3 f )RU ERWK VSHFLHV WKH KLJKHVW SURSRUWLRQ RI FDFKHG VHHGV RFFXUUHG LQ WKH ILUVW WKUHH ZHHNV )LJ f $IWHU WKDW WLPH WKH SURSRUWLRQ RI FDFKHG NXQWKLDQD GHFOLQHG DV PRVW VHHGV ZHUH GXJ XS DQG HDWHQ ZKLOH WKH SURSRUWLRQ RI JODEUD UHPDLQHG UHODWLYHO\ FRQVWDQW DV PRVW FDFKHG VHHGV ZHUH QRW UHWULHYHG 7KH IDWH RI FDFKHG NXQWKLDQD VHHGV ZDV LQIOXHQFHG E\ GLVWDQFH WR SDUHQW WUHHV &DFKHG VHHGV WKDW ZHUH HYHQWXDOO\ GXJ XS DQG HDWHQ ZHUH FORVHU WR SDUHQW WUXQNV WKDQ ZHUH VHHGV WKDW ZHUH QRW UHWULHYHG :LOFR[RQ 5DQN 6XP WHVW 3 f 2I VHHGV WKDW ZHUH FDFKHG DQG ODWHU HDWHQ ZHUH GLUHFWO\ EHQHDWK WKH FURZQV RI SDUHQW WUHHV )RU JODEUD GLVWDQFH IURP FRQVSHFLILFV GLG QRW LQIOXHQFH WKH IDWH RI FDFKHG VHHGV :LOFR[RQ 5DQN 6XP WHVW 3 f

PAGE 112

F k R LB k D *XDUHD JODEUD Â’ 127 5(029(' (6 %85,(' Â’ ($7(1 F k R k D NXQWKODQD Â’ 127 5(029(' % %85,(' ($7(1 :((. )LJXUH &XPXODWLYH SHUFHQW RI VHHGV HDWHQ GLDJRQDO OLQHVf EXULHG VWLSSOHGf RU QRW UHPRYHG RSHQ VSDFH DW WRSf RYHU WKH FRXUVH RI WKH IUXLWLQJ VHDVRQ IRU JODEUD DQG NXQWKLDQD 1RWH WKDW WLPH ]HUR LV QRW WKH VDPH GD\ IRU DOO VHHGV WKH LQGLYLGXDO SKHQRORJLHV KDYH EHHQ DGMXVWHG IRU HDVH RI FRPSDULVRQ

PAGE 113

7KH GLVWDQFH EHWZHHQ WKH LQLWLDO GLVSHUVDO VLWH DQG WKH VLWH RI EXULDO ZDV JUHDWHU IRU NXQWKLDQD s Pf WKDQ IRU JODEUD s P :LOFR[RQ UDQN 6XPV 3 f 7KH QHW RXWFRPH RI VHFRQGDU\ GLVSHUVDO ZDV D VOLJKW LQFUHDVH LQ WKH DYHUDJH GLVWDQFH IURP SDUHQW WUHHV 7DEOH f 7KH VHHG VKDGRZ RI FDFKHG VHHGV ZDV DOVR WUXQFDWHG EHFDXVH PRVW VHHGV GLUHFWO\ XQGHU WKH SDUHQWV DV ZHOO DV WKRVH IDU Pf IURP SDUHQWV ZHUH HDWHQ UDWKHU WKDQ FDFKHG )LJ f 7KH LQLWLDO GLVWDQFH IURP WKH WUXQN ZDV QRW FRUUHODWHG ZLWK KRZ IDU VHHGV ZHUH PRYHG GXULQJ VHFRQGDU\ GLVSHUVDO IRU NXQWKLDQD 6SHDUPDQnV 5KR 3 f RU JODEUD 5KR 3 f NXQWKLDQD VHHGV ZHUH RIWHQ FDFKHG QHDU ORJV VHFRQGDU\ GLVSHUVDO VLWHV ZHUH FORVHU WR ORJV WKDQ ZHUH WKH LQLWLDO GLVSHUVDO ORFDWLRQV SDLUHG WWHVW W GI 3 7DEOH f )RU JODEUD VHFRQGDU\ GLVSHUVDO VLWHV WHQGHG WR KDYH OHVV OHDI OLWWHU W GI 3 f DQG ORZHU YHJHWDWLRQ GHQVLW\ I GI 3 f WKDQ WKH LQLWLDO GLVSHUVDO ORFDWLRQV 7KH VDPH WUHQGV ZHUH HYLGHQW IRU NXQWKLDQD OHDI OLWWHU W GI 3 YHJHWDWLRQ GHQVLW\ W GI 3 7DEOH f *HUPLQDWLRQ 1RQH RI WKH PDUNHG JODEUD VHHGV JHUPLQDWHG LQ WKH ILHOG 2I WKH FDFKHG VHHGV WKDW UHPDLQHG DIWHU ZHHNV b RI f DSSHDUHG PHDO\ RU URWWHQ 0RVW RI WKH UHVW bf ZHUH LQIHVWHG ZLWK LQVHFWV ZKLOH WKH UHPDLQGHU bf ZHUH VROLG DQG DSSHDUHG YLDEOH 7KH FDFKHG NXQWKLDQD VHHGV UHPDLQLQJ DW WKH HQG RI WKH VWXG\ DSSHDUHG YLDEOH ,Q WKH JUHHQKRXVH JHUPLQDWLRQ ZDV KLJKHU IRU EXULHG WKDQ IRU VXUIDFH VHHGV ; GI 3 )LJ f 6HHGV LQ WKH EXULHG WUHDWPHQW HVWDEOLVKHG PRUH VHHGOLQJV WKDQ VXUIDFH VHHGV ; GI 3 )LJ f ,Q ORJLVWLF UHJUHVVLRQ PRGHOV LQFOXGLQJ VHHG ZLGWK OHQJWK DQG PDVV LQ DGGLWLRQ WR D WUHDWPHQW HIIHFW

PAGE 114

7DEOH +DELWDW FKDUDFWHULVWLFV DYHUDJH s 6'f RI *XDUHD NXQWKLDQD DQG JODEUD GLVSHUVDO ORFDWLRQV EHIRUH SULPDU\ GLVSHUVDO VLWHf DQG DIWHU VFDWWHUKRDUGLQJ VHFRQGDU\ GLVSHUVDO VLWHf 1 IRU NXQWKLDQD DQG IRU JODEUD :LWKLQ HDFK YDULDEOH YDOXHV IRU HDFK VLWH EHIRUH DQG DIWHU VFDWWHUKRDUGLQJ ZHUH FRPSDUHG ZLWK SDLUHGWWHVWV ZLWK DQ DOSKD YDOXH RI DGMXVWHG IRU PXOWLSOH FRPSDULVRQV 6HH WH[W IRU GHWDLOV RI WHVW UHVXOWV NXQWKLDQD JODEUD SULPDU\ VHFRQGDU\ SULPDU\ VHFRQGDU\ FRQVSHFLILF WUHH Pf f rr f f rr f FP WUHH Pf f f f r f ORJ Pf f rr f f f OHDI OLWWHU f f r f f rr f YHJHWDWLRQ GHQVLW\ f r f f rr f FDQRS\ FRYHU bf f f f f r 3 PDUJLQDOO\ VLJQLILFDQWf rr3A

PAGE 115

180%(5 2) 6(('6 EXULHG 2 VXUIDFH JHUPLQDWHG QRW HVWDEOLVKHG QRW )LJXUH 1XPEHU RI VHHGV EXULHG FP RU SODFHG RQ WKH VRLO VXUIDFH WKDW JHUPLQDWHG DQG HVWDEOLVKHG VHHGOLQJV LQ D JUHHQKRXVH 1 VHHGV LQ HDFK WUHDWPHQW

PAGE 116

EXULHG RU VXUIDFHf RQO\ WUHDWPHQW ZDV D VLJQLILFDQW SUHGLFWRU RI WKH SUREDELOLW\ RI JHUPLQDWLRQ 5 ; 3 f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b IDUWKHU ZD\ IURP WKH WUHH GLUHFWO\ DW WKH WUXQN WR DSSUR[LPDWHO\ b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f IRU NXQWKLDQD :KLOH JODEUD ZDV VRPHWLPHV FDFKHG QHDU ORJV PRVW RIWHQ WKH FDFKHV ZHUH QRW REYLRXVO\ DVVRFLDWHG ZLWK DQ\ REMHFW ,Q DGGLWLRQ NXQWKLDQD VHHGV WHQGHG WR EH PRYHG ORQJHU GLVWDQFHV GXULQJ VHFRQGDU\ GLVSHUVDO 7KHVH GLIIHUHQFHV EHWZHHQ WKH VSHFLHV VXJJHVW WKDW PRUH WKDQ RQH DQLPDO VSHFLHV ZDV LQYROYHG LQ VFDWWHUKRDUGLQJ RU WKDW RQH VSHFLHV RI GLVSHUVHU WUHDWV WKH WZR VHHG VSHFLHV GLIIHUHQWO\ )RUJHW f SRLQWV RXW WKDW WKH RQO\ 1HRWURSLFDO VSHFLHV NQRZQ WR EXU\ VHHGV LQ WKH VRLO DUH FDYLRPRUSK URGHQWV

PAGE 117

'DV\SURFWD 0\RSURFWDf ,I RWKHU VSHFLHV DUH LQYROYHG LQ VFDWWHUKRDUGLQJ WKH OLNHO\ FDQGLGDWHV LQ 0RQWHYHUGH DUH 6FLXUXV GHSSHL 3HURP\VFXV QXGLSHV DQG +HWHURP\V GHVPDUHVWLDQXV 6SHFLHV LQ DOO WKUHH RI WKHVH JHQHUD DUH SRVVLEOH VFDWWHUKRDUGHUV LQ RWKHU DUHDV 9DQGHU:DOO f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f VKRZHG ZHUH HDWHQ E\ SHFFDULHV DQG LQVHFWV LI QRW VFDWWHUKRDUGHG E\ DJRXWLV ,Q DGGLWLRQ WR VHHG SUHGDWRUV VHHGV RQ WKH VXUIDFH IDFH JUHDWHU IOXFWXDWLRQV LQ WHPSHUDWXUH DQG PRLVWXUH WKDQ GR EXULHG VHHGV %XULHG VHHGV PD\ DOVR KDYH ORQJHU LQGXFHG GRUPDQF\ WKDQ VXUIDFH VHHGV )HQQHU f 3RRU JHUPLQDWLRQ RI JODEUD LQ WKH ILHOG PD\ H[SODLQ WKH ORZ UHWULHYDO UDWHV bf RI FDFKHG VHHGV FRPSDUHG WR NXQWKLDQD bf :LWK VRPH VSHFLHV VFDWWHUKRDUGHG E\ DJRXWLV WKH IRRG UHZDUG LV WKH GHYHORSLQJ VHHGOLQJ (PPRQV )RUJHW DQG 0LOOHURQ f SDUWLFXODUO\ WKH IOHVK\ FRW\OHGRQV RI VSHFLHV ZLWK HSLJHDO JHUPLQDWLRQ HJ *DUZRRG f 3ODQWV PD\ VKLIW GHIHQVLYH FRPSRXQGV IURP FRW\OHGRQV WR OHDYHV DIWHU JHUPLQDWLRQ WKXV UHQGHULQJ WKH FRW\OHGRQ PRUH SDODWDEOH -DQ]HQ f ,I VXFK

PAGE 118

FRPSRXQGV LQ JODEUD OLPLW VHHG FRQVXPSWLRQ E\ DJRXWLV SHUKDSV WKH\ FDFKH VHHGV WR FRQVXPH OHVVGHIHQGHG SDUWV RI VHHGOLQJV DIWHU JHUPLQDWLRQ &RQFOXVLRQV 7KH DOWHUDWLRQ RI WKH VHHG VKDGRZV GXULQJ VHFRQGDU\ GLVSHUVDO LV OLNHO\ DQ LPSRUWDQW VWDJH LQ UHFUXLWPHQW IRU *XDUHD DQG RWKHU VSHFLHV DV ZHOO 9HU\ IHZ VWXGLHV KDYH H[DPLQHG VHHG GLVSHUVDO E\ VFDWWHUKRDUGLQJ URGHQWV -XGJLQJ IURP WKRVH DYDLODEOH 6P\WKH )RUJHW &LQWUD E 3HUHV DQG %DLGHU f DQG WKH XELTXLW\ RI FDYLRPRUSK URGHQWV LQ 1HRWURSLFDO IRUHVWV :ULJKW HW DO f WKH HIIHFWV RI VFDWWHUKRDUGLQJ RQ IRUHVW FRPSRVLWLRQ FRXOG EH FRQVLGHUDEOH 3HUHV DQG %DLGHU f VXJJHVWHG WKDW VFDWWHUKRDUGLQJ E\ DJRXWLV FDQ OHDG WR FOXPSHG GLVWULEXWLRQV RI WUHHV )UDJRVR f DOVR IRXQG D FOXPSHG GLVWULEXWLRQ RI SDOPV WKDW ZHUH GLVSHUVHG SULPDULO\ E\ WDSLUV DQG VHFRQGDULO\ E\ DJRXWLV +H DWWULEXWHG WKH FOXPSHG GLVWULEXWLRQ KRZHYHU WR WDSLUV UDWKHU WKDQ WR DJRXWLV EHFDXVH WKH WDSLUV GHSRVLW PDQ\ VHHGV LQ KDELWXDO ODWULQHV XS WR NP IURP H[LVWLQJ SDOP SDWFKHV ZKHUHDV WKH DJRXWLV WKHQ VSDFH RXW WKH VHHGV IURP WKH ODWULQHV )UDJRVR VHH DOVR %RGPHU f 2WKHU VWXGLHV KRZHYHU KDYH QRW IRXQG FOXPSHG SDWWHUQV RI WUHH VSHFLHV VFDWWHUKRDUGHG E\ FDYLRPRUSK URGHQWV )RUJHW f 7KH UHODWLRQVKLSV RI SULPDU\ GLVSHUVDO VHFRQGDU\ GLVSHUVDO VHHG SUHGDWLRQ KDELWDW HIIHFWV DQG VHHGOLQJ VXUYLYDO DUH MXVW EHJLQQLQJ WR EH XQUDYHOHG DQG UHTXLUH IXUWKHU VWXG\ WR EHWWHU XQGHUVWDQG WKH UROH RI PXOWLVWDJH GLVSHUVDO LQ IRUHVW G\QDPLFV

PAGE 119

&+$37(5 $'9$17$*(6 2) ',63(56$/ $ 5((9$/8$7,21 2) ',5(&7(' ',63(56$/ %XW WKH UHDO LPSRUWDQFH RI D ODUJH QXPEHU RI HJJV RU VHHGV LV WR PDNH XS IRU PXFK GHVWUXFWLRQ DW VRPH SHULRG RI OLIH DQG WKLV SHULRG LQ WKH JUHDW PDMRULW\ RI FDVHV LV DQ HDUO\ RQH &KDUOHV 'DUZLQ f KDYH QR GRXEW WKDW SHU JUDP HDWHQ WKH VHHG SUHGDWRUV KDYH WKH ODUJHVW LPSDFW RQ WURSLFDO IRUHVW VWUXFWXUH RI DQ\ DQLPDO OLIH IRUP 'DQLHO + -DQ]HQ Ef ,W PDWWHUV ZKR GHIHFDWHV ZKDW ZKHUH + -DQ]HQ f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f LQ ZKLFK FDVH D SDUWLFXODU GLVSHUVDO DJHQW KDV D GLVSURSRUWLRQDWH HIIHFW RQ SODQW UHFUXLWPHQW DQG GHPRJUDSK\ ,I WUXH WKHQ SDWWHUQV RI GLVSHUVDO QRW MXVW GLVSHUVDO PD\ KDYH D JUHDWHU UROH LQ WKH VWUXFWXUH DQG FRPSRVLWLRQ RI SODQW FRPPXQLWLHV WKDQ FXUUHQWO\ EHOLHYHG 7R GRFXPHQW GLUHFWHG GLVSHUVDO WZR FRQGLWLRQV PXVW EH PHW f QRQUDQGRP PRYHPHQW RI VHHGV DQG f GLVSURSRUWLRQDWH DUULYDO LQ VLWHV ZLWK NQRZQ FKDUDFWHULVWLFV WKDW DUH HVSHFLDOO\ IDYRUDEOH IRU VXUYLYDO +RZH DQG 6PDOOZRRG +RZH f $OWKRXJK

PAGE 120

,OO QRQUDQGRP PRYHPHQW SDWWHUQV DQG KDELWDW VHOHFWLRQ E\ DQLPDOV DQG YDULDWLRQ LQ PLFURKDELWDW VXLWDELOLW\ IRU SODQW JURZWK DUH ZHOO GRFXPHQWHG H[DPSOHV RI GLUHFWHG GLVSHUVDO DUH WKRXJKW WR EH UDUH DQG XQXVXDO +RZH f VXJJHVW WKDW GLUHFWHG GLVSHUVDO LV PRUH FRPPRQ DQG HFRORJLFDOO\ VLJQLILFDQW WKDQ SUHYLRXVO\ EHOLHYHG EXW KDV EHHQ RYHUORRNHG IRU VHYHUDO UHDVRQV )LUVW WKH FXUUHQW SDUDGLJP RI GLIIXVH FRHYROXWLRQ EHWZHHQ SODQWV DQG WKHLU GLVSHUVHUV GRHV QRW SUHGLFW WKH HYROXWLRQ RI DGDSWDWLRQV IRU GLUHFWHG GLVSHUVDO 6HFRQG WKH DOWHUQDWLYH DGYDQWDJHV RI GLVSHUVDO HVFDSH FRORQL]DWLRQ GLUHFWHG GLVSHUVDO VHH EHORZf DUH QRW PXWXDOO\ H[FOXVLYH DQG FDQ EH GLIILFXOW WR GLVWLQJXLVK WKH GHPRQVWUDWLRQ RI RQH DGYDQWDJH GRHV QRW SUHFOXGH WKH RWKHUV ,Q SDUWLFXODU WKH HPSKDVLV RI UHVHDUFK RQ HVFDSH REVFXUHV WKH RFFXUUHQFH RI GLUHFWHG GLVSHUVDO 7KLUG IHZ VWXGLHV KDYH LQWHJUDWHG SDWWHUQV RI VHHG GLVWULEXWLRQ VHHG VKDGRZVf DQG WKH FRQVHTXHQFHV RI VXFK SDWWHUQV IRU SODQW UHFUXLWPHQW LQ D ZD\ WKDW FDQ GHWHFW GLUHFWHG GLVSHUVDO 3ODQW UHFUXLWPHQW LV D PXOWLVWDJH SURFHVV WKDW KDV PRVW RIWHQ EHHQ VWXGLHG LQ D VWDJHVSHFLILF PDQQHU +RXOH -RUGDQR DQG +HUUHUD 6FKXSS DQG )XHQWHV f =RRORJLVWV KDYH VWXGLHG WKH IUXLW UHPRYDO DQG JXW WUHDWPHQW VWDJHV ZKLOH ERWDQLVWV KDYH IRFXVHG RQ WKH VHHGOLQJ VWDJHV +RZH Ef )RXUWK ILQGLQJ GLVSHUVHG VHHGV HVSHFLDOO\ WKRVH GLVSHUVHG E\ YHUWHEUDWHV LV GLIILFXOW ,GHQWLI\LQJ SDUHQWV RI GLVSHUVHG VHHGV LV DOVR GLIILFXOW DQG VR IDU KDV EHHQ OLPLWHG WR VWXGLHV RI LVRODWHG WUHHV *ODGVWRQH $XJVSXUJHU $XJVSXUJHU DQG .LWDMLPD f RU LQGLUHFWO\ E\ JHQHWLF VWXGLHV *LEVRQ DQG :KHHOZULJKW +DPULFN DQG 1DVRQ f )LIWK VHYHUDO GLIIHUHQW QDPHV KDYH EHHQ XVHG IRU WKH VDPH RU FORVHO\ UHODWHG LGHDV 7KH WHUP GLUHFWHG GLVSHUVDO ZDV LQWURGXFHG E\ +RZH DQG 6PDOOZRRG f DW WKH VDPH WLPH WKH LQLWLDO IUDPHZRUN RI VHHG GLVSHUVDO WKHRU\ ZDV EHLQJ FULWLFDOO\ H[DPLQHG :KHHOZULJKW DQG 2ULDQV +HUUHUD f 7KXV EHFDXVH GLUHFWHG GLVSHUVDO ZDV QRW H[SHFWHG XQGHU WKH GLIIXVH PXWXDOLVP YLHZ WKH WHUP ZDV QRW ZLGHO\ DGRSWHG DQG LQVWHDG WHUPV VXFK DV nVDIH VLWHVn nQXUVH SODQWVn nUHFUXLWPHQW IRFLn nVXFFHVVLRQ IDFLOLWDWLRQf DQG nWDUJHWHG GLVSHUVDOn ZHUH XVHG IRU FRQFHSWV UHODWHG WR GLUHFWHG GLVSHUVDO *UXEE +DUSHU 6WLOHV %D]]D] 9LHLUD HW D/ f

PAGE 121

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n GLVSHUVDO IDWH $GYDQWDJHV RI 'LVSHUVDO 3ODQWV GLVSOD\ PDQ\ PRUSKRORJLFDO IHDWXUHV DVVRFLDWHG ZLWK GLIIHULQJ PHWKRGV RI VHHG GLVSHUVDO HJ %HDO 5LGOH\ YDQ GHU 3LMO f &RQVLGHULQJ WKH FRVW WR WKH SODQW RI SURGXFLQJ VXFK GLVSHUVDOUHODWHG VWUXFWXUHV LW LV UHDVRQDEOH WR H[SHFW VRPH DGYDQWDJH WR GLVSHUVDO +DPLOWRQ DQG 0D\ 7KRPSVRQ DQG :LOOVRQ +RZH DQG 6PDOOZRRG f 7KUHH K\SRWKHVHV FRQFHUQLQJ WKH VSDWLDO DVSHFWV RI GLVSHUVDO KDYH EHHQ SURSRVHG f HVFDSH IURP KLJK PRUWDOLW\ FDXVHG E\ GLVWDQFH RU GHQVLW\GHSHQGHQW IDFWRUV QHDU FRQVSHFLILFV HVFDSH K\SRWKHVLVf f FRORQL]DWLRQ RI UDUH XQSUHGLFWDEOH HSKHPHUDO VLWHV VXFK DV WUHHIDOO JDSV FRORQL]DWLRQ K\SRWKHVLVf DQG f GLUHFWHG GLVSHUVDO WR SDUWLFXODU PLFURKDELWDWV HVSHFLDOO\ VXLWDEOH IRU VXUYLYDO +RZH DQG 6PDOOZRRG +RZH :LOOVRQ f %HFDXVH WKHVH K\SRWKHVHV DUH QRW PXWXDOO\ H[FOXVLYH WKH\ FDQ EH DVVHVVHG RQO\ LI GLVSHUVDO VLWHV DQG SRVWGLVSHUVDO IDWHV DUH NQRZQ ,Q DGGLWLRQ WR WKH WKUHH K\SRWKHVHV PHQWLRQHG DERYH VHYHUDO RWKHU DGYDQWDJHV RI GLVSHUVDO KDYH EHHQ SURSRVHG DQG ZLOO EH PHQWLRQHG KHUH EXW QRW GLVFXVVHG LQ GHWDLO EHFDXVH WKH\ DUH QRW FRQFHUQHG ZLWK WKH VSDWLDO GLVWULEXWLRQ RI VHHGV LQ UHODWLRQ WR DGXOW WUHHV *HQH IORZ LV DQRWKHU DGYDQWDJH RI GLVSHUVDO /HYLQ DQG .HUVWHU f *HQH IORZ

PAGE 122

KHOSV DYRLG LQEUHHGLQJ DQG PD\ RFFXU GXULQJ ERWK SROOLQDWLRQ DQG GLVSHUVDO +DPULFN DQG 1DVRQ f 7KH FRQWULEXWLRQV RI SROOLQDWLRQ DQG GLVSHUVDO WR JHQH IORZ FDQ EH GLVWLQJXLVKHG E\ UHFHQWO\ GHYHORSHG WHFKQLTXHV DQG IXWXUH VWXGLHV VKRXOG EH KHOSIXO LQ HYDOXDWLQJ WKH LQIOXHQFH RI GLIIHUHQW GLVSHUVDO YHFWRUV RQ JHQHWLF VWUXFWXUH RI SODQW SRSXODWLRQV +DPULFN DQG /RYHOHVV +DPULFN HW DO *LEVRQ DQG :KHHOZULJKW :HLEOHQ DQG 7KRPVRQ $OYDUH]%X\OOD HW DO f 1RWH WKDW LQEUHHGLQJ LV QRW QHFHVVDULO\ D SUREOHP IRU DOO SODQW VSHFLHV DQG LQ VRPH FDVHV VKRUWGLVWDQFH GLVSHUVDO LV IDYRUHG RYHU ORQJGLVWDQFH +RZH DQG 6PDOOZRRG f $OVR WKH VSDWLDO GLVWDQFH DPRQJ FRQVSHFLILF WUHHV LV QRW QHFHVVDULO\ D JRRG LQGLFDWLRQ RI WKH JHQHWLF GLVWDQFH DPRQJ LQGLYLGXDOV RU WKH GLVWDQFH WUDYHOHG GXULQJ GLVSHUVDO HJ *LEVRQ DQG :KHHOZULJKW f (IIHFWV RQ JHUPLQDWLRQ KDYH DOVR EHHQ SURSRVHG DV DGYDQWDJHV RI GLVSHUVDO UHYLHZHG E\ 7UDYHVHW DQG :LOOVRQ f $OWKRXJK DQLPDOV PD\ VFDULI\ VHHGV DQG LQIOXHQFH WKH UDWH RU SURSRUWLRQ RI VHHGV JHUPLQDWLQJ WKH HIIHFWV RI JXW SDVVDJH KDYH OLWWOH WR GR ZLWK GLVSHUVDO SHU VH +RZH DQG 6PDOOZRRG f &RQVXPSWLRQ RI IOHVK\ IUXLWV E\ YHUWHEUDWHV LV JHQHUDOO\ DGYDQWDJHRXV IRU UHPRYDO RI WKH SXOS EXW RIWHQ GRHV QRW RWKHUZLVH DIIHFW JHUPLQDWLRQ &KDSWHU -DQ]HQ E +RZH HW DO -DFNVRQ HW DO f (VFDSH 7KH HVFDSH K\SRWKHVLV LV H[SHFWHG WR EH DQ DGYDQWDJH IRU PRVW SODQWV DQG LV VXSSRUWHG E\ QXPHURXV VWXGLHV WKDW KDYH VKRZQ GHQVLW\ RU GLVWDQFHGHSHQGHQW PRUWDOLW\ QHDU SDUHQW WUHHV HJ &ODUN DQG &ODUN +RZH f 1RWH WKDW KLJK OHYHOV RI VHHG SUHGDWLRQ RU VHHGOLQJ PRUWDOLW\ DUH QRW VXIILFLHQW WR GHPRQVWUDWH EHQHILWV RI HVFDSH ,I b RI D SODQWnV SURJHQ\ LV NLOOHG UHJDUGOHVV RI ORFDWLRQ WKHQ HVFDSH LV QRW DQ DGYDQWDJH RI GLVSHUVDO ,QVWHDG WKH HYLGHQFH VKRXOG EH GLVSURSRUWLRQDWH PRUWDOLW\ FRUUHODWHG QHJDWLYHO\ ZLWK GLVWDQFH RU SRVLWLYHO\ ZLWK GHQVLW\ +RZH f (YHQ LI HVFDSH LV VKRZQ WR EH DQ DGYDQWDJH IRU D SDUWLFXODU SODQW VSHFLHV KRZHYHU FRORQL]DWLRQ RU GLUHFWHG GLVSHUVDO FRXOG DOVR EH LPSRUWDQW IRU WKH VHHGV WKDW GR HVFDSH

PAGE 123

&RORQL]DWLRQ 7KH FRORQL]DWLRQ K\SRWKHVLV LV PRVW UHOHYDQW ZKHQ VXLWDEOH VLWHV IRU HVWDEOLVKPHQW DUH XQSUHGLFWDEOH RU UDQGRPO\ GLVWULEXWHG DV LV WKRXJKW WR EH WKH FDVH IRU QHZ WUHHIDOO JDSV LQ WURSLFDO IRUHVWV +DUWVKRUQ %URNDZ 9DQ GHU 0HHU DQG %RQJHUV f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f 7KH LPSRUWDQFH RI VPDOO JDSV DQG HYHQ VXEFDQRS\ JDSV LV QRW ZHOO NQRZQ &RQQHOO HW DO f 'LUHFWHG 'LVSHUVDO $OWHUQDWLYHO\ GLUHFWHG GLVSHUVDO FDQ UHVXOW LI GLVSHUVDO DJHQWV GHSRVLW VHHGV GLVSURSRUWLRQDWHO\ LQ VXLWDEOH ORFDWLRQV 7KXV GLUHFWHG GLVSHUVDO KDV WZR FRPSRQHQWV QRQUDQGRP DUULYDO DQG LQFUHDVHG VXUYLYDO LQ SUHGLFWDEOH ORFDWLRQV +RZH DQG 6PDOOZRRG +RZH 6FKXSS HW DO f 7R EHQHILW IURP GLUHFWHG GLVSHUVDO D SODQW PXVW KDYH IUXLWV ZLWK FKDUDFWHULVWLFV WKDW DWWUDFW FHUWDLQ GLVSHUVHUV PRUH WKDQ RWKHUV RU PXVW KDYH D PRUSKRORJ\ WKDW HQDEOHV WKH SURSDJXOHV WR DUULYH LQ FHUWDLQ KDELWDW SDWFKHV PRUH RIWHQ WKDQ H[SHFWHG E\ FKDQFH 9HQDEOH DQG %URZQ f 1RWH KRZHYHU WKDW WKH H[DPSOHV EHORZ VKRZ WKDW SODQWV QHHG QRW EH DGDSWHG VSHFLILFDOO\ IRU GLUHFWHG GLVSHUVDO DV ZDV RULJLQDOO\ WKHRUL]HG ,QVWHDG DUJXH WKDW GLUHFWHG GLVSHUVDO RFFXUV LQ PDQ\ VLWXDWLRQV

PAGE 124

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f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f QRZ GLVFXVV WKH UHDVRQV IRU WKLV GLIILFXOW\ DQG VXJJHVW SRWHQWLDO VROXWLRQV 0RVW GLVSHUVDO GRHV QRW UHVXOW LQ VHHGOLQJ HVWDEOLVKPHQW HYHQ E\ KLJK TXDOLW\ GLVSHUVHUV WKXV GLUHFWHG GLVSHUVDO LV OLNHO\ WR EH VXEWOH %HFDXVH PRVW VHHGV GLH UHVHDUFK RQ HVFDSH KDV RYHUVKDGRZHG FRORQL]DWLRQ DQG GLUHFWHG GLVSHUVDO 0XFK ODUJHU VDPSOH VL]HV DUH QHHGHG WR GHWHFW IDFWRUV FRUUHODWHG ZLWK VHHGOLQJ VXUYLYDO WKDQ IRU VHHG UHPRYDO LQ D QDWXUDO VHWWLQJ +RZH Ef ,Q DGGLWLRQ PDQ\ VWXGLHV HLWKHU GR QRW DWWHPSW WR GLVWLQJXLVK WKH WKUHH K\SRWKHVHV FRPELQH FRORQL]DWLRQ DQG GLUHFWHG GLVSHUVDO LQWR FRORQL]DWLRQ RU LJQRUH GLUHFWHG GLVSHUVDO DOWRJHWKHU +HUUHUD DQG -RUGDQR 'LU]R DQG 'RPLQJXH] 0DUWLQH]5DPRV DQG $OYDUH]%X\OOD /HYH\ HW DO f )XUWKHUPRUH DOO WKH GDWD WR HYDOXDWH GLUHFWHG GLVSHUVDO DUH VHOGRP FROOHFWHG LQ RQH VWXG\ ,Q SDUWLFXODU WKH HFRORJLFDO OLWHUDWXUH LV GRPLQDWHG E\ VWDJHVSHFLILF VWXGLHV ZKRVH UHVXOWV FDQQRW QHFHVVDULO\ EH XVHG WR GLVWLQJXLVK DPRQJ WKH K\SRWKHVHV /RQJWHUP VWXGLHV WKDW GRFXPHQW VXUYLYDO WKURXJK VHTXHQWLDO VWDJHV LQFOXGLQJ GLVSHUVDO JHUPLQDWLRQ VHHGOLQJ

PAGE 125

HVWDEOLVKPHQW DQG JURZWK WR PDWXULW\ DUH QHHGHG WR HYDOXDWH WKH FRQVHTXHQFHV RI GLVSHUVDO DQG WKHUHE\ WHDVH DSDUW WKH K\SRWKHVHV +HUUHUD HW DO +RXOH 6FKXSS DQG )XHQWHV f 6XFK VWXGLHV DUH GLIILFXOW IRU ORQJOLYHG VSHFLHV ,Q PRVW FDVHV WKH GLVWULEXWLRQ RI GLVSHUVHG VHHGV LV KLJKO\ OHSWRNXUWLF DQG ULJKW VNHZHG ZLWK PRVW VHHGV QHDU WKH SDUHQW DQG SURJUHVVLYHO\ IHZHU IDUWKHU DZD\ -DQ]HQ /HYLQ DQG .HUVWHU :LOOVRQ +HUUHUD HW DO +RXOH 6FKXSS DQG )XHQWHV f 7KH VKDSH RI WKH FXUYH LV EHVW GHVFULEHG E\ HLWKHU QHJDWLYH H[SRQHQWLDO RU LQYHUVH SRZHU IXQFWLRQV 3RUWQR\ DQG :LOOVRQ :LOOVRQ /DPDQ D EXW VHH 0XUUD\ f 7KHUHIRUH VHHG VKDGRZV DUH DOZD\V QRQUDQGRP ZLWK UHVSHFW WR GLVWDQFH IURP WKH SDUHQW SODQW 3ORWWLQJ WKH DEXQGDQFH RI VHHGV RYHU GLVWDQFH DV D OLQHDU IXQFWLRQ KRZHYHU FDQ REVFXUH WKH JUHDW KHWHURJHQHLW\ ZLWKLQ WKH WDLOV RI WKH GLVWULEXWLRQ -DQ]HQ D 6WLOHV 3RUWQR\ DQG :LOOVRQ :LOOVRQ f ,W LV LPSRUWDQW WR QRWH WKDW QRQUDQGRP GLVSHUVDO DVVRFLDWHG ZLWK VXFK KHWHURJHQHLW\ LV D QHFHVVDU\ EXW QRW VXIILFLHQW FRQGLWLRQ IRU GLUHFWHG GLVSHUVDO )RU GLUHFWHG GLVSHUVDO WR RFFXU QRQUDQGRP GLVSHUVDO PXVW EH WR VXLWDEOH VLWHV VR WKDW WKH RYHUDOO UHVXOW LV D SRVLWLYH GLVSURSRUWLRQDWH HIIHFW RQ UHFUXLWPHQW 6LPLODUO\ PDQ\ VWXGLHV VKRZ KHWHURJHQHLW\ LQ VHHGOLQJ GLVWULEXWLRQV
PAGE 126

GLVSHUVDO HJ /DPDQ Ef +RZHYHU D VXEVHW RI WKH GLVSHUVHU DVVHPEODJH PD\ EH UHVSRQVLEOH IRU GLVSHUVLQJ PRVW RI WKH VHHGV WKDW HYHQWXDOO\ HVWDEOLVK (YHQ IRU SODQW VSHFLHV ZLWK D UHVWULFWHG VHW RI GLVSHUVHUV LW KDV EHHQ VKRZQ WKDW GLVSHUVHU VSHFLHV GLIIHU LQ WKH SHU VHHG SUREDELOLW\ RI HVWDEOLVKPHQW 5HLG f 7KH RWKHU FULWLFDO DVSHFW QHFHVVDU\ WR DVVHVV GLUHFWHG GLVSHUVDO LV GHWHUPLQLQJ SRVWn GLVSHUVDO SODQW IDWH ,W LV ZHOO NQRZQ WKDW SODQW PRUWDOLW\ LV JHQHUDOO\ KLJKHVW LQ WKH VHHG DQG VHHGOLQJ VWDJHV 'DUZLQ +DUSHU f (YHU\ SRWHQWLDO UHFUXLW UHTXLUHV D VSDFH LQ ZKLFK WR JURZ LQ DGGLWLRQ WR ZDWHU QXWULHQWV DQG OLJKW 6XFK VSDFHV DUH UDUH LQ WKH HQYLURQPHQW 7KHUHIRUH WKH SUREDELOLW\ RI HVWDEOLVKPHQW IRU DQ\ VHHG LV YHU\ ORZ ,W IROORZV WKHQ WKDW DQ\ VXUYLYDO EHQHILW DW HDUO\ VWDJHV LV OLNHO\ WR LQFUHDVH WKH FKDQFHV RI IXUWKHU VXUYLYDO 6FKXSS DQG )XHQWHV f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nV ILWQHVV DV HVFDSH DQG FRORQL]DWLRQ ,Q VXPPDU\ PRUH GHWDLOHG VWXGLHV RQ ZKHUH VHHGV DUH GLVSHUVHG KRZ WKH\ JHW WKHUH DQG ZKDW KDSSHQV DIWHUZDUGV DUH QHFHVVDU\ WR

PAGE 127

GLVWLQJXLVK WKH K\SRWKHVHV NQRZ RI RQO\ RQH VWXG\ WKDW KDV DWWHPSWHG WR H[DPLQH WKH WKUHH K\SRWKHVHV 0DVDNL HW DO f 7KH 'LIIXVH 0XWXDOLVP 3DUDGLJP 7KH HYROXWLRQDU\ DVSHFWV RI WKH IUXLWIUXJLYRUH LQWHUDFWLRQ QRWHG E\ 5LGOH\ f DQG YDQ GHU 3LMO f ZHUH ILUVW FOHDUO\ GHYHORSHG E\ 6QRZ f 7KHVH LGHDV ZHUH ODWHU H[SDQGHG LQWR WKH VSHFLDOLVWJHQHUDOLVW GLFKRWRP\ RI SODQW WUDLWV DQG GLVSHUVDO TXDOLW\ 0F.H\ +RZH DQG (VWDEURRN f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f 2YHU WKH QH[W GHFDGH KRZHYHU WKH YLHZ WKDW HPHUJHG IURP PRUH GHWDLOHG VWXGLHV PDLQO\ RQ DYLDQ IRUDJLQJ EHKDYLRUf ZDV WKDW FRHYROXWLRQ EHWZHHQ SODQWV DQG WKHLU GLVSHUVHUV ZDV QRW VSHFLHV VSHFLILF EXW LQVWHDG ZDV GLIIXVH LQYROYLQJ DGDSWDWLRQV EHWZHHQ ODUJH JURXSV RI SODQWV DQG JURXSV RI GLVSHUVHUV :KHHOZULJKW DQG 2ULDQV +RZH +HUUHUD /HYH\ HW DO f ,QGHHG DOWKRXJK SODQW VSHFLHV YDU\ WUHPHQGRXVO\ LQ WKH QXPEHU RI SRWHQWLDO GLVSHUVHUV WKH\ DWWUDFW DQG D IHZ KDYH YHU\ UHVWULFWHG VHWV RI GLVSHUVDO DJHQWV -DQ]HQ DQG 0DUWLQ &KDSPDQ HW DO f WKHUH DUH QR XQHTXLYRFDO H[DPSOHV RI D SODQW VSHFLHV HQWLUHO\ GHSHQGHQW RQH VSHFLHV RI GLVSHUVHU :LWPHU DQG &KHNH f 6RPH LQWHUSUHWDWLRQV RI WKH GLIIXVH PXWXDOLVP IUDPHZRUN VHHP WR H[FOXGH WKH SRVVLELOLW\ RI GLUHFWHG GLVSHUVDO 7ZR RI WKH PDLQ REMHFWLRQV WR WKH VSHFLDOLVWJHQHUDOLVW

PAGE 128

GLFKRWRP\ DUH f WKDW SODQWV UHZDUG GLVSHUVHUV IRU UHPRYLQJ IUXLWV EXW QRW IRU GLVSHUVLQJ WKHP WR DSSURSULDWH VLWHV DQG WKXV FDQ GR OLWWOH WR LQIOXHQFH ZKHUH GLVSHUVHUV WDNH VHHGV DQG f VXLWDEOH VLWHV IRU SODQW HVWDEOLVKPHQW VDIH VLWHVf DUH VSDWLDOO\ DQG WHPSRUDOO\ XQSUHGLFWDEOH :KHHOZULJKW DQG 2ULDQV -DQ]HQ Ff 6WXGLHV RQ SROOLQDWLRQ ZHUH ZHOO DKHDG RI WKRVH RQ VHHG GLVSHUVDO ZKHQ VSHFLDOLVWJHQHUDOLVW IUDPHZRUN ZDV EHLQJ GHYHORSHG +RZH Ef DQG ERWK WKH VSHFLDOLVWJHQHUDOLVW GLFKRWRP\ DQG WKH FULWLTXHV WKDW IROORZHG ZHUH EDVHG RQ FRPSDULVRQV WR SROOLQDWLRQ 7KLV FRPSDULVRQ WR SROOLQDWLRQ LV H[HPSOLILHG E\ WKH VWDWHPHQW SODQWV FDQQRW GLUHFW WKH GLVSHUVDO RI VHHGV WR D SDUWLFXODU ORFDWLRQ ZLWK D GHJUHH RI H[DFWQHVV FRPSDUDEOH WR SROOHQ GLVSHUVDO WKRXJK SRVVLEO\ WKH\ FRXOG IDYRU DQLPDO YHFWRUV ZLWK SDUWLFXODU IRUDJLQJ EHKDYLRUV KDELWDW SUHIHUHQFHV RU SUREDELOLVWLF SDWWHUQV RI VHHG GLVSHUVDO :KHHOZULJKW DQG 2ULDQV f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f DQG RWKHUV ZHUH QRW WR UHMHFW GLUHFWHG GLVSHUVDO SHU VH EXW UDWKHU WR UHIXWH WKH LGHD RI VSHFLHVVSHFLILF FRHYROXWLRQ EHWZHHQ SODQWV DQG GLVSHUVHUV ,QWHUHVW LQ VHHG GLVSHUVDO PXWXDOLVPV SDUDOOHOHG WKDW LQ IDFWRUV SURPRWLQJ WURSLFDO IRUHVW GLYHUVLW\ HVSHFLDOO\ VHHG SUHGDWLRQ DQG VHHGOLQJ PRUWDOLW\ LQ UHODWLRQ WR DGXOW GLVWULEXWLRQV &RQQHOO -DQ]HQ +XEEHOO f 7KH HVFDSH K\SRWKHVLV LV DOVR UHIHUUHG WR DV WKH -DQ]HQ&RQQHOO K\SRWKHVLV DOWKRXJK WKH WZR DUH QRW HTXLYDOHQW +RZH DQG 6PDOOZRRG &ODUN DQG &ODUN %XUNH\ &LQWUD Ef $V QRWHG DERYH WKH HVFDSH K\SRWKHVLV SUHGLFWV GLVSURSRUWLRQDWH PRUWDOLW\ DV D IXQFWLRQ RI GLVWDQFH IURP SDUHQW SODQWV ZKLOH WKH -DQ]HQ&RQQHOO K\SRWKHVLV VWDWHV WKDW VXFK GLVSURSRUWLRQDWH PRUWDOLW\ IRU RQH VSHFLHV DOORZV RWKHU VSHFLHV D KLJKHU SUREDELOLW\ RI HVWDEOLVKPHQW DQG

PAGE 129

WKXV SURPRWHV VSHFLHV FRH[LVWHQFH %RWK K\SRWKHVHV SUHGLFW DQ DGYDQWDJH RI ORFDO GLVSHUVDO 6XEVHTXHQWO\ VHHG SUHGDWLRQ H[SHULPHQWV KDYH GRPLQDWHG WKH OLWHUDWXUH ZLWK UHODWLYHO\ OLWWOH HIIRUW SODFHG RQ GRFXPHQWLQJ VHHG VXUYLYDO DV D IXQFWLRQ RI DQ\ RWKHU YDULDEOH EHVLGH GLVWDQFH IURP SDUHQWV EXW VHH 6FKXSS D .LWDMLPD DQG $XJVSXUJHU +RZH E /DPDQ &LQWUD Df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f WKH\ DUH DFWXDOO\ FRQVLVWHQW ZLWK WKH WKHRU\ DV DOO WKUHH H[DPSOHV LQYROYH HDFK VSHFLHV RI SODQW EHLQJ GLVSHUVHG E\ PRUH WKDQ RQH VSHFLHV RI GLVSHUVHU ZKLFK LQ WXUQ GLVSHUVH VHYHUDO RWKHU VSHFLHV RI SODQWV HJ 5HVWUHSR f 7KH\ PD\ EH PRUH VSHFLDOL]HG WKDQ PRVW RWKHU SODQWGLVSHUVHU V\VWHPV DQG SHUKDSV LOOXVWUDWH RQH HQGSRLQW RI D FRQWLQXXP RI GLVSHUVDO VWUDWHJLHV &RUYLGV $SSUR[LPDWHO\ VSHFLHV RI SLQHV 3LQXV PRVW LQ WKH VXEJHQXV 6WUREXVf DUH GLVSHUVHG E\ MD\V DQG QXWFUDFNHUV &RUYLGDH 7RPEDFN DQG /LQKDUW %HQNPDQ f 0RVW RWKHU SLQHV DUH ZLQGGLVSHUVHG DOWKRXJK WKH\ PD\ EHQHILW IURP VHFRQGDU\ GLVSHUVDO E\ VFDWWHUKRDUGLQJ URGHQWV 9DQGHU :DOO f 7KHVH SLQHV RFFXU LQ ZHVWHUQ 1RUWK $PHULFD DQG DFURVV (XUDVLD DQG DUH GLVWLQJXLVKHG IURP WKH ZLQGGLVSHUVHG VSHFLHV E\ KDYLQJ ODUJH VHHGV WKDW ODFN D ZHOOGHYHORSHG ZLQJHG DSSHQGDJH IRU ZLQG GLVSHUVDO DQG

PAGE 130

FRQHV WKDW GR QRW UHOHDVH WKH VHHGV 7RUDEDFN DQG /LQKDUW f 7KH PRVW VSHFLDOL]HG GLVSHUVHUV DUH WKH WZR VSHFLHV RI QXWFUDFNHUV 1XFLIUDJDf 1XWFUDFNHUV KDYH D VXEOLQJXDO SRXFK XQLTXH DPRQJ ELUGVf LQ ZKLFK WKH\ FDUU\ XS WR VHHGV 9DQGHU :DOO DQG %DLGD 7RPEDFN f 7KH\ EXU\ VHHGV LQ WKH JURXQG LQ VKDOORZ VXUIDFH FDFKHV RI VHHGV 3UHVXPDEO\ WKH\ GLVSHUVH VHHGV FRQVLGHUDEOH GLVWDQFHV /DQQHU f (DFK QXWFUDFNHU FDFKHV WKRXVDQGV RI VHHGV HDFK \HDU H[FHHGLQJ LWV GLHWDU\ UHTXLUHPHQWV WLPHV 9DQGHU :DOO DQG %DLGD 7RPEDFN f 6HYHUDO RWKHU FRUYLGV FDFKH SLQH VHHGV DQG DUH LPSRUWDQW GLVSHUVHUV LQFOXGLQJ SLQ\RQ MD\V *\PQRUKLQXV F\DQRFHSKDOXVf ZHVWHUQ VFUXE MD\V $SKHORFRPD FDOLIRPLFDf DQG 6WHOOHUnV MD\V &\DQRFLWWD VWHOOHQf 7KHVH ELUGV KDYH KLJKO\ GHYHORSHG VSDWLDO PHPRU\ %DLGD DQG .DPLO .DPLO DQG -RQHV f EXW GR QRW UHWULHYH DOO FDFKHV HYHU\ \HDU /DQQHU f 0RVW VHHGV WKDW DUH QRW WDNHQ E\ ELUGV DQG IDOO EHQHDWK DGXOW 3LQXV PRQRSK\OOD WUHHV DUH KDUYHVWHG E\ URGHQWV ZKLFK DUH OHVV HIIHFWLYH GLVSHUVHUV WKDQ FRUYLGV EHFDXVH WKH\ ODUGHUKRDUG PDQ\ VHHGV LQ QHDUE\ EXUURZV 9DQGHU :DOO f 0RVW RI WKH FRUYLGGLVSHUVHG SLQH VSHFLHV DUH IRXQG LQ [HULF KDELWDWV ZKHUH EXULDO SURWHFWV VHHGV IURP GHVLFFDWLRQ /DQQHU f +RZHYHU ERWK WKH FRUYLGV DQG WKH URGHQWV VFDWWHUKRDUG VRPH VHHGV LQ VLWHV XQGHU EXVKHV WKDW SURYLGH VKDGH DQG DUH WKXV EHQHILFLDO IRU JHUPLQDWLRQ DQG HVWDEOLVKPHQW 9DQGHU :DOO f ,Q DGGLWLRQ VRPH RI WKH SLQHV DUH HDUO\ VXFFHVVLRQDO VSHFLHV SLRQHHUVf DQG WKH ELUGV QXWFUDFNHUV DQG SLQ\RQ MD\V LQ SDUWLFXODU DUH NQRZQ WR PDNH IUHTXHQW FDFKHV LQ RSHQ DUHDV /DQQHU f &RUYLGV DQG VTXLUUHOV DOVR GLVSHUVH RDNV DQG EHHFKHV )DJDFHDHf E\ VFDWWHUKRDUGLQJ $OWKRXJK WKH SUHVXPHG PXWXDO DGDSWDWLRQV LQ WKHVH VSHFLHV DUH OHVV FOHDU WKDQ IRU QXWFUDFNHUV DQG SLQHV LQ LV FOHDU WKDW PDQ\ VHHGV DUH FDFKHG LQ VXLWDEOH VLWHV IRU JHUPLQDWLRQ DQG HVWDEOLVKPHQW %RVVHPD 'DUOH\+LOO DQG -RKQVRQ 6RUN -RKQVRQ HW DO f ,Q DGGLWLRQ GLVSHUVDO E\ MD\V KDV EHHQ LPSOLFDWHG LQ WKH UDSLG SRVW 3OHLVWRFHQH QRUWKZDUG VSUHDG RI RDNV DQG RWKHU SODQW VSHFLHV -RKQVRQ DQG :HEE :LONLQVRQ D &ODUN HW DO f

PAGE 131

0LVWOHWRHV 0LVWOHWRHV DERXW JHQHUD 9LVFDFHDH /RUDQWKDFHDH (UHPROHSLGDFHDHf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f ,Q IRUHVWHG UHJLRQV RI &HQWUDO $PHULFD DQG 6RXWK $PHULFD WKH\ DUH GLVSHUVHG PDLQO\ E\ HXSKRQLDV DQG RWKHU WDQDJHUV 7KUDXSLGDHf DQG WR D OHVVHU H[WHQW IO\FDWFKHUV HJ =LPPHULXV YLOLVVLPXV 7\UDQQLGDH 'DYLGDU 5HVWUHSR 6WLOHV DQG 6NXWFK 0RQWHLUR HW DO 6DUJHQW f ,Q D &KLOHDQ GHVHUW VLWH PRFNLQJELUGV ^0LPDV WKHQFD 0LPLGDHf DUH WKH PDLQ GLVSHUVHUV RI PLVWOHWRH 0DUWLQH] GHO 5LR HW DO f ZKLOH LQ 1RUWK $PHULFD ZD[ZLQJV DQG SKDLQRSHSODV %RUDE\FLOOLGDHf DUH WKH PDLQ PLVWOHWRH GLVSHUVHUV :DOVEHUJ 6NHDWH /DUVRQ f 0LVWOHWRH VHHGV LQ IUXLWV QRW HDWHQ E\ ELUGV DQG VHHGV GLVSHUVHG WR WKH JURXQG KDYH QR FKDQFH RI VXUYLYDO 6WXGLHV E\ 5HLG f DQG 6DUJHQW f LOOXVWUDWH GLUHFWHG GLVSHUVDO RI PLVWOHWRHV 6XLWDEOH HVWDEOLVKPHQW VLWHV DUH QRW VLPSO\ WZLJV DQG EUDQFKHV EXW OLYH WZLJV RI D FHUWDLQ VL]H GHSHQGLQJ RQ WKH VSHFLHV ,Q $XVWUDOLD $P\HPD TXDQGDQJ PLVWOHWRHV HVWDEOLVKHG EHVW RQ PP GLDPHWHU WZLJV 5HLG f ZKLOH LQ &RVWD 5LFD 3KRUDGHQGURQ UREXVWLVVLPXP HVWDEOLVKHG EHVW RQ PP WZLJV 6DUJHQW f /DUJHU WZLJV DQG EUDQFKHV KDYH WKLFNHU EDUN WKDW PD\ SUHFOXGH HVWDEOLVKPHQW DQG VPDOOHU WZLJV DUH PRUH OLNHO\ WR GLH WKDQ ODUJHU WZLJV DIWHU PLVWOHWRH LQIHFWLRQ 6DUJHQW f 6DUJHQW f VKRZHG WKDW HXSKRQLDV WHQG WR SHUFK RQ WZLJV RI WKH DSSURSULDWH VL]H IRU

PAGE 132

HVWDEOLVKPHQW DQG WKHUHIRUH ZLWKLQ D WUHH HVWDEOLVKPHQW ZDV RQ D QRQUDQGRP VHW RI WKH DYDLODEOH WZLJV DQG EUDQFKHV 5HLG f VLPLODUO\ VKRZHG WKDW PLVWOHWRHELUGV 'LFDHXP KLUXQLQDFHXPf ZHUH PRUH OLNHO\ WR GHSRVLW VHHGV RQ WKH DSSURSULDWHVL]HG WZLJV WKDQ ZHUH KRQH\HDWHUV $FDQWKDJHQ\V UXIRJXODULVf ,Q ERWK FDVHV D UHVWULFWHG VHW RI GLVSHUVHUV LV HQWLUHO\ UHVSRQVLEOH IRU VXFFHVVIXO PLVWOHWRH HVWDEOLVKPHQW 1RQUDQGRP GLVWULEXWLRQ RI PLVWOHWRHV DPRQJ WKH DYDLODEOH KRVW SODQWV KRVW SUHIHUHQFHVf KDYH EHHQ VKRZQ LQ VHYHUDO RWKHU VWXGLHV 0RQWHLUR HW DO 0DUWLQH] GHO 5LR HW DO /DUVRQ f +RVW SUHIHUHQFHV KDYH EHHQ QRWHG IRU RWKHU HSLSK\WHV HJ )LFXV 'DQLHOV /DPDQ 3DWHO 3XW] f EXW LW LV XQFOHDU LI WKH\ DUH UHODWHG WR WKH SDWWHUQ RI GLVSHUVDO RU RI VXUYLYDO 'LVSURSRUWLRQDWH HVWDEOLVKPHQW LQ FHUWDLQ VLWHV GRHV QRW LPSO\ GLVSURSRUWLRQDWH GLVSHUVDO WR WKRVH VLWHV 6FKXSS DQG )XHQWHV f DOWKRXJK $XJXVW f GLG VXJJHVW WKDW EDWV SURYLGHG GLUHFWHG GLVSHUVDO RI ILJ VHHGV WR SDOP WUHHV LQ 9HQH]XHOD /DPDQ Ef KRZHYHU SURSRVHG WKDW KRVW SUHIHUHQFHV RI %RUQHDQ )LFXV DUH FDXVHG E\ GLIIHUHQFHV LQ HVWDEOLVKPHQW UHTXLUHPHQWV QRW E\ GLIIHUHQWLDO GLVSHUVDO DV )LFXV VHHGV DUH GLVSHUVHG E\ PDQ\ VSHFLHV RI ELUGV DQG PDPPDOV DQG SUREDEO\ DFKLHYH ZLGHVSUHDG GLVVHPLQDWLRQ $QWV $ ZLGH YDULHW\ RI WURSLFDO DQG WHPSHUDWH SODQWV SURGXFH VHHGV ZLWK OLSLGULFK HODLVRPHV WKDW DWWUDFW DQWV %HDWWLH DQG &XOYHU +RUYLW] DQG 6FKHPVNH E 2KNDZDUD HW DO f 7KLV DQWSODQW PXWXDOLVP LV ZRUOGZLGH RFFXUULQJ LQ KDELWDWV UDQJLQJ IURP GHVHUWV WR WURSLFDO IRUHVWV DQG LQYROYLQJ UHSUHVHQWDWLYHV RI RYHU SODQW IDPLOLHV DQG QXPHURXV DQW WD[D %HDWWLH f 7\SLFDOO\ DQWV WDNH WKH VHHGV EDFN WR WKHLU QHVWV DQG GLVFDUG WKH VHHG DIWHU WKH QXWULHQWULFK HODLVRPH KDV EHHQ FRQVXPHG 6HHGV DUH GLVSHUVHG ZKHQ SODFHG RQ UHIXVH SLOHV RU DEDQGRQHG LQ QHVWV %HDWWLH DQG &XOYHU f %HFDXVH WKH QHVWV RU UHIXVH SLOHV KDYH KLJKHU FRQFHQWUDWLRQV RI QXWULHQWV DYDLODEOH WR VHHGOLQJV FRPSDUHG WR WKH VXUURXQGLQJ VRLO VHHGOLQJ JURZWK RU HVWDEOLVKPHQW LV RIWHQ

PAGE 133

KLJKHU LQ VXFK VLWHV WKDQ LQ UDQGRP VLWHV 'DYLGVRQ DQG 0RUWRQ +RUYLW] /HYH\ DQG %\PH f ,Q RQH RI WKH IHZ VWXGLHV WR GRFXPHQW WKH ILWQHVV EHQHILWV RI GLVSHUVDO E\ DQLPDOV +DQ]DZD HW DO f VKRZHG WKDQ D FRKRUW RI DQWGLVSHUVHG &RU\GDOLV DXUHD SURGXFHG b PRUH RIIVSULQJ WKDW D FRQWURO FRKRUW $UULYDO WR QXWULHQWULFK VLWHV KRZHYHU LV DSSDUHQWO\ QRW WKH H[SODQDWLRQ IRU P\UPHFRFKRU\ LQ VRPH DUHDV 5LFH DQG :HVWRE\ %RQG DQG 6WRFN f ,Q &DOLIRUQLD FKDSDUUDO DQW QHVWV DUH ORFDWHG LQ WKH JDSV EHWZHHQ VKUXEV UDWKHU WKDQ XQGHU VKUXEV VR WKDW VHHGV GLVFDUGHG LQ QHVWV PD\ IDFH ORZHU FRPSHWLWLRQ IURP RWKHU SODQWV %R\G f ,Q FRQWUDVW WR YHUWHEUDWH IUXJLYRUHV ZKLFK XVXDOO\ HDW WKH UHZDUG IUXLW SXOSf EHIRUH GLVSHUVLQJ VHHGV DQWV W\SLFDOO\ WDNH VHHGV EDFN WR WKHLU QHVWV EHIRUH HDWLQJ WKH HODLVRPH &XOYHU DQG %HDWWLH +RUYLW] f ,Q VRPH VSHFLHV UHPRYDO RI WKH HODLVRPH LQGXFHV JHUPLQDWLRQ DQG WKXV WKH DQWGLVSHUVDO SURYLGHV D SUHGLFWDEOH FXH WKDW D VXLWDEOH VLWH KDV EHHQ UHDFKHG +RUYLW] f 6HHGV QRW WDNHQ E\ DQWV DUH OLNHO\ HDWHQ E\ URGHQWV DQG RWKHU VHHG SUHGDWRUV 2n'RZG DQG +D\ +HLWKDXV EXW VHH 6PLWK HW DO f EXW PDQ\ VSHFLHV UHOHDVH WKH VHHGV GXULQJ WKH PRUQLQJ ZKHQ DQWV DUH DFWLYH EXW URGHQWV DUH QRW *LEVRQ 2KNDZDUD HW DO f $QWV DOVR DFW DV VHFRQGDU\ GLVSHUVHUV RI VHHGV LQ ELUG GURSSLQJV VHH EHORZf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

PAGE 134

'LVSHUVDO EY :LQG DQG %LUGV WR *DSV &DQRS\ JDSV FDXVHG E\ GHDG RU IDOOHQ WUHHV RU EUDQFKHV DUH FUXFLDO UHFUXLWPHQW VLWHV LQ PRVW IRUHVW HFRV\VWHPV 7\SLFDOO\ b RI WKH IRUHVW FDQRS\ LV RSHQHG DQQXDOO\ DQG DERXW b RI WKH IRUHVW DUHD LV LQ JDSV IRUPHG LQ WKH ODVW \HDUV /DZWRQ DQG 3XW] 0DUWLQH]5DPRV HW DO 0XUUD\ f %HFDXVH PDQ\ SODQW VSHFLHV UHTXLUH JDSV IRU JHUPLQDWLRQ DQG VHHGOLQJ JURZWK VHHG DUULYDO LQ JDSV ZRXOG EH KLJKO\ DGYDQWDJHRXV IRU WKHVH VSHFLHV 6HHGV FDQ DUULYH YLD WKH VHHG EDQN SDVW GLVSHUVDOf RU YLD VHHG UDLQ FXUUHQW GLVSHUVDOf 7KH IRUPHU FDVH ZRXOG EH FRORQL]DWLRQ ZKLOH WKH ODWWHU FRXOG EH FRORQL]DWLRQ LI DUULYDO LV LQ SURSRUWLRQ WR WKH DUHD DYDLODEOH RU GLUHFWHG GLVSHUVDO LI DUULYDO LV PRUH RIWHQ WKDQ H[SHFWHG E\ FKDQFH $LU FXUUHQWV DURXQG JDSV PD\ SXOO LQ ZLQGGLVSHUVHG VHHGV IURP WKH VXUURXQGLQJ DUHD 6FKXSS HW DO f ,Q VXSSRUW RI WKLV K\SRWKHVLV DW OHDVW WKUHH VWXGLHV KDYH IRXQG GLVSURSRUWLRQDWH DUULYDO LQ JDSV RI ZLQGGLVSHUVHG VSHFLHV $XJVSXUJHU DQG )UDQVRQ 'HQVORZ DQG *RPH] 'LD] /RLVHOOH HW DO f 6LPLODUO\ %H WXOD OHQWD VHHGV GLVSHUVHG E\ ZLQG DFURVV VQRZ DFFXPXODWHG LQ LQGHQWDWLRQV UHVXOWLQJ LQ QRQUDQGRP VHHG GLVWULEXWLRQV 0DWODFN f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f DQG WKDW IUXJLYRURXV ELUGV DUH DOVR HVSHFLDOO\ DFWLYH LQ DQG DURXQG JDSV HYHQ WKRXJK

PAGE 135

IHZ VSHFLHV FDQ EH FRQVLGHUHG JDS VSHFLDOLVWV 6FKHPVNH DQG %URNDZ :LOOVRQ HW DO /HYH\ Ef %DVHG RQ WKHVH VWXGLHV 6FKXSS HW DO f K\SRWKHVL]HG WKDW VHHG UDLQ RI VPDOOVHHGHG DQLPDOGLVSHUVHG VHHGV ZRXOG EH KLJKHU LQ WURSLFDO IRUHVW JDS HGJHV EXW QR VWXGLHV KDYH WHVWHG WKLV LGHD ,Q ,OOLQRLV KRZHYHU +RSSHV f GLG ILQG VXFK D SDWWHUQ VHHG UDLQ ZDV RIWHQ ELPRGDO ZLWK PRVW VHHGV ODQGLQJ QHDU WKH SDUHQWV DQG D VHFRQG VPDOOHU SHDN DW JDS HGJHV 2YHUDOO +RSSHV f HVWLPDWHG b RI ELUGGLVSHUVHG VHHGV ODQGHG LQ JDSV DQG JDS HGJHV ZKLFK WRJHWKHU FRPSULVHG RQO\ b RI WKH VWXG\ DUHD ,Q -DSDQ VHHG GLVSHUVDO RI &RPXV FRQWURYHUVD E\ ELUGV ZDV QRW GLVSURSRUWLRQDWH WR JDSV EXW JDS DQG JDS HGJH VLWHV ZHUH QRW GLVWLQJXLVKHG 0DVDNL HW DO f %RWK +RSSHV f DQG 'HQVORZ DQG *RPH]'LD] f QRWHG WKH OLNHO\ RFFXUUHQFH RI VHHG GLVSHUVDO IURP SODQWV LQ RQH JDS WR DQRWKHU JDS 7KH PRVW OLNHO\ SODQW VSHFLHV WR EHQHILW IURP GLUHFWHG GLVSHUVDO WR JDSV DUH WKRVH WKDW FDQ HVWDEOLVK QRW LQ QHZ JDSV EXW LQ EXLOGLQJ SKDVH JDSV D IHZ \HDUV ROG 'HQVORZ %UDQGDQL HW DO f &DQRS\ JDSV DUH IRUPHG ZKHQ D WUHH RU EUDQFK IDOOV RU ZKHQ D WUHH GLHV VWDQGLQJ /RFDWLRQV RI QHZ JDSV DUH WKRXJKW WR EH UDQGRPO\ GLVWULEXWHG DQG XQSUHGLFWDEOH +DUWVKRUQ %URNDZ 0DUWLQH]5DPRV HW DO 9DQ GHU 0HHU DQG %RQJHUV f 7KH IDFW WKDW WUHHV RQ JDS HGJHV DUH PRUH OLNHO\ WR IDOO WKDQ DUH WUHHV LQ FORVHG FDQRS\ IRUHVW /DZWRQ DQG 3XW]
PAGE 136

+DELWXDO 3HUFKHV DQG +DELWXDO 'HIHFDWLRQ 6LWHV 0DQ\ IUXJLYRURXV WURSLFDO ELUGV KDYH OHN EUHHGLQJ V\VWHPV LQ ZKLFK WKH PDOHV KDYH GLVSOD\ SHUFKHV ZKHUH WKH\ VSHQG WKH PDMRULW\ RI WKH GD\ GXULQJ WKH EUHHGLQJ VHDVRQ 6XFK VSHFLHV UDQJH IURP PDQDNLQV 3LSULGDHf WR WKH PXFK ODUJHU EHOOELUGV XPEUHOODELUGV DQG FRFNVRIWKHURFN &RWLQJGDHf LQ 1HRWURSLFDO IRUHVWV DQG ELUGVRISDUDGLVH 3DUDGLVDHLGDHf LQ 1HZ *XLQHD 0DQ\ RI WKHVH VSHFLHV DUH FRPPRQ DQG FRQVSLFXRXV 7KH PDOHV RI WZR VSHFLHV RI PDQDNLQV DQG WZR VSHFLHV RI EHOOELUGV VSHQW b RI WKH GD\ DW GLVSOD\ SHUFKHV OHDYLQJ RQO\ IRU EULHI IRUDJLQJ ERXWV ': 6QRZ E D %. 6QRZ f 0RVW RI WKH VHHGV GLVSHUVHG E\ PDOHV RI WKHVH VSHFLHV DUH SUREDEO\ GHSRVLWHG LQ WKH YLFLQLW\ RI WKH OHNV RU GLVSOD\ SHUFKHV ,I SHUFK RU OHN DUHDV DUH DOVR VXLWDEOH HVWDEOLVKPHQW VLWHV WKHQ WKHVH VSHFLHV PD\ SURYLGH GLUHFWHG GLVSHUVDO )RU H[DPSOH WKUHH ZDWWOHG EHOOELUGV 3URFQLDV WULFDUXQFXODWDf LQ &RVWD 5LFD W\SLFDOO\ GLVSOD\ RQ WDOO H[SRVHG SHUFKHV VXFK DV GHDG EUDQFKHV RU GHDG WUHHV 6QRZ f ,Q D VWXG\ RQ VHHG GLVSHUVDO RI WKH FDQRS\ WUHH 2FRWHD HQGUHVLDQD /DXUDFHDHf b RI WKH VHHGV GLVSHUVHG E\ EHOOELUGV ODQGHG LQ JDSV XQGHU GLVSOD\ SHUFKHVf FRPSDUHG WR RQO\ b RI WKH VHHGV GLVSHUVHG E\ IRXU RWKHU VSHFLHV &KDSWHU f 2YHUDOO LQFOXGLQJ VHHGV ZKRVH GLVSHUVDO DJHQW ZDV XQNQRZQ rb RI WKH VHHGV ODQGHG LQ JDSV ZKLFK ZDV PXFK JUHDWHU WKDQ WKH b H[SHFWHG EDVHG RQ WKH DUHD RI WKH IRUHVW LQ JDSV \U 6HHGOLQJV LQ JDSV KDG DOPRVW WZLFH WKH FKDQFH RI VXUYLYLQJ RQH \HDU FRPSDUHG WR VHHGOLQJV LQ FORVHG FDQRS\ IRUHVW 7KXV EHOOELUGV SURYLGHG GLUHFWHG GLVSHUVDO GLVSURSRUWLRQDWH GLVSHUVDO WR HVSHFLDOO\ VXLWDEOH VLWHV 2WKHU VSHFLHV RI EHOOELUGV DQG FRWLQJDV KDYH VLPLODU EHKDYLRU LQ SHUFKLQJ RQ H[SRVHG OLPEV RU LQ WUHHV ZLWK VSDUVH IROLDJH 6QRZ f DQG FRQVLGHULQJ WKH ZLGH YDULHW\ RI IUXLWV WKH\ HDW EHOOELUGV DUH OLNHO\ WR SURYLGH GLUHFWHG GLVSHUVDO IRU RWKHU VSHFLHV DV ZHOO 2WKHU OHNNLQJ VSHFLHV VXFK DV PDQDNLQV DQG FRFNVRI WKHURFN GLVSOD\ LQ WKH IRUHVW XQGHUVWRU\ $OWKRXJK WKLV SDWWHUQ PD\ OHDG WR FRORQL]DWLRQ YLD DFFXPXODWLRQ RI GRUPDQW VHHGV LQ WKH VRLO .ULMJHU HW DO f DQG VXSSUHVVHG VHHGOLQJV (QGOHU DQG 7KU\ f KDYH VKRZQ WKDW *XLDQDQ FRFNRIWKHURFN 5XSLFROD UXSLFRODf DQG WZR VSHFLHV RI

PAGE 137

PDQDNLQV &RUDSLSR JXWWXUDOLV /HSLGRWKUL[ VHUHQDf VHOHFW OHN DUHDV LQ VSHFLILF OLJKW HQYLURQPHQWV LQ WKH IRUHVW XQGHUVWRU\ VR DV WR PD[LPL]H FRQVSLFXRXVQHVV GXULQJ GLVSOD\V 6RPH VSHFLHV RI PDQDNLQV HYHQ PDLQWDLQ WKHLU OHNV E\ UHPRYLQJ OHDYHV WKDW REVWUXFW WKH YLHZ DW OHN KHLJKW 6QRZ f 7KH LPSRUWDQFH RI GLIIHUHQFHV DV VPDOO DV b LQ OLJKW DYDLODELOLW\ KDV EHHQ GHPRQVWUDWHG IRU VRPH VKDGHWROHUDQW VHHGOLQJV +RZH HW DO +RZH Ef DQG LW LV SRVVLEOH WKDW VRPH RI WKH VSHFLHV GLVSHUVHG E\ OHNNLQJ VSHFLHV EHQHILW IURP KLJKHU OLJKW OHYHOV DURXQG OHNV ,Q FRFNRIWKHURFN OHNV LQ )UHQFK *XLDQD b RI WKH SODQWV SUHVHQW DV VHHGOLQJV DQG VDSOLQJV ZHUH OLNHO\ GLVSHUVHG E\ 5XSLFROD ZKLOH RQO\ b RI WKH VSHFLHV LQ D IRUHVW XQGHUVWRU\ VLWH ZHUH VKDUHG ZLWK WKH OHN 7KU\ DQG /DUSLQ f 7KXV WKH 5XSLFROD OHNV FDQ KDYH D JUHDW LPSDFW RQ WKH YHJHWDWLRQ VWUXFWXUH DQG PD\ FRQWULEXWH WR FOXPSHG DQG SDWFK\ GLVWULEXWLRQV RI IUXLWLQJ SODQWV ,Q FRQWUDVW PDQDNLQV WKDW GLVSHUVH VHHGV RI VKDGHLQWROHUDQW 0HODVWRPDWDFHDH EXW GLVSOD\ LQ WKH XQGHUVWRU\ FRQWULEXWH KHDYLO\ WR WKH VRLO VHHG EDQN EXW DSSDUHQWO\ QRW WR WKH YHJHWDWLRQ DURXQG WKH OHN .ULMJHU HW DO f $ VLPLODU VLWXDWLRQ KDV EHHQ IRXQG IRU VOHHSLQJ WUHHV RI KRZOHU PRQNH\V $ORXDWWD VHQLFXOXVf LQ )UHQFK *XLDQD -XOOLRW f 2I VL[ SODQW VSHFLHV VWXGLHG DOO ZHUH PRUH FRPPRQ DV VHHGOLQJV LQ WKH YLFLQLW\ RI VOHHSLQJ URRVWV WKDQ RQ FRQWURO VLWHV 7KLV GLIIHUHQFH FRXOG EH D UHVXOW RI KLJKHU VHHG LQSXW DV VXJJHVWHG E\ -XOOLRW f RU E\ GLIIHUHQWLDO VXUYLYDO RI VHHGOLQJV XQGHU URRVWV $OWKRXJK WKH SUHVHQFH RI VHHGOLQJV GHPRQVWUDWHV VXFFHVVIXO HVWDEOLVKPHQW WKH HIIHFWV RI VHHG SUHGDWRUV DQG VHFRQGDU\ GLVSHUVDO ZHUH QRW TXDQWLILHG +DELWXDO VOHHSLQJ VLWHV KDYH EHHQ QRWHG IRU VHYHUDO RWKHU IUXJLYRURXV SULPDWH VSHFLHV &KDSPDQ E +H\PDQQ f DQG IRUNLQNDMRXV 3RWRV IODYXV -XOLHQ/DIHUULHUH f /RZODQG JRULOODV *RULOOD JRULOODf LQ *DERQ GHIHFDWH DURXQG QHVW VLWHV ZKLFK DUH W\SLFDOO\ ORFDWHG LQ RSHQ DUHDV RI WKH IRUHVW 7XULQ HW DO f *RULOODV ZHUH WKH PRVW LPSRUWDQW GLVSHUVHU RI WKH PRVW FRPPRQ WUHH LQ WKH VWXG\ DUHD &ROD OL]DH 6WHUFXOLDFHDHf &ROD VHHGOLQJ VXUYLYDO VL[ PRQWKV DIWHU GLVSHUVDO DW QHVW VLWHV LQ RSHQ DUHDV ZDV b EXW

PAGE 138

RQO\ b LQ IRUHVW $OWKRXJK WKH VHHGOLQJV LQ IHFDO FOXPSV PD\ IDFH LQWHQVH VLEOLQJ FRPSHWLWLRQ WKH RSHQ FDQRS\ DW QHVW VLWHV SURYLGHV D IDYRUDEOH JURZWK HQYLURQPHQW 7XWLQ HW DO f *RULOODV HDW IUXLWV RI DW OHDVW SODQW VSHFLHV DQG PD\ GLVSHUVH PDQ\ RWKHU VSHFLHV LQ D VLPLODU SDWWHUQ 7XWLQ HW DO f 6LPLODUO\ EDERRQV 3DSLR DQXELVf LQ *KDQD FRQJUHJDWH GDLO\ RQ URFN\ RXWFURSV ZKHUH WKH\ RIWHQ GHIHFDWH VHHGV /LHEHUPDQ HW DO f IRXQG VHHGV IURP VSHFLHV LQ GXQJ FROOHFWHG IURP WKH RXWFURSV 7KH\ QRWHG WKDW DOWKRXJK WKH URFNV WKHPVHOYHV ZHUH QRW JRRG HVWDEOLVKPHQW VLWHV PDQ\ VHHGV ZHUH OLNHO\ ZDVKHG E\ UDLQ WR WKH VXUURXQGLQJ VRLO DQG WKH YHJHWDWLRQ DURXQG WKH RXWFURSV ZDV GRPLQDWHG E\ WKH VDPH SODQW VSHFLHV EDERRQV FRPPRQO\ GLVSHUVHG 5KLQRFHURV 5KLQRFHURV XQLFRUQLVf LQ 1HSDO DQG WDSLUV 7DSLUXV WHUUHVWULVf LQ %UD]LO KDYH KDELWXDO ODWULQH VLWHV 'LQHUVWHLQ DQG :HPPHU 'LQHUVWHLQ )UDJRVR f 5KLQRFHURV ODWULQHV LQ IORRGSODLQ JUDVVODQGV SURYLGH FUXFLDO UHFUXLWPHQW VLWHV IRU WKH VKDGH LQWROHUDQW WUHH 7UHZLD WKDW GRPLQDWHV WKH ULYHULQH IRUHVW 6HHGV LQ GXQJ SLOHV DUH QRW VHFRQGDULO\ GLVSHUVHG E\ URGHQWV EXW PD\ EH VFDWWHUHG E\ UHSHDWHG XVH RI ODWULQHV DQG RFFDVLRQDO IORRGV 'LQHUVWHLQ f 7KH VHHGOLQJV PD\ IDFH LQWHUVSHFLILF FRPSHWLWLRQ EXW DUH ORFDWHG LQ QXWULHQWULFK VLWHV 6HHGV LQ WDSLU ODWULQHV DUH VFDWWHUKRDUGHG E\ DJRXWLV VHH EHORZ )UDJRVR f /DWULQH XVH KDV DOVR EHHQ GRFXPHQWHG IRU (XURSHDQ EDGJHUV 0HOHV PHOHVf LQ ,WDO\ 3LJQR]]L f %DGJHUV KDYH ODWULQHV DURXQG WKH SHULSKHU\ RI WKHLU WHUULWRULHV DQG GLVSHUVHG WKH VHHGV RI DW OHDVW SODQW VSHFLHV DOWKRXJK VHHGOLQJ VXUYLYDO ZDV QRW GRFXPHQWHG 2WKHU H[DPSOHV RI VHHG GLVSHUVDO FRQFHQWUDWHG DW ODWULQHV RU EXUURZ VLWHV RI KLJKO\ IUXJLYRURXV PDPPDOV GHVHUYH IXUWKHU VWXG\ HJ SDOP FLYHWV 3DUDGR[XUXV KHUPDSKURGLWXV -RVKL HW DO f ,Q DGGLWLRQ VSHFLHV VXFK DV UDEELWV 2U\FWRODJXV FXQLFXOXVf DUH NQRZQ WR KDYH ODWULQHV 6QHGGRQ f EXW WKHLU UROH DV VHHG GLVSHUVHUV LV QRW ZHOO NQRZQ EXW VHH 1RJDOHV HW D/ 6FKXSS HW DO D Ef ,Q VRPH FDVHV GLVSHUVDO FRQFHQWUDWHG LQ FHUWDLQ VLWHV ZLOO QR GRXEW OHDG WR H[WUHPHO\ KLJK VHHG RU VHHGOLQJ

PAGE 139

PRUWDOLW\ GXH WR LQDSSURSULDWH JURZLQJ FRQGLWLRQV LQWHQVH VHHGOLQJ FRPSHWLWLRQ RU DWWUDFWLRQ RI SUHGDWRUV RU SDWKRJHQV 6QRZ +RZH f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f 6RLO XQGHU QXUVH SODQWV HVSHFLDOO\ QLWURJHQIL[LQJ VSHFLHV LV VRPHWLPHV KLJKHU LQ QXWULHQWV WKDQ WKH VXUURXQGLQJ VRLO %DUQHV DQG $UFKHU f EXW QXWULHQW DYDLODELOLW\ LV QRW QHFHVVDULO\ WKH PDLQ OLPLWLQJ IDFWRU )UDQFR3L]DD HW DO f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f ,Q VRPH FDVHV VHHG SUHGDWLRQ XQGHU VKUXEV LV ORZHU WKDQ LQ WKH RSHQ &DOODZD\ f EXW RWKHU VWXGLHV KDYH VXJJHVWHG WKH RSSRVLWH UHVXOW 2ZHQV HW DO +XOPH f 'LVSHUVDO E\ ELUGV WR QXUVH SODQWV LQ DULG DQG VHPLDULG HFRV\VWHPV SUHVHQWV D VWURQJ FDVH IRU GLUHFWHG GLVSHUVDO EXW YLUWXDOO\ DOO VWXGLHV KDYH IRFXVHG RQ VHHGOLQJ JURZWK DQG LQWHUVSHFLILF SODQW DVVRFLDWLRQV ZLWKRXW GRFXPHQWLQJ SDWWHUQV RI GLVSHUVDO 1XUVH SODQWV PD\ EH SDUWLFXODUO\ LPSRUWDQW IRU PDLQWDLQLQJ SRSXODWLRQV RI UDUH SODQW VSHFLHV HJ &DUGHO HW DO f

PAGE 140

(IRUJJ /RJV 6LPLODU WR QXUVH SODQWV GHFRPSRVLQJ ORJV PD\ VHUYH DV LPSRUWDQW UHFUXLWPHQW VLWHV IRU VRPH VSHFLHV :KHWKHU VHHGV DUULYH DW ORJV GLVSURSRUWLRQDWHO\ KDV QRW EHHQ GRFXPHQWHG EXW WKH LQIOXHQFH RI QXUVH ORJ DEXQGDQFH RQ IRUHVW FRPSRVLWLRQ KDV EHHQ VXJJHVWHG 6FRZFURIW +RIJDDUG f ,Q DGGLWLRQ WKH SURSHQVLW\ RI FHUWDLQ ELUG VSHFLHV WR SHUFK RQ IDOOHQ ORJV KDV EHHQ QRWHG HJ 7XUGXV DOELFROLV &KDUOHV'RPLQLTXH f ,Q D WHPSHUDWH ZRRGODQG b RI WKH VSHFLHV RQ ORJV ZHUH SUREDEO\ DQWGLVSHUVHG DQG b ZHUH IOHVK\IUXLWHG VSHFLHV W\SLFDOO\ GLVSHUVHG E\ ELUGV DQG PDPPDOV 7KRPSVRQ f $V DQW QHVWV DUH RIWHQ ORFDWHG LQ RU QHDU ORJV DQG ELUGV DQG PDPPDOV RIWHQ SHUFK RQ ORJV HVSHFLDOO\ LQ WUHHIDOO JDSV 7KRPSVRQ f VXJJHVWHG WKDW ORJV PD\ GLIIHUHQWLDOO\ DFFXPXODWH VHHGV 6LPLODUO\ VHYHUDO 0HODVWRPDWDFHDH VSHFLHV KDYH EHHQ QRWHG LQ DVVRFLDWLRQ ZLWK ORJV LQ WZR WURSLFDO IRUHVWV /DZWRQ DQG 3XW] /DFN f 7KHVH PHODVWRPH VSHFLHV SURGXFH IOHVK\ IUXLWV W\SLFDOO\ GLVSHUVHG E\ ELUGV DQG VRPHWLPHV UHPRYHG IURP ELUG GURSSLQJV DQG VHFRQGDULO\ GLVSHUVHG E\ DQWV /HYH\ DQG %\PH f 6RPH RWKHU VSHFLHV RQ IDOOHQ ORJV KRZHYHU FOHDUO\ ZHUH HVWDEOLVKHG RQ WKH WUHHV EHIRUH WKH\ IHOO /DZWRQ DQG 3XW] f $QRWKHU SRVVLELOLW\ IRU GLUHFWHG GLVSHUVDO WR QXUVH ORJV LV WKH GLVSHUVDO RI IXQJDO VSRUHV E\ URGHQWV HJ &RUN DQG .HQDJ\ -RKQVRQ f 3HUFKHV LQ 6XFFHVVLRQDO /DQGVFDSHV 1XPHURXV VWXGLHV KDYH GRFXPHQWHG KLJK UDWHV RI VHHG GLVSHUVDO E\ DQLPDOV WR SHUFKHV LQ SDVWXUHV 2OGILHOGV RU RWKHU KXPDQGLVWXUEHG ODQGVFDSHV 'HEXVVFKH HW DO *XHYDUD DQG /DERUGH 0F&ODQDKDQ DQG :ROIH 5RELQVRQ DQG +DQGHO 'D 6LOYD HW DO 1HnHPDQ DQG ,]KDNL f 6XFK VLWHV DUH UHIHUUHG WR DV UHFUXLWPHQW IRFL 0F'RQQHOO 0\VWHU DQG 3LFNHWW f RU VXFFHVVLRQ IDFLOLWDWRUV 9LHLUD HW DO f ,Q WKHVH VLWXDWLRQV WKH GLVSHUVDO SDWWHUQ LV FOHDUO\ QRQUDQGRP DV VHHG UDLQ EHQHDWK SHUFKHV LV RIWHQ RU RUGHUV RI PDJQLWXGH JUHDWHU WKDQ LQ RSHQ DUHDV :LOOVRQ

PAGE 141

DQG &URPH 5RELQVRQ DQG +DQGHO 1HSVWDG HW DO 'XQFDQ f $OWKRXJK WKH SDWWHUQ RI VHHG UDLQ LV VLPLODU WR WKDW IRU QXUVH SODQWV LQ DULG HFRV\VWHPV WKH VXLWDELOLW\ RI VXFK VLWHV IRU JURZWK DQG VXUYLYDO LV OHVV FOHDU ,Q DFWLYH SDVWXUHV ZKHUH JUD]LQJ DQLPDOV ZRXOG HDW RU WUDPSOH VHHGOLQJV LQ WKH RSHQ GLVSHUVDO EHQHDWK VKUXEV IHQFHURZV RU E\ URFNV PD\ UHSUHVHQW WKH EHVW UHFUXLWPHQW VLWHV ZLWKLQ WKH SDVWXUH /LYLQJVWRQ +HUUHUD Df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f RU SODQWHG VKUXEV RU WUHHV 5RELQVRQ DQG +DQGHO 'HEXVVFKH DQG ,VHQPDQQ f ,I WUXH WKLV DOVR LOOXVWUDWHV WKH LGHD WKDW SODQWV PD\ QRW EH DGDSWHG IRU GLUHFWHG GLVSHUVDO EXW FDQ WDNH DGYDQWDJH RI LW LQ VRPH VLWXDWLRQV ,Q VXFFHVVLRQDO ODQGVFDSHV WKH VSHFLHV PRVW OLNHO\ WR EHQHILW IURP GLUHFWHG GLVSHUVDO DUH WKH VPDOOHUVHHGHG JHQHUDOLVW SODQWV WKDW DWWUDFW PDQ\ GLVSHUVHUV UDWKHU WKDQ WKH ODUJHUVHHGHG IUXLWV ZLWK D PRUH UHVWULFWHG GLVSHUVHU DVVHPEODJH 'D 6LOYD HW DO f ,W IROORZV WKHQ WKDW WKH DGYDQWDJHV RI GLUHFWHG GLVSHUVDO LQ VXFFHVVLRQDO ODQGVFDSHV ZRXOG GLPLQLVK IRU WKH JHQHUDOLVW SODQWV DV VXFFHVVLRQ DSSURDFKHV WKH FOLPD[ VWDJH DQG WKHQ FRORQL]DWLRQ RI XQSUHGLFWDEOH VLWHV LQFUHDVHV LQ LPSRUWDQFH VHH DOVR 'HEXVVFKH DQG ,VHQPDQQ f 6HFRQGDU\ 'LVSHUVDO EY $JRXWLV $QWV DQG 'XQJ %HHWOHV 5HFHQW VWXGLHV LQ WURSLFDO IRUHVWV LQGLFDWH WKDW GLVSHUVDO E\ YHUWHEUDWHV LV RIWHQ RQO\ WKH ILUVW RI WZR VWDJHV RI GLVSHUVDO 6HHGV LQ IUXJLYRUH GHIHFDWLRQV DUH UHPRYHG E\ DQWV 5REHUWV DQG +HLWKDXV .DVSDUL /HYH\ DQG %\UQH f DQG URGHQWV -DQ]HQ 0RHJHQEXUJ )UDJRVR f ,Q ERWK FDVHV WKH VHFRQGDU\ GLVSHUVHUV DFW

PAGE 142

PDLQO\ DV D VHHG SUHGDWRUV EXW WKH EHQHILFLDO WUHDWPHQW RI WKH VHHGV WKH\ GR GLVSHUVH PD\ OHDG WR D ODUJH LQIOXHQFH RQ SODQW UHFUXLWPHQW $QWV DUH XELTXLWRXV LQ 1HRWURSLFDO IRUHVW XQGHUVWRU\ HQYLURQPHQWV DQG UDSLGO\ UHPRYH VPDOO VHHGV IURP ELUG GURSSLQJV 7KH DQWV WDNH VHHGV EDFN WR WKHLU QHVWV DQG PRVW VHHGV DUH HYHQWXDOO\ HDWHQ ,Q RQH VWXG\ DERXW b RI WKH VHHGV ZHUH GLVFDUGHG RQ UHIXVH SLOHV EXW SHUKDSV PRUH LPSRUWDQWO\ QHVW WZLJV DQG WKH VHHGV ZLWKLQ WKHP ZHUH DEDQGRQHG RQ DYHUDJH HYHU\ WKUHH PRQWKV /HYH\ DQG %\UQH f 6HHGOLQJV KDG KLJKHU VXUYLYDO RQ H[SHULPHQWDO UHIXVH SLOHV XQGHU OLJKW OHYHOV W\SLFDO RI VPDOO JDSV 7KXV DQWV UHPRYH VHHGV IURP D KLJKGHQVLW\ SDWWHUQ LQ D IUXJLYRUH GHIHFDWLRQ ZKHUH LQWHUVSHFLILF FRPSHWLWLRQ PD\ EH LQWHQVH +RZH /RLVHOOH f DQG SODFH D IHZ VHHGV LQ ORZGHQVLW\ KLJKQXWULHQW VLWHV IDYRUDEOH IRU VHHGOLQJ JURZWK 6RPH SODQW VSHFLHV KDYH EHHQ QRWHG LQ DVVRFLDWLRQ ZLWK URWWLQJ ORJV 7KRPSVRQ /DZWRQ DQG 3XW] /DFN f DQG SHUKDSV DQWV DUH WKH PHFKDQLVP RI VHHG DUULYDO DW VXFK QXUVH ORJV .DXIIPDQ HW DO f VXJJHVWHG WKDW VRPH )LFXV VHHGV KDYH DQ HODLVRPH DQG DUH DGDSWHG IRU VHFRQGDU\ GLVSHUVDO E\ DQWV EXW %\UQH DQG /HYH\ f IRXQG QR REYLRXV HODLVRPHV RQ WKH 0HODVWRPDWDFHDH VHHGV FRPPRQO\ KDUYHVWHG E\ DQWV IURP ELUG GURSSLQJV 7DSLUV LQ $PD]RQLDQ %UD]LO GLVSHUVH SDOP VHHGV ^0D[LPLOLDQR PDULSDf WR ODWULQH VLWHV ZKLFK DUH XVHG UHSHDWHGO\ )UDJRVR f ,Q WHUUH ILUPH IRUHVW b RI WDSLU ODWULQHV ZHUH ORFDWHG E\ EXWWUHVVHV RI WKH HPHUJHQW WUHH &RXUDWDUL PXOWLIORUD /HF\WKLGDFHDHf 0RVW VHHGV QRW GLVSHUVHG E\ WDSLUV DUH NLOOHG E\ LQVHFWV 7KH PRVW UHPDUNDEOH DVSHFW RI WKLV V\VWHP LV WKDW DJRXWLV UHPRYHG VHHGV IURP WDSLU ODWULQHV DQG EXULHG WKHP LQ VKDOORZ VXUIDFH FDFKHV 7KH GHQVLW\ RI VHHGOLQJV DQG VDSOLQJV \U ROG ZDV VLJQLILFDQWO\ KLJKHU DURXQG ODWULQH VLWHV WKDQ DURXQG HLWKHU SDUHQW WUHHV RU QRQSDOP FRQWURO WUHHV )UDJRVR f VXJJHVWHG WKDW WKH FRPELQDWLRQ RI WDSLU DQG DJRXWL GLVSHUVDO ZDV UHVSRQVLEOH IRU WKH FOXPSHG GLVWULEXWLRQ RI WKH SDOPV REVHUYHG RYHU ODUJH DUHDV RI WKH IRUHVW &OXPSHG GLVWULEXWLRQV RI RWKHU VSHFLHV RI WUHHV KDYH EHHQ DWWULEXWHG WR DJRXWLV LQ RWKHU 1HRWURSLFDO IRUHVWV DV ZHOO 3HUHV DQG % DLGHU f 6FDWWHUKRDUGLQJ E\ DJRXWLV FDQ

PAGE 143

EH HLWKHU SULPDU\ RU VHFRQGDU\ GLVSHUVDO GHSHQGLQJ RQ WKH SODQW VSHFLHV DOWKRXJK WKH EHQHILWV RI VFDWWHUKRDUGLQJ WR VHHGV DUH VLPLODU LQ ERWK FDVHV LQ DGGLWLRQ WR HVFDSH IURP VRPH LQVHFW DQG PDPPDOLDQ VHHG SUHGDWRUV EXULDO PD\ SUHYHQW GHVLFFDWLRQ DQG IDFLOLWDWH JHUPLQDWLRQ DQG HVWDEOLVKPHQW +DOOZDFKV 6P\WKH )RUJHW 9DQGHU :DOO f $JRXWLV KDYH EHHQ IUHTXHQWO\ QRWHG FDFKLQJ VHHGV E\ ORJV RU RWKHU REMHFWV 6P\WKH +DOOZDFKV )RUJHW &KDSWHU f EXW WKH H[WHQW WR ZKLFK WKHVH VLWHV UHSUHVHQW VXLWDEOH HVWDEOLVKPHQW VLWHV KDV QRW EHHQ VWXGLHG ,Q DGGLWLRQ .LOWLH f QRWHG WKDW SHFFDULHV RIWHQ IRUDJH QHDU REMHFWV SHUKDSV WR ORFDWH DJRXWL FDFKHV $QRWKHU H[DPSOH RI VHFRQGDU\ GLVSHUVDO LV UHPRYDO RI PDPPDOLDQ GXQJ DQG WKH VHHGV LW PD\ FRQWDLQ E\ GXQJ EHHWOHV (VWUDGD DQG &RDWHV(VWUDGD (VWUDGD HW DO 6KHSKHUG DQG &KDSPDQ LQ SUHVVf ,Q WKLV FDVH WKH GXQJ EHHWOHV KDYH QR LQWHUHVW LQ WKH VHHGV DQG PD\ HYHQ GLVFDUG ODUJHU VHHGV $QGUHVHQ LQ SUHVVf 7KH EHHWOHV EXU\ VPDOO EDOOV RI GXQJ VHYHUDO FP GHHS 7KH IHPDOHV RYLSRVLW RQ WKH GXQJ WKH ODUYDH FRQVXPH PRVW RI WKH GXQJ EDOO DQG OHDYH WKH VHHGV EHKLQG (VWUDGD DQG &RDWHV(VWUDGD f IRXQG WKDW b RI WKH SODQW VSHFLHV GLVSHUVHG E\ KRZOHU PRQNH\V EHQHILWHG IURP VHFRQGDU\ GLVSHUVDO E\ GXQJ EHHWOHV 7KH\ VXJJHVWHG WKDW EXULDO E\ EHHWOHV DOORZ HVFDSH IURP URGHQW VHHG SUHGDWRUV EXW WKH OLNHO\ EHQHILWV RI EXULDO LQ QXWULHQWULFK GXQJ KDYH QRW EHHQ H[DPLQHG 1XPHURXV VSHFLHV RI GXQJ EHHWOHV PD\ EH LQYROYHG LQ VHFRQGDU\ GLVSHUVDO DQG WKH\ YDU\ LQ WKH DPRXQW RI GXQJ WDNHQ KRZ IDU LW LV WDNHQ DQG KRZ GHHS LW LV EXULHG (VWUDGD HW DO $QGUHVHQ LQ SUHVVf &RQFOXVLRQV *RLQJ WKH 'LVWDQFH DQG %HYRQG 7KH H[DPSOHV DERYH LOOXVWUDWH WKH SRWHQWLDO IRU GLUHFWHG GLVSHUVDO LQ D ZLGH YDULHW\ RI V\VWHPV DQG VXJJHVW LWV LPSRUWDQFH LQ UHVWRUDWLRQ HFRORJ\ FRQVHUYDWLRQ DQG LQ XQGHUVWDQGLQJ SODQWDQLPDO LQWHUDFWLRQV $ WHVW RI WKH GLUHFWHG GLVSHUVDO K\SRWKHVLV LV ZLWKLQ UHDFK EXW WKH ODFN RI GHWDLOHG GDWD RQ SDWWHUQV RI GLVSHUVDO JHQHUDWHG E\ SDUWLFXODU

PAGE 144

YHFWRUV LQWHJUDWHG ZLWK SRVWGLVSHUVDO VHHG DQG VHHGOLQJ IDWH OLPLWV RXU XQGHUVWDQGLQJ RI WKH UROH RI GLUHFWHG GLVSHUVDO LQ SODQW UHFUXLWPHQW ,Q DGGLWLRQ VHHG UDLQ LQWR GLIIHUHQW SDWFKHV DW WKH FRPPXQLW\ OHYHO LV VWLOO SRRUO\ NQRZQ 6XUSULVLQJO\ WKH PRGHOV RI 6FKXSS HW DO f KDYH QRW EHHQ DGHTXDWHO\ WHVWHG 7HVWLQJ WKHLU SUHGLFWLRQV RI VHHG UDLQ SDWWHUQV LQWR JDSV JDS HGJHV DQG IRUHVW XQGHUVWRU\ VKRXOG EH IDLUO\ VWUDLJKWIRUZDUG DQG VKRXOG LGHQWLI\ FHUWDLQ SODQW VSHFLHV DQG WKHLU GLVSHUVHUV VXLWDEOH IRU PRUH GHWDLOHG VWXG\ WKDW ZLOO \LHOG LQVLJKW LQWR SODQW FRPPXQLW\ G\QDPLFV 7HVWV RI WKH HVFDSH K\SRWKHVLV KDYH VKRZQ WKDW VRPH PRUWDOLW\ DJHQWV UHVSRQG WR GLVWDQFH DQGRU GHQVLW\ DQG RWKHU GR QRW +RZH HW DO 0HUJ f 1RZ ZH QHHG VWXGLHV WKDW JR EH\RQG WKH GLVWDQFH IDFWRU DQG H[DPLQH WKH FKDUDFWHULVWLFV RI VLWHV LQ ZKLFK VHHGV DFWXDOO\ ODQG DQG WKH VXEVHTXHQW IDWH RI WKRVH VHHGV ([SHULPHQWDO VWXGLHV DUH XVHIXO LQ GHWHUPLQLQJ FDXVDWLRQ EXW WKH\ QHHG WR EH EDVHG RQ VRPH UHDOLVWLF LQLWLDO SDWWHUQ (YHQ PRUH LPSRUWDQW LV WKH QHHG IRU VWXGLHV WKDW H[DPLQH WKH VHTXHQWLDO VWDJHV RI UHFUXLWPHQW IURP GLVSHUVDO WKURXJK HVWDEOLVKPHQW HJ +HUUHUD HW DO f 6WXGLHV WKDW LJQRUH WKH SRVVLELOLW\ RI VHFRQGDU\ GLVSHUVDO DUH OLNHO\ WR OHDG WR PLVOHDGLQJ FRQFOXVLRQV DERXW WKH LPSRUWDQFH RI GLVSHUVDO SDWWHUQV /HYH\ DQG %\UQH f 7KXV DQRWKHU XVHIXO VWXG\ ZRXOG EH WR H[DPLQH WKH LQIOXHQFH RI DUHDV RI KLJK VHHG UDLQ VXFK DV KDELWXDO SHUFKHV OHNV SULPDWH VOHHSLQJ VLWHV ODUJHPDPPDO ODWULQHV JDS HGJHV RU QXUVH SODQWV RQ SODQW GLVWULEXWLRQV DQG FRPPXQLW\ FRPSRVLWLRQ $UH GHQVH FRQFHQWUDWLRQV RI VHHGV DZD\ IURP WKH SDUHQW SODQWV IRFL RI UHFUXLWPHQW IRU VHHGOLQJV VHHG SUHGDWLRQ RU VHFRQGDU\ GLVSHUVDO" ,Q DGGLWLRQ WR SHUFKHV LQ VXFFHVVLRQDO ODQGVFDSHV UHFUXLWPHQW IRFL PD\ EH SUHVHQW ZLWKLQ IRUHVWHG DUHDV DV ZHOO (PHUJHQW WUHHV VWDQGLQJ GHDG WUHHV RU SHUKDSV WUHHV ZLWK DEXQGDQW IUXLW FURSV HJ )LFXVf PD\ DWWUDFW YHUWHEUDWH VHHGGLVSHUVHUV DQG KDYH KLJKHU VHHG UDLQ WKDQ EHQHDWK WUHHV QHDUE\ 6PLWK &RDWHV(VWUDGD DQG (VWUDGD 0DVDNL HW DO f ,Q D SUHOLPLQDU\ WHVW )OHPLQJ f IRXQG QR GLIIHUHQFH LQ WKH GLYHUVLW\ RU GHQVLW\ RI YHUWHEUDWHGLVSHUVHG SODQWV DURXQG )LFXV VSS WUHHV DQG ZLQGGLVSHUVHG

PAGE 145

&DO\FRSK\OOXP FDQGLGLVVLPXP WUHHV EXW WKH VFDOH P SORWVf VPDOO VDPSOH VL]H DQG WKH GLIIHUHQW VL]HV RI WKH WUHHV DUH FRQIRXQGLQJ IDFWRUV 9HQDEOH DQG %URZQ f VXJJHVWHG WKDW GLUHFWHG GLVSHUVDO LQYROYHV EHLQJ GHSRVLWHG LQ GXQJ ZKLFK DFWV DV IHUWLOL]HU EXW KLJK UDWHV RI SRVWGLVSHUVDO VHHG UHPRYDO IURP GHIHFDWLRQV FDVW GRXEW RQ WKLV LGHD -DQ]HQ &KDSPDQ D /HYH\ DQG %\UQH )UDJRVR f 1HYHUWKHOHVV 'LQHUVWHLQ DQG :HPPHU f DQG 'LQHUVWHLQ f VKRZHG WKDW WKH GLVWULEXWLRQ RI 7UHPD QXGLIORUD LQ ULSDULDQ IRUHVWV LQ 1HSDO LV DOPRVW HQWLUHO\ DWWULEXWDEOH WR GLVSHUVDO E\ UKLQRFHURV DQG VXJJHVWHG WKDW UKLQRFHURV GXQJ SLOHV SURYLGHG QXWULHQWV QHFHVVDU\ IRU WKH VHHGOLQJV 7KH LPSRUWDQFH RI GXQJ IRU VHHGV EXULHG E\ GXQJ EHHWOHV KDV QRW EHHQ H[DPLQHG DQG VKRXOG EH DPHQDEOH WR H[SHULPHQWDWLRQ 2QH RI WKH GLIILFXOWLHV LQ HYDOXDWLQJ GLUHFWHG GLVSHUVDO LV WKDW VXLWDEOH ORFDWLRQV IRU HVWDEOLVKPHQW LH VDIH VLWHVf DUH RIWHQ GLIILFXOW WR VSHFLI\ RU DUH SRRUO\ NQRZQ IRU PRVW VSHFLHV *UXEE )RZOHU f ,Q DGGLWLRQ VDIH VLWHV IRU VHHGV PD\ QRW EH WKH VDPH DV WKRVH IRU VHHGOLQJV RU DGXOWV 6PLWK /DPRQW HW DO 6FKXSS f OHDGLQJ VRPH UHVHDUFKHUV WR VWDWH WKDW VDIH VLWHV EHFRPH PRUH VSHFLILF RYHU WLPH 2VZDOG DQG 1HXHQVFKZDQGHU f )RU H[DPSOH VLWHV RFFXSLHG E\ VDSOLQJV PD\ D VSHFLILF VXEVHW RI WKRVH RFFXSLHG E\ VHHGOLQJV ZKLFK LQ WXUQ DUH D VXEVHW RI WKRVH RFFXSLHG E\ VHHGV 2VZDOG DQG 1HXHQVFKZDQGHU f )RU WURSLFDO IRUHVW VSHFLHV WKH K\SRWKHVL]HG SDUWLWLRQLQJ RI JDSV DPRQJ VHHGOLQJV LQ WHUPV RI JDS DJH DQG SRVLWLRQ ZLWKLQ WKH JDS VXJJHVWV WKDW UHFUXLWPHQW FRQGLWLRQV FRXOG EH GHILQHG +DUWVKRUQ 'HQVORZ %UDQGDQL HW DO 3RSPD HW DO %URZQ DQG :KLWPRUH f ,I VR WKH QH[W VWHS ZRXOG EH WR H[DPLQH VHHG DUULYDO WR VXFK VLWHV HJ 6FKXSS HW DO f $QRWKHU WKHPH LOOXVWUDWHG E\ WKH H[DPSOHV DERYH LV WKDW GLUHFWHG GLVSHUVDO PD\ RIWHQ EH D IRUWXLWRXV RXWFRPH RI GLVSHUVHU EHKDYLRU UDWKHU WKDQ DQ DGDSWLYH VWUDWHJ\ E\ SODQWV 7KXV LQ VLWXDWLRQV ZKHUH GLVSHUVHU EHKDYLRU FDQ EH SUHGLFWHG GLVSHUVDO FDQ EH PDQLSXODWHG WR LQFUHDVH WKH SUREDELOLW\ RI GLUHFWHG GLVSHUVDO 5HJHQHUDWLRQ RI GHJUDGHG ODQG LV RIWHQ OLPLWHG E\ ORZ VHHG LQSXW $LGH DQG &DYHOLHU 'XQFDQ f $FFRUGLQJO\

PAGE 146

PDQLSXODWLRQ RI GLUHFWHG GLVSHUVDO KDV SRWHQWLDO WR SOD\ D ODUJH UROH LQ UHVWRUDWLRQ HFRORJ\ 'H 3LHWUL 0F&ODQDKDQ DQG :ROIH 5RELQVRQ DQG +DQGHO f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f )RU KLJKO\ LQYDVLYH VSHFLHV ZLWK KLJK VHHGVHHGOLQJ VXUYLYDO GXH WR ODFN RI VHHG SUHGDWRUV SDWKRJHQV RU KHUELYRUHV WKH LQLWLDO SUHPLVH RI WKH FRORQL]DWLRQ K\SRWKHVLV LH UDUH DQG XQSUHGLFWDEOH VDIH VLWHVf PD\ QRW EH PHW ,Q DGGLWLRQ WKH UROHV RI LQWURGXFHG DQLPDOV DV VHHG GLVSHUVHUV KDYH EHHQ YLUWXDOO\ LJQRUHG 0H\HU DQG )ORUHQFH f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

PAGE 147

$33(1',; 2&27($ (1'5(6,$1$ /2&$7,21 '$7$ 7KLV DSSHQGL[ FRQWDLQV WKH ORFDWLRQV RI DOO WKH GLVSHUVHG W\SH 'f DQG QRQGLVSHUVHG 1f 2FRWHD HQGUHVLRQD VHHGV IRXQG LQ DQG 7UHH QXPEHUV FRUUHVSRQG ZLWK WKRVH LQ )LJXUH &RPSDVV GLUHFWLRQV DUH JLYHQ IURP WKH WUHH WRZDUG WKH VLWH DQG GLVWDQFHV DUH IURP WKH WUXQN RI WKH FORVHVW IUXLWLQJ FRQVSHFLILF 7KLV GDWD VHW UHSUHVHQWV DQ HVWLPDWHG b RI WKH WRWDO VHHG FURS DSSUR[LPDWHO\ f SURGXFHG LQ WKH VWXG\ VLWH LQ ERWK \HDUV FRPELQHG \HDU WUHH W\SH FRPSDVV GLVWDQFH ' ' ' ' ' 8 8 8 8 8 8 8 8 8 8 8 8 ' ' '

PAGE 148

' ' 8 8 X X X X X X X X X X X X X X X X X X X X X X X X X X X X ' ' ' ' ' ' ' ' ' 8 8

PAGE 149

:2-!`8-:8!:8!8!8!8!8-8f8f8!8!1!WV!1!W2W2W22W21!W2W21!.!W2W21!W2W2WW2W21!W21!W2W2W21!W2W21!1!W2W2W21!W2 8 8 X X X X X X X X X X X X X X X X X X ' ' ' 8 X X X X X X X X X X ' ' ' ' ' ' ' '

PAGE 150

' ' ' ' ' ' ' 8 X X X X X X X X X X X X X X X X X X ' ' ' ' ' ' ' ' ' 8 8 8

PAGE 151

8 8 8 X X X X X ' 8 X X ' ' ' ' ' ' ' ' 8 X X X X X X X X X X X X X X ' ' ' '

PAGE 152

X X X X X X X X X X X X X X X X X X X X X X X X ' ' ' ' ' ' ' ' 8 8 8 X X X X X ' ' '

PAGE 153

nn' 2V 2V 2V 2V 2V 2V 2V 2V 2V 2V 2V ? 2V 2V 2V 2V 2V 2V 2V2V2V2V2V2V2V2V2V2V2V2V2V2V2V2V2V222V22222V2V2V2V2V2V2V2V2V2V2V2V ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' 8 8 8 8 8 8 8 8 8 8 X X X '

PAGE 154

' 8 8 8 8 8 8 8 ' ' ' ' ' ' ' ' ' 8 8 8 X X X X X X X X X X X X X X X X ' ' ' '

PAGE 155

' ' ' ' ' ' ' ' 8 8 8 X X X X X X X X X ' ' ' ' ' ' ' ' ' ' ' ' '

PAGE 156

' ' ' ' ' ' ' ' ' ' ' ' ' 8 8 8 X X ' ' ' 8 8 8 8 8 8 8 8 X X X X X X X

PAGE 157

' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' 8 8 8 8 X X X X X X X X X X X X X X X X X X

PAGE 158

8 ' ' 8 8 8 8 ' ' ' ' ' ' 8 8 8 8 X X X X X X X X X X ' ' ' 8 8 8 8 8 8 X X X '

PAGE 159

' 8 8 8 X X ' 8 X X X X X X 8 X ' 8 X X ' ' 8 ' ' ' ' ' ' 8 ' ' ' '

PAGE 160

' ' ' 8 8 8 8

PAGE 161

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£QGH] 2 DQG ( 9HODUGH 7KH GLHW RI UHVSOHQGHQW TXHW]DO 3KDURPDFKUXV PRFLQQR PRFLQQR 7URJRQLGDHf LQ D 0H[LFDQ FORXG IRUHVW %LRWURSLFD

PAGE 162

%DOGD 5 3 DQG $ & .DPLO $ FRPSDUDWLYH VWXG\ RI FDFKH UHFRYHU\ E\ WKUHH FRUYLG VSHFLHV $QLPDO %HKDYLRU %DUQHV 3 : DQG 6 $UFKHU ,QIOXHQFH RI DQ RYHUVWRUH\ WUHH 3URVRSLV JODQGXORVDf RQ DVVRFLDWHG VKUXEV LQ D VDYDQQD SDUNODQG LPSOLFDWLRQV IRU SDWFK G\QDPLFV 2HFRORJLD %D]]D] ) $ +DELWDW VHOHFWLRQ LQ SODQWV $PHULFDQ 1DWXUDOLVW 66 %D]]D] ) $ DQG 6 7 $ 3LFNHWW 3K\VLRORJLFDO HFRORJ\ RI WURSLFDO VXFFHVVLRQ D FRPSDUDWLYH UHYLHZ $QQXDO 5HYLHZ RI (FRORJ\ DQG 6\VWHPDWLFV %HDO : 6HHG GLVSHUVDO *LQQ t &R %RVWRQ 0DVVDFKXVHWWV %HDWWLH $ 7KH HYROXWLRQDU\ HFRORJ\ RI DQWSODQW PXWXDOLVPV &DPEULGJH 8QLYHUVLW\ 3UHVV &DPEULGJH %HDWWLH $ DQG & &XOYHU 7KH JXLOG RI P\UPHFRFKRUHV LQ WKH KHUEDFHRXV IORUD RI :HVW 9LUJLQLD IRUHVWV (FRORJ\ %HDWWLH $ DQG & &XOYHU ,QKXPDWLRQ KRZ DQWV DQG RWKHU LQYHUWHEUDWHV KHOS VHHGV 1DWXUH %HDWWLH $ DQG & &XOYHU 7KH QHVW FKHPLVWU\ RI WZR VHHGGLVSHUVLQJ DQW VSHFLHV 2HFRORJLD %HFNHU 3 / : /HH ( 5RWKPDQ DQG : +DPLOWRQ 6HHG SUHGDWLRQ DQG WKH FRH[LVWHQFH RI WUHH VSHFLHV +XEEHOOnV PRGHOV UHYLVLWHG 2LNRV %HQNPDQ & : :LQG GLVSHUVDO FDSDFLW\ RI SLQH VHHGV DQG WKH HYROXWLRQ RI GLIIHUHQW VHHG GLVSHUVDO PRGHV LQ SLQHV 2LNRV %RGPHU 5 ( )UXLW SDWFK VL]H DQG IUXJLYRU\ LQ WKH ORZODQG WDSLU 7DSLUXV WHUUHVWULVf -RXUQDO RI =RRORJ\ %RGPHU 5 ( 6WUDWHJLHV RI VHHG GLVSHUVDO DQG VHHG SUHGDWLRQ LQ $PD]RQLDQ XQJXODWHV %LRWURSLFD %RQG : DQG : 6WRFN 7KH FRVWV RI OHDYLQJ KRPH DQWV GLVSHUVH P\UPHFRFKRURXV VHHGV WR ORZ QXWULHQW VLWHV 2HFRORJLD %RVVHPD -D\V DQG RDNV DQ HFRHWKRORJLFDO VWXG\ RI D V\PELRVLV %HKDYLRXU %RZHUV 0 $ DQG / 'RROH\ 3UHGDWLRQ KD]DUG DQG VHHG UHPRYDO E\ VPDOO PDPPDOV PLFURKDELWDW YHUVXV SDWFK VFDOH HIIHFWV 2HFRORJLD %R\G 5 6 $QWPHGLDWHG VHHG GLVSHUVDO RI WKH UDUH FKDSDUUDO VKUXE )UHPRQWRGHQGURQ GHFXPEHQV 6WHUFXOLDFHDHf 0DGURR %UDQGDQL $ 6 +DUWVKRUQ DQG + 2ULDQV ,QWHUQDO KHWHURJHQHLW\ RI JDSV DQG VSHFLHV ULFKQHVV LQ &RVWD 5LFDQ WURSLFDO ZHW IRUHVW -RXUQDO RI 7URSLFDO (FRORJ\

PAGE 163

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f %XUNH\ 7 9 7URSLFDO WUHH VSHFLHV GLYHUVLW\ D WHVW RI WKH -DQ]HQ&RQQHOO PRGHO 2HFRORJLD %XVWDPDQWH 5 2 DQG 0 &DQDOV / 'LVSHUVDO TXDOLW\ LQ SODQWV KRZ WR PHDVXUH HIILFLHQF\ DQG HIIHFWLYHQHVV RI D VHHG GLVSHUVHU 2LNRV %\UQH 0 0 DQG /HYH\ 5HPRYDO RI VHHGV IURP IUXJLYRUH GHIHFDWLRQV E\ DQWV LQ D &RVWD 5LFDQ UDLQ IRUHVW 9HJHWDWLR &DOODZD\ 5 0 (IIHFW RI VKUXEV RQ UHFUXLWPHQW RI 4XHUFXV GRXJODVLL DQG 4XHUFXV OREDWD LQ &DOLIRUQLD (FRORJ\ &DOODZD\ 5 0f ( + 'H/XFLD 0RRUH 5 1RZDN DQG : + 6FKOHVLQJHU &RPSHWLWLRQ DQG IDFLOLWDWLRQ FRQWUDVWLQJ HIIHFWV RI $UWHPLVLD WULGHQWDWD RQ GHVHUW YV PRQWDQH SLQHV (FRORJ\ &DQKDP & 'LIIHUHQW UHVSRQVHV WR JDSV DPRQJ VKDGHWROHUDQW WUHHV (FRORJ\ &DUGHO < 9 5LFR*UD\ *DUFD)UDQFR DQG / % 7KLHQ (FRORJLFDO VWDWXV RI %HDXFDPHD JUDFLOLV DQ HQGHPLF VSHFLHV RI WKH VHPLDULG 7HKXDF£Q 9DOOH\ 0[LFR &RQVHUYDWLRQ %LRORJ\ &DVWUR 2 & &KHPLFDO DQG ELRORJLFDO H[WUDFWLYHV RI /DXUDFHDH VSHFLHV LQ &RVWD 5LFDQ WURSLFDO IRUHVWV 5HFHQW $GYDQFHV LQ 3K\WRFKHPLVWU\ &DYHOLHU (QYLURQPHQWDO IDFWRUV DQG HFRSK\VLRORJLFDO SURFHVVHV DORQJ DOWLWXGLQDO JUDGLHQWV LQ ZHW WURSLFDO PRXQWDLQV 3DJHV LQ 6 6 0XONH\ 5 / &KD]GRQ DQG $ 3 6PLWK HGLWRUV 7URSLFDO IRUHVW SODQW HFRSK\VLRORJ\ &KDSPDQ DQG +DOO 1HZ
PAGE 164

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
PAGE 165

5HLGnV SDUDGR[ RI UDSLG SODQW PLJUDWLRQ GLVSHUVDO WKHRU\ DQG LQWHUSUHWDWLRQ RI SDOHRHFRORJLFDO UHFRUGV %LR6FLHQFH &OLIIRUG + 7 DQG % 0RQWHLWK $ WKUHH SKDVH VHHG GLVSHUVDO PHFKDQLVP LQ $XVWUDOLDQ TXLQLQH EXVK 3HOWRVWLJPD SXEHVFHQV 'RPLQf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f 0 /RZPDQ DQG 5 1REOH 6XEFDQRS\ JDSV LQ WHPSHUDWH DQG WURSLFDO IRUHVWV $XVWUDOLDQ -RXUQDO RI (FRORJ\ &RUN 6 DQG .HQDJ\ 5DWHV RI JXW SDVVDJH DQG UHWHQWLRQ RI K\SRJHRXV IXQJDO VSRUHV LQ WZR IRUHVWGZHOOLQJ URGHQWV -RXUQDO RI 0DPPDORJ\ &RUOHWW 5 7 DQG 3 : /XFDV $OWHUQDWLYH VHHGKDQGOLQJ VWUDWHJLHV LQ SULPDWHV VHHGVSLWWLQJ E\ ORQJWDLOHG PDFDTXHV 0DFDFD IDVFLFXODULVf 2HFRORJLD &UDZOH\ 0 6HHG SUHGDWRUV DQG SODQW SRSXODWLRQ G\QDPLFV 3DJHV LQ 0 )HQQHU HGLWRU 6HHGV WKH (FRORJ\ RI 5HJHQHUDWLRQ LQ 3ODQW &RPPXQLWLHV &$% ,QWHUQDWLRQDO :DOOLQJIRUG 8. &URPH ) + 7KH HFRORJ\ RI IUXLW SLJHRQV LQ WURSLFDO QRUWKHUQ 4XHHQVODQG $XVWUDOLDQ :LOGOLIH 5HVHDUFK &UX] $ %LUG DFWLYLW\ DQG VHHG GLVSHUVDO RI D PRQWDQH IRUHVW WUHH 'XQDOLD DUERUHVFHQVf LQ -DPDLFD %LRWURSLFD 6 &XOYHU & DQG $ %HDWWLH 0\UPHFRFKRU\ LQ 9LROD G\QDPLFV RI VHHGDQW LQWHUDFWLRQV LQ VRPH :HVW 9LUJLQLD VSHFLHV -RXUQDO RI (FRORJ\ 'D 6LOYD 0 & & 8KO DQG 0XUUD\ 3ODQW VXFFHVVLRQ ODQGVFDSH PDQDJHPHQW DQG WKH HFRORJ\ RI IUXJLYRURXV ELUGV LQ DEDQGRQHG $PD]RQLDQ SDVWXUHV &RQVHUYDWLRQ %LRORJ\ 'DUOH\+LOO 6 DQG : & -RKQVRQ $FRUQ GLVSHUVDO E\ WKH EOXH MD\ &\DQRFLWWD FULVWDWDf 2HFRORJLD 'DUZLQ & 7KH RULJLQ RI VSHFLHV E\ PHDQV RI QDWXUDO VHOHFWLRQ 'RXEOHGD\ 1HZ
PAGE 166

'DYLGDU 3 %LUGV DQG QHRWURSLFDO PLVWOHWRHV HIIHFWV RQ VHHGOLQJ UHFUXLWPHQW 2HFRORJLD 'DYLGVRQ : DQG 6 5 0RUWRQ 0\UPHFRFKRU\ LQ VRPH SODQWV ) FKHQRSRGLDFHDHf RI WKH $XVWUDOLDQ DULG ]RQH 2HFRORJLD 'H 3LHWUL ( $OLHQ VKUXEV LQ D QDWLRQDO SDUN FDQ WKH\ KHOS LQ WKH UHFRYHU\ RI QDWXUDO GHJUDGHG IRUHVW" %LRORJLFDO &RQVHUYDWLRQ 'HDQ : 5 DQG 6 0LOWRQ 'LVSHUVDO RI VHHGV E\ UDSWRUV $IULFDQ -RXUQDO RI (FRORJ\ 'HEXVVFKH 0f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f DQG WKH IORUD RI 5KLQRFHURV ODWULQHV 0DPPDOLD 'LQHUVWHLQ ( DQG & 0 :HPPHU )UXLWV 5KLQRFHURV HDW GLVSHUVDO RI 7UHZLD QXGLIORUD (XSKRUELDFHDHf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

PAGE 167

(QGOHU $ DQG 0 7KU\ ,QWHUDFWLQJ HIIHFWV RI OHN SODFHPHQW GLVSOD\ EHKDYLRU DPELHQW OLJKW DQG FRORU SDWWHUQV LQ WKUHH QHRWURSLFDO IRUHVWGZHOOLQJ ELUGV $PHULFDQ 1DWXUDOLVW (QJOXQG 5 )UXLW UHPRYDO LQ 9LEXUQXP RSXOXV FRSLRXV VHHG SUHGDWLRQ DQG VSRUDGLF PDVVLYH VHHG GLVSHUVDO LQ D WHPSHUDWH VKUXE 2LNRV (VWUDGD $ DQG 5 &RDWHV(VWUDGD +RZOHU PRQNH\V $ORXDWWD SDOOLDWDf GXQJ EHHWOHV 6FDUDEDHLGDHf DQG VHHG GLVSHUVDO HFRORJLFDO LQWHUDFWLRQV LQ WKH WURSLFDO UDLQ IRUHVW RI /RV 7X[WODV 0H[LFR -RXUQDO RI 7URSLFDO (FRORJ\ (VWUDGD $ +DOIIWHU 5 &RDWHV(VWUDGD DQG $ 0HULWW -U 'XQJ EHHWOHV DWWUDFWHG WR PDPPDOLDQ KHUELYRUH $ORXDWWD SDOOLDWDf DQG RPQLYRUH 1DVXD QDULFDf GXQJ LQ WKH WURSLFDO UDLQ IRUHVW RI /RV 7X[WODV 0H[LFR -RXUQDO RI 7URSLFDO (FRORJ\ )HQQHU 0 6HHG HFRORJ\ &KDSPDQ DQG +DOO /RQGRQ 8. )LVFKHU ( $ 7KH UROH RI SOXPHV LQ (ULRWKHFD SHQWDSK\OOD %RPEDFDFHDHf VHHG VXUYLYDO LQ VRXWKHDVWHUQ %UD]LO -RXUQDO RI 7URSLFDO (FRORJ\ )OHPLQJ 7 + 7KH VKRUWWDLOHG IUXLW EDW D VWXG\ LQ SODQWDQLPDO LQWHUDFWLRQV 8QLYHUVLW\ RI &KLFDJR 3UHVV &KLFDJR ,OOLQRLV )RJGHQ 0 $Q DQQRWDWHG FKHFNOLVW RI WKH ELUGV RI 0RQWHYHUGH DQG 3HDV %ODQFDV *UHHQ 0RXQWDLQ 3XEOLFDWLRQV 0RQWHYHUGH &RVWD 5LFD )RUJHW 30 6HHGGLVSHUVDO RI 9RXDFDSRXD DPHULFDQD &DHVDOSLQLDFHDHf E\ FDYLRPRUSK URGHQWV LQ )UHQFK *XLDQD -RXUQDO RI 7URSLFDO (FRORJ\ )RUJHW 30 6FDWWHUKRDUGLQJ RI $VWURFDU\XP SDUDPDFD E\ 3URHFKLP\V LQ )UHQFK *XLDQD FRPSDULVRQ ZLWK 0\RSURFWD H[LOLV 7URSLFDO (FRORJ\ )RUJHW 30 3RVWGLVSHUVDO SUHGDWLRQ DQG VFDWWHUKRDUGLQJ RI 'LSWHU\[ SDQDPHQVLV 3DSLOLRQDFHDHf VHHGV E\ URGHQWV LQ 3DQDPD 2HFRORJLD )RUJHW 30 5HPRYDO RI VHHGV RI &DU DSD SURFHUD 0HOLDFHDHf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

PAGE 168

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
PAGE 169

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f LQKHULWRUV RI JXDSLQRO +\PHQDHD FRXUEDULO /HJXPLQRVDHf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

PAGE 170

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n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f LQ QRUWKHDVW 3HUX -RXUQDO RI =RRORJ\ +RIJDDUG $ \HDUV RI FKDQJH LQ D 6ZHGLVK ERUHDO ROGJURZWK 3LFHD DELHV IRUHVW -RXUQDO RI 9HJHWDWLRQ 6FLHQFH +ROO 'f DQG 0 ( /XORZ (IIHFW RI VSHFLHV KDELWDW DQG GLVWDQFH IURP HGJH RQ SRVWGLVSHUVDO VHHG SUHGDWLRQ LQ D WURSLFDO UDLQIRUHVW %LRWURSLFD +ROP 6 $ VLPSOH VHTXHQWLDOO\ UHMHFWLYH PXOWLSOH WHVW SURFHGXUH 6FDQGLQDYLDQ -RXUQDO RI 6WDWLVWLFV +ROWKXLM]HQ $ 0 $ 7 / 6KDULN DQG )UDVHU 'LVSHUVDO RI HDVWHUQ UHG FHGDU -XQLSHUXV YLUJLQLDQDf LQWR SDVWXUHV DQ RYHUYLHZ &DQDGLDQ -RXUQDO RI %RWDQ\ +RSSHV : 3UH DQG SRVWIRUDJLQJ PRYHPHQWV RI IUXJLYRURXV ELUGV LQ DQ HDVWHUQ GHFLGXRXV IRUHVW ZRRGODQG 86$ 2LNRV +RSSHV : 6HHGIDOO SDWWHUQ RI VHYHUDO VSHFLHV RI ELUGGLVSHUVHG SODQWV LQ DQ ,OOLQRLV ZRRGODQG (FRORJ\

PAGE 171

+RUYLW] & & $QDO\VLV RI KRZ DQW EHKDYLRUV DIIHFW JHUPLQDWLRQ LQ D WURSLFDO P\UPHFRFKRUH &DODWKHD PLFURFHSKDOD 3 t (f .RHPLFNH 0DUDQWDFHDHf PLFURVLWH VHOHFWLRQ DQG DULO UHPRYDO E\ QHRWURSLFDO DQWV 2GRQWRPDFKXV 3DFK\FRQG\OD DQG 6ROHQRSVLV )RUPLFLGDHf 2HFRORJLD +RUYLW] & & DQG : 6FKHPVNH D $QWQHVW VRLO DQG VHHGOLQJ JURZWK LQ D QHRWURSLFDO DQWGLVSHUVHG KHUE 2HFRORJLD +RUYLW] & & DQG : 6FKHPVNH E 6HHG GLVSHUVDO DQG HQYLURQPHQWDO KHWHURJHQHLW\ LQ D QHRWURSLFDO KHUE D PRGHO RI SRSXODWLRQ DQG SDWFK G\QDPLFV 3DJHV LQ $ (VWUDGD DQG 7 + )OHPLQJ HGLWRUV )UXJLYRUHV DQG VHHG GLVSHUVDO 'U : -XQN 3XEOLVKHUV 'RUGUHFKW 1HWKHUODQGV +RXOH 6HHG GLVSHUVDO DQG VHHGOLQJ UHFUXLWPHQW WKH PLVVLQJ OLQNVf ‹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nWKH SDUDGLJPn VWDQG" 9HJHWDWLR +RZH + ) DQG 'H6WHYHQ )UXLW SURGXFWLRQ PLJUDQW ELUG YLVLWDWLRQ DQG VHHG GLVSHUVDO RI *XDUHD JODEUD LQ 3DQDPD 2HFRORJLD +RZH + ) DQG ) (VWDEURRN 2Q LQWUDVSHFLILF FRPSHWLWLRQ IRU DYLDQ GLVSHUVHUV LQ WURSLFDO WUHHV $PHULFDQ 1DWXUDOLVW +RZH + ) DQG : 0 5LFKWHU (IIHFWV RI VHHG VL]H RQ VHHGOLQJ VL]H LQ 9LUROD VXULQDPHQVLV D ZLWKLQ DQG EHWZHHQ WUHH DQDO\VLV 2HFRORJLD +RZH + ) ( : 6FKXSS DQG / & :HVWOH\ (DUO\ FRQVHTXHQFHV RI VHHG GLVSHUVDO IRU D QHRWURSLFDO WUHH 9LUROD VXULQDPHQVLVf (FRORJ\

PAGE 172

+RZH + ) DQG 6PDOOZRRG (FRORJ\ RI VHHG GLVSHUVDO $QQXDO 5HYLHZ RI (FRORJ\ DQG 6\VWHPDWLFV +RZH + ) DQG $ 9DQGH .HUFNKRYH 1XWPHJ GLVSHUVDO E\ WURSLFDO ELUGV 6FLHQFH +RZH + ) DQG $ 9DQGH .HUFNKRYH 5HPRYDO RI ZLOG QXWPHJ 9LUROD VXULQDPHQVLVf FURSV E\ ELUGV (FRORJ\ +XEEHOO 6 3 6HHG SUHGDWLRQ DQG WKH FRH[LVWHQFH RI WUHH VSHFLHV LQ WURSLFDO IRUHVWV 2LNRV +XOPH 3 ( 3RVWGLVSHUVDO VHHG SUHGDWLRQ E\ VPDOO PDPPDOV 6\PSRVLXP RI WKH =RRORJLFDO 6RFLHW\ RI /RQGRQ +XOPH 3 ( 1DWXUDO UHJHQHUDWLRQ RI \HZ 7D[XV EDFFDWD /f PLFURVLWH VHHG RU KHUELYRUH OLPLWDWLRQ -RXUQDO RI (FRORJ\ +XOPH 3 ( 3RVWGLVSHUVDO VHHG SUHGDWLRQ DQG WKH HVWDEOLVKPHQW RI YHUWHEUDWH GLVSHUVHG SODQWV LQ 0HGLWHUUDQHDQ VFUXEODQGV 2HFRORJLD +XQWHU 5 6HHG GLVSHUVDO DQG JHUPLQDWLRQ RI (QWHURORELXP F\FORFDUSXP -DFTf *ULVHE /HJXPLQRVDH 0LPRVRLGHDHf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n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

PAGE 173

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fV 7DSLUf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f LPSOLFDWLRQV IRU WUHH GHPRJUDSK\ $PHULFDQ 0LGODQG 1DWXUDOLVW -RKQVRQ : & DQG 7 :HEE 7KH UROH RI EOXH MD\V &\DQRFLWWD FULVWDWD /f LQ WKH SRVWJODFLDO GLVSHUVDO RI IDJDFHRXV WUHHV LQ HDVWHUQ 1RUWK $PHULFD -RXUQDO RI %LRJHRJUDSK\ -RUGDQR 3 )UXLWV DQG IUXJLYRU\ 3DJHV LQ 0 )HQQHU HGLWRU 6HHGV WKH HFRORJ\ RI UHJHQHUDWLRQ LQ SODQW FRPPXQLWLHV &$% ,QWHUQDWLRQDO :DOOLQJIRUG 8. -RUGDQR 3 6SDWLDO DQG WHPSRUDO YDULDWLRQ LQ WKH DYLDQIUXJLYRUH DVVHPEODJH RI 3UXQXV PDKDOHE SDWWHUQV DQG FRQVHTXHQFHV 2LNRV -RUGDQR 3 DQG & 0 +HUUHUD 6KXIIOLQJ WKH RIIVSULQJ XQFRXSOLQJ DQG VSDWLDO GLVFRUGDQFH RI PXOWLSOH VWDJHV LQ YHUWHEUDWH VHHG GLVSHUVDO ‹FRVFLHQFH -RVKL $ 5 / 6PLWK DQG ) &XWKEHUW ,QIOXHQFH RI IRRG GLVWULEXWLRQ DQG SUHGDWLRQ SUHVVXUH RQ VSDFLQJ EHKDYLRU LQ SDOP FLYHWV -RXUQDO RI 0DPPDORJ\ -XOLHQ/DIHUULHUH 5DGLRWUDFNLQJ REVHUYDWLRQV RQ UDQJLQJ DQG IRUDJLQJ SDWWHUQV E\ NLQNDMRXV ^3RWRV IODYXVf LQ )UHQFK *XLDQD -RXUQDO RI 7URSLFDO (FRORJ\

PAGE 174

-XOOLRW & 6HHG GLVSHUVDO E\ UHG KRZOLQJ PRQNH\V $ORXDWWD VHQLFXOXVf LQ WKH WURSLFDO UDLQ IRUHVW RI )UHQFK *XLDQD ,QWHUQDWLRQDO -RXUQDO RI 3ULPDWRORJ\ -XOOLRW & ,PSDFW RI VHHG GLVSHUVDO E\ UHG KRZOHU PRQNH\V $ORXDWWD VHQLFXOXV RQ WKH VHHGOLQJ SRSXODWLRQ LQ WKH XQGHUVWRUH\ RI WURSLFDO UDLQ IRUHVW -RXUQDO RI (FRORJ\ .DPLO $ & DQG ( -RQHV 7KH VHHG VWRULQJ FRUYLG &ODUNnV QXWFUDFNHU OHDUQV JHRPHWULF UHODWLRQVKLSV DPRQJ ODQGPDUNV 1DWXUH .DVSDUL 0 5HPRYDO RI VHHGV IURP 1HRWURSLFDO IUXJLYRUH GURSSLQJV 2HFRORJLD .DXIIPDQ 6 % 0F.H\ 0 +RVVDHUW0F.H\ DQG & & +RUYLW] $GDSWDWLRQV IRU D WZRSKDVH VHHG GLVSHUVDO V\VWHP LQYROYLQJ YHUWHEUDWHV DQG DQWV LQ D KHPLHSLSK\WLF ILJ )LFXV PLFURFDUSD 0RUDFHDHf $PHULFDQ -RXUQDO RI %RWDQ\ .HOO\ & DQG $ 3XUYLV 6HHG VL]H DQG HVWDEOLVKPHQW FRQGLWLRQV LQ WURSLFDO WUHHV 2HFRORJLD .LOWLH 5 $ 'LVWULEXWLRQ RI SDOP IUXLWV RQ D UDLQIRUHVW IORRU ZK\ ZKLWHOLSSHG SHFFDULHV IRUDJH QHDU REMHFWV %LRWURSLFD .LWDMLPD (FRSK\VLRORJ\ RI WURSLFDO WUHH VHHGOLQJV 3DJHV LQ 6 6 0XONH\ 5 / &KD]GRQ DQG $ 3 6PLWK HGLWRUV 7URSLFDO IRUHVW SODQW HFRSK\VLRORJ\ &KDSPDQ DQG +DOO 1HZ
PAGE 175

/DPRQW % % ( 7 ) :LWNRZVNL DQG 1 (QULJKW 3RVWILUH OLWWHU PLFURVLWHV VDIH IRU VHHGV XQVDIH IRU VHHGOLQJV (FRORJ\ /DQJWLPP & $ 6SHFLDOL]DWLRQ IRU YHUWLFDO KDELWDWV ZLWKLQ D FORXG IRUHVW FRPPXQLW\ RI PLFH 'LVVHUWDWLRQ 8QLYHUVLW\ RI )ORULGD *DLQHVYLOOH )ORULGD /DQQHU 5 0 0DGH IRU HDFK RWKHU D V\PELRVLV RI ELUGV DQG SLQHV 2[IRUG 8QLYHUVLW\ 3UHVV 1HZ
PAGE 176

/HYH\ DQG ) 6WLOHV (YROXWLRQDU\ SUHFXUVRUV RI ORQJGLVWDQFH PLJUDWLRQ UHVRXUFH DYDLODELOLW\ DQG PRYHPHQW SDWWHUQV LQ QHRWURSLFDO ODQGELUGV $PHULFDQ 1DWXUDOLVW /HYH\ DQG ) 6WLOHV %LUGV HFRORJ\ EHKDYLRU DQG WD[RQRPLF DIILQLWLHV 3DJHV LQ / $ 0F'DGH 6 %DZD + $ +HVSHQKHLGH DQG 6 +DUWVKRUQ HGLWRUV /D 6HOYD HFRORJ\ DQG QDWXUDO KLVWRU\ RI D QHRWURSLFDO UDLQIRUHVW 8QLYHUVLW\ RI &KLFDJR 3UHVV &KLFDJR ,OOLQRLV /HYLQ $ DQG + : .HUVWHU *HQH IORZ LQ VHHG SODQWV (YROXWLRQDU\ %LRORJ\ /L 0 0 /LHEHUPDQ DQG /LHEHUPDQ 6HHGOLQJ GHPRJUDSK\ LQ XQGLVWXUEHG WURSLFDO ZHW IRUHVW LQ &RVWD 5LFD 3DJHV LQ 0 6ZDLQH HGLWRU (FRORJ\ RI WURSLFDO IRUHVW WUHH VHHGOLQJV 81(6&23DUWKHQRQ 3DULV&DPIRUWK /LHEHUPDQ 'HPRJUDSK\ RI WURSLFDO WUHH VHHGOLQJV D UHYLHZ 3DJHV LQ 0 6ZDLQH HGLWRU (FRORJ\ RI WURSLFDO IRUHVW WUHH VHHGOLQJV 81(6&23DUWKHQRQ 3DULV&DPIRUWK /LHEHUPDQ % +DOO 0 6ZDLQH DQG 0 /LHEHUPDQ 6HHG GLVSHUVDO E\ EDERRQV LQ WKH 6KDL +LOOV *KDQD (FRORJ\ /LYLQJVWRQ 5 % ,QIOXHQFH RI ELUGV VWRQHV DQG VRLO RQ WKH HVWDEOLVKPHQW RI SDVWXUH MXQLSHU -XQLSHUXV FRPPXQLV DQG UHG FHGDU YLUJLQLDQD LQ 1HZ (QJODQG SDVWXUHV (FRORJ\ /RLVHOOH % $ 6HHGV LQ GURSSLQJV RI WURSLFDO IUXLWHDWLQJ ELUGV LPSRUWDQFH RI FRQVLGHULQJ VHHG FRPSRVLWLRQ 2HFRORJLD /RLVHOOH % $ ( 5LEEHQV DQG 2 9DUJDV 6SDWLDO DQG WHPSRUDO YDULDWLRQ RI VHHG UDLQ LQ D WURSLFDO ORZODQG ZHW IRUHVW %LRWURSLFD /RYHORFN & (f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

PAGE 177

0DUWQH]5DPRV 0 DQG ( 5 $OYDUH]%X\OOD 6HHG GLVSHUVDO DQG SDWFK G\QDPLFV LQ WURSLFDO UDLQ IRUHVWV D GHPRJUDSKLF DSSURDFK ‹FRVFLHQFH 0DUWQH]5DPRV 0 ( 5 $OYDUH]%X\OOD 6DUXNKDQ DQG 3LHUR 7UHHIDOO DJH GHWHUPLQDWLRQ DQG JDS G\QDPLFV LQ D WURSLFDO IRUHVW -RXUQDO RI (FRORJ\ 0DUWQH]5DPRV 0 DQG $ 6RWR&DVWUR 6HHG UDLQ DQG DGYDQFHG UHJHQHUDWLRQ LQ D WURSLFDO UDLQ IRUHVW 9HJHWDGR 0DVDNL 7 < .RPLQDPL DQG 7 1DNDVKL]XND 6SDWLDO DQG VHDVRQDO SDWWHUQV RI VHHG GLVVHPLQDWLRQ RI &RPXV FRQWURYHUVD LQ D WHPSHUDWH IRUHVW (FRORJ\ 0DWODFN 5 6HFRQGDU\ GLVSHUVDO RI VHHG DFURVV VQRZ LQ %HWXOD OHQWD D JDS FRORQL]LQJ WUHH VSHFLHV -RXUQDO RI (FRORJ\ 0D]HU 6 DQG 1 7 :KHHOZULJKW )UXLW VL]H DQG VKDSH DOORPHWU\ DW GLIIHUHQW WD[RQRPLF OHYHOV LQ ELUGGLVSHUVHG SODQWV (YROXWLRQDU\ (FRORJ\ 0F&DQQ\ 6 $OWHUQDWLYHV LQ SDUHQWRIIVSULQJ UHODWLRQVKLSV LQ SODQWV 2LNRV 0F&ODQDKDQ 7 5 DQG 5 : :ROIH $FFHOHUDWLQJ IRUHVW VXFFHVVLRQ LQ D IUDJPHQWHG ODQGVFDSH WKH UROH RI ELUGV DQG SHUFKHV &RQVHUYDWLRQ %LRORJ\ 0F'RQQHOO 0 2OG ILHOG YHJHWDWLRQ KHLJKW DQG WKH GLVSHUVDO SDWWHUQ RI ELUG GLVVHPLQDWHG ZRRG\ SODQWV %XOOHWLQ RI WKH 7RUUH\ %RWDQLFDO &OXE 0F'RQQHOO 0 DQG ( : 6WLOHV 7KH VWUXFWXUDO FRPSOH[LW\ RI ROG ILHOG YHJHWDWLRQ DQG WKH UHFUXLWPHQW RI ELUGGLVSHUVHG SODQW VSHFLHV 2HFRORJLD 0F.H\ 7KH HFRORJ\ RI FRHYROYHG VHHG GLVSHUVDO V\VWHPV 3DJHV LQ / ( *LOEHUW DQG 3 + 5DYHQ HGLWRUV &RHYROXWLRQ RI 3ODQWV DQG $QLPDOV 8QLYHUVLW\ RI 7H[DV 3UHVV $XVWLQ 7H[DV 86$ 0HUJ ) 8WLOLW\ RI VHHG GLVSHUVDO LQ HVFDSH IURP VHHG SUHGDWRUV DQ H[SHULPHQW LQ D ORZODQG UDLQIRUHVW RI 1HZ *XLQHD 06 7KHVLV 8QLYHUVLW\ RI )ORULGD *DLQHVYLOOH 0H\HU -< DQG )ORUHQFH 7DKLWLnV QDWLYH IORUD HQGDQJHUHG E\ WKH LQYDVLRQ RI 0LFRQLD FDOYHVFHQV '& 0HODVWRPDWDFHDHf -RXUQDO RI %LRJHRJUDSK\ 0RHJHQEXUJ 6 0 6DEDO SDOPHWWR VHHGDQLPDO LQWHUDFWLRQV 06 7KHVLV )ORULGD *DLQHVYLOOH 0RHJHQEXUJ 6 0 6DEDO SDOPHWWR VHHG VL]H FDXVHV RI YDULDWLRQ FKRLFHV RI SUHGDWRUV DQG FRQVHTXHQFHV IRU VHHGOLQJV 2HFRORJLD 0RHUPRQG 7 & DQG 6 'HQVORZ )UXLW FKRLFH LQ QHRWURSLFDO ELUGV HIIHFWV RI IUXLW W\SH DQG DFFHVVLELOLW\ RQ VHOHFWLYLW\ -RXUQDO RI $QLPDO (FRORJ\

PAGE 178

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
PAGE 179

1RJDOHV 0 ) 0 0HGLQD DQG $ 9DOLGR ,QGLUHFW VHHG GLVSHUVDO E\ WKH IHUDO FDWV )HOLV FDPV LQ LVODQG HFRV\VWHPV &DQDU\ ,VODQGVf (FRJUDSK\ 1RJDOHV 0 $ 9DOLGR DQG ) 0 0HGLQD )UXJLYRU\ RI 3ORFDPD SQGXOD 5XELDFHDHf E\ WKH UDEELW 2U\FWRODJXV FXQLFXOXVf LQ [HURSK\WLF ]RQHV RI 7HQHULIH &DQDU\ ,VODQGVf $FWD 2HFRORJLFD 2n'RZG DQG 0 ( +D\ 0XWXDOLVP EHWZHHQ KDUYHVWHU DQWV DQG D GHVHUW HSKHPHUDO VHHG HVFDSH IURP URGHQWV (FRORJ\ 2KNDZDUD 0 2KDUD DQG 6 +LJDVKL 7KH HYROXWLRQ RI DQWGLVSHUVDO LQ D VSULQJHSKHPHUDO &RU\GDOLV DPELJXD 3DSDYHUDFHDHf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f LQ VRXWKHDVWHUQ $PD]RQLD -RXUQDO RI 7URSLFDO (FRORJ\ 3HUHV & $ / & 6FKLHVDUL DQG & / 'LDV/HPH 9HUWHEUDWH SUHGDWLRQ RI %UD]LO QXWV %HUWKROOHWLD H[FHOVD /HF\WKLGDFHDHf DQ DJRXWLGLVSHUVHG $PD]RQLDQ VHHG FURS D WHVW RI WKH HVFDSH K\SRWKHVLV -RXUQDO RI 7URSLFDO (FRORJ\ 3LJQR]]L )UXJLYRU\ DQG VHHG GLVSHUVDO E\ WKH (XURSHDQ EDGJHU LQ D 0HGLWHUUDQHDQ KDELWDW -RXUQDO RI 0DPPDORJ\ 3RSPD -f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

PAGE 180

3\NH $ DQG 1 7KRPSVRQ 6WDWLVWLFDO DQDO\VLV RI VXUYLYDO DQG UHPRYDO UDWH H[SHULPHQWV (FRORJ\ 5HGER7RUVWHQVVRQ 3 DQG $ 7HOHQLXV 3ULPDU\ DQG VHFRQGDU\ GLVSHUVDO E\ ZLQG DQG ZDWHU LQ 6SHUJXDOULD VDOLQD (FRJUDSK\ 5HLG 1 'LVSHUVDO RI PLVWOHWRHV E\ KRQH\HDWHUV DQG IORZHUSHFNHUV FRPSRQHQWV RI VHHG GLVSHUVDO TXDOLW\ (FRORJ\ 5HLG 1 &RHYROXWLRQ RI PLVWOHWRHV DQG IUXJLYRURXV ELUGV" $XVWUDOLDQ -RXUQDO RI (FRORJ\ 5HVWUHSR & $VSHFWRV HFROJLFRV GH OD GLVHPLQDFLQ GH FLQFR HVSHFLHV GH PXUGDJRV SRU DYHV +XPEROWLD 5LFH % / DQG 0 :HVWRE\ (YLGHQFH DJDLQVW WKH K\SRWKHVLV WKDW DQWGLVSHUVHG VHHGV UHDFK QXWULHQWHQULFKHG PLFURVLWHV (FRORJ\ 5LGOH\ + 1 7KH GLVSHUVDO RI SODQWV WKURXJKRXW WKH ZRUOG 5HHYH $VKIRUG 8. 5LOH\ & 0 DQG 6PLWK 6H[XDO GLPRUSKLVP DQG IRUDJLQJ EHKDYLRU RI (PHUDOG 7RXFDQHWV $XODFRUK\QFKXV SUDVLQXV LQ &RVWD 5LFD 2PLV 6FDQGLQDYLFD 5REHUWV 7 DQG ( 5 +HLWKDXV $QWV UHDUUDQJH WKH YHUWHEUDWHJHQHUDWHG VHHG VKDGRZ RI D QHRWURSLFDO ILJ WUHH (FRORJ\ 5RELQVRQ 5 DQG 6 1 +DQGHO )RUHVW UHVWRUDWLRQ RQ D FORVHG ODQGILOO UDSLG DGGLWLRQ RI QHZ VSHFLHV E\ ELUG GLVSHUVDO &RQVHUYDWLRQ %LRORJ\ 5XQNOH 5 *DS G\QDPLFV LQ DQ 2KLR $FHU)DJXV IRUHVW DQG VSHFXODWLRQV RQ WKH JHRJUDSK\ RI GLVWXUEDQFH &DQDGLDQ -RXUQDO RI )RUHVW 5HVHDUFK 6DQWDQD & ( DQG % 0LOOLJDQ %HKDYLRU RI WRXFDQHWV EHOOELUGV DQG TXHW]DOV IHHGLQJ RQ ODXUDFHRXV IUXLWV %LRWURSLFD 6DUJHQW 6 6HHG IDWH LQ D WURSLFDO PLVWOHWRH WKH LPSRUWDQFH RI KRVW WZLJ VL]H )XQFWLRQDO (FRORJ\ 6$6 ,QVWLWXWH 6$667$7 XVHUnV JXLGH 6$6 ,QVWLWXWH ,QF &DU\ 1RUWK &DUROLQD 6$6 ,QVWLWXWH -03 XVHUnV JXLGH YHUVLRQ 6$6 ,QVWLWXWH ,QF &DU\ 1RUWK &DUROLQD 6FKHPVNH : DQG 1 9 / %URNDZ 7UHHIDOOV DQG WKH GLVWULEXWLRQ RI XQGHUVWRU\ ELUGV LQ D WURSLFDO IRUHVW (FRORJ\ 6FKXSS ( : D )DFWRUV DIIHFWLQJ SRVWGLVSHUVDO VHHG VXUYLYDO LQ D WURSLFDO IRUHVW 2HFRORJLD 6FKXSS ( : E 6HHG DQG HDUO\ VHHGOLQJ SUHGDWLRQ LQ WKH IRUHVW XQGHUVWRU\ DQG LQ WUHHIDOO JDSV 2LNRV

PAGE 181

6FKXSS ( : 4XDQWLW\ TXDOLW\ DQG WKH HIIHFWLYHQHVV RI VHHG GLVSHUVDO E\ DQLPDOV 9HJHWDGR 6FKXSS ( : 6HHGVHHGOLQJ FRQIOLFWV KDELWDW FKRLFH DQG SDWWHUQV RI SODQW UHFUXLWPHQW $PHULFDQ -RXUQDO RI %RWDQ\ 6FKXSS ( : DQG ( )URVW 'LIIHUHQWLDO SUHGDWLRQ RI :HOILD JHRUJLL VHHGV LQ WUHHIDOO JDSV DQG WKH IRUHVW XQGHUVWRU\ %LRWURSLFD 6FKXSS ( : DQG 0 )XHQWHV 6SDWLDO SDWWHUQ RI VHHG GLVSHUVDO DQG WKH XQLILFDWLRQ RI SODQW SRSXODWLRQ HFRORJ\ ‹FRVFLHQFH 6FKXSS ( : 0 *PH] ( -LPQH] DQG 0 )XHQWHV D 'LVSHUVDO RI -XQLSHUXV RFFLGHQWDOV ZHVWHUQ MXQLSHUf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f 6PLWKVRQLDQ &RQWULEXWLRQV WR =RRORJ\ 6P\WKH 1 6HHG VXUYLYDO LQ WKH SDOP $VWURFDU\XP VWDQGOH\DQXP HYLGHQFH IRU GHSHQGHQFH XSRQ LWV VHHG GLVSHUVHUV %LRWURSLFD

PAGE 182

6QHGGRQ / $ /DWULQH XVH E\ WKH (XURSHDQ UDEELW 2U\FWRODJXV FXQLFXOXVf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f %XOOHWLQ RI WKH %ULWLVK 0XVHXP 1DWXUDO +LVWRU\f 6QRZ : 7KH ZHE RI DGDSWDWLRQ 4XDGUDQJOH 1HZ
PAGE 183

6WLOHV ) DQG $ ) 6NXWFK $ JXLGH WR WKH ELUGV RI &RVWD 5LFD &RUQHOO 8QLYHUVLW\ 3UHVV ,WKDFD 1HZ
PAGE 184

9DOLHQWH%DQXHW $ ) 9LWH DQG $ =DYDOD+XUWDGR ,QWHUDFWLRQ EHWZHHQ WKH FDFWXV 1HREX[EDXPLD WHWH]R DQG WKH QXUVH VKUXE 0LPRVD OXVLDQD -RXUQDO RI 9HJHWDWLRQ 6FLHQFH 9DQ GHU 0HHU 3 DQG ) %RQJHUV 3DWWHUQV RI WUHHIDOO DQG EUDQFKIDOO LQ D WURSLFDO IRUHVW LQ )UHQFK *XLDQD -RXUQDO RI (FRORJ\ YDQ GHU 3LMO / 3ULQFLSOHV RI GLVSHUVDO LQ KLJKHU SODQWV 6SULQJHU9HUODJ %HUOLQ 9DQGHU :DOO 6 % )RRG KRDUGLQJ LQ DQLPDOV 8QLYHUVLW\ RI &KLFDJR 3UHVV &KLFDJR ,OOLQRLV 9DQGHU :DOO 6 % 7KH UROH RI DQLPDOV LQ GLVSHUVLQJ D ZLQGGLVSHUVHG SLQH (FRORJ\ 9DQGHU :DOO 6 % 'LVSHUVDO RI VLQJOHOHDI SLQ SLQH 3LQXV PRQRSK\OODf E\ VHHGn FDFKLQJ URGHQWV -RXUQDO RI 0DPPDORJ\ 9DQGHU :DOO 6 % DQG 5 3 %DLGD &RDGDSWDWLRQV RI WKH &ODUNnV QXWFUDFNHU DQG WKH SLQ SLQH IRU HIILFLHQW VHHG KDUYHVW DQG GLVSHUVDO (FRORJLFDO 0RQRJUDSKV 9£VTXH] 5 $ 3DWFK XWLOL]DWLRQ E\ WKUHH VSHFLHV RI &KLOHDQ URGHQWV GLIIHULQJ LQ ERG\ VL]H DQG PRGH RI ORFRPRWLRQ (FRORJ\ 9HHQHQGDDO ( 0 0 6ZDLQH 9 $J\HPDQ %OD\ $EHEUHVH DQG & ( 0XOOLQV 'LIIHUHQFHV LQ SODQW DQG VRLO ZDWHU UHODWLRQV LQ DQG DURXQG D IRUHVW JDS LQ :HVW $IULFD GXULQJ WKH GU\ VHDVRQ PD\ LQIOXHQFH VHHGOLQJ HVWDEOLVKPHQW DQG VXUYLYDO -RXUQDO RI (FRORJ\ 9HQDEOH / DQG 6 %URZQ 7KH SRSXODWLRQG\QDPLF IXQFWLRQV RI VHHG GLVSHUVDO 9HJHWDGR 9LHLUD & *f & 8KO DQG 1 7KH UROH RI WKH VKUXE &RUGLD PXOWLVLSFDWD &KDP DV D nVXFFHVVLRQ IDFLOLWDWRUn LQ DQ DEDQGRQHG SDVWXUH 3DUDJRPLQDV $PD]RQLD 9HJHWDGR :DKDM 6 $ /HYH\ $ 6DQGHUV DQG 0 / &LSROOLQL LQ SUHVV &RQWURO RI JXW UHWHQWLRQ WLPH E\ VHFRQGDU\ PHWDEROLWHV LQ ULSH 6RODQXP IUXLWV (FRORJ\ [[[[ :DOVEHUJ ( 'LJHVWLYH DGDSWDWLRQV RI 3KDLQRSHSOD QLWHQV DVVRFLDWHG ZLWK WKH HDWLQJ RI PLVWOHWRH EHUULHV &RQGRU :HLEOHQ DQG 7KRPVRQ 6HHG GLVSHUVDO LQ (U\WKURQLXP JUDQGLIORUXP /LOLDFHDHf 2HFRORJLD :KHHOZULJKW 1 7 )UXLWV DQG WKH HFRORJ\ RI 5HVSOHQGHQW 4XHW]DOV $XN :KHHOZULJKW 1 7 D &RPSHWLWLRQ IRU GLVSHUVHUV DQG WKH WLPLQJ RI IORZHULQJ DQG IUXLWLQJ LQ D JXLOG RI WURSLFDO WUHHV 2LNRV :KHHOZULJKW 1 7 E )UXLW VL]H JDSH ZLGWK DQG WKH GLHWV RI IUXLWHDWLQJ ELUGV (FRORJ\

PAGE 185

:KHHOZULJKW 1 7 $ VHYHQ\HDU VWXG\ RI LQGLYLGXDO YDULDWLRQ LQ IUXLW SURGXFWLRQ LQ WURSLFDO ELUGGLVSHUVHG WUHH VSHFLHV LQ WKH IDPLO\ /DXUDFHDH 3DJHV LQ $ (VWUDGD DQG 7 + )OHPLQJ HGLWRUV )UXJLYRUHV DQG VHHG GLVSHUVDO 'U : -XQN 3XEOLVKHUV 'RUGUHFKW 7KH 1HWKHUODQGV :KHHOZULJKW 1 7 )RXU FRQVWUDLQWV RQ FRHYROXWLRQ EHWZHHQ IUXLWHDWLQJ ELUGV DQG IUXLWLQJ SODQWV D WURSLFDO FDVH VWXG\ 3URFHHGLQJV RI WKH ,QWHUQDWLRQDO 2UQLWKRORJLFDO &RQJUHVV ;,; :KHHOZULJKW 1 7 +RZ ORQJ GR IUXLWHDWLQJ ELUGV VWD\ LQ WKH SODQWV ZKHUH WKH\ IHHG" %LRWURSLFD :KHHOZULJKW 1 7 : $ +DEHU 0XUUD\ DQG & $ *XLQGRQ 7URSLFDO IUXLWHDWLQJ ELUGV DQG WKHLU IRRG SODQWV D VXUYH\ RI D &RVWD 5LFDQ ORZHU PRQWDQH IRUHVW %LRWURSLFD :KHHOZULJKW 1 7 DQG + 2ULDQV 6HHG GLVSHUVDO E\ DQLPDOV FRQWUDVWV ZLWK SROOHQ GLVSHUVDO SUREOHPV RI WHUPLQRORJ\ DQG FRQVWUDLQWV RQ FRHYROXWLRQ $PHULFDQ 1DWXUDOLVW :KHODQ & 0 / 'LOJHU 5REVRQ 1 +DOO\Q DQG 6 'LOJHU (IIHFWV RI ROIDFWRU\ FXHV RQ DUWLILFLDOQHVW H[SHULPHQWV $XN :KLWPRUH 7 & $ UHYLHZ RI VRPH DVSHFWV RI WURSLFDO UDLQ IRUHVW VHHGOLQJ HFRORJ\ ZLWK VXJJHVWLRQV IRU IXUWKHU HQTXLU\ 3DJHV LQ 0 6ZDLQH HGLWRU (FRORJ\ RI WURSLFDO IRUHVW WUHH VHHGOLQJV 81(6&23DUWKHQRQ 3DULV&DPIRUWK :LONLQVRQ 0 D 3ODQW FRORQL]DWLRQ DUH ZLQG GLVSHUVHG VHHGV UHDOO\ GLVSHUVHG E\ ELUGV DW ODUJHU VSDWLDO DQG WHPSRUDO VFDOHV" -RXUQDO RI %LRJHRJUDSK\ :LONLQVRQ 0 E 7KH UROH RI VHHG GLVSHUVDO LQ WKH HYROXWLRQ RI P\FRUUKL]DH 2LNRV :LOOVRQ 0 ) 7KH HFRORJ\ RI VHHG GLVSHUVDO 3DJHV LQ 0 )HQQHU HGLWRU 6HHGV WKH HFRORJ\ RI UHJHQHUDWLRQ LQ SODQW FRPPXQLWLHV &$% ,QWHUQDWLRQDO :DOOLQJIRUG 8. :LOOVRQ 0 ) 'LVSHUVDO PRGH VHHG VKDGRZV DQG FRORQL]DWLRQ SDWWHUQV 9HJHWDGR :LOOVRQ 0 ) DQG ) + &URPH 3DWWHUQV RI VHHG UDLQ DW WKH HGJH RI D WURSLFDO 4XHHQVODQG UDLQ IRUHVW -RXUQDO RI 7URSLFDO (FRORJ\ :LOOVRQ 0 )f $ ,UYLQH DQG 1 :DOVK 9HUWHEUDWH GLVSHUVDO V\QGURPHV LQ VRPH $XVWUDOLDQ DQG 1HZ =HDODQG SODQW FRPPXQLWLHV ZLWK JHRJUDSKLF FRPSDULVRQV %LRWURSLFD :LOOVRQ 0 ) ( $ 3RUWHU DQG 5 &RQGLW $YLDQ IUXJLYRUH DFWLYLW\ LQ UHODWLRQ WR IRUHVW OLJKW JDSV &DULEEHDQ -RXUQDO RI 6FLHQFH :LWPHU 0 & DQG $ 6 &KHNH 7KH GRGR DQG WKH WDPEDODFRTXH WUHH DQ REOLJDWH PXWXDOLVP UHFRQVLGHUHG 2LNRV

PAGE 186

:ULJKW 6 0 ( *RPSSHU DQG % 'H/HRQ $UH ODUJH SUHGDWRUV NH\VWRQH VSHFLHV LQ 1HRWURSLFDO IRUHVWV" 7KH HYLGHQFH IURP %DUUR &RORUDGR ,VODQG 2LNRV
PAGE 187

%,2*5$3+,&$/ 6.(7&+ %RUQ LQ 3RXJKNHHSVLH 1HZ
PAGE 188

DQG OHVV LQMXU\SURQH $IWHU JUDGXDWLQJ IURP (DUOKDP ZRUNHG RQH VXPPHU LQ 1RUWK 'DNRWD DV D ILHOG DVVLVWDQW IRU &RUQHOO 8QLYHUVLW\ JUDGXDWH VWXGHQW &DUROD +DDV QRZ DW 9LUJLQLD 7HFKf 7KHQ PRYHG RQ WR D VXUSULVLQJO\ TXLFN DQG HIILFLHQW PDVWHUn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

PAGE 189

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

PAGE 190

/2 " :+ 81,9(56,7< 2) )/25,'$


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ENYHIK4XU_0JWY4P INGEST_TIME 2014-06-05T17:07:06Z PACKAGE AA00021353_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES