Citation
The biology of Orbignya martiana (PALMAE), a tropical dry forest dominant in Brazil

Material Information

Title:
The biology of Orbignya martiana (PALMAE), a tropical dry forest dominant in Brazil
Creator:
Anderson, Anthony B
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Forests ( jstor )
Fruits ( jstor )
Germination ( jstor )
Inflorescences ( jstor )
Leaves ( jstor )
Pastures ( jstor )
Primary forests ( jstor )
Productivity ( jstor )
Secondary forests ( jstor )
Seedlings ( jstor )
Babassu
Genre:
Palms--Brazil--Maranhão

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
30349357 ( ALEPH )
11437530 ( OCLC )

Downloads

This item has the following downloads:


Full Text











THE BIOLOGY OF Orbignya martiana (PALMAE), A TROPICAL DRY FOREST DOMINANT IN BRAZIL




















BY

ANTHONY B. ANDERSON



















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

UNIVERSITY OF FLORIDA 1983




























This study is dedicated to the late Harold E. Moore, Jr., who inspired it.















ACKNOWLEDGEMENTS

My field research and data analysis were financed by Grant No. 5107-09-07 to John Ewel from the USDA Forest Service, Consortium for the

Study of Man's Relationship with the Global Environment. Additional support was provided by a fellowship from the Superintendencia de Desenvolvimento Cientifico e Tecnol6gico (CNPq). I am grateful to the Museu

Paraense Emilio Goeldi and its Director, Dr. Jose Seixas Lourenco, for invaluable logistic as well as moral support during the course of my research.

Special thanks are due to Bernardo da Silva and Mario Costa, who gave unfailing assistance in the field, and to Michael Balick and William Overal, who collaborated with me on various aspects of the research. I am also grateful to the following people and their associated institutions: Jos6 Mdrio Frazao and Claddio Pinheiro of the Instituto Estadual do Babapu (INEB) for their collaboration in studies on germination, phenology, and taxonomy of babassu; Heraclito Aquino and

Walbert Carvalho of the Empresa Maranhense de Pesquisa Agropecuiria (EMAPA) for providing field workers; S6rgio Alves, Mirio Dantas, Elizabeth de Oliveira, Raimundo Rego, and Waldemar Ferreira of the Centro de Pesquisa Agropecuaria do Tropico Umido (CPATU) for invaluable logistic support; and Raimundo Bittencourt, Waldemar Franca, Nelson Rosa, and Sr. Milton of the Museu Goeldi for their assistance in the field.


iii











The excellent drawings were by George Fuller, and my battery of dedicated typists included Bob Epting, Donna Epting, and Leslie Rigg. Less tangible but.nonetheless meaningful assistance was provided by Helda Lenz Cesar, Warwick Kerr, Mario Leal, Peter May, and Ghillean Prance.

I would like to express my sincere appreciation to the people who made my dedication to the research possible: Clemente Rodrigues and his extended family; Isa and Nelson Carvalho; my parents, Martha and Seneca Anderson; and especially my wife, Suely Anderson.

Finally, very special thanks go to my supervisory committee, Drs. Peter Feinsinger, Walter Judd, Hugh Popenoe, Francis Putz, and especially my chairman, John Ewel, who visited the field sites and provided patient guidance in all phases of the research.




























iv
















TABLE OF CONTENTS



ACKNOWLEDGEMENTS . . . . . . . . . . . .

LIST OF TABLES . . . . . . . . . a a vii

LIST OF FIGURES . . . . . . . . . . . . ix

ABSTRACT . . . . . * * * * * * *xii

CHAPTER

1 INTRODUCTION . . . . . . . . . . 1

2 TAXONOMY AND PHYTOGEOGRAPHY . . . . . . 27
Methods * * . * . . e 29 Results and Discussion . . . . . 32

3 REPRODUCTI EBIOOLO... .. .. .. . ... 38
Methods . . . . . . . . . . 39
Results . . . . . . . . . . 46
Discussion . . . . . . . . . 74

4 ESTABLISHMENT . . . . . . . . . . 88
Me thods . . . . . . . . . . 90
Discussion . . . . . . . . . 112

5 GROWTH AND PRODUCTIVITY ... .. .. . .. 115
Results . . . . . . ... 1216
Discussion. . * * e * 9 . . 129

6 POPULATION STRUCTURE AND DYNAMICS . . . . . 140


Discussion., 154

7 IMPLICATIONS FOR MANAGEMENT . . . . . . 157
.People and the Palm Forest . . 157 Limiting Factors. .....* *159
Possible Solutions ............. . 163



v












REFERENCES CITED . . . . . . . . . . . 168

APPENDIX

A SPECIES ABUNDANCE, FREQUENCY, DOMINANCE, AND
IMPORTANCE IN 1 HA OF PRIMARY FOREST, LAGO VERDE,
MARANHA0 . . . . . . . . 179

B SPECIES ABUNDANCE, FREQUENCY, DOMINANCE, AND
IMPORTANCE IN 1 HA OF SECONDARY FOREST, LAGO VERDE,
MARANHAO . . . . . . . . . . .. 184

C MORPHOLOGICAL DESCRIPTION OF Orbignya martiana
Barb. Rodr. .. . . .. .. . . . 187

D SELECTIVE COMPARISON OF GROSS MORPHOLOGICAL
CHARACTERS IN Orbignya martiana, "0.
teixeirana", and 0. eichleri . . . . . . 190

E SELECTIVE COMPARISON OF GROSS MORPHOLOGICAL
CHARACTERS IN Orbignya martiana, "Markeleya
dahlgreniana", and Maximiliana maripa . . 192

BIOGRAPHICAL SKETCH. . . . . . . . . . .. 194




























vi














LIST OF TABLES

Table

1 Uses of occupied lands in Maranhio during 1950 and 1975 . 22

2 Nomenclatural history of babassu and its suspected hybrids. 30

3 Fruit yields of babassu per ha and per palm during 1980-81
and 1981-82 at three sites. .0. . . . . . . .. 60

4 Sources of increased fruit yields per ha between 1980-81
and 1981-82 at three sites ................ 61

5 Attack of babassu fruits and seeds by the seed-predaceous
beetle, Pachymerus nucleorum .. . . . . .. 68

6 Germination of babassu in four experiments. .*. . . 73

7 Preliminary checklist of insects that feed on babassu .... 80

8 Hypothesized success of babassu in four experiments .... 89

9 Palms per m2 (a), leaf area per palm (b), and leaf area
index (c) of babassu at end of irrigation experiment.. . .102

10 Palms per m2 (a), leaf area per palm (b), and leaf area
index (c) of babassu at end of fertilizer experiment.. . .103

11 Palms per m2 (a), leaf area per palm (b), and leaf area
index (c) of babassu at end of insecticide experiment . .104

12 Palms per m2 (a), leaf area per palm (b), and leaf area
index (c) of babassu at end of climate experiment, with
climate and weeding as main effects 0 . . 0 . . .105 13 Palms per m2 (a), leaf area per palm (b), and leaf area index (c) of babassu at end of climate experiment, with
climate and ecotype as main effects . . . 0 0 0 $106

14 Leaf area index (LAI) of babassu and competitors in four
experiments 0 . . . . .. .. .. . .107

15 Dry mass of babassu and competitors in four experiments . .108




vii











16 Soil chemistry of four treatments at end of nutrient
experiment . . . . . . . . . . . . .109

17 Soil chemistry of experimental sites at Lago Verde and
Belim . . . . . . . . . . . . . *110

18 Definition of 12 life stages of babassu used in study . .117 19 Estimated growth rates and ages of babassu palms on the
primary forest site at Lago Verde . . . . . . 122

20 Estimated growth rates and ages of babassu palms on the
secondary forest site at Lago Verde . . . . . .123

21 -Partitioning of above-ground dry matter production of
babassu over an estimated life span of 184 yr . . . .126 22 Annual above-ground dry matter production of mature babassu
palms on three study sites at Lago Verde. . ... . . 127 23 Annual above-ground dry matter production of all babassu
palms on the primary and secondary forest sites at Lago
Verde . . . . . . . . . . . .. .128

24 Life table of the babassu population on the 1-ha primary
forest site at Lago Verde . . . . . . . . .144

25 Annual changes in number of palms per life stage on the 1ha primary forest site at Lago Verde. . . . . . .145

26 Annual changes in number of palms per life stage on the 1ha secondary forest site at Lago Verde. . . . . .. .146

27 Annual changes in number of palms per life stage on the 1ha pasture site at Lago Verde . . . ... . . .147

















viii















LIST OF FIGURES

Figure

1 Approximate distribution of high-density stands of babassu
(Orbignya martiana) in Brazil . . . . . . . . 2

2 Aerial view of babassu stands in the county of Bacabal,
Maranhao. . . . . . . . . . . . . .. 3

3 Specimen of babassu (0. martiana) in Pirapora, Minas
Gerais. .0. . . . . . . . . . . . . 3

4 Deforestation in the Tropical Moist Forest zone of
Maranh o. . . . . . . . . . . . . 4

5 Regrowth of babassu on a deforested site in the Tropical
Moist Forest zone of Maranhao . . . . . . . . 4

6 Production of babassu kernels in Maranhio and Brazil, 19201979. . . . . . . . . . . . . . . 5

7 Flowchart of products derived from babassu fruits by industries in Bacabal, Maranhao, during 1980 . . . . . 6

8 Flowchart of products derivable from babassu fruits via
current (1983) technology . . . . . . . . 7

9 Maranhao and neighboring states, with locations of research
sites . . . . . . . . . . . . 10
10 Major soil groups of Maranhao, according to the Brazilian
Classification System . . . . . . . . . 11

11 Principal ecological zones of Maranhao. . . . . .. 12

12 Coverage of babassu stands in Maranhao. . . . . . 13

13 Monthly rainfall at four research sites . . . . . 18

14 Distribution of land and households according to size of
landholding in rural areas of Maranhao during 1950 and
1975. . . . . . . . . . . . . . . 20

15 Distribution of specimens of Orbignya martiana and its
suspected hybrid complexes. . . . . . ....... 31


ix












16 Mean monthly gain and loss of leaves per palm on three
sites in Maranhao . . . . .. . . . .. . 47

17 Mean monthly gain and loss of leaves per palm at three
study sites at Lago Verde . . . . . . . . .. 48

18 Mean monthly production of male and functionally female
inflorescences per palm at three sites in Maranhao.... . 50 19 Mean monthly production of male and functionally female
inflorescences per palm at three sites at Lago Verde . . 51 20 Distributions of female inflorescences per palm in six
populations of babassu. . . . . . . . . .. 54

21 Mean monthly number of immature and mature infructescences
per palm at three sites in Maranhio . .. . ... . 55

22 Mean monthly number of immature and mature infructescences
per palm at three sites at Lago Verde . . .. . .. 56

23 Monthly fruit production per ha on three study sites in
Lago Verde. . . . . . . . . . . . 57

24 Monthly fruit production per palm on three study sites at
Lago Verde. . .. . . . . . . . . 58

25 Removal of babassu fruits on three study sites at Lago
Verde . . 0. . . . . . . . . . 66

26 Attack of babassu fruits and seeds by the Bruchid beetle,
Pachymerus nucleorum, over time . . . . . . . 67

27 Seed fate over time in three study sites at Lago Verde . 69 28 Cumulative germination of babassu fruits over time. . . 71 29 Germination in fruits obtained from 10 babassu palms . . 72 30 Force required to break palm fruits . . . . . .. 84

31 Cryptogeal germination in babassu . . . . . . .. 86

32 Leaf area as a function of leaf length in seedlings of
babassu . . . . . . . . . . . . 93

33 Palms per m2 (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in irrigation experiment . 96


x










34 Palms per m2 (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in fertilizer experiment . 97 35 Palms per m2 (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in insecticide experiment. . 98 36 Palms per m2 (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in climate experiment ..... 99 37 Absolute allocation of annual above-ground dry matter
production per babassu palm as a function of age.. ... .130 38 Relative allocation of annual above-ground dry matter
production per babassu palm as a function of age. ... 0131 39 Mean annual number of inflorescences produced per palm as a
function of mean age per life stage . . . . . . .134

40 Mean annual number of seeds produced per palm as a function
of mean age per life stage. . . . . . * e .135

41 Stage distributions of babassu on the primary forest,
secondary forest, and pasture sites at Lago Verde .. 148 42 Age distribution of babassu on the primary forest site at
Lago Verde . . . . . 0 0 0 0 .149

43 Age distribution of babassu on the secondary forest site
at Lago Verde . .. . . . .. . .. 0 .150
44 Population dynamics of babassu on the 1-ha primary forest,

secondary forest, and pasture sites at Lago Verde .. . .151 45 Shifting cultivation site just prior to burn at Lago Verde. .160 46 Clearcut of palm forest prior to pasture establishment,
near Bacabal, Maranhdo . . ............ 160














xi















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

THE BIOLOGY OF Orbignya martiana (PALMAE), A TROPICAL DRY FOREST DOMINANT IN BRAZIL

By

Anthony B. Anderson

December, 1983

Chairman: John J. Ewel
Major Department: Botany


Monospecific stands of the babassu palm (Orbignya martiana) occupy widespread areas of Brazil, attaining their greatest coverage (102,970 km2) in the state of Maranhio. Near Lago-Verde, Maranhio, I conducted comparative studies of babassu's life history to determine how the palm is maintained in mature forests as well as how it dominates disturbed sites.

Flowering in babassu peaked during the rainy season (January-March) and coincided with a flush of new leaves (x = 4.1 per yr); fruiting peaked during the following dry season (October-December). Outcrossing was predominant and was effected by wind and nitidulid beetles (MystroDs

sp.). Dispersal of the massive ( >100 g) fruits was mediated primarily by gravity and by rodents. Due to the hard fruit endocarp, seed predation was virtually limited to bruchid larvae (Pachymerus nucleorum), which entered through the germination pores; ca. 40% of seeds suffered



xii











predation. Germination required 3 mo and was enhanced in shade. Stemless palms persisted in the understory of a primary forest for about 50 yr, permitting build-up of high densities (6,000 per ha). Escape from seed predators, shade tolerance, and an estimated lifespan of 184 yr account for babassu's importance in such forests. Due to burial of

their apical meristems via cryptogeal germination, stemless palms survive when sites are burned for shifting cultivation. During the subsequent fallow, these palms are released and form virtually monospecific stands. One such stand was still increasing in numbers after a fallow of 33 yr. Annual above-ground dry matter production of this stand was ca. 25 t'ha-1"yr-1, 70% of which was allocated to leaves. With fallows of 4-5 yr, enough fuel accumulates in the form of leaves to permit the hot fire necessary for weed control and nutrient release. Thus shifting cultivation can be practiced with minimal thinning of the stands.

Babassu's dominance of the landscape results from its persistence over a wide gradient of disturbances. Because of its capacity to form monospecific stands, its low maintenance requirements, its integration into major land uses, and its current importance in market and subsistence economies, babassu seems most amenable for use as an extensive, moderate-yielding crop.













xiii















CHAPTER 1
INTRODUCTION

Numerous species of palms dominate the landscape over extensive areas of the lowland tropics (Beard 1944, Richards 1952, Moore 1973a). Palm-dominated stands often occur on edaphically limited sites where competition is apparently reduced (Myers 1981). Several species, however, form nearly monospecific stands on sites originally covered by forests of high biological diversity. Such secondary palm forests

originate from and are apparently maintained by human activities. People, in turn, depend on the palms as a source of food, fuel, shelter and income.

This study concerns one of the most economically important and certainly the most ubiquitous of such secondary forest palms: Orbignya martiana, known in Brazil as babassu. The tree-sized babassu palm forms high-density stands over extensive areas of Brazil (Figs. 1-3). The stands attain their greatest extent in the state of Maranhdo, where they

cover an area estimated at 102,970 km2 (Anonymous 1981), or roughly equivalent to the U.S. state of Virginia. Babassu flourishes on sites subject to intensive disturbances such as cutting and burning (Figs. 45). The palm is a crucial resource for rural poor, who utilize the fruits, leaves, and stem for a wide range of subsistence products. Manual extraction of kernels from the extremely hard fruits represents the state's largest cottage industry (involving over 400,000 families) and its most important income source for women. The kernels are util1





2













75 70 65 60 C.5 50 45 40 3
h I ti I
5

ATLANTIC OCEAN



MARANHAO 0



AMAZONAS
PARA- E
PIU


10
MATO GOSSO GOIAS
-. ;. r ...- ',, --






MINAS GERAIS"' a oo 000k / 20





125




-0







Figure 1. Approximate distribution of high-density stands of
babassu (Orbignya martiana) in Brazil. States containing stands are labelled. Adapted from Mendes
and Carioca (1981).
























PL4




.Q. Ca cc U

Wcac























cac





















ca 0 ~ca
r p



cu a




































Figure 4. Deforestation in the Tropical Mo11ist Forest zone of Iflaranhao.




-C










.. ~. ..












Figure 5. Regrowth of babassu on a deforested site in the Tropical
Moist Forest zone of Maranhdo.





5




















240


200

0= BRAZIL
160
z

0
120

a.
80
.J
w
Z
w 40




1920 1930 1940 1950 1960 1970 1980




Figure 6. Production of babassu kernels in Maranhio and Brazil,
1920-79. Sources: IBGE (1981a and previous volumes to
1916).






6













1 3 8 S T A- R C H



MESOCARP 9.2 -023.0 23.0 -- FERTILIZER
2.2

< 8.8

EPICARP 8.8
I 1. 1.0 PRIMARY
COMBUSTION


ENDOCARP 18.64
59.0

1 4 .7 -. C O K E


4.20 OIL





KERNELS
7.0
2.38 -FEEDCKE






Figure 7. Flowchart of products derived from babassu fruits by industries
in Bacabal, Maranhao, during 1980. Diamonds represent "eitheror" options. Numbers are percentages of fruit weight. Adapted
from Mendes and Carioca (1981) and Empresa Industrial de Bacabal (personal communication).






7
















138 STARCH ETHANOL


FERTILIZER
MESOCARP
23.0
3.0 9.2 ANIMAL
-RATION
110 FIBERS

-EPICARP 10 .PRMAY
O COMBUSTION 3 REACTIVATED
CHARCOAL


CHARCOAL 1 7 COKE
_COMBUSTIBLE GASES
4.0 ACETIC 4.3 ACID
COMBUTISLEACETATES O GASES CEO

7 ENDOCARP2.2 OIL M GAIN
CONDENSED

4 FEEDCAKEOL


0.2
EDIBLE44 KERNELS 4R2 OIL MARGARINE
7.0

46 CRUDE

50AP


0 G LYCERIN





Figure 8. Flowchart of products derivable from babassu fruits via
current (1983) technology. Diamonds represent "either-or"
options. Numbers are percentages of fruit weight. Adapted from Abreu (1940), Viveiros (1943), Escola Tecnica Federal
do Maranhao (1976), IPT (1979), and Mendes and Carioca
(1981).





8






ized by industries for production of vegetable oil, soap, and feedcake. These industries represent the largest single source of tax revenues in the Maranhao economy. Over the past 60 yr, kernel production has steadily increased (Fig. 6), and as of 1979 babassu provided 89% of all vegetable oil obtained from non-domesticated sources in Brazil (IBGE 1981 a). Recently introduced machinery capable of mechanically breaking the hard fruits has enabled additional babassu-derived products (e.g., charcoal, coke) to enter national markets (Fig. 7), and technology presently exists to derive a host of additional products from the fruits (F ig 8).

Despite its economic importance, babassu is not currently domesticated, and formation of the stands remains a spontaneous phenomenon. In the past, stands of babassu were generally preserved and formed an integral part of shifting cultivation systems. In recent years, however, population pressures and changing land use pa'Vterns are inducing both shifting cultivators and ranchers to clearcut the stands (Anderson and Anderson 1983). Sound management of the palm forest requires basic information concerning the biology of babassu. The purpose of this study is to provide such information.

The study covers babassu's taxonomy (Chapter 2), life history (Chapters 3-.5), and population biology (Chapter 6); levels of inquiry change respectively from the species to individual palms to stands of




9





babassu. The study concludes with a discussion of the implications for management (Chapter 7).

The Setting

Maranhd.o

The Brazilian state of Maranh&o (Fig. 9) encompasses a wide range of ecological conditions. To the south lies a geologically old (Mesozoic), elevated peneplain, broken by abrupt plateaus and canyons, and by low, eroded mountain ranges. The predominant soils in this region are deeply weathered and nutrient poor, and annual rainfall rarely exceeds 1,000 mm. Human population densities are consequently low (<5 inhabitants per kin2), and the principal land use is grazing on unimproved range.

In central and northern Maranh~o, elevations decline and the terrain gently undulates as one enters a broad plain comprised of geologically young (Tertiary) sediments, dissected by many of the state's major rivers. Soils are'extremely variable (Fig. 10), with an abundance of nutrient-rich alluvium, and annual rainfall ranges from 1,000 to over 2,000 mm. Agriculture (most commonly shifting cultivation) and grazing constitute the principal land uses. Outside the capital (Siio Luis), population densities in this region average >10 inhabitants per km

The babassu forests of Maranha-o are largely confined to the central and northern portions of the state, where they are associated with three broadly defined ecological zones (Fig. 11): seasonally wet savanna

(baixada); woodland, scrub, or dry savanna (cerrado); and moist or dry forest (ma ta).





10







48 46 44 42
0 I I1 1
0
ATLANTIC OCEAN


BELEM TRACUATEUA

2


PARA

44 PINDARE
MIRIM.

4 LAGO VERDE
+ BACABAL 44 CAXIAS MARANHAO *

6 -.



PARNAIASA



8- PIAUI

GOIAS
50 0 50 100 150 200 250 KILOMETERS



10




Figure 9. Maranhio and neighboring states, with locations of
research sites.











co

Pd

L
Pd




QQ

Ppd

Ppd
































Figure 10. iAajor soil groups of Maranhao, according to the Brazilian
classification system (tEIBRAPA 1976). H = Hydromorphic L Pe

















Laterite, Pd = Dystrophic Red-Yellow Podzolic soils, Pe =
















Eutrophic Red-Yellow Podzolic soils, Ppd = Plinthic PodLr Lr
















zolic soils, L = Yellow Latosol, Lr Pe= Red-Yellow Latosol,
Pe































and Q = Quartz sands. Unlabelled areas = other soil types.
Source: eBRAPA (1981).
QQ
tr

Lr
Pd
LrQ


Lr .Lr


Lr 0 100 200 km






Lr


Figure 10. Major soil groups of Maranhio, according to the Brazilian
classification system (EMBRAPA 1976). H =Hydromorphic
Laterite, Pd = Dystrophic Red-Yellow Podzolic soils, Pe= Eutrophic Red-Yellow Podzolic soils, Ppd = Printhic Podzolic soils, L = Yellow Latosol, Lr = Red-Yellow Latosol, and Q = Quartz sands. Unlabelled areas = other soil types.
Source: EMBRAPA (1981) .





12











* MANGROVE AND BEACH
VEGETATION
SEASONALLY WET
SAVANNA
WOODLAND,SCRUB,
AND DRY SAVANNA MOIST FOREST






























Figure 11. Principal ecological zones of i1aranhao.
Source: Kuhlmann (1977).





13






COVERAGE OF
BASASSU STANDS (%) < 5O
5- 33
34-67 '

>67






<'2
>e..-.




............... x .













0 00 2CCkm








Figure 12. Coverage of babassu stands in Maranhao. Source:
Anonymous (1981).





14





Seasonally wet savanna. An extensive area traversed by the lower reaches of Maranhdo's major rivers is flooded during the December-May rainy season and burned during the dry season. This low-lying area is called baixada. The predominant soils (ground-water laterites and gleys) are hydromorphic, the major vegetation type is a low savanna dominated by sedges and grasses, and the principal land use is grazing

of cattle and buffalo on unimproved range. Throughout this region, forests of babassu are generally limited to topographically elevated islands, which are also the principal sites for shifting cultivation.

Woodland, scrub, or dry savanna. The term cerrado refers to both an extensive phytogeographic region of Brazil and the region's predominant vegetation. Within Maranh~o, cerrado is characterized by annual rainfall >1500 mm, a dry season of 6 mo or longer, and well- to excessively

drained soils. The vegetation consists of deciduous woodland, scrub, and occasional savanna. The predominant land use is grazing of cattle, goats, and sheep on unimproved range. Shifting cultivation and permanent plantations of sugarcane and cotton are generally confined to the moist bottomlands, which are also the habitats occupied by babassu forests.

Moist or dry forest. This zone is characterized by moderate to high rainfall and a dry season of 6 mo. It extends across both Tropical Moist Forest and Tropical Dry Forest Life Zones (sensu Holdridge 1967). The vegetation was originally tall, diverse forest ranging from evergreen to deciduous. Due to widespread deforestation during the past 60 yr, the original forest has been virtually eliminated, and today stands





15





of babassu dominate the landscape. During the past 25 yr, rapid influx of settlers from the arid Brazilian northeast--combined with an ambitious road-building program throughout the state--have pushed human settlement deep into Tropical Moist Forest, and extensive areas are presently occupied by young stands of babassu. The principal land uses throughout the moist or dry forest zone are shifting cultivation and cattle ranching. In contrast to other ecological zones in Maranhio, ranching here is carried out primarily on improved pastures.

The palm forest. In the major ecological zones of Maranhao described above, babassu stands are generally found on the sites most intensively utilized by people. On such sites, crops are established and pastures maintained by periodic cutting and burning. These practices account for the origin of the babassu stands in Maranhao.

Prior to the arrival of Europeans, Maranhio had long been occupied

by numerous indigenous groups (described in Steward 1963). Among the dominant sedentary tribes (e.g., Tupinamba, Timbira), agriculture was generally confined to the moist bottomlands, and villages were established nearby. During the dry months, bottomland stands of babassu and

buriti (Mauritia flexuosa) palms provided an important dietary supplement. As with the more nomadic tribes, territorial rights were identified with access to these stands, which served as a source of fiber and fuel, as well as a hedge against famine. High densities of babassu were probably limited to these bottomland sites. The original nature of the far more extensive upland forests can be discerned in the remnants that





16





still exist; babassu was commonly present but invariably associated with a high diversity of hardwoods.

With the arrival of Europeans during the 1600s, the landscape of Maranhi.o changed. In response to the needs of the colonial economy, plantations of rice, cotton, and sugarcane were established on bottomlands over an extensive area of the state. Yet more than two centuries elapsed before the first historical reference was made to the babassu forests of Maranhi.o (Prazeres 1820, cited in Abreu 1940: 11), which suggests that these forests were not as extensive as they are today. With the abolition of slavery in 1888, the plantation economy collapsed afid vast areas of cultivated land were abandoned. These sites were recolonized by high-density stands of babassu. Like the Indians before them, the freed slaves and their descendants were dependent on the palm for food, fuel, and fiber. As their populations increased, agriculture

in the form of shifting cultivation moved out of the bottomlands into the mixed upland forests, and stands of babassu gradually spread across

the landscape. The size of these stands today (Fig. 12) indicates the extent and nature of human impact on the landscape of Maranh~o. The Study Region

Climate and soils. The ecological research was located primarily in the central Mearim Valley of north-central Maranhao (Fig. 9), which I shall henceforth refer to as the study region. Climate and soils in this region are extremely favorable for agriculture. Meteorological data from the region's principal town, Bacabal, show a mean annual precipitation of 157.5 cm; the 6-mo wet season (December-May) is suffi-





17





cient for rain-fed agriculture on upland sites (Fig. 13). Mean monthly temperature and relative humidity are subject to moderate variations in

response to rainfall; the former ranges from 25.7C (February) to 28.4C (November), and the latter from 67.8% (October) to 85.2% (JanuaryFebruary). Throughout the region, the terrain is flat to gently rolling

and is dissected by numerous streams that feed into the Mearim River. The predominant soil (Fig. 10) is the Eutrophic Red-Yellow Podzolic type (Brazilian classification system), roughly equivalent to a Ferric Luvisol (FAO legend), or an Ustalf or Udalf (U.S.D.A. Soil Taxonomy). This soil is characterized by high base saturation ( 50%) and relatively high fertility compared to other regional soil types; soil structure is generally excellent (EMBRAPA 1975).

Settlement. Even after the first wave of non-indigenous settlers arrived during the 1920s, the region remained thinly populated. At that

time, Babacal was a small village, named after an adjacent stand of bacaba palms (0enocarpus bacaba) that has long since disappeared. Hunting, fishing, gathering, and shifting cultivation continued to be the principal economic activities. During the 1950s and 1960s, droughts in

the Brazilian northeast and roadbuilding throughout the state brought waves of migrants into the region. Populations soared, and by 1980 the

town of Bacabal contained 43,229 inhabitants (IBGE 1981b) and was the fifth largest urban center in Maranhao. A roadbuilding program begun

during the 1960s linked the major towns of the Mearim Valley with a well-developed system of roads. Influx of settlers, enhanced access, ample rainfall, and excellent soils transformed The region into one of





18


6 0 BELEM o PINDARE MIRIM

50 o (273.9cm) (206.7cm)
50- o

40. oo
o
30"


20- *

10- .
I0 e

E 0 1 F MAIM I IIAIS1O1NID J FaMA I iIi M AI S0 N 0



<60
z BACABAL/LAGO VERDE CAXIAS
(157.5cm) (129.Ocm)
50
0

40- a

30-*
2 0





10
0
0
20*
***

0- 'AM J' JA ND JA MJ J S ON

Figure 13. Monthly rainfall at four research sites. Numbers are longterm annual means, bars are long-term monthly means, shaded
circles are monthly data for 1981, and open circles are monthly data for 1982. Period over which long-term means were calculated and source of unpublished data for each
site are: Belem, 1931-60, Centro de Pesquisa Agropecuaria do Tropico Ifmido (CPATU); Pindare Mirim, 1966-80, Superintendincia de Desenvolvimento do Nordeste (SUDENE); Bacabal,
1972-80, Instituto Nacional de Heteorologia (INMET); and
Caxias, 1966-80, INMET.





19





the most agriculturally productive areas in Maranh-ao. Yet agricultural productivity has not brought prosperity to most of the people living there.

Land tenure. The main front of agricultural expansion in Maranh&o swept westward through Bacabal during the 1950s. But the pattern by which the study region was settled can still be seen today in other areas of frontier expansion throughout Brazil (Velho 1972). As settlement pushes forward, once isolated areas become increasingly accessible, and the process of land concentration soon commences. The aftermath of this process can be seen from the data on land distribution in MaranhAo (Fig. 14). In 19,50, 61.8% of the state's occupied lands were owned by 2.0% of its households. Despite a rapid expansion of area under occupation, there had been no discernible improvement by 1975, when 41.4% of the land was controlled by 0.5% of the households. In absolute terms, the amount of land in smallholdings (<100 ha) increased by a factor of 4.0; however, the number of small-holdings increased by a factor of 5.6. Smallholdings became, on the average, smaller, and the number of landless increased.

As throughout most of Maranh&o, land tenure in the study region is now well defined. The vast majority of rural inhabitants live in latched huts strung like beads along the roads, footpaths, and watercourses that form a network throughout the region. Behind these hamlets and villages, fences are invariably present, and beyond these fences lie the resources upon which people depend.





20








2.0 LAND (%)


5 .9





~- .. .. .
922.





1950
(95,112k9) 17
4 6.1( ............







1950.......
( 95,2k9)1-7







88.4

SIZE OF LANDHOLDING

D < IOha
E] 0-100 ha

100-1000 ha 1975
(n= 494,596) 1000-10,000 ha

S> 10,000 ha



Figure 14. Distribution of land and households according to size of
landholding in rural areas of Maranhao during 1950 and
1975. Size of holding defined by effective occupation of the land. Amount of land increased from 1950 to 1975 due
to frontier expansion. Sources: IBGE 1956, 1979.





21





Shifting cultivation. Throughout Maranhio, sharecropping is the means by which the vast majority of cultivators gain access to land. Crops are raised principally by shifting cultivation (Table 1). The forest is felled by axes and burned toward the end of the dry season. Planting usually takes place in December with the first wet season rains. In the study region, the principal crops are upland rice, cassava, maize, and beans. As rice is sensitive to drought, and cassava is intolerant of prolonged waterlogging, farmers often plant both, and intercropping is the rule. Cultivated areas are frequently weeded until

the April-June rice harvest; cassava may be harvested up to 2 yr after planting. Cultivated plots on upland sites are virtually never used for

two crops in a row, so new plots are required each year. Short cropping periods make this form of land use extremely extensive. Shifting cultivation of this sort can only remain viable in areas of low population densities. Over large portions of Maranh&o, this is no longer the case.

From 1950 to 1975, land area in Maranhao utilized for temporary cropping increased threefold, while the ratio of fallowed to cropped land was cut in half (Table 1). The immediate implications of shortened fallows for crop yields--and long-term implications for site quality-are obvious. Increased population densities are the principal cause of shortened fallows throughout Maranhio. But exceptionally short fallows in the study region are a consequence of widespread conversion to pasture.





22















Table 1. Uses of occupied lands in Maranhio during 1950 and 1975.
Numbers represent ha (percent of total classified in
parentheses). Sources: IBGE 1956, 1979.


Year
Land Use 1950 1975

Grazing
Range 3,41541,400 (410.2%) 2,590,600 (21.7%)
Improved Pasture 110,800 0.5%) 1,218,200 (10.2%)

Cropping
Temporary 3141,500 (3.6%) 1,0141,100 ( 8.5%)
Permanent 14,200 (0.2%) 111,900 ( 0.4%)

Fallow 2,4107,800 (28.0%) 4,058,100 (341.0%)

Permanent Forest
Natural 2,361,700 (27.5%) 3,0141,300 (25.2%)
Planted 41,000 ( 0.0%) 1100 ( 0.0%)

Total Classified 8,597,4100 (100.0%) 11,937,600 (100.0%)

Not Classified 9111,500 471,500

Total Occupied 9,538,900 12,4109,100





23





Grazing. In contrast to the rest of Maranh&o, grazing in the study region is carried out predominantly on improved pastures. This has been a recent development; in 1950, improved pastures comprised only 0.5% of Maranhio's total occupied area (Table 1). The formation of improved pastures has enabled grazing to encroach into areas once used exclusively for shifting cultivation.

Conversion to pasture is generally carried out as part of shifting

cultivation. On upland plantations of crops such as rice and maize (but rarely cassava, due to its relatively long cropping period), seeds of the African grass, hyarrhenia ruf'a (known locally as lajeado), are introduced just prior to harvest (March-May). Use of fertilizers is rare, and legume-grass associations are apparently unknown. Pastures are usually burned during the dry season, and the burn often gets out of control; wildfires are a conspicuous feature of the landscape throughout Maranh.o.

Babassu. The study region is characterized by exceptionally high coverage of babassu. Spontaneous occurrence of the palm appears to be limited only on steep hilltops (where fruit dispersal may limit colonization) and waterlogged soils (where root respiration may be inhibited). Elsewhere the palm forest virtually blankets the landscape, except where it is thinned or clearcut by shifting cultivators and ranchers. Although clearcutting of babassu is illegal in Maranhio, the practice appears to be increasingly common, especially in areas where pasture conversion is widespread.





24





Study Sites

Lago Verde. The ecological research was centered on an 800-ha farm in the county of Lago Verde, ca. 25 km NW of Bacabal (Fig. 9). The farm is flat and dissected by several intermittent streams. Soils are generally fertile and well-drained. The predominant soil type is Eutrophic Red-Yellow Podzolic. Approximately 75% of the farm has been converted to pasture, and the rest is covered by various ages of forest fallow.

I selected three 1-ha sites for ecological studies, representative of undisturbed primary forest, old secondary forest, and annually burned pasture. The forest sites were located in an 11-ha reserve maintained for hunting and for harvest of fallen trees as a source of lumber. Each

study site measured 80 x 125 m. For orientation, wooden stakes were used to define twenty 20 x 25 m blocks; on the forested sites, the blocks were delimited by 2-in wide paths. Community characteristics of each site were obtained by measuring abundance, frequency per block, and diameter at breast height (dbh) of all trees with dbh 10 cm. Unknown species were collected and identified in the herbarium of the Museu Paraense Emilio Goeldi in Bel~m. Community characteristics of the

forested sites are summarized in Appendices A and B.

The primary forest site is probably representative of the original vegetation that once covered the region. Radiocarbon dating of charcoal fragments found at ca. 15 cm depth indicate that the site was burned 650 250 yr before present (letter dated 4 August 1983 from Richard R. Pardi, Radiocarbon Laboratory, Queens College, Flushing, New York, USA). I believe that the site has suffered no major disturbance since. Local





25





informants reported that the site was still covered by structurally complex, species-rich forest when they first arrived in the region during the 1930s. Since this time, the site has been subjected to minimal disturbance. At the time of my study, emergents attained 30-35 m, and the canopy was more or less continuous at 15-25 m. Most of the emergent species were deciduous during all or part of the dry season. Vines were relatively abundant, whereas epiphytes were rare. Species richness, at 63 species (dbh >10 cm) per ha, was lower than that of upland forests in Amazonia (Anderson and Benson 1980). With relatively high abundance (20.2%), frequency (9.2%), and dominance (27.4%), babassu was the most important species on the site.

The secondary forest was originally cut and burned for shifting cultivation in 1948; the site had suffered no major disturbance since.

Emergents attained 20-25 m, and the canopy--comprised largely of a uniform stand of babassu--was relatively continuous at 10-15 m. Species

richness was 31 species (dbh >10 cm) per ha. The percentage of total species that occurred in both the primary and secondary forest sites (coefficient of community: Whittaker 1975) was 31.9%. In abundance (66.7%), frequency (21.7%), and dominance (84.1%), babassu was far and away the most important species on the site.

The pasture was established through introduction in 1974 of the exotic grass, Hyparrhenia rufa, on a site that had already undergone several cycles of shifting cultivation. The abundant understory competitors (including Lantana camara, Stachytarpheta cajanenesis, Sida spp.,





26





Cyperus spp., and juveniles of babassu) were annually cut and burned. The 10-15 m high canopy was comprised exclusively of babassu, which occurred at a density of ca. 100 stems per ha.

Satellite sites. In addition to the principal field site at Lago

Verde, satellite sites were established at Bel~m, Pindar6 Mirim, and Caxias (Fig. 7). The Bel6m site was used for experiments involving establishment (Chapter 4); the Pindar6 Mirim and Caxias sites were used for phenological observations (Chapter 3). Meteorological data (Fig. 11) show a wide range of rainfall at these sites. Information concerning land uses and experimental design at each site is provided in the appropriate chapters.













CHAPTER 2

TAXONOMY AND PHYTOGEOGRAPHY

Babassu belongs to the cocosoid group, one of 15 major groups within the palm family (Moore 1973b). Cocosoid palms are distinguished from the other groups by their fruits, which contain a thick, generally three-pored endocarp. The principal center of cocosoid diversity is South America, where 24 of 28 genera occur (Moore 1973b). Two wellknown representatives of the group are the coconut (Cocos nucifera) and the African oil palm (Elaeis guineensis).

The taxonomy of babassu has been a source of confusion since the palm was first described well over a century ago. This confusion begins on the generic level. Although babassu is generally assigned to the genus Orbignva, the status of this and the four to five other genera in the Attalea alliance has been questioned by Wessels Boer (1965). These genera are distinguished solely on the basis of staminate flower morphology. No correlation with other characters has been observed, and staminate flower morphology runs across a separation based on other characters such as endocarp pores. Lines between genera in the Attalea alliance are further blurred by collections of intermediate staminate flower types, which have prompted the description of two new genera; one of these, Markleva, was described by Bondar (1957) as a possible intergeneric hybrid involving babassu.

A classification based solely on staminate flower morphology is of little use in the field, as most species flower only for a short time.



27





28




Thus it is no surprise that babassu is often confused with other species and even genera, notably Attalea. Oil palms commercially referred to as babassu include Orbignva agrestis in Pars, Attalea oleifera in Goias and Minas Gerais, A. eraensis in Minas Gerais, and A. Dindobassu in Bahia (Markley 1971). Botanists are not immune to generic confusion. Aboriginal dispersal was evoked to account for the disjunct occurrence of babassu in the state of Sdo Paulo (Ferri 1974, 1980), although the population has been identified as pertaining to Attalea (J. T. de M. Costa, pers. comm.).

To resolve intergeneric confusion in the Attalea alliance, Wessels Boer (1965) recommended reducing Orbignva and related genera to the single genus Attalea, but his suggestion has not been followed by other contemporary taxonomists (Moore 1973b, Glassman 1977). Countervailing arguments include the weight of taxonomic tradition, and the observation that Wessels Boer's scheme merely reduces the confusion to the specific level (H. E. Moore, pers. comm.). Due to the uncertain limits between genera, the possible occurrence of intergeneric hybrids, and the practical difficulty of distinguishing genera in the field, I believe that the current splitting of the Attalea alliance into five genera (Moore 1973b) is not justified. Final resolution of the confused generic limits within the Attalea alliance will require exhaustive study of the entire group. As such a study is not on the horizon at present, I shall follow taxonomic tradition and consider babassu within the genus Orbignva.





29




On the specific level, at least 10 names associated with babassu have been published within the genus Orbignva (Table 2). Most of this proliferation has resulted from species descriptions based on inadequate botanical specimens, a problem inherent to palm taxonomy in general (Tomlinson 1979). In the more recent taxonomic literature (e.g. Rizzini 1963; Bondar 19641; Glassman 1977, 1978), most of the names listed in Table 2 are considered synonymous, and it is generally agreed that the babassu complex consists of a maximum of three species. Due to

potential confusion among these species, I embarked on a taxonomic study of the babassu complex. My objectives were (1) to identify the species that was the subject of the ecological studies described in the following chapters and (2) to distinguish it from morphologically similar species.

Methods

Lack of adequate specimens has been the principal impediment to resolving the taxonomy of the babassu complex. To remedy this situation, I collected representatives of the complex over a substantial portion of its presently known range (Fig. 15). Due to the unusual structure and bulkiness of large palms such as those in the babassu complex, complete botanical collections require extensive field documentation; recommended procedures are described at length in Balick et al. (1982). Specimens of all collections were deposited at the Museu Paraense Emilio Goeldi in Bel~m, Brazil; duplicates were sent to other Brazilian institutions (CENARGEN, INPA, CPATU) and the New York Botanical Garden, Bronx, N.Y., U.S.A.







30






L, c c c


cii 61 il







41 41 41 41

















cd c; t o
61



.11
cd ol 1 .2 2

44
0





61 61 c;l
61 61 cd


Z 7 61 41 61 c;l 1. c
ca







C4 61 61



Cd
E

I o o .






lo Ac

im
r


l
Si. m





31












80 70 60 50 40
I I I I
444

10
o".. ,:GUYANA ,0
r SURINAME
t- FRENCH
COLOMBIA \ <*-- GUIANA


0
ECUADOR /.







0I BRAZl L -.-o
20-0 0' -"



SBOLIVIA 1


20-V
:i ', .-., --.
s -,,. 2
102





SARGENTINA
20I I I I
80 70 60 50 40



Figure 15. Distribution of specimens of Orbignya martiana and its
suspected hybrid complexes. Circles represent 0. martiana; squares represent the Orbignya-Maximiliana complex; triangles represent the 0. martiana-0. eichleri
complex. Sources: Wessels Boer (1965), Glassman (1977),
and M. J. Balick and A. B. Anderson (unpublished).





32




A first-hand familiarity with the taxa in the field provided a sound basis for reviewing the extensive and often contradictory literature on nomenclature. Results pertaining to the morphology and distribution of taxa in the babassu complex represent a synthesis of data already published, data obtained in the field, and morphological comparisons of specimens in the laboratory.

Results and Discussion

Examination of data on specimens obtained over a substantial area indicates that the babassu complex is comprised of one highly variable, widely distributed species and two associated species that appear to hybridize with the former. Babassu is henceforth used exclusively in reference to the principal species of the complex, which was the subject of the ecological research described in the following chapters. Nomenclature

The scientific name of babassu tentatively utilized in this study is Orbiznva martiana Barb. Rodr. (Table 2). Its valid publication by Barbosa Rodrigues (1898) represents the first taxonomic description that can be definitely attributed to babassu. Previously published names associated with babassu rAttalea speciosa (Mart.) Barb. Rodr. and Orbiinva lvdiae Drude] were based on incomplete collections, and their accompanying descriptions were either insufficient or incorrect; neither can be definitely attributed to babassu. Barbosa Rodrigues (1903) subsequently reduced O. martiana to synonymy under the new combination, O. speciosa (Mart.) Barb. Rodr. Although widely accepted in Brazil, the latter name is invalid because the same combination (Q_ speciosa Barb.





33




Rodr.) had been previously published by Barbosa Rodrigues (1891) in reference to another species (cf. Article 64 International Code of Botanical Nomenclature: Stafleu 1972, p. 59). Orbignva martiana is therefore the earliest, validly published name definitely referable to this taxon and it maintains precedence over subsequently published names associated with babassu (Table 2).

My use of 0. martiana is tentative, due to an older name that remains to be considered. Martius (1844) published a relatively complete description of D phalerata Mart., which he designated as the type species of the genus. Subsequent collections of this species have apparently not been obtained (Glassman 1977), and Martius' published description and illustration provide no basis for distinguishing it from 0. martiana. The lack of specimens of 0. chalerata prompted a recent (July-August 1982) expedition, sponsored by the New York Botanical Garden, to the type locality in eastern Bolivia. Final judgment

concerning the relationship of these two taxa awaits examination of the

specimens obtained.

Description and Differentiation

A technical description of babassu, based on specimens obtained to date, is provided in Appendix C. This description illustrates the extreme variability of the species. Despite its variability, babassu possesses a number of diagnostic features that together enable one to distinguish it from related species. These features include an erect, solitary habit; large, erect leaves recurved at the apex, with rigid leaflets disposed in a vertical plane and often with striations on the





34




outer (abaxial) surface of the sheath; large inflorescences that are either exclusively staminate or hermaphroditic, each containing a persistent, woody bract with a prominent apex; staminate inflorescences with numerous, erect branches (rachillae), each branch bearing two to four rows of staminate flowers on its lower (abaxial) surface only; staminate flowers with a small calyx, two or rarely three dentate-tipped petals, approximately 24 irregularly coiled and twisted stamens, and a pistillode; hermaphroditic inflorescences with numerous branches, each branch bearing one to rarely three pistillate flowers at the base and a variable number of staminate flowers at the tip, the latter either inviable or opening after the pistillate flowers; and a massive, oblong drupe with a thick, woody indehiscent endocarp and usually two to five oily seeds.

Due to its apparent propensity to hybridize, the distinction between babassu and at least two closely related species (Orbi~nva eichleri and Maximiliana maria) is somewhat blurred. I suspect that these latter species hybridize with babassu. Morphological comparisons (Appendices D and E) illustrate the intermediate morphology of the

purported hybrids ("Orbi~nva teixeirana" and "Markleva dahlgreniana"). Germination experiments have shown that both of the latter produce viable seeds and thus are capable of regenerating; however, they are only known to occur in association with their apparent progenitors. The distribution of babassu and its purported hybrids is shown in Figure 15. Anatomical, palynological, and electrophoretic studies of these complexes have been initiated by colleagues at the New York Botanical





35




Garden in the U.S. and the Centro Nacional de Recursos Gengticos (CENARGEN) in Brazil.

Distribution and Origin

Babassu has a widespread distribution in South America (Fig. 15). Dissemination of the large, heavy fruits has probably been exclusively by mammals (either living or extinct) or water. Among living mammals, rodents such as pacas (Agouti Daca) and agoutis (DasvDrocta Dunctata) are effective, short-range dispersal agents (see Chapter 3); humans and possibly monkeys disseminate the fruits over greater distances. Coevolution between hard-fruited palms such as babassu and large, Pleistocene mammals seems likely (Janzen and Martin 1982), but a paucity of paleontological evidence makes it impossible to assess the potential contribution of these now extinct animals to the current distribution of babassu.

Markley (1971) argued convincingly that the babassu palm originated on the Brazilian Shield in the state of Goias, a major center of diversity for the entire Attalea alliance. Since its origin during the Precambrian, the Brazilian Shield has undergone considerable erosion by numerous river systems. With few exceptions (discussed below), babassu areas from the Paraguay River to the Amazon Basin are connected by watersheds to the Brazilian Shield in Goi~s (Figs. 1 and 15). Throughout its range, babassu is largely confined to floodplains, river valleys, and former deltas, which lends support to the idea of dissemination by rivers. Although their high specific gravity (ca. 1.4 g fresh weight/cm3) prevents the fruits from floating, they can be





36




carried considerable distances by runoff' and rivers in flood. Lighter fruits would tend to be carried farther, which could explain why babassu fruits are relatively light (80-250 g fresh mass) over most of the

species' range, while the heaviest fruits known (up to 800 g fresh mass) are found in its presumed center of origin in Goias (Fonseca 19211; M. J. Balick and A. B. Anderson, unpublished data).

There are three important exceptions to the links between the

Brazilian Shield and babassu's present range. The first is its disjunct occurrence in Suriname and Guyana (Fig. 15). Until the early Pleistocene, the eastern portions of the Brazilian and Guayana Shields were connected, which probably accounts for the close floristic ties between the two regions (Maguire 1970). The Guianas represent a second center of diversity of the entire Attalea alliance (Wessels Boer 1965).

A second exception is babassu's occurrence in the Brazilian states of Maranhao and Piaui' (Fig. 1). Markley (1971) suggested that the Tocantins River may have once served as a link for dissemination of the palm to these states. The Tocantins has its headwaters in the Brazilian Shield and presently drains into the Amazon River. During the

Cretaceous, however, it may have flowed across the northern part of Maranhao and built up the huge delta that presently covers most of the northern third of the state. It seems unlikely that this delta could have been built by the relatively insignificant rivers of Maranhao that exist today (Pindar6, Grajau, Mearim, Itapecuru, and Parnai"ba), most of which have their headwaters only a short distance from the delta. Uplift of the Serra do Gurupfi in western Maranhao (probably during the





37




lower Tertiary) may have forced the Tocantins to change to its present course.

A final anomaly in babassu's distribution is its disjunct occurrence in the Brazilian state of Ceari (Fig. 1). Here the palm is confined to isolated and geologically very old uplifts, and there are no river systems that run between these structures and the highlands of Goigs. The uplifted areas to which babassu is confined in Ceard have long served as favorable sites for human settlement and agriculture, due to their relatively cool temperatures, high rainfall, and fertile soils. Markley (1971) argued that babassu's occurrence in Cearg is due to dissemination by humans. Early aboriginal groups may have introduced fruits obtained from the Parnafba valley ca. 200-300 km distant.















CHAPTER 3
REPRODUCTIVE BIOLOGY


In autecological studies of plants, the various components of

reproduction (e.g., phenology, pollination, dispersal, predation, germination) are often analyzed as separate entities. Yet these events are

obviously interdependent. The timing of flowering, for example, may be related to competition for animal pollinators (e.g., Rathcke in press). Tight interdependence between the often conflicting demands of seed dispersal, escape from seed predators, germination, and ultimately establishment has been repeatedly documented (e.g., Stebbins 1971, Harper 1977, Silvertown 1981, Willson 1983).

This chapter commences the life history of babassu with an analysis of the palm's reproduction under a range of ecological conditions. Thus, comparative phenological observations were conducted in the major ecological zones of Maranhao where babassu forests occur. The widespread changes in land uses currently underway in Maranhao are likely to have a profound impact on the reproductive biology of local populations of the palm (e.g., see Moore 1977, Janzen 1978a). I therefore carried out studies of babassuls phenology, floral biology, dispersal, predation, germination, and seed fate under a range of land uses associated with babassu.







38





39





Methods

Phenology

I carried out monthly phenological observations of babassu from January 1981 to March 1982. All adult palms at each site were labelled and mapped. On the intensive study sites at Lago Verde, the sample consisted of all adults in a 0.5-ha area; in each of the satellite sites (Pindar6 Mirim and Caxias), 60 adults were randomly selected for observation. The former site had been recently converted to pasture, whereas

the latter was in fallow following its use for shifting cultivation in 1979-80. These sites are representative of current management practices

of babassu stands in the seasonally wet savanna (Pindar Mirim) and dry woodland (Caxias) ecological zones of Maranhao. On the site at Pindar6 Mirim, leaves of babassu had been harvested for thatch ca. 2 yr prior to initiation of the study.

To permit observations of leaf production and loss, the palms were climbed and green leaves were labelled by spraying paint on the undersides of the rachises. To enhance the visibility of newly emerged leaves, ca. 10 basal leaflets per side (representing ca. 5% of the total) were removed from marked leaves. I calculated leaf production per month by counting the newly formed (unlabelled) leaves that had completely expanded in the crown; loss per month was determined by counting the number of labelled leaves that had fallen to the ground. Fallen leaves were cut in half after being counted so as to prevent being recounted in subsequent months, and the leaves of adjacent palms were




40





painted with different colors and/or patterns to avoid confusing the sources of fallen leaves.

As in other palms, the reproductive structures (inflorescences and infructescences) in babassu are relatively few, large, and permanent. In addition to obtaining separate data on each structure, I could thus deduce its phenological activity during the ca. 1-mo interval since the

last observation. To prevent confusion among several flowering and fruiting structures borne simultaneously on a palm, phenological observations were accompanied by a measurement of the compass bearing of each structure in relation to the palm stem. In the case of functionally female inflorescences (containing at least one pistillate flower and a

variable number of staminate flowers that were usually inviable), I distinguished the following categories: (1) inflorescence in flower (pistillate flowers yellow); (2) infructescence with immature fruits (fruits tan to brown and not falling); (3) infructescence with mature fruits (fruits falling); and (4) infructescence with no fruits (all fruits fallen). In the case of male inflorescences (containing exclusively staminate flowers), I distinguished the following categories:

(1) inflorescence in flower (flowers intact or falling); (2) inflorescence in flower during the past month (flowers fallen, inflorescence axes yellow to light tan); and (3) inflorescence in flower prior to past

month (inflorescence axes tan).

On the three study sites at Lago Verde, fruit production per month was measured from September 1980 to April 1982. Mature infructescences were harvested, and the number of fruits produced per infructescence was





41





determined by counting fruits and/or calyces (in the case of abscised fruits). A subsample of ca. 30 fruits from each infructescence was airdried and weighed to the nearest gram.

Considerable variability in the ratio of male to female inflorescences (sex ratio) was observed among the palms in the phenological study. To compare sex ratio variability on the population level, I measured the distributions of inflorescence types in six populations. Distributions were obtained by determining the percentage of each inflorescence type per palm. The inflorescences tend to remain attached

for several years following anthesis; my counts thus provided a longterm picture of reproductive behavior. Data were obtained from two geographically isolated populations located in Tocantin6polis, Goias, and Pirapora, Minas Gerais. Additional data were obtained from four populations located within my phenological study sites at Lago Verde, Pindar4 Mirim, and Caxias; all of the latter sites occur within the more or less continuous band of babassu stands that stretches across northcentral Maranhao (Fig. 12).

Floral Biology

To determine duration of flowering in individual inflorescences, I carried out daily phenological observations from 1-15 March 1982 on 38

palms on the secondary forest site and 15 palms on the pasture site. Palms with bracts about to open were selected for observation. Flowering was judged to begin upon opening of the bract. Proportion of

flowers that ultimately became fruits was determined through examination





42





of a random sample of 10 infructescences obtained in each of the two habitats.

Studying pollination in babassu required recognition of female inflorescences before they opened. As no morphological differences could be discerned between the two inflorescence types prior to anthesisq I distinguished them by opening a small slit in the bract and

examining the flowers within. In female infloreseences, the slit was subsequently sealed with tape to prevent contamination.

To determine whether self-pollination occurs in babassu, sleeves were designed to exclude external sources of pollen from female inflorescences. Each sleeve was made of light-weight canvas in the form of a tapered cylinder with a length of ca. 2.5 m and a maximum diameter of ca. 0.5 m. Circular wires were sewn on the inside of the sleeve to provide -support, and a small polyethylene window was placed in the middle to allow observation of the inflorescence. Entry by insects was prevented by painting bands of insect trapping adhesive over the base and tip of the inflorescence bract; one end of the sleeve was placed over each band and securely shut with wire.

Small bags were designed to isolate individual pistillate flowers from pollen sources. To reduce chances of contamination, the entire inflorescence was covered with a sleeve prior to opening. Each flower was individually bagged after the inflorescence had opened, and the sleeves were subsequently removed. Time of pollination was estimated by

exposing pistillate flowers to external sources of pollen at different periods following opening of the bract. To test whether parthenogenesis





43





occurs in babassu, flowers were left bagged throughout the flowering period, and I removed the pistils from others; both sets of flowers were subsequently checked for fruit set.

Abundance and behavior of insect visitors to both inflorescence types were determined by direct observation and collection. Sticky traps were used to monitor insects between observations. Insect specimens were identified at the Museu Paraense Emiltio Goeldi in Bel6m; unidentified specimens were sent to specialists in Brazil and the U.S.

To determine whether wind pollination occurs in babassu, sleeves were designed to exclude insect visitors from female inflorescences while permitting penetration of wind-borne pollen. The sleeves were made out of nylon netting with a ca. 0.5-mm mesh, sufficient to exclude all observed insect visitors. Dispersal of pollen by wind was monitored through the use of pollen traps, which were distributed at regular intervals on both the secondary forest and pasture sites. The trap consisted of a microscope slide covered with a thin coat of vaseline; the slide was placed inside a horizontally oriented section of 15-cmwide PVC pipe secured to a stake or stem. After a 1 -wk exposure, the slides were removed, covered, and subsequently examined in the laboratory.

Fruit Biology

Dispersal. I monitored removal of babassu fruits in the pasture and secondary forest, as well as on a nearby field. This latter site had been cut, burned, and planted at the end of 1980. During the course of the study, the site remained virtually abandoned except for sporadic





44





harvesting of cassava, and dense regrowth became established. At each of the three sites, 10 one-layered clusters of 25 fruits each were set up, with a stake marking the center of each custer. Fruit movement was measured at 1- to 2-wk intervals. The search for dispersed fruits was conducted to a distance of 4 m from the stake. To label fruits, a 14-mm staple was driven into the apex of each fruit. To eliminate human odor, fruits were handled with gloves that had been thoroughly rubbed with leaves and soil. I tested potential effects of the labelling technique

on dispersal agents by setting up four pairs of fruit clusters (25 fruits per cluster) on the secondary forest site. Dispersal was arbitrarily defined as movement of more than 50 cm. No significant difference (p = .19) was found in dispersal of labelled and unlabelled fruits, according to the randomization test for matched pairs (Siegel 1956).

Predation. To monitor colonization by seed-predaceous beetles, six

fruit clusters were set up on the secondary forest site. I prevented dispersal by stapling the apices of the fruits to a nylon line (10

fruits per line) attached to a stake. Twenty fruits were gathered from each cluster at intervals of 0, 109 209 /09 80, and 160 d following initiation of the experiment (27 October 1981). The fruits gathered from each cluster were isolated in closed sacks to prevent further predator colonization. After air drying, the fruits were opened to determine total number of seeds, number of attacked seeds, and number of beetle larvae per fruit.

Seed fate. To monitor the fate of seeds over time under various environmental conditions, five fruit clusters were set up in each of





45





three sites: secondary forest, pasture, and field. I prevented dispersal by stapling the fruits to nylon lines attached to a stake. Twenty fruits per pile were gathered at the outset of the experiments (27 October 1981) and at 3-mo intervals during 1 yr.

I obtained a long-term (ca. 18 mo) measure of seed fate from fruits used in the irrigation experiment (described in Chapter 4). At the end of this experiment, 20 fruits were gathered from each of five weeded (sun) and five non-weeded (shade) plots that had not been irrigated. In both seed fate experiments, gathered fruits were stored in closed sacks,

air-dried, and opened to determine numbers of eaten, defective (aborted or rotted), germinated, and quiescent (viable but not germinated) seeds.

Germination. An experiment was set up on the secondary forest site to monitor germination over time and to test for between-palm variability in fruit germination. Recently f allen f ruits (100 f rom each of 10 palms) were dehusked to inhibit dispersal and planted in 40 Plots (25 fruits per plot, four plots per palm). At ca. 1-mo intervals, I measured germination by checking the base of each fruit for emergence of one or more embryonic axes (i.e., plumule, cotyledonary node, and radicle).

I conducted additional experiments to test the effects of fire, herbivory, and shading on germination of babassu fruits. These experiments were established on a cleared, periodically weeded site ca. 50 m from secondary forest on well-drained soil. The fruits used in each of these experiments (100 per treatment) were obtained from a single palm.

Unless otherwise specified, fruits were dehusked and placed on plots shaded with suspended frames upon which babassu leaves were tied. At 1-





46





wk intervals, the frames were lifted to check for germination. To test the effects of fire, I exposed fruits to each of the following treatments: (1) a prolonged, high-intensity fire, simulating a forest fallow burn; (2) a rapid, low-intensity fire, simulating a pasture burn; and

(3) no fire. To test the effect of consumption of the fruit mesocarp by

herbivores, I compared germination in fruits that had been dehusked (epicarp and mesocarp removed, simulating the effect of herbivory) and fruits that had not been dehusked. To test the effect of shading, I planted 100 fruits on an exposed plot and 100 fruits on a plot shaded as

described as above.


Results

Phenology

Leaf turnover. Figure 16 shows data on leaf turnover of babassu at the wet (Pindar Mirim), intermediate (Lago Verde), and dry (Gaxias) sites; similar data from the primary forest, secondary forest, and pasture sites are shown in Figure 17. At all sites, leaf production peaked during the two rainy seasons and leaf loss peaked during the single dry season of observations. As in palms generally, leaf production in babassu is synonymous with growth; the rainy season is, therefore, the growing season for the palm. Over a 1-yr period after initiation of observations, no difference was detected among leaf production rates per palm at the wet (_x = 3-9) intermediate (5i = 4.1 ), and dry (_x = 3.7) sites. By contrast, Duncan's Multiple Range test revealed significant differences (p :S.05) in rates of leaf production per palm at the three Lago Verde sites; the order was pasture (Rc = 5.1) > primary forest




47







1981 1982J F M A MJ J A S O N D J FM
+1.5 WET SITE
+ 1.0 (N= 60)
+0.5

O _0 0 5.
-0.5

U + 1.5 INTERMEDIATE SITE
+1.0 (N= 143

Z +0.5


o -0.5
o 1.0H
0
g -1.5
CL
+1.5 DRY SITE
U- +1.0 (N 60)
< +0.5


-0.5
-1.0


Figure 16. Mean monthly gain and loss of leaves per palm on three
sites in Maranhao. Wet site = Pindare Mirim, intermediate
site = Lago Verde, and dry site = Caxias. N= number of
palms observed per site. ND = no data.





48






1981 1 1982J F M A M J J A S 0 N D J F M

+1.5 -PRIMARY FOREST
+1.0 (N=21)
+0.5
+ 0 N

-< -0.5
-1.0

X -I.5
Ll.
1 +1.5 SECONDARY FOREST

+1.0 N=75)
O +0.5
O = -i = -=
o -0.5
0 -1.0
CL -I.5
PASTURE
La_ +1.0 -(N= 44)
+0-5- 1 .0




Figure 17. Mean monthly gain and loss of leaves per palm on three
study sites at Lago Verde. N = number of palms observed
per site. ND = no data.





49






4-4) > secondary forest (R = 3.7). These results can probably be attributed to differential shading of adult palms on the three sites. Shading was lowest in the pasture. Potentially high shading in the primary forest may have been reduced by the relatively pronounced canopy

roughness (which thus permitted greater lateral penetration of light) and the emergent status of many of the adult palms. In the secondary forest, shading was highest because the dense, uniform babassu stand formed a closed, interlocking canopy.

Except at the wet site, leaf production and loss rates were approximately equal after 1 yr of observations; net gain for this period averaged .05 leaves per palm, which is virtually steady state. Assuming steady state conditions, one can divide the number of leaves borne per palm (R = 18.2, n = 199) by annual production per palm (R = 4.1 n = 199) to obtain an average leaf lifespan of 4.4 yr. At the wet site, production rates were far greater than loss rates; the net gain per year was 3.3 leaves per tree. The low losses were probably due to leaf harvesting for thatch 2 yr prior to initiation of the study. Yet gross leaf production (X_ = 3-9) was not significantly different from the combined mean obtained from the other sites (3Z = 4.1 ). I thus suspect that the single defoliation had little effect on leaf production rate at the wet site.

Flowering and fruiting. Data on inflorescence production for both

sets of sites are shown in Figures 18 and 19. Like leaf production, flowering of babassu peaked during the rainy season and was lowest during the dry season. The similarity of inflorescence and leaf Dheno-





50






1981 '1982J FMAMJ J AS ONDJ FM

0.5 WET SITE
(N= 60)
-j 0 o U


C 0.5


1.0
z
0 0.5 INTERMEDIATE SITE
(N= 143)

o 0-L

0.5


Z 1.0
w
S0.5 DRY SITE
(N=60)


0.5



S1.0-


Figure 18. Mean monthly production of male and functionally female
inflorescences per palm at three sites in Maranh~o. Wet
site = Pindar& Mirim, intermediate site = Lago Verde,
and dry site = Caxias. N = number of palms observed per
site.





51





1982J F MA MJ J AS 0 N D J FM
0.5 PRIMARY FOREST
(N=24)



S0.5


w .
CL
0.5 SECONDARY FOREST
z O -(N=75)
0 0



0 0.5
CL
1.0
w
C.)
z 1.5

c 0.5 PASTURE
w
O : 0(N:44) 7
O -- t-
U-H
z
0.5


1.0



Figure 19. Mean monthly production of male and functionally female
inflorescences per palm at three study sites at Lago
Verde. N = number of palms observed per site.






52






logies is probably due to the physical association of' the two structures; each inflorescence is attached to the inner (adaxial) surface of' a leaf' sheath. However, a ca. 1-yr lag occurred between a leaf's em ergence in the crown and the opening of' its associated inflorescence. Earlier inflorescence expansion was apparently prohibited by the tight packing of' newly emerged leaves near the top of' the crown. Only continued growth of' the palm during the following rainy season loosened this packing sufficiently to permit inflorescence expansion.

Dissection of' crowns showed that each leaf' base contains a single inflorescence. Under suitable conditions, all inf'lorescences ultimately

develop to anthesis, thus resulting in an inf'lorescence-to-leaf' production ratio of' 1. Production data over the entire period of' observations revealed high ratios f'or the dry site at Caxias (1.02) and the pasture at Lago Verde (1.04). (The possibility of' obtaining inflorescence-toleaf' production ratios greater than 1 results from counting these structures simultaneously. A more accurate ratio would be obtained by comparing production of' leaves in a given year and production of' their corresponding inflorescences during the following year.) Shading probably caused intermediate ratios on the primary forest (.92) and secondary forest (.89) sites at Lago Verde. The low ratio at the wet site (.69) can be attributed to the leaf' harvest 2 yr previously. Babassu's apparent response to defoliation was to increase inflorescence abortion rather than to decrease leaf' production.

No difference in the phenology of' male and female inf'lorescences was detected. Approximately 20% of' total inf'lorescences were female in





53






all populations observed, despite high variability (0-74%) among individual palms.

Variability on the population level was analyzed by comparing the distributions of female inflorescences per palm in six populations (Fig. 20). Despite a wide range of ecological conditions, the four contiguous populations (forest and pasture at Lago Verde, wet site at Pindarg Mirim, and dry site at Caxias) showed no differences in distributions of female inflorescences per palm, according to the Kolmogorov-Smirnov Test (Siegel 1956). However, distributions in the two disjunct populations (northerly site in Goigs and southerly site in Minas Gerais) were significantly different from the first four populations, as well as from each other (p .01 in all cases). These results suggest that the distribution of female inflorescences per palm in populations of babassu is governed by genetic factors. Potential implications of this finding are

discussed in Chapter 7.

Fruit production. At all sites, frequencies of immature infructescences peaked as flowering ended in April-June and were lowest when fruiting peaked in October-December (Figs. 21-22). This pattern was consistent despite the wide range of climatic conditions among sites (Fig. 13). According to phenological observations, there was a ca. 9-mo lag between the flowering peak (January-March, Figs. 18-19) and the fruiting peak (October-December, Figs. 23-24). Observations on individual palms confirmed that fruits require an average of 9 mc to reach maturity.





54





50
FOREST PASTURE
40 (N=75) (N=41)

3020

I0
0-

50 WET SITE DRY SITE
40 (N=60) (N=60)

30Cn 2010

CL 0 I I I I i F1

50
NORTHERLY SITE SOUTHERLY SITE
40 (N=25) (N=25)
30

20

10-

0 I 1 7
0 20 40 60 80 100 0 20 40 60 80 100
FEMALE INFLORESCENCES
PER PALM (%)


Figure 20. Distributions of female inflorescences per palm in six
populations of babassu. Forest and pasture sites = Lago
Verde; wet site = Pindar Mirim; dry site = Caxias;
northerly site = Tocantin6polis, Goias; and southerly site = Pirapora, Minas Gerais. N = number of palms observed per site.





55







1981 1982J F M A M J J A S 0 N D J FM
S0.5 WET SITE
-"(N=60)

cr 0


0.5

1. 0.5 INTERMEDIATE SITE
-(N= 143)
O---=--- -=-----------0 -'- -- 7-rw z


0.5
z Li


w
O 1.0
Lii
w 0.5 DRY SITE
pt(N=60)

LL 0per palm at three sites in Maranho. Wet site = Pindar
.Mirim, int numbermediate site = Lago Verde, and dry site
Caxias. N = number of palms observed per site.





56






1981 I 1982J F M A M J J A SONDJ FM
-" 0.5 PRIMARY FOREST
I

0
i~~~o.~ ~ LllIU iTDLI


n 0.5







1.0
w 0. SECONDARY FOREST









a 0.5 -PASTURE
S(N=4475)
0
U)
L w






S0.5
1.0 E
w 0.5 PASTURE
(N=44)

z

w0.5

1.0


1.5



Figure 22. Mean monthly number of immature and mature inflorescences
per palm at three study sites at Lago Verde. N = number
of palms observed per site.





57







400- PRIMARY FOREST

200



40SECONDARY FOREST 400
)- 200-


a 0- rf l 1 r
0

W800
80 PASTURE

- 6005

IL 400- -200
20 1 F

o0 n n ] l l
J F M A M J JASONDJ FMA 1981 1982


Figure 23. Monthly fruit production per ha on three study sites
at Lago Verde. Values represent air-dried mass.




58







8 -PRIMARY FOREST
7 -(N=24)
6

-~ 4


I F

c0
W SECONDARY FOREST
a. 3 (N = 75)


0
S9
o PASTURE
0 8 (N=44)
a: 7
6
F- 5


2





1981 1982

Figue 2. Man mnthy fuitproduction per palm on three study
stsat Lago Verde. Values are air-dried mass. N
numbr ofpalm persite.






59





Fruit yields per ha and per palm (Table 3) were highest in the pasture,, probably due to the moderate density and relative openness of' the babassu stand on this site. Per-ha yields were lowest in the primary forest and intermediate in the secondary forest. This is probably due to differences in palm densities, which were low in the primary forest (48 mature palms/ha) and high in the-secondary forest (150 mature palms/ha). Conversely, per-palm yields were intermediate in the primary forest and low in the secondary forest, probably due to increased shading of mature palms on the latter site.

All three sites showed substantial increases in yields between the 1980-81 and 1981 -82 fruiting seasons. On all sites the principal source of these higher yeilds was an increase in the number of palms that bore fruit; increased numbers of infructescences per fruiting palm constituted the second most important source (Table 4). Floral Biology

Fruit set. In the secondary forest, 78.4+% (standard deviation= 24.3%) of flowers became fruits, 19.9% (24.3%) either aborted or were not fertilized, and 1.7% (1.7%) were apparently attacked by the larvae of an unidentified weevil. In the pasture, the figures were 82.1% (20.7%)v 17.1% (18.8%), and 0.8% (2.1%) respectively. These numbers indicate high breeding success of babassu in both habitats.

Parthenogenesis. Fruit set consistently failed to occur in flowers that were not exposed to pollen sources (n = 31 on three inflorescences), as well as in flowers in which the stigmas had been removed





60



















Table 3. Fruit yields of babassu per ha and per palm during 1980-81 and 1981-82 on three sites. Values are
oven-dried mass in kg.




SITE PER HECTARE PER PALM
1980-81 1981-82 1980-81 1081-82

Primary Forest 4l73 732 9.9 15.2


Secondary Forest 1328 1650 8.8 10.9


Pasture 1356 2866 15.4 32.6





61
















Table 4. Sources of increased fruit yields of babassu per
ha between 1980-81 and 1981-82 on three sites.
Numbers represent percent of total increase in
yields.



SOURCE S I T E
Primary Forest Secondary Forest Pasture Yield per Infructescence -6.5 -26.5 +12.3


Number of Infructescences +39.4 +35.6 +24.7
per Fruiting Palm



SUBTOTAL +32.9 +9.1 +37.0
(Yield per Fruiting Palm)


Number of Fruiting Palm +67.1 +90.9 +63.0


TOTAL 100.0 100.0 100.0





62





(n = 8 on one inflorescence). These results indicate that parthenogenesis does not occur in babassu.

Self-pollination. Inflorescences in babassu are generally either male (containing exclusively staminate flowers) or female (containing at least one pistillate flower and a variable number of staminate flowers that are usually inviable). The former are far more numerous than the latter. During a 3-wk period in March 1981, 21 (12%) of 173 inflorescences examined contained pistillate flowers. Of these 21 inflorescences, only three (14%) contained staminate flowers that shed pollen and thus appeared to be viable. Consequently, most inflorescences that contain pistillate flowers are exclusively female in a functional sense.

In a pistillate flower bagging experiment, fruit set occurred only in flowers exposed 0-48 h after opening of the inflorescence; subsequently exposed flowers did not set fruit. Thus pistillate flowers appear to be receptive for a maximum of 48 h after inflorescences open. In the three female inflorescences with functional staminate flowers that I observed, pollen release invariably began more than 48 h after opening of the inflorescence. In one of these inflorescences, fruit set did not occur when external sources of pollen were excluded by bagging. Thus pollen transfer from viable staminate flowers to receptive pistillate flowers on the same inflorescence appears to be prevented by protogyny.

Pollen transfer from male inflorescences to receptive female inflorescences on the same palm likewise appears to be rare. On a given palm, simultaneous opening of the two inflorescence types is unlikely.





63





In 21 female inflorescences observed, overlap of receptivity (0-48 h after bract opening) with pollen release from a male inflorescence on the same palm occurred only once.

The possibilities of inducing self-pollination in babassu are unknown. As inflorescences containing pistillate flowers are either functionally female or (more rarely) protogynous--and opportunities for pollen transfer between inflorescences on the same palm are rare --selfpollination is probably unusual under natural conditions.

Insect pollination. Presentation of complete data on abundance-and

behavior of insect visitors to inflorescences of babassu awaits identification of the specimens sent to specialists. A high diversity and abundance of insects were observed visiting male infloresences, which are odoriferous and produce copious pollen. Conversely, relatively low insect diversity and abundance were found on female inflorescences, the pistillate flowers of which give off no noticeable scent and do not produce nectar. The only insect that occurred in abundance on both inflorescence types was an undescribed species of Mystrops. This ca. 2mm long nitidulid beetle was found in sticky traps placed among the leaves of palms that were not bearing inflorescences, which suggests that it lives in the crowns of babassu. Prior to opening of either inflorescence type, the beetles were observed to gather in the crown by the hundreds or sometimes thousands. In a sample of flowers obtained from an open male inflorescence, I consistently found more than one beetle per flower; subsequent calculations produced an estimate of 48,300 staminate flowers on the inflorescence. The beetles were ob-





64





served to feed on pollen and probably use both flower types as breeding sites. Although they were far less abundant on female inflorescences, more than one beetle was consistently observed entering each pistillate flower. Entry was made either along the petal margins or, more commonly, between the basally fused stigmas at the tip of the flower. Pollen of babassu was found in isolated collections of Mystroa sp. obtained from female inflorescences that contained no viable staminate flowers, indicating pollen transport from another inflorescence. The beetles were observed in both secondary forest and pasture, and similar specimens of Mystrops were collected from the crown of a babassu palm in Pirapora, Minas Gerais, indicating that the association is widespread.

Wind pollination. Transport of babassu pollen by wind was repeatedly confirmed by visual observations in both the pasture and secondary forest sites. Preliminary analysis of pollen trapping data indicates a higher abundance of wind-borne pollen in the open pasture than in the closed secondary forest. Variances in quantities of pollen trapped in both habitats were extremely high. Insect exclusion from one female inflorescence in each habitat resulted in 57% fruit set in the secondary forest (total = 180 flowers) and 92% fruit set in the pasture (total = 167 flowers). These results indicate that wind pollination occurs in both habitats but may be more effective on the relatively open pasture site.

Fruit Biology

Dispersal. The proportion of babassu fruits removed > 4 m was significantly higher (p < .01 one-way ANO VA) in the secondary forest





65





than in the field or pasture (Fig. 25). In the secondary forest, fruits were probably removed by pacas (Aouti paca) and agoutis (2,asyprocta punc ta ta). In the field and pasture, removal was effected either by domesticated animals or by runoff.

Predation. The only non-human seed predator of babassu observed at my study sites was a species of bruchid beetle, Pachyjrners nucleorum. Attack of fruits and seeds by this predator (Fig. 26) occurred only after fruit abscission (0 d). The rate of colonization increased sharply at 10 d and leveled off at 40 d. Although ca. 70% of fruits were colonized, only ca. 40% of seeds were destroyed; an average of 1.9 out of a total of 3.1 seeds per fruit escaped predation.

Among fruits collected 40 d or longer after abscission, 20.1%

suffered predation of all seeds. Colonization of these fruits by P. nucleorum was approximately the same regardless of the number of seeds per fruit (Table 5). Because each larva was limited to one seed, however, lower proportions of seeds were eaten as the number of seeds per fruit increased (Table 5). These results suggest that selection by seed predaceous beetles favors many-seeded fruits in babassu.

Seed fate. In my experiment on seed fate over time (Fig. 27), the percentage of defective (aborted or rotted) seeds did not vary significantly among sites. After 1 yr, most of the seeds on the secondary forest and pasture sites had succumbed to predation; removal of palm trees that harbor seed-predaceous bruchid beetles probably resulted in significantly lower levels of predation in the field (p _".05, one-way ANO VA). Largely due to differences in predation, percentage of quies-





66













30
0

25


20
o

I..i. I 5I0
U)

5

LL

FOREST FIELD PASTURE



Figure 25. Removal of babassu fruits on three study sites at Lago
Verde. Shaded portions of bars show percentage of fruits
removed >4 m.






67










80


70



60
a
w
50


40



- /
Ll / --+ FRUITS
20- SEEDS


10-1
0
z/








0 40 80 120 160

TIME (d)



Figure 26. Attack of babassu fruits and seeds by the hruchid beetle,
Pachymerus nucleorum, over time. Bars represent 1 standard
error.





68













Table 5. Attack of babassu fruits and seeds by the seedpredaceous beetle, Pachymerus nucleorum.


SEEDS/FRUIT FRUITS SEEDS
Total N Colonized Total N Colonized
N % N %

1 16 10 62.5 16 10 62.5

2 63 50 79.4 126 67 53.2

3 124 86 69.4 372 157 42.2

4 113 85 75.2 452 198 43.8

5 23 19 82.6 115 37 32.2
6 4 3 75.0 24 6 25.0





69

100 FOREST

80 60

40 20



PASTURE
1-00 ir80 60

Cj 40 ~-20

0

FIELD

80 60

40QUECN

20



1981 1982


Figure 27. Seed fate over time in three study sites at Lago Verde.





70





cent seeds was significantly greater in the field after 1 yr (p S .05, one-way ANOVA). The relative lack of shading in the pasture probably accounted for the significantly lower percentage of germinated fruits in this habitat (p : .01, one-way ANOVA).

In the experiment on long-term (18 mo) seed fate, the percentage of fruits with 2 1 germinated seed(s) was significantly higher (p '_' .01,V Chi-square test) on non-weeded plots (62%) than on weeded plots (1%). Conversely, the percentage of fruits with : l quiescent seed(s) was significantly higher (p .05, Chi-square test) on weeded plots (78%) than on non-weeded plots (28%). I suspect that these latter seeds were quiescent rather than dormant; their germination was probably low due to

insufficient moisture on the non-weeded plots. These results suggest that under relatively exposed conditions, babassu seeds can maintain their viability for extended periods.

Germination. Figure 28 shows babassu's cumulative germination over time, beginning at the peak of fruiting during the latter part of the dry season. Most germination occurred ca. 3 mo after fruit fall; high soil moisture conditions prevalent during the rainy season are probably necessary for successful germination. Although this seasonal pattern of

germination appears to be relatively constant in babassu, germination success of fruits obtained from different palm trees varied, considerably and in some cases significantly (Fig. 29). As all fruits were planted under identical conditions, these results suggest that observed intraspecific variability in babassu's germination may be under genetic control.








71





























50- 20








00







3 0 ....

.. .. .. .

C-E



20 0....0.. 50..00..5
......T.M (days) ...



Figure. 28. Cumulativ germiatio ofbbas.futsoertme. ntr
vals ~ ~ ~ ~ ~ ..... rersnz tnaderr hddbr ersn
rainfall. ~ ~ ~ ~ ~ ~ ~ .. Daa.eeoband.rm.4Noebr.91. o2
April 1982..




72





PALM NUMBER




2







5

6


7

8


9 10


0 20 40 60 80 100
GERMINATION

Figure 29. Germination in fruits obtained from 10 babassu
palms. Each hatched bar represents one palm;
intervals represent 1 standard error.





73


















Table 6. Germination of babassu fruits in four experiments. Asterisks indicate significant
differences (* for p -S05, ** for p .01).


EXPERIMENT G E R M I N A T 0 N )
Treatment Control

Removal of Mesocarp 54 34


Intensive Burn 10 54


Moderate Burn 58 54


Shading 66 0





74





The abundance of' seedlings and juveniles on burned sites has led to

the suggestion that fire promotes germination in babassu (Smith 1974). Removal of' the mesocarp by dispersal agents and shading could also stimulate germination in the palm. Results of' experiments designed to test these hypotheses are shown in Table 6. 1 found that a moderate fire, such as occurs annually in pastures, had no effect on germination success. However, fruits exposed to an intensive fire, such as occurs when forest fallow is burned in shifting cultivation, exhibited significantly reduced germination (p,-: .01, Chi-square test). Removal of the mesocarp, simulating the effects of dispersal agents, was found to promote germination of babassu (p ...05). Likewise, germination was significantly higher in shaded fruits than in fruits left in the sun (p <.O1). This final response probably results from enhancement of germination by higher soil moisture levels in the shade.

Discussion

Phenology

Babassu's phenology showed remarkable constancy over a wide range of ecological conditions. The general pattern was a pronounced peak in leaf production and flowering during the wet season, while leaf loss and fruiting peaked during the dry season. Despite its dry-season peak, flowering within a given population occurred in most months. Male inflorescences were generally more abundant than female inflorescences, and I could detect no seasonal variation in sex ratios within the populations observed.




75





Leaf and inflorescence production in babassu are strongly correlated due to the close physical association of the two structures. Nevertheless, most phenological studies of palms in their natural habitats (e.g., Mora Urpi and Solis 1980, Bullock 1981, Myers 1981, Beach in press) have dealt exclusively with reproductive components. To my

knowledge, studies of both leaf and inflorescence production in palms have been limited to cultigens under plantation conditions (e.g., Child 1974, Hartley 1977). In the African oil palm, for example, leaf production occurs throughout the year, which results in nearly continuous flowering (Corley et al. 1976); in babassu, leaf production and flowering tend to be seasonal. Likewise, the minimum annual rate of leaf production in oil palm recorded in the literature is ca. 18 leaves (Corley et al. 1976), compared to a mean of ca. four leaves in babassu. Probably as a result of its high rate of leaf production, minimum leaf lifespan in wild stands of oil palm in Nigeria is only 2.1 yr (Zeven 1967); in babassu, mean leaf lifespan is 4.4 yr. I believe that this latter figure has important implications for management of babassu under shifting cultivation, a topic explored further in Chapter 7.

As has been documented in many other species (e.g., Harper 1977, Janzen 1978b, Pifiero and Sarukhan 1982), fruit yields in babassu vary considerably from year to year. As is the case in oil palm (Hartley 1977), I suspect that climatic factors such as rainfall may have a significant effect on annual variability in babassu's fruit yields. The influence is probably most pronounced at the time of inflorescence differentiation rather than at fruit set. Examination of developmental




76





series shows that inflorescence types (male or female) can be clearly distinguished up to 12 leaves prior to the newest emerged leaf. Given an average production of ca. four leaves per year, this means that sex determination occurs at least 3 yr prior to leaf expansion in the crown, 4 yr prior to flowering, and nearly 5 yr prior to fruiting. Detecting possible relationships between climate and yield will therefore require

long-term studies of babassu's phenology. Pollination

My results indicate that reproduction in babassu is predominantly or perhaps exclusively via outcrossing, and that pollination is effected by nitidulid beetles and wind. Such a combination of pollination systems has been observed in other species of palms (Uhl and Moore 1977, Moore and Uhl 1982). The coconut (Cocos nucifera) appears to be pollinated by either honey bees or wind in Hawaii (Shodt and Mitchell 1967).

The oil palm was thought to be exclusively wind-pollinated (Hartley 1977), but a recent study (Syed 1979) has implicated weevils as pollinators within the palm's natural range in West Africa and thrips on the Malay peninsula. Likewise, Myers (1981) suggested that both wind and insects effect pollination in Raphia taedigera in Costa Rica. Nitidulid beetles are rarely cited as pollinators (Faegri and van der Pijl 1979),

but their role in the pollination of palms has either been documented (Essig 1971) or implicated (Dransfield 1982). Wind pollination in palms has been reported since antiquity, according to Faegri and van der Pijl (1979).





77





In babassu, insect pollination is probably predominant in forests, while wind pollination is likely to be enhanced in open habitats such as

pastures. Neither system is exclusive of the other, and both may have come into play during babassu's evolutionary history. Predominance of

male inflorescences; small (20 pm in diameter), monosulcate pollen (personal observation); and lack of nectaries all point to pollination

by wind (Moore and Uhl 1982). Yet the strong scent produced by babassu's staminate flowers and the bright inner surface of its inflorescence bracts suggest adaptation to insect pollinators as well. The combination of two pollination systems clearly enhances the palm's adaptability to a wide range of ecological conditions. Dispersal

The fruits of many palms are fleshy and colored, and obviously adapted for dispersal by animals (Corner 1966, van der Pijl 1969, Kiltie

1981). Birds are probably the principal dispersal agents of palms (McKey 1975), although mammals such as rodents and primates (including humans) also feed on their fruits and disperse their seeds (Burtt 1929, Janzen 1971).

The dispersal of babassu fruits by animals other than humans occurs almost exclusively within forested habitats, where pacas (Agouti paca)

and agoutis (Dasyrocta punctata) are abundant. These rodents carry fruits to less exposed micro-sites (often > 4 m distant, Fig. 25), where the starchy mesocarp is partially or completely consumed and the fruits abandoned. Scatterhoarding of babassu fruits by these rodents apparently does not take place. Other mammals that I observed or were reported




78





by local inhabitants to feed on the mesocarp (but not the seeds) of babassu fruits include porcupines (Ooendu prehensilis and Coendu sp.), spiny rats (Proechimys longicaudatus and Mesomys hispidus), and peccaries (Taassu tajacu and T. pecari, both locally extinct). None of these species appears to have a significant role in dispersal of babassu fruits.

My field observations indicate that water-mediated dispersal is a significant factor on relatively open sites such as pastures. Here soil

compaction results in increased surface runoff, by which fruits were observed to be carried distances of ca. 25 m. The potential role of water-mediated dispersal is far greater. Dissemination by rivers largely accounts for the present distribution of babassu over nearly half a

continent (see Chapter 2).

Although it remains undocumented, the role of humans in the dispersal of babassu is potentially significant. As discussed in Chapter 2, the present occurrence of babassu in the Brazilian state of Ceara may be due to dissemination by indigenous peoples. Today settlers throughout the Amazon region commonly transplant babassu to areas where it did not

formerly occur. On a local scale in Maranhiio, gathering of babassu fruits is a widespread activity, particularly among women. Intensity of fruit gathering is especially high on relatively open, accessible sites

such as pastures. Although the seeds from almost all gathered fruits are ultimately consumed, some accidental dispersal occurs during gathering and transport. More significantly, perhaps, is the role of human gatherers of babassu fruits as agents of selection. People prefer-




79





entially gather large, many-seeded fruits for extraction of the oil-rich seeds; smaller, few-seeded fruits are left to regenerate.

Although rodents, runoff, and people play significant roles as

dispersal agents, babassu's massive fruits (usually >100 g) are relatively immobile. This immobility constitutes a crucial limiting factor in the palm's ecology and largely accounts for its current distribution

(see Chapter 2). Reduced dispersal is the price babassu pays for enhanced protection from seed predators. Predation

Of 28 insect species known to feed on babassu (Table 7), nine are potential seed predators. Most of the latter feed on stored seeds; the only seed-predaceous insects under natural conditions are two species of

bruchid beetle: Pachymerus nucleorum and Carbruchus lipismatus. Of these, only the former was present at all of my study sites in Maranhao. Popularly known in Brazil as bicho do coco, the larvae of this beetle prey on the seeds of a wide variety of palms, including economic species such as carnauba (Copernicia prunifera), licuri (Syagrus coronata), piassava (Attalea funifera), tucum (Astrocaryum vulgare), coconut (Cocos nucifera), and African oil palm (Elaeis guineensis). The penetration of

babassu fruits by this beetle was described by Bondar (1936). Ovipositing beetles remain in the crown of babassu during the day; at night

they deposit eggs on the surfaces of fallen fruits. The odor of the fermenting fruit mesocarp probably attracts ovipositing beetles (see Janzen 1971). Such attractants appear to persist for a maximum of 40 d after fruit abscission, after which colonization of babassu fruits by





80







+ + + +
0 4)
43 4)
ca





.f.1 CO 4->



ON









L4
0




0

Cd

4..)
>




CO

0
0 CL

4J C/I


L4
cd


0
0 0 cd
L. 0 4-) 0 Cd
cd 0 0 a 0
0 0. :s 4.3 "-f 0 m -H c
(D Go L4 0 >
.0 4J r. 9.. :$I
cd ca :3 r-4 a :s (D

w ca (a Cd
L. co Go .0 4.3 r. V m
a4 0 :s :3 0 a 0 0
4.) ca -H L. L :3 0 a Cd
cu r-I 0 u 0 m L. 0 0 .,4 -0
0 -H L. a 0 (D v 4-1 fn
0 0 > r-4 0 a) q 0 4)
a) -H > 0 4 0 0 0 0 L4 0
0 cd 0
a) cd &0 I'd -H -0 cd -H o 4) Cd 0
44 0 ca a) CA O14 -H -4 1: r-4 u z v 94
0 q Cd 10 cd .0 a (D 0 r4
Cd -I V L. -H a 0 >% 8 ca -0
E-1 cd 0 q C\J 0 401 M ---T -H U) %0 El- 00 4 ON.0 0 0 10 CJ 0 M
Lo S4 a .0 -H m L4 a a] -H 4)
0 0 m a 43 Cd L. 8
E- .4.3 .0 co 0 m L. (D L.
cl. ul r4 v 4 0 f. 0 0) -4 w
0 q a4 0 w pa 0 0 u Cl
a C4 r-I
0 0 0




81







L. V 0 0
43 (D V2 cil















z C)
0




















Go

4J + + +
C40






4-) :3
a :3 U3 L4
ca 0 ca
0 L4 S
.rj ca -H r-I
;.4 4z ca a
:3 q a
ul U2 0 Q ca m
ul 0 Cd U3
U3 co a z m
:s o Cd (d r-I m 0 m m 9. 0 m
4.3 0 -H 0 4.3 -H m 0 a PC 0 r-q
010 0 m 0 u 0 0 z N 0 0 =
13 0 -H m 0 0 -ri 0 x 4) -H 0 in
=s m Q > 'I > :3 0 L,


4) -H r-I 0 4-) co a m -H 0 .0 = 0 in
ca 4) m r-I co .11 m 0 S. 0 0 a L4 0 0 0 0
0 N ca (D 0 > 0 0 S. 4J 0 01 0 0 0 a
. x 0 > 10 0) c El r-q 0 r--q -H > = 4) L.
0 0 -H M 0 (D 0 A C 0 ca 0) ca as
E-4 M :14 =4 0 01 0 rn I
Cd CA mi
0 -H r. 0 rq
(0 'a -H a 1-1 -I r-I .
-.T r-A -,q Ln r-I ko t--- cd 00 ON C:) C\j 0 m X:r Ln %.o t- 0 w 0 S, 0 4.) N 5.. C4 CM CM N CV -0 CM
4J 0 Cd 0) >1 H
CL C pq 0 04 04 4.)
0 0 z -H

0




82





beetle larvae apparently ceased (Fig. 26). Approximately 15-18 d following oviposition, tiny (0.75 x 0.10 mm) larvae emerge and proceed to the base of' the fruits. Penetration takes place through the germination

pores, which offer minimal resistance. Prior to abscission, fruits are attached at the bases and their germination pores are thus protected, which explains why fruits are only attacked after they fall. Approximately 2-3 d after penetration begins, the larvae reach the seeds. Observations of fruit colonization revealed a maximum of one larva per seed chamber (locule), suggesting that late arrivals are probably cannibalhzed (see Janzen 1971); movement between seed chambers is prohibited by the thickness of the inner endocarp walls. One seed provides sufficient food for a larva's complete development. Adults emerge ca. 2.5 mo after oviposition by burrowing a ca. 8-mm-wide exit hole through the endocarp.

In a closely related palm (Scheelea rostrata) in Costa Rica, Bradford and Smith (1977) found that ca. 60% of the fruits lost all their seeds to a predaceous bruchid beetle (Caryobruchus buscki) or to rodents; in babassu, the total figure was ca. 20% (excluding potential losses to humans). A key component of this high predator escape is the predominance of many-seeded fruits in babassu (Table 5); in contrast, most (ca. 95%) fruits in the Costa Rican population of S. rostrata were one-seeded (Bradford and Smith 1977). In both species, the proportion of seeds eaten by predators declined as the number of seeds per fruit increased. Thus, in S. rostrata mean seed losses to predators in one-, two-, and three-seeded fruits were 61%, 46%, and 40%, respectively




83





(Bradford and Smith 1977); the corresponding figures for babassu--62%, 53%, and 42% (Table 5)--are remarkably similar. The trend suggests that in both species, selection by seed predators favors many-seeded fruits. However, such selection probably operates within strict constraints. Bradford and Smith (1977) found that in S. rostrata, individual seeds in one-seeded fruits contained 46% and 123% higher energy content than those in two- and three-seeded fruits,, respectively. Seedlings from one-seeded fruits should therefore have a competitive advantage, In addition, seedlings emerging from the same many-seeded fruit compete intensively with each other, further increasing the competitive advantage of seedlings from one-seeded fruits. The number of seeds per fruit thus represents a compromise between opposing selective forces.

The endocarp of babassu is exceptionally thick (R 49.8 mm,, s.e.

1.1) and resistant to crushing (Fig. 30) compared to other species of palms. It would thus appear to provide effective deterrence against seed predators. Yet the figures cited above indicate that levels of predation in one- to three-seeded fruits are almost identical in babassu and S. rostrata, despite the latter's comparatively thin endocarp (4-5 mm,, according to Bradford and Smith 1977). As described above, seedpredaceous beetles penetrate the fruits of babassu via the germination pores, where the endocarp is least resistant; entry is effected in 2-3 d

(Bondar 1936). The thick endocarp of babassu thus appears to provide little deterrence against the palm's most effective seed predator. Conversely, the endocarp completely deters other potential predators such as rodents and squirrels; only monkeys (e.g., Cebu~ apla and




84











Jessenia bataua (20)


Astrocaryum cf. mocrocalyx (34)


Socrotea exhorrhiza (35)


Iriartea ventricosa (32)


Mauritia flexuoso (7)


Scheelea sp. (19)


Phytelephas microcarpa (17)


Orbignya martiana (6)


I I i 1 I i i I I i I I I I I I I 1Ill
0.05 0.1 0.5 1.0 5.0 10.0
FORCE TO BREAK (t) Figure 30. Force required to break palm fruits. Intervals represent +
1 standard deviation; values in parentheses are sample sizes.
Data for Orbignya martiana from Fonseca (1924); data for all
other species from Kiltie (1982).




85





Aotus trikingatus, according to local informants) successfully prey on babassu by pulling off immature fruits and consuming the liquid endosperm within (see Izawa and Mizuno 1977). Thus seed predators either are completely deterred or successfully bypass babassuls thick endocarp. I suspect that there is no predator currently exerting sufficient selective pressure to account for the evolution of this extraordinary structure. Coevolution with large, Pleistocene mammals (Janzen and Martin 1982) thus provides the most satisfactory explanation for the origin of

the thick, massive fruits characteristic of babassu. Germination

The results of my experiments indicate that germination in babassu

is least successful on disturbed sites such as pastures and shifting cultivation plots, where shading is minimal, fires either frequent or intensive, and mesocarp-re moving dispersal agents rare. While germination success of babassu is thus reduced, its peculiar mode of germination confers upon the palm an extraordinary capacity to recover from the stresses inherent to these sites.

Germination in babassu occurs when one or more embryonic axes emerge from the germination pores at the base of the fruit. Each axis

enters the soil and penetrates to a variable depth, where it subsequently differentiates into roots and leaves (Fig. 31). This so-called

cryptogeal mode of germination is found throughout the Attalea alliance and in distantly related palm genera such as Acanthococos and Sabal (Corner 1966), as well as in many other plant families (Jackson 1974). In plants that exhibit cryptogeal germination, the shoot arises from





86








































"' i:

-7; /













igure 31. Cryptogeal germination in babassu. From left to right,
seedlings were 1-, 4-, 8-, and 14-wk old. Scale = 15 cm.





87





below ground even though the seed germinates on the surface. In the case of babassu, the underground meristem in seedlings and stemless juveniles enables them to regenerate on sites that are subject to cutting and burning. Many species exhibiting cryptogeal germination occur in fire-prone habitats, and Jackson (1974) proposed that it represents an evolutionary response to fire. However, I suspect that babassu--as well as many of its relatives in the Attalea alliance-- evolved in moist, forested habitats where natural fires were infrequent (see Chapter 2). The presence of cryptogeal germination in these species is more likely due to the general lack of secondary growth throughout the family. Lateral expansion of the palm stem is accomplished by so-called establishment growth (Tomlinson and Zimmermann 1966), during which the stem takes on a highly unstable, obconical shape. One of the means by which palms provide support for this structure is to bury their shoot axes underground.

Regardless of the selective pressures that brought about its evolution, there is no doubt that cryptogeal germination plays an important role in the current proliferation of babassu on sites subjected to human disturbances.




Full Text
PRODUCTIVITY (%)
131
LENGTH OF STEM (m)
O I 25 10 15 20 30
0 20 40 60 80 iOO 120 140 160 180
AGE (yr)
Figure 38.
Relative allocation of
production per babassu
annual above-ground dry matter
palm as a function of age.


114
This is a clear case of facilitation, in which the presence of an
earlier plant community promotes the recruitment of a subsequent
species. Despite unrealistically high densities of planted fruits
p
(l/m^), seedling recruitment of babassu was extremely low on weeded
plots. This suggests that, under natural conditions, initial
recruitment of babassu requires the presence of previously established
competitors.
Facilitation of recruitment does not always translate into net
facilitation, because inhibition later in the life cycle may be stronger
than the initial facilitation (Turner 1983). The experiments described
in this chapter show that whereas competitors facilitate recruitment in
babassu, they also inhibit subsequent growth. In the following chapter,
I shall examine growth of babassu under conditions of high competitor
abundance.


129
Discussion
Growth and Productivity
Partitioning of dry matter production changes considerably during
the potential life of a babassu palm growing on the primary forest site
(Figs. 37-38). Although root productivity was not measured, excavations
revealed that most dry matter production is initially allocated to
roots. Following germination, a root system forms prior to expansion of
the first leaf (Fig. 31). The system continues to enlarge during the
period of so-called establishment growth (Tomlinson and Zimmermann
1966), when the underground stem gradually increases in girth. The base
of the stem forms an inverted cone from which adventitious roots arise.
These roots provide crucial support for the mechanically unstable stem
during establishment growth (Holttum 1955, Zimmermann 1973); as dis
cussed in Chapter 3, the underground placement of the apical meristem
via cryptogeal germination provides additional support in babassu.
Establishment growth continues until the stem approaches maximum girth,
whereupon stem growth becomes predominantly vertical (Waterhouse and
Quinn 1978).
At the beginning of the seedling stage, most above-ground produc
tion (ca. 70%) is allocated to photo synthetic tissues (leaflets). As
the palm grows, however, increased allocation goes to support in the
form of leaf axes and, ultimately, the stem. Prior to commencement of
emergence of the stem from the soil, allocation to leaf axes comprises
over 70% of total productivity. These axes provide crucial support, as
the leaves at this time approach their maximum length and are oriented


LEAF AREA (cm
93
Figure 32.
LEAF LENGTH (cm)
Leaf area as a function of leaf length in seedlings of
babassu. The best fit was: leaf area (cm'1) = 0.0741
leaf length (cm) + 1.1302 leaf length (cm) + 19.9644.


:Table 27. Annual changes in number of palms per life stage on the 1-ha pasture site
at Lago Verde. Life stages are defined in Table 18.
LIFE STAGE
INITIAL N
GAINS
LOSSES
NET CHANGE
FINAL N
Number and Description
Recruitment
from Last Stage
Recruitment
to Next Stage
Mortality
1 Seeds on ground
1,620
86,720
980
85,390
+350
1,970
2 Seedlings
4,900
980
80
1,880
-380
4,520
3 Pre-established juveniles
4,460
80
30
520
-470
3,990
4 Established juveniles
370
30
10
40
-20
350
5
32
10
5
5
0
32
6
15
5
2
3
0
15
7 Mature palms
83
2
15
0
-13
68
8
3
15
0
0
+15
18
9
3
0
0
0
0
3
10
0
0
0
0
0
0
11
0
0
0
0
0
0
12
0
0
0
0
0
0
147


PALMS(no./ha)
150
Figure 43. Age distribution of babassu on the secondary forest site
at Lago Verde.


173
. 1981b. Sinopse preliminar do censo demogrfico: Maranhao. Fun-
dapo Instituto Brasileiro de Geografia e Estatistica, Rio de
Janeiro, Brazil.
IPT. 1979. Anlise tecnolgica, econmica e social do aproveitamento
integral do coco de babacu. Two volumes. Instituto de Pesquisas
Tecnolgicas do Estado de Sao Paulo, Sao Paulo, Brazil.
Izawa, K., and A. Mizuno. 1977. Palm-fruit cracking behavior of wild
black-capped capuchin (Cebus apella). Primates 18: 773-792.
Jackson, G. 1974 Cryptogeal germination and other seedling adapta
tions to the burning of vegetation in savanna regions: The origin
of the pyrophytic habit. New Phytologist 73s 771-780.
Janzen, D. H. 1971. The fate of Scheelea rostrata fruits beneath the
parent tree: predispersal attack by Bruchids. Principes 15: 89-
101.
1978a. Reduction of seed predation on Bauhinia pauletia
(Leguminosae) through habitat destruction in a Costa Rican
deciduous forest. Brenesia 14-15: 325-335.
. 1978b. Seeding patterns of tropical trees. Pages 83-128
in P. B. Tomlinson and M. H. Zimmermann, editors. Tropical trees
as living systems. Cambridge University Press, Cambridge, England.
, and P. S. Martin. 1982. Neotropical anachronisms: the
fruits the gomphotheres ate. Science 215: 19-27.
Kiltie, R. A. 1981. Distribution of palm fruits on a rain forest
floor: why white-lipped peccaries forage near objects. Biotropica
13: 141-145.
. 1982. Bite force as a basis for niche differentiation
between rain forest peccaries (Tayas su tajacu and T¡_ pcari).
Biotropica 14: 188-195.
Kira, T. 1975. Primary production of forests. Pages 5-40 in J. P.
Cooper, editor. Photosynthesis and productivity in different
environments. Cambridge University Press, Cambridge, England.
Kuhlmann, E. 1977. Vegetaco. Pages 85-110 in IBGE. Geografia do
Brasil 2: Regio nordeste. Fundapo Instituto Brasileiro de Geo
grafia e Estatistica, Rio de Janeiro, Brazil.
Lefkovitch, L. P. 1965. The study of population growth in organisms
grouped by stages. Biometrics 21: 1-18.


151
PRIMARY FOREST
1,620 20 3
^ vl' 'l'
SECONDARY FOREST
1,820 20 17
\J, 'l' 'l/
PASTURE
1,880 520 48
^ ^ ^
SEEDLINGS
PRE-ESTABLISHED ESTABLISHED
JUVENILES JUVENILES
MATURE
PALMS
Figure 44. Population dynamics of babassu on the 1-ha primary forest,
secondary forest, and pasture sites at Lago Verde. Numbers
in boxes are N at t = 0, numbers on horizontal arrows re
present recruitment over 1 yr, and numbers on vertical ar
rows represent mortality over 1 yr.


139
allocation to reproductive structures is directed to protection, either
of seeds by fruit walls (66.5%) or of developing female flowers by
bracts (6.6%). Morphological protection of reproductive components is a
characteristic feature of palms (Uhl and Moore 1973). Selective pres
sures that may have resulted in babassu's particularly high investment
were discussed in Chapter 3. High investment in protection of reproduc
tive components may ultimately lead to high seedling recruitment, a
topic that I shall examine further in the following chapter.


70
cent seeds was significantly greater in the field after 1 yr (p .05,
one-way ANOVA). The relative lack of shading in the pasture probably
accounted for the significantly lower percentage of germinated fruits in
this habitat (p .01, one-way ANOVA).
In the experiment on long-term (18 mo) seed fate, the percentage of
fruits with -1 germinated seed(s) was significantly higher (p i .01,
Chi-square test) on non-weeded plots (62%) than on weeded plots (1%).
Conversely, the percentage of fruits with -1 quiescent seed(s) was
significantly higher (p .05, Chi-square test) on weeded plots (78%)
than on non-weeded plots (28%). I suspect that these latter seeds were
quiescent rather than dormant; their germination was probably low due to
insufficient moisture on the non-weeded plots. These results suggest
that under relatively exposed conditions, babassu seeds can maintain
their viability for extended periods.
Germination. Figure 28 shows babassu's cumulative germination over
time, beginning at the peak of fruiting during the latter part of the
dry season. Most germination occurred ca. 3 mo after fruit fall; high
soil moisture conditions prevalent during the rainy season are probably
necessary for successful germination. Although this seasonal pattern of
germination appears to be relatively constant in babassu, germination
success of fruits obtained from different palm trees varied-considerably
and in some cases significantly (Fig. 29). As all fruits were planted
under identical conditions, these results suggest that observed intra-
specific variability in babassu's germination may be under genetic
control.


THE BIOLOGY OF Orbignya martiana (PALMAE),
A TROPICAL DRY FOREST DOMINANT IN BRAZIL
BY
ANTHONY B. ANDERSON
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REOUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1983


120
from each of 30 randomly selected palms was harvested, oven-dried, and
weighed; leaflets and leaf axes (sheath, petiole, and rachis) were
weighed separately. For each of the 30 palms, I multiplied the mass of
its leaf by the mean number of leaves produced per year in the corres
ponding life stage. The estimated leaf productivity of each palm was
then plotted against its age (estimated separately for each site) to
produce the following regressions for the primary and secondary forests,
respectively:
logtleaf production (kg/yr)] = -0.001[age (yr)]2
+ 0.125 age (yr) + 0.042, r2 = 0.98;
log[leaf production (kg/yr)] = -0.001[age (yr)]2
+ 0.146 age (yr) + 0.033, r2 = 0.99.
Using these regressions, annual leaf production was calculated for palms
in the sampled area and corrected to a per-ha basis.
Inflorescence and fruit production in mature palms was monitored
over the 1-yr period (see Chapter 3). Samples of three staminate and
three hermaphroditic inflorescences were oven dried to obtain mean mass
of bracts, peduncles, rachises, and rachillae. Mean oven-dried mass of
staminate flowers per inflorescence was estimated from a sample of three
inflorescences. Mature infructescences were harvested monthly, and the
number of fruits produced per infructescence was determined by counting
fruits and/or calyces. A sample of ca. 30 fruits from each infruces-
cence was air-dried and weighed. A correction factor for calculating
oven dried mass was obtained from a sample of five fruits.


90
as nutrient-poor soils (Connell 1978), flooding (Hoopes 1974), intensive
herbivory (Churchill et al. 1964), and increasing dryness (Gentry 1982)
have repeatedly been implicated as important causal agents of reduced
diversitywith concomitant increases in dominance by one or a few
speciesin plant communities.
Methods
Experiments involving irrigation, fertilization, and insecticide
applications were set up in Lago Verde, Maranhso; the climate experiment
was carried out at Lago Verde (dry site) and Belm, Para (wet site).
The Lago Verde site consisted of ca. 10-yr-old secondary forest
dominated by juvenile babassu. An area of ca. 5,400 m2 was cut on 15
September and burned on 30 October 1980. To promote homogeneity,
remaining unburned material was subsequently gathered into piles and
burned again. To prevent future confusion between planted and previ
ously established seedlings and juveniles, the latter were killed either
by uprooting (in the case of seedlings) or by driving stakes into the
apical meristems following defoliation (in the case of juveniles). This
latter technique often had to be repeated to achieve the desired effect.
On 26-27 November 1980, eighty 5 x 2 m plots were set up and 160 babassu
fruits planted in each. With the exception of the climate experiment
(described below), fruits were obtained locally and virtually all had
ripened (i.e. abscised) within the past month, according to local infor
mants. The husks were removed to promote germination and inhibit
dispersal by animals. Fruits were partially buried to prevent dispersal
by runoff.


61
Table 4. Sources of increased fruit yields of babassu per
ha between 1980-81 and 1981-82 on three sites.
Numbers represent percent of total increase in
yields.
SOURCE
SITE
Primary Forest
Secondary Forest
Pasture
Yield per Infructescence
-6.5
-26.5
+12.3
Number of Infructescences
per Fruiting Palm
+39.4
+35.6
+24.7
SUBTOTAL
(Yield per Fruiting Palm)
+32.9
+9.1
+37.0
Number of Fruiting Palm
+67.1
+90.9
+63.0
TOTAL
100.0
100.0
100.0


49
(x = 4.4) > secondary forest (x = 3.7). These results can probably be
attributed to differential shading of adult palms on the three sites.
Shading was lowest in the pasture. Potentially high shading in the
primary forest may have been reduced by the relatively pronounced canopy
roughness (which thus permitted greater lateral penetration of light)
and the emergent status of many of the adult palms. In the secondary
forest, shading was highest because the dense, uniform babassu stand
formed a closed, interlocking canopy.
Except at the wet site, leaf production and loss rates were approx
imately equal after 1 yr of observations; net gain for this period
averaged .05 leaves per palm, which is virtually steady state. Assuming
steady state conditions, one can divide the number of leaves borne per
palm (x = 18.2, n = 199) by annual production per palm (x = 4*1 n =
199) to obtain an average leaf lifespan of 4-4 yr. At the wet site,
production rates were far greater than loss rates; the net gain per year
was 3*3 leaves per tree. The low losses were probably due to leaf
harvesting for thatch 2 yr prior to initiation of the study. Yet gross
leaf production (x = 3.9) was not significantly different from the
combined mean obtained from the other sites (x = 4.1 ) I thus suspect
that the single defoliation had little effect on leaf production rate at
the wet site.
Flowering and fruiting. Data on inflorescence production for both
sets of sites are shown in Figures 18 and 19. Like leaf production,
flowering of babassu peaked during the rainy season and was lowest
during the dry season. The similarity of inflorescence and leaf pheno-


14
Seasonally wet savanna An extensive area traversed by the lower
reaches of Maranhao's major rivers is flooded during the December-May
rainy season and burned during the dry season. This low-lying area is
called baixada. The predominant soils (ground-water laterites and
gleys) are hydromorphic, the major vegetation type is a low savanna
dominated by sedges and grasses, and the principal land use is grazing
of cattle and buffalo on unimproved range. Throughout this region,
forests of babassu are generally limited to topographically elevated
islands, which are also the principal sites for shifting cultivation.
Woodland, scrub, or dry savanna. The term cerrado refers to both an
extensive phytogeographic region of Brazil and the region's predominant
vegetation. Within Maranho, cerrado is characterized by annual rain
fall >1500 mm, a dry season of 6 mo or longer, and well- to excessively
drained soils. The vegetation consists of deciduous woodland, scrub,
and occasional savanna. The predominant land use is grazing of cattle,
goats, and sheep on unimproved range. Shifting cultivation and perma
nent plantations of sugarcane and cotton are generally confined to the
moist bottomlands, which are also the habitats occupied by babassu
forests.
Moist or dry forest. This zone is characterized by moderate to high
rainfall and a dry season of 6 mo. It extends across both Tropical
Moist Forest and Tropical Dry Forest Life Zones (sensu Holdridge 1967).
The vegetation was originally tall, diverse forest ranging from ever
green to deciduous. Due to widespread deforestation during the past 60
yr, the original forest has been virtually eliminated, and today stands


102
O
Table 9. Palms per m (a), leaf area per palm (b),
and leaf area index (c) of babassu at end of
irrigation experiment. Asterisks indicate
significant differences (* for p £ .05)
between main effect means. Where there is
interaction between main effects, super
scripts are used to compare treatment means.
Values that do not share the same superscript
(a)
differ significantly (p £ .05).
Palms Der m2
No Irrigation
Irrigation
Mean
Not Weeded
28.1a
26.4a
27.2
Weeded
3.6b
l8.4b
11.0
Mean
15.8
22.4
(b)
Leaf Area Der Palm (cm2)
No Irrigation
Irrigation
Mean
Not Weeded
256.0
259.0
257.5
£
Weeded
358.6
329.0
343-8
Mean
307.3
294.0
(c)
Leaf Area Index
No Irrigation
Irrigation
Mean
Not Weeded
.26
.31
.29
*
Weeded
. 10
.08
.14
Mean
.18
.24


184
APPENDIX B. SPECIES ABUNDANCE, FREQUENCY, DOMINANCE, AND IMPORTANCE
IN 1 HA OF SECONDARY FOREST, LAGO VERDE, MARANHAO.
s p
E C I E S
ABUNDANCE
stems/ha
FREQUENCY
%
DOMINANCE
mVha 1
IMPORTANCE
%
1.
Orbicnva martiana Barb. Rodr.*
244
66.67
21.74
21.703
84.11
57.51
2.
SDondias mombim Urb.*
20
5.46
9.78
1.110
4.30
6.52
3.
Cordia sellowiana Chamb.
16
4.37
6.52
0.778
3.01
4.64
4.
CecroDia concolor Willd.
13
3.55
7.61
0.335
1.30
4.15
5.
Astrocarvum vulcare Mart.*
19
5.19
5.43
0.244
0.95
3.86
6.
Guateria chrvsoDetala (Stend.) Mia.
10
2.73
7.61
0.255
0.99
3-78
7.
Hieronima alchornioides (Freire) Allem.*
6
1.64
3.26
0.314
1.22
2.04
8.
Aoeiba tibourbor Aubl.
4
1.09
3.26
0.162
0.63
1.66
9.
Sterculia sd.*
4
1.09
3.26
0.133
0.52
1.62
10.
SaDium lanceolatum Hub.
3
0.82
3.26
0.102
0.40
1.49
11.
Cocoloba latifolia Lam.*
3
0.82
3.26
0.088
0.34
1.47
12.
Jacaratia SDinosa A.DC.*
2
0.55
2.17
0.059
0.23
0.98
13.
Vitex cf. flavens H.B.K.
2
0.55
2.17
0.039
0.15
0.96
14.
Casearia svlvestris Sw.*
2
0.55
2.17
0.024
0.09
0.94


121
Individual productivity. Productivity of individual palms as a
function of age was calculated using the primary forest population,
where long-lived palms were represented. I calculated the productivity
of each above-ground component separately. For leaves, I used the
regression describing leaf productivity as a function of age provided
above. For stems, I first converted the mean annual height increment of
each stage (Table 19) to mass using the mass-height regression provided
above. I then plotted stem biomass increment against the mean age of
each stage. For reproductive components, I likewise plotted each
stage's mean annual production against its mean age.
Lifetime productivity refers to an individual palm's production of
above-ground biomass over the maximum lifespan of babassu. Maximum
lifespan was assumed to be equal to the estimated age of the tallest
palm on the primary forest site. Lifetime productivity of leaves and
reproductive structures was obtained by integrating their respective
functions describing annual productivity as a function of age. Lifetime
stem productivity was assumed to be equal to the stem mass of the
tallest palm on the primary forest site, which was estimated using the
mass-height regression provided above.
Results
Growth
Data from the primary and secondary forests (Tables 19 and 20,
respectively) indicate that babassu's initial growth was extremely slow.
According to the mean growth rate measurements, an average of ca. 38 yr
in the primary forest and ca. 29 yr in the secondary forest elapsed


33
Rodr.) had been previously published by Barbosa Rodrigues (1891) in
reference to another species (cf. Article 64, International Code of
Botanical Nomenclature: Stafleu 1972, p. 59). Orbignva martiana is
therefore the earliest, validly published name definitely referable to
this taxon and it maintains precedence over subsequently published names
associated with babassu (Table 2).
My use of martiana is tentative, due to an older name that
remains to be considered. Martius (1844) published a relatively
complete description of Q¡_ phalerata Mart., which he designated as the
type species of the genus. Subsequent collections of this species have
apparently not been obtained (Glassman 1977), and Martius* published
description and illustration provide no basis for distinguishing it from
0. martiana. The lack of specimens of 0. phalerata prompted a recent
(July-August 1982) expedition, sponsored by the New York Botanical
Garden, to the type locality in eastern Bolivia. Final judgment
concerning the relationship of these two taxa awaits examination of the
specimens obtained.
Description and Differentiation
A technical description of babassu, based on specimens obtained to
date, is provided in Appendix C. This description illustrates the
extreme variability of the species. Despite its variability, babassu
possesses a number of diagnostic features that together enable one to
distinguish it from related species. These features include an erect,
solitary habit; large, erect leaves recurved at the apex, with rigid
leaflets disposed in a vertical plane and often with striations on the


40
painted with different colors and/or patterns to avoid confusing the
sources of fallen leaves.
As in other palms, the reproductive structures (inflorescences and
infructescences) in babassu are relatively few, large, and permanent.
In addition to obtaining separate data on each structure, I could thus
deduce its phenological activity during the ca. 1-mo interval since the
last observation. To prevent confusion among several flowering and
fruiting structures borne simultaneously on a palm, phenological obser
vations were accompanied by a measurement of the compass bearing of each
structure in relation to the palm stem. In the case of functionally
female inflorescences (containing at least one pistillate flower and a
variable number of staminate flowers that were usually inviable), I
distinguished the following categories: (1) inflorescence in flower
(pistillate flowers yellow); (2) infructescence with immature fruits
(fruits tan to brown and not falling); (3) infructescence with mature
fruits (fruits falling); and (4) infructescence with no fruits (all
fruits fallen). In the case of male inflorescences (containing exclu
sively staminate flowers), I distinguished the following categories:
(1) inflorescence in flower (flowers intact or falling); (2) inflores
cence in flower during the past month (flowers fallen, inflorescence
4
axes yellow to light tan); and (3) inflorescence in flower prior to past
month (inflorescence axes tan).
On the three study sites at Lago Verde, fruit production per month
was measured from September 1980 to April 1982. Mature infructescences
were harvested, and the number of fruits produced per infructescence was


Appendix E continued.
CHARACTER
Orbienva martiana
"Markleva dahlcreniana"
(possible hybrid)
Maximiliana mariDa
INFLORESCENCE (cont.)
Peduncle
length (cm)
56-185
170
100-150
Rachis
length (cm)
48-130
70-110
40-80
Rachillae
length of subtending
2-4
4-5
10-11
bracteole (mm)
Starainate flowers
length of petals
14
10
3-7
number of stamens
(22-)24-26(-30)
(7)9(10)
6
orientation of thecae
irregularly coiled
and twisted
twisted or hamate
straight
fusion of thecae
separate
partially united
united
Pistillate flowers
number per rachilla
12(3)
3-4
3-12
size of sepals
unequal
unequal
equal
FRUIT
Shape
broadly elliptic
to oblong
oblong
ovoid-oblong
Length (mm)
74-125
70
50-70
Width (mm)
49-99
30
20-30
Thickness of endocarp (mm)
35-76
40-60
30-50
193


79
entially gather large, many-seeded fruits for extraction of the oil-rich
seeds; smaller, few-seeded fruits are left to regenerate.
Although rodents, runoff, and people play significant roles as
dispersal agents, babassu's massive fruits (usually >100 g) are rela
tively immobile. This immobility constitutes a crucial limiting factor
in the palm's ecology and largely accounts for its current distribution
(see Chapter 2). Reduced dispersal is the price babassu pays for en
hanced protection from seed predators.
Predation
Of 28 insect species known to feed on babassu (Table 7), nine are
potential seed predators. Most of the latter feed on stored seeds; the
only seed-predaceous insects under natural conditions are two species of
bruchid beetle: Pachymerus nucleorum and Carybruchus lipismatus. Of
these, only the former was present at all of my study sites in Maranhao.
Popularly known in Brazil as bicho do coco, the larvae of this beetle
prey on the seeds of a wide variety of palms, including economic species
such as ca rnauba (Copernicia prunifera), li curi (Sya grus cor on ata),
piassava (Attalea funifera), tucum (Astrocaryum vulgare), coconut (Cocos
nucfera), and African oil palm (Elaeis guineensis). The penetration of
babassu fruits by this beetle was described by Bondar (1936). Ovi
positing beetles remain in the crown of babassu during the day; at night
they deposit eggs on the surfaces of fallen fruits. The odor of the
fermenting fruit mesocarp probably attracts ovipositing beetles (see
Janzen 1971). Such attractants appear to persist for a maximum of 40 d
after fruit abscission, after which colonisation of babassu fruits by


LEAF PRODUCTION PER PALM
47
1981 1 1982-
J FMAMJJ A SONDJ FM
Figure 16. Mean monthly gain and loss of leaves per palm on three
sites in Maranhao. Wet site = Pindar Mirim, intermediate
site = Lago Verde, and dry site = Caxias. N= number of
palms observed per site. ND = no data.


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.
Johif J. Ewel, Chairman
Frofessor of Botany
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.
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.
Associate Professor of Botany
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.


36
carried considerable distances by runoff and rivers in flood. Lighter
fruits would tend to be carried farther, which could explain why babassu
fruits are relatively light (80-250 g fresh mass) over most of the
species' range, while the heaviest fruits known (up to 800 g fresh mass)
are found in its presumed center of origin in Goias (Fonseca 1924; M. J.
Balick and A. B. Anderson, unpublished data).
There are three important exceptions to the links between the
Brazilian Shield and babassu's present range. The first is its disjunct
occurrence in Suriname and Guyana (Fig. 15). Until the early Pleisto
cene, the eastern portions of the Brazilian and Guayana Shields were
connected, which probably accounts for the close floristic ties between
the two regions (Maguire 1970). The Guianas represent a second center
of diversity of the entire Attalea alliance (Wessels Boer 1965).
A second exception is babassu's occurrence in the Brazilian states
of Maranho and Piaui (Fig. 1). Markley (1971) suggested that the
Tocantins River may have once served as a link for dissemination of the
palm to these states. The Tocantins has its headwaters in the Brazilian
Shield and presently drains into the Amazon River. During the
Cretaceous, however, it may have flowed across the northern part of
Maranho and built up the huge delta that presently covers most of the
northern third of the state. It seems unlikely that this delta could
have been built by the relatively insignificant rivers of Maranho that
exist today (Pindar, Grajau, Mearim, Itapecuru, and Parnaiba), most of
which have their headwaters only a short distance from the delta.
Uplift of the Serra do Gurupi in western Maranho (probably during the


FRUIT PRODUCTION PER PALM (kg)
1981 1982
Figure 24. Mean monthly fruit production per palm on three study
sites at Lago Verde. Values are air-dried mass. N =
number of palms per site.



PAGE 1

7+( %,2/2*< 2) 2UELJQ\D PDUWLDQD 3$/0$(f $ 7523,&$/ '5< )25(67 '20,1$17 ,1 %5$=,/ %< $17+21< % $1'(5621 $ ',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(28,5(0(176 )25 7+( '(*5(( 2) '2&725 2) 3+,/2623+< 81,9(56,7< 2) )/25,'$

PAGE 2

7KLV VWXG\ LV GHGLFDWHG WR WKH ODWH +DUROG ( 0RRUH -U ZKR LQVSLUHG LW

PAGE 3

$&.12:/('*(0(176 0\ ILHOG UHVHDUFK DQG GDWD DQDO\VLV ZHUH ILQDQFHG E\ *UDQW 1R WR -RKQ (ZHO IURP WKH 86'$ )RUHVW 6HUYLFH &RQVRUWLXP IRU WKH 6WXG\ RI 0DQnV 5HODWLRQVKLS ZLWK WKH *OREDO (QYLURQPHQW $GGLWLRQDO VXSSRUW ZDV SURYLGHG E\ D IHOORZVKLS IURP WKH 6XSHULQWHQGHQFLD GH 'HVHQn YROYL HQWR &LHQWILFR H 7HFQROJLFR &13Tf DP JUDWHIXO WR WKH 0XVHX 3DUDHQVH (PLOLR *RHOGL DQG LWV 'LUHFWRU 'U -RV 6HL[DV /RXUHQSR IRU LQYDOXDEOH ORJLVWLF DV ZHOO DV PRUDO VXSSRUW GXULQJ WKH FRXUVH RI P\ UHVHDUFK 6SHFLDO WKDQNV DUH GXH WR %HUQDUGR GD 6LOYD DQG 0DULR &RVWD ZKR JDYH XQIDLOLQJ DVVLVWDQFH LQ WKH ILHOG DQG WR 0LFKDHO %DOLFN DQG :LOOLDP 2YHUDO ZKR FROODERUDWHG ZLWK PH RQ YDULRXV DVSHFWV RI WKH UHVHDUFK DP DOVR JUDWHIXO WR WKH IROORZLQJ SHRSOH DQG WKHLU DVVRn FLDWHG LQVWLWXWLRQV -RV 0£ULR )UD]£R DQG &ODXGLR 3LQKHLUR RI WKH ,QVWLWXWR (VWDGXDO GR %DEDSX ,1(%f IRU WKHLU FROODERUDWLRQ LQ VWXGLHV RQ JHUPLQDWLRQ SKHQRORJ\ DQG WD[RQRP\ RI EDEDVVX +HU£FOLWR $TXLQR DQG :DOEHUW &DUYDOKR RI WKH (PSUHVD 0DUDQKHQVH GH 3HVTXLVD $JURSHFX£ULD (0$3$f IRU SURYLGLQJ ILHOG ZRUNHUV 6HUJLR $OYHV 0£ULR 'DQWDV (OL]DEHWK GH 2OLYHLUD 5DLPXQGR 5HJ DQG :DOGHPDU )HUUHLUD RI WKH &HQWUR GH 3HVTXLVD $JURSHFXDULD GR 7USLFR 8PLGR &3$78f IRU LQYDOXDEOH ORJLVWLF VXSSRUW DQG 5DLPXQGR %LWWHQFRXUW :DOGHPDU )UDQFD 1HOVRQ 5RVD DQG 6U 0LOWRQ RI WKH 0XVHX *RHOGL IRU WKHLU DVVLVWDQFH LQ WKH ILHOG LLL

PAGE 4

7KH H[FHOOHQW GUDZLQJV ZHUH E\ *HRUJH )XOOHU DQG P\ EDWWHU\ RI GHGLFDWHG W\SLVWV LQFOXGHG %RE (SWLQJ 'RQQD (SWLQJ DQG /HVOLH 5LJJ /HVV WDQJLEOH EXW mQRQH WKHOHV V PHDQLQJIXO DVVLVWDQFH ZDV SURYLGHG E\ +HOGD /HQ] &HVDU :DUZLFN .HUU 0DULR /HDO 3HWHU 0D\ DQG *KLOOHDQ 3UDQFH ZRXOG OLNH WR H[SUHVV P\ VLQFHUH DSSUHFLDWLRQ WR WKH SHRSOH ZKR PDGH P\ GHGLFDWLRQ WR WKH UHVHDUFK SRVVLEOH &OHPHQWH 5RGULJXHV DQG KLV H[WHQGHG IDPLO\ ,VD DQG 1HOVRQ &DUYDOKR P\ SDUHQWV 0DUWKD DQG 6HQHFD $QGHUVRQ DQG HVSHFLDOO\ P\ ZLIH 6XHO\ $QGHUVRQ )LQDOO\ YHU\ VSHFLDO WKDQNV JR WR P\ VXSHUYLVRU\ FRPPLWWHH 'UV 3HWHU )HLQVLQJHU :DOWHU -XGG +XJK 3RSHQRH )UDQFLV 3XW] DQG HVSHFLDOn O\ P\ FKDLUPDQ -RKQ (ZHO ZKR YLVLWHG WKH ILHOG VLWHV DQG SURYLGHG SDWLHQW JXLGDQFH LQ DOO SKDVHV RI WKH UHVHDUFK ,9

PAGE 5

7$%/( 2) &217(176 $&.12:/('*(0(176 LLL /,67 2) 7$%/(6 YLL /,67 2) ),*85(6 L[ $%675$&7 [LL &+$37(5 ,1752'8&7,21 7$;2120< $1' 3+,72*(2*5$3+< 0HWKRGV 5HVXOWV DQG 'LVFXVVLRQ 5(352'8&7,9( %,2/2*< 0HWKRGV 5HVXOWV 'LVFXVVLRQ (67$%/,6+0(17 0HWKRGV 5HVXOWV 'LVFXVVLRQ *52:7+ $1' 352'8&7,9,7< 0HWKRGV 5HVXOWV 'LVFXVVLRQ 3238/$7,21 6758&785( $1' '<1$0,&6 0HWKRGV 5HVXOWV 'LVFXVVLRQ ,03/,&$7,216 )25 0$1$*(0(17 3HRSOH DQG WKH 3DOP )RUHVW /LPLWLQJ )DFWRUV 3RVVLEOH 6ROXWLRQV Y

PAGE 6

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

PAGE 7

/,67 2) 7$%/(6 8VHV RI RFFXSLHG ODQGV LQ 0DUDQKDR GXULQJ DQG 1RPHQFDWXUDO KLVWRU\ RI EDEDVVX DQG LWV VXVSHFWHG K\EULGV )UXLW \LHOGV RI EDEDVVX SHU KD DQG SHU SDOP GXULQJ DQG DW WKUHH VLWHV 6RXUFHV RI LQFUHDVHG IUXLW \LHOGV SHU KD EHWZHHQ DQG DW WKUHH VLWHV $WWDFN RI EDEDVVX IUXLWV DQG VHHGV E\ WKH VHHGSUHGDFHRXV EHHWOH 3DFK\PHUXV QFOHR UXP *HUPLQDWLRQ RI EDEDVVX LQ IRXU H[SHULPHQWV 3UHOLPLQDU\ FKHFNOLVW RI LQVHFWV WKDW IHHG RQ EDEDVVX +\SRWKHVL]HG VXFFHVV RI EDEDVVX LQ IRXU H[SHULPHQWV 3DOPV SHU PA Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI LUULJDWLRQ H[SHULPHQW 3DOPV SHU PA Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI IHUWLOL]HU H[SHULPHQW 3DOPV SHU PA Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI LQVHFWLFLGH H[SHULPHQW 3DOPV SHU PA Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI FOLPDWH H[SHULPHQW ZLWK FOLPDWH DQG ZHHGLQJ DV PDLQ HIIHFWV 3DOPV SHU PA Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI FOLPDWH H[SHULPHQW ZLWK FOLPDWH DQG HFRW\SH DV PDLQ HIIHFWV /HDI DUHD LQGH[ /$,f RI EDEDVVX DQG FRPSHWLWRUV LQ IRXU H[SHULPHQWV 'U\ PDVV RI EDEDVVX DQG FRPSHWLWRUV LQ IRXU H[SHULPHQWV YLL

PAGE 8

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

PAGE 9

)LJXUH /,67 2) ),*85(6 $SSUR[LPDWH GLVWULEXWLRQ RI KLJKGHQVLW\ VWDQGV RI EDEDVVX 2UELJQ\D PDUWLDQDf LQ %UD]LO $HULDO YLHZ RI EDEDVVX VWDQGV LQ WKH FRXQW\ RI %DFDEDO 0DUDQKDR 6SHFLPHQ RI EDEDVVX B PDUWLDQDf LQ 3LUDSRUD 0LQDV *HUDLV 'HIRUHVWDWLRQ LQ WKH 7URSLFDO 0RLVW )RUHVW ]RQH RI 0DUDQKDR 5HJURZWK RI EDEDVVX RQ D GHIRUHVWHG VLWH LQ WKH 7URSLFDO 0RLVW )RUHVW ]RQH RI 0DUDQKDR 3URGXFWLRQ RI EDEDVVX NHUQHOV LQ 0DUDQKDR DQG %UD]LO )ORZFKDUW RI SURGXFWV GHULYHG IURP EDEDVVX IUXLWV E\ LQGXVn WULHV LQ %DFDEDO 0DUDQKDR GXULQJ )ORZFKDUW RI SURGXFWV GHULYDEOH IURP EDEDVVX IUXLWV YLD FXUUHQW f WHFKQRORJ\ 0DUDQKDR DQG QHLJKERULQJ VWDWHV ZLWK ORFDWLRQV RI UHVHDUFK VLWHV 0DMRU VRLO JURXSV RI 0DUDQKDR DFFRUGLQJ WR WKH %UD]LOLDQ &ODVVLILFDWLRQ 6\VWHP 3ULQFLSDO HFRORJLFDO ]RQHV RI 0DUDQKDR &RYHUDJH RI EDEDVVX VWDQGV LQ 0DUDQKDR 0RQWKO\ UDLQIDOO DW IRXU UHVHDUFK VLWHV 'LVWULEXWLRQ RI ODQG DQG KRXVHKROGV DFFRUGLQJ WR VL]H RI ODQGKROGLQJ LQ UXUDO DUHDV RI 0DUDQKDR GXULQJ DQG 'LVWULEXWLRQ RI VSHFLPHQV RI 2UELJQ\D PDUWLDQD DQG LWV VXVSHFWHG K\EULG FRPSOH[HV L[

PAGE 10

0HDQ PRQWKO\ JDLQ DQG ORVV RI OHDYHV SHU SDOP RQ WKUHH VLWHV LQ 0DUDQK£R 0HDQ PRQWKO\ JDLQ DQG ORVV RI OHDYHV SHU SDOP DW WKUHH VWXG\ VLWHV DW /DJR 9HUGH 0HDQ PRQWKO\ SURGXFWLRQ RI PDOH DQG IXQFWLRQDOO\ IHPDOH LQIORUHVFHQFHV SHU SDOP DW WKUHH VLWHV LQ 0DUDQK£R 0HDQ PRQWKO\ SURGXFWLRQ RI PDOH DQG IXQFWLRQDOO\ IHPDOH LQIORUHVFHQFHV SHU SDOP DW WKUHH VLWHV DW /DJR 9HUGH 'LVWULEXWLRQV RI IHPDOH LQIORUHVFHQFHV SHU SDOP LQ VL[ SRSXODWLRQV RI EDEDVVX 0HDQ PRQWKO\ QXPEHU RI LPPDWXUH DQG PDWXUH LQIUXFWHVFHQFHV SHU SDOP DW WKUHH VLWHV LQ 0DUDQK£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f OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX RYHU WLPH LQ LUULJDWLRQ H[SHULPHQW [

PAGE 11

3DOPV SHU PA Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX RYHU WLPH LQ IHUWLOL]HU H[SHULPHQW 3DOPV SHU PA Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX RYHU WLPH LQ LQVHFWLFLGH H[SHULPHQW 3DOPV SHU PA Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX RYHU WLPH LQ FOLPDWH H[SHULPHQW $EVROXWH DOORFDWLRQ RI DQQXDO DERYHJURXQG GU\ PDWWHU SURGXFWLRQ SHU EDEDVVX SDOP DV D IXQFWLRQ RI DJH 5HODWLYH DOORFDWLRQ RI DQQXDO DERYHJURXQG GU\ PDWWHU SURGXFWLRQ SHU EDEDVVX SDOP DV D IXQFWLRQ RI DJH 0HDQ DQQXDO QXPEHU RI LQIORUHVFHQFHV SURGXFHG SHU SDOP DV D IXQFWLRQ RI PHDQ DJH SHU OLIH VWDJH 0HDQ DQQXDO QXPEHU RI VHHGV SURGXFHG SHU SDOP DV D IXQFWLRQ RI PHDQ DJH SHU OLIH VWDJH 6WDJH GLVWULEXWLRQV RI EDEDVVX RQ WKH SULPDU\ IRUHVW VHFRQGDU\ IRUHVW DQG SDVWXUH VLWHV DW /DJR 9HUGH $JH GLVWULEXWLRQ RI EDEDVVX RQ WKH SULPDU\ IRUHVW VLWH DW /DJR 9HUGH $JH GLVWULEXWLRQ RI EDEDVVX RQ WKH VHFRQGDU\ IRUHVW VLWH DW /DJR 9HUGH 3RSXODWLRQ G\QDPLFV RI EDEDVVX RQ WKH KD SULPDU\ IRUHVW VHFRQGDU\ IRUHVW DQG SDVWXUH VLWHV DW /DJR 9HUGH 6KLIWLQJ FXOWLYDWLRQ VLWH MXVW SULRU WR EXUQ DW /DJR 9HUGH &OHDUFXW RI SDOP IRUHVW SULRU WR SDVWXUH HVWDEOLVKPHQW QHDU %DFDEDO 0DUDQK£R [L

PAGE 12

$EVWUDFW RI 'LVVHUWDWLRQ 3UHVHQWHG WR WKH *UDGXDWH 6FKRRO RI WKH 8QLYHUVLW\ RI )ORULGD LQ 3DUWLDO )XOILOOPHQW RI WKH 5HTXLUHPHQWV IRU WKH 'HJUHH RI 'RFWRU RI 3KLORVRSK\ 7+( %,2/2*< 2) 2UELJQ\D PDUWLDQD 3$/0$(f $ 7523,&$/ '5< )25(67 '20,1$17 ,1 %5$=,/ %\ $QWKRQ\ % $QGHUVRQ 'HFHPEHU &KDLUPDQ -RKQ (ZHO 0DMRU 'HSDUWPHQW %RWDQ\ 0RQRVSHFLILF VWDQGV RI WKH EDEDVVX SDOP 2UELJQ\D PDUWLDQDf RFFXS\ ZLGHVSUHDG DUHDV RI %UD]LO DWWDLQLQJ WKHLU JUHDWHVW FRYHUDJH NPAf LQ WKH VWDWH RI 0DUDQK£R 1HDU /DJR 9HUGH 0DUDQKDR FRQGXFWHG FRPSDUDWLYH VWXGLHV RI EDEDVVXnV OLIH KLVWRU\ WR GHWHUPLQH KRZ WKH SDOP LV PDLQWDLQHG LQ PDWXUH IRUHVWV DV ZHOO DV KRZ LW GRPLQDWHV GLVWXUEHG VLWHV )ORZHULQJ LQ EDEDVVX SHDNHG GXULQJ WKH UDLQ\ VHDVRQ -DQXDU\0DUFKf DQG FRLQFLGHG ZLWK D IOXVK RI QHZ OHDYHV [ SHU \Uf IUXLWLQJ SHDNHG GXULQJ WKH IROORZLQJ GU\ VHDVRQ 2FWREHU'HFHPEHUf 2XWFURVVLQJ ZDV SUHGRPLQDQW DQG ZDV HIIHFWHG E\ ZLQG DQG QLWLGXOLG EHHWOHV 0\VWURSV VSf 'LVSHUVDO RI WKH PDVVLYH Jf IUXLWV ZDV PHGLDWHG SULPDULO\ E\ JUDYLW\ DQG E\ URGHQWV 'XH WR WKH KDUG IUXLW HQGRFDUS VHHG SUHGDn WLRQ ZDV YLUWXDOO\ OLPLWHG WR EUXFKLG ODUYDH 3DFK\PHUXV QFOHR UXPf ZKLFK HQWHUHG WKURXJK WKH JHUPLQDWLRQ SRUHV FD b RI VHHGV VXIIHUHG [LL

PAGE 13

SUHGDWLRQ *HUPLQDWLRQ UHTXLUHG PR DQG ZDV HQKDQFHG LQ VKDGH 6WHP OHVV SDOPV SHUVLVWHG LQ WKH XQGHUVWRU\ RI D SULPDU\ IRUHVW IRU DERXW \U SHUPLWWLQJ EXLOGXS RI KLJK GHQVLWLHV SHU KDf (VFDSH IURP VHHG SUHGDWRUV VKDGH WROHUDQFH DQG DQ HVWLPDWHG OLIHVSDQ RI \U DFFRXQW IRU EDEDVVXnV LPSRUWDQFH LQ VXFK IRUHVWV 'XH WR EXULDO RI WKHLU DSLFDO PHULVWHPV YLD FU\SWRJHDO JHUPLQDWLRQ VWHPOHVV SDOPV VXUn YLYH ZKHQ VLWHV DUH EXUQHG IRU VKLIWLQJ FXOWLYDWLRQ 'XULQJ WKH VXEVHn TXHQW IDOORZ WKHVH SDOPV DUH UHOHDVHG DQG IRUP YLUWXDOO\ PRQRVSHFLILF VWDQGV 2QH VXFK VWDQG ZDV VWLOO LQFUHDVLQJ LQ QXPEHUV DIWHU D IDOORZ RI \U $QQXDO DERYHJURXQG GU\ PDWWHU SURGXFWLRQ RI WKLV VWDQG ZDV FD WfKD \U b RI ZKLFK ZDV DOORFDWHG WR OHDYHV :LWK IDOORZV RI \U HQRXJK IXHO DFFXPXODWHV LQ WKH IRUP RI OHDYHV WR SHUPLW WKH KRW ILUH QHFHVVDU\ IRU ZHHG FRQWURO DQG QXWULHQW UHOHDVH 7KXV VKLIWLQJ FXOWLYDWLRQ FDQ EH SUDFWLFHG ZLWK PLQLPDO WKLQQLQJ RI WKH VWDQGV %DEDVVXnV GRPLQDQFH RI WKH ODQGVFDSH UHVXOWV IURP LWV SHUVLVWHQFH RYHU D ZLGH JUDGLHQW RI GLVWXUEDQFHV %HFDXVH RI LWV FDSDFLW\ WR IRUP PRQRVSHFLILF VWDQGV LWV ORZ PDLQWHQDQFH UHTXLUHPHQWV LWV LQWHJUDWLRQ LQWR PDMRU ODQG XVHV DQG LWV FXUUHQW LPSRUWDQFH LQ PDUNHW DQG VXEVLVWHQFH HFRQRPLHV EDEDVVX VHHPV PRVW DPHQDEOH IRU XVH DV DQ H[WHQVLYH PRGHUDWH\LHOGLQJ FURS [LLL

PAGE 14

&+$37(5 ,1752'8&7,21 1XPHURXV VSHFLHV RI SDOPV GRPLQDWH WKH ODQGVFDSH RYHU H[WHQVLYH DUHDV RI WKH ORZODQG WURSLFV %HDUG 5LFKDUGV 0RRUH Df 3DOPGRPLQDWHG VWDQGV RIWHQ RFFXU RQ HGDSKLFDOO\ OLPLWHG VLWHV ZKHUH FRPSHWLWLRQ LV DSSDUHQWO\ UHGXFHG 0\HUV f 6HYHUDO VSHFLHV KRZn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f 7KH VWDQGV DWWDLQ WKHLU JUHDWHVW H[WHQW LQ WKH VWDWH RI 0DUDQK£R ZKHUH WKH\ FRYHU DQ DUHD HVWLPDWHG DW NPA $QRQ\PRXV f RU URXJKO\ HTXLYDOHQW WR WKH 86 VWDWH RI 9LUJLQLD %DEDVVX IORXULVKHV RQ VLWHV VXEMHFW WR LQWHQVLYH GLVWXUEDQFHV VXFK DV FXWWLQJ DQG EXUQLQJ )LJV f 7KH SDOP LV D FUXFLDO UHVRXUFH IRU UXUDO SRRU ZKR XWLOL]H WKH IUXLWV OHDYHV DQG VWHP IRU D ZLGH UDQJH RI VXEVLVWHQFH SURGXFWV 0DQXDO H[WUDFWLRQ RI NHUQHOV IURP WKH H[WUHPHO\ KDUG IUXLWV UHSUHVHQWV WKH VWDWHnV ODUJHVW FRWWDJH LQGXVWU\ LQYROYLQJ RYHU IDPLOLHVf DQG LWV PRVW LPSRUWDQW LQFRPH VRXUFH IRU ZRPHQ 7KH NHUQHOV DUH XWLO

PAGE 15

? R )LJXUH $SSUR[LPDWH GLVWULEXWLRQ RI KLJKGHQVLW\ VWDQGV RI EDEDVVX 2UELJQ\D PDUWLDQDf LQ %UD]LO 6WDWHV FRQn WDLQLQJ VWDQGV DUH ODEHOOHG $GDSWHG IURP 0HQGHV DQG &DULRFD f

PAGE 16

)LJXUH $HULDO YLHZ RI EDEDVVX VWDQGV QHDU %DFDEDO +DUDQKDR )LJXUH 6SHFLPHQ RI EDEDVVX &0 PDUWLDQDf LQ 3LUDSRUD 0LQDV *HUDLV X!

PAGE 17

)LJXUH 5HJURZWK RI EDEDVVX RQ D GHIRUHVWHG VLWH LQ WKH 7URSLFDO 0RLVW )RUHVW ]RQH RI 0DUDQKDR

PAGE 18

.(51(/ 352'8&7,21 ,222,f )LJXUH 3URGXFWLRQ RI EDEDVVX NHUQHOV LQ 0DUDQK£R DQG %UD]LO 6RXUFHV ,%*( D DQG SUHYLRXV YROXPHV WR f

PAGE 19

)LJXUH )ORZFKDUW RI SURGXFWV GHULYHG IURP EDEDVVX IUXLWV E\ LQGXVWULHV LQ %DFDEDO 0DUDQKDR GXULQJ 'LDPRQGV UHSUHVHQW HLWKHU RU RSWLRQV 1XPEHUV DUH SHUFHQWDJHV RI IUXLW ZHLJKW $GDSWHG IURP 0HQGHV DQG &DULRFD f DQG (PSUHVD ,QGXVWULDO GH %DFDn EDO SHUVRQDO FRPPXQLFDWLRQf

PAGE 20

)LJXUH )ORZFKDUW RI SURGXFWV GHULYDEOH IURP EDEDVVX IUXLWV YLD FXUUHQW f WHFKQRORJ\ 'LDPRQGV UHSUHVHQW HLWKHURU RSWLRQV 1XPEHUV DUH SHUFHQWDJHV RI IUXLW ZHLJKW $GDSWHG IURP $EUHX f 9LYHLURV f (VFROD 7FQLFD )HGHUDO GR 0DUDQKDR f ,37 f DQG 0HQGHV DQG &DULRFD f

PAGE 21

L]HG E\ LQGXVWULHV IRU SURGXFWLRQ RI YHJHWDEOH RLO VRDS DQG IHHGFDNH 7KHVH LQGXVWULHV UHSUHVHQW WKH ODUJHVW VLQJOH VRXUFH RI WD[ UHYHQXHV LQ WKH 0DUDQKDR HFRQRP\ 2YHU WKH SDVW \U NHUQHO SURGXFWLRQ KDV VWHDGn LO\ LQFUHDVHG )LJ f DQG DV RI EDEDVVX SURYLGHG b RI DOO YHJHWDEOH RLO REWDLQHG IURP QRQGRPHVWLFDWHG VRXUFHV LQ %UD]LO ,%*( Df 5HFHQWO\ LQWURGXFHG PDFKLQHU\ FDSDEOH RI PHFKDQLFDOO\ EUHDNLQJ WKH KDUG IUXLWV KDV HQDEOHG DGGLWLRQDO EDEDVVXGHULYHG SURGXFWV HJ FKDUFRDO FRNHf WR HQWHU QDWLRQDO PDUNHWV )LJ f DQG WHFKQRORJ\ SUHVHQWO\ H[LVWV WR GHULYH D KRVW RI DGGLWLRQDO SURGXFWV IURP WKH IUXLWV )LJ f 'HVSLWH LWV HFRQRPLF LPSRUWDQFH EDEDVVX LV QRW FXUUHQWO\ GRPHVn WLFDWHG DQG IRUPDWLRQ RI WKH VWDQGV UHPDLQV D VSRQWDQHRXV SKHQRPHQRQ ,Q WKH SDVW VWDQGV RI EDEDVVX ZHUH JHQHUDOO\ SUHVHUYHG DQG IRUPHG DQ LQWHJUDO SDUW RI VKLIWLQJ FXOWLYDWLRQ V\VWHPV ,Q UHFHQW \HDUV KRZn HYHU SRSXODWLRQ SUHVVXUHV DQG FKDQJLQJ ODQG XVH SDEWHUQV DUH LQGXFLQJ ERWK VKLIWLQJ FXOWLYDWRUV DQG UDQFKHUV WR FOHDUFXW WKH VWDQGV $QGHUVRQ DQG $QGHUVRQ f 6RXQG PDQDJHPHQW RI WKH SDOP IRUHVW UHTXLUHV EDVLF LQIRUPDWLRQ FRQFHUQLQJ WKH ELRORJ\ RI EDEDVVX 7KH SXUSRVH RI WKLV VWXG\ LV WR SURYLGH VXFK LQIRUPDWLRQ 7KH VWXG\ FRYHUV EDEDVVXnV WD[RQRP\ &KDSWHU f OLIH KLVWRU\ &KDSWHUV f DQG SRSXODWLRQ ELRORJ\ &KDSWHU f OHYHOV RI LQTXLU\ FKDQJH UHVSHFWLYHO\ IURP WKH VSHFLHV WR LQGLYLGXDO SDOPV WR VWDQGV RI

PAGE 22

EDEDVVX 7KH VWXG\ FRQFOXGHV ZLWK D GLVFXVVLRQ RI WKH LPSOLFDWLRQV IRU PDQDJHPHQW &KDSWHU f 7KH 6HWWLQJ 0DUDQK£R 7KH %UD]LOLDQ VWDWH RI 0DUDQK£R )LJ f HQFRPSDVVHV D ZLGH UDQJH RI HFRORJLFDO FRQGLWLRQV 7R WKH VRXWK OLHV D JHRORJLFDOO\ ROG 0HVRn ]RLFf HOHYDWHG SHQHSODLQ EURNHQ E\ DEUXSW SODWHDXV DQG FDQ\RQV DQG E\ ORZ HURGHG PRXQWDLQ UDQJHV 7KH SUHGRPLQDQW VRLOV LQ WKLV UHJLRQ DUH GHHSO\ ZHDWKHUHG DQG QXWULHQW SRRU DQG DQQXDO UDLQIDOO UDUHO\ H[FHHGV PP +XPDQ SRSXODWLRQ GHQVLWLHV DUH FRQVHTXHQWO\ ORZ LQKDELn WDQWV SHU NPAf DQG WKH SULQFLSDO ODQG XVH LV JUD]LQJ RQ XQLPSURYHG UDQJH ,Q FHQWUDO DQG QRUWKHUQ 0DUDQK£R HOHYDWLRQV GHFOLQH DQG WKH WHUn UDLQ JHQWO\ XQGXODWHV DV RQH HQWHUV D EURDG SODLQ FRPSULVHG RI JHRORJLn FDOO\ \RXQJ 7HUWLDU\f VHGLPHQWV GLVVHFWHG E\ PDQ\ RI WKH VWDWHnV PDMRU ULYHUV 6RLOV DUH H[WUHPHO\ YDULDEOH )LJ f ZLWK DQ DEXQGDQFH RI QXWULHQWULFK DOOXYLXP DQG DQQXDO UDLQIDOO UDQJHV IURP WR RYHU PP $JULFXOWXUH PRVW FRPPRQO\ VKLIWLQJ FXOWLYDWLRQf DQG JUD]LQJ FRQVWLWXWH WKH SULQFLSDO ODQG XVHV 2XWVLGH WKH FDSLWDO 6£R /XLVf 2 SRSXODWLRQ GHQVLWLHV LQ WKLV UHJLRQ DYHUDJH LQKDELWDQWV SHU NP 7KH EDEDVVX IRUHVWV RI 0DUDQK£R DUH ODUJHO\ FRQILQHG WR WKH FHQWUDO DQG QRUWKHUQ SRUWLRQV RI WKH VWDWH ZKHUH WKH\ DUH DVVRFLDWHG ZLWK WKUHH EURDGO\ GHILQHG HFRORJLFDO ]RQHV )LJ f VHDVRQDOO\ ZHW VDYDQQD EDL[DGDf ZRRGODQG VFUXE RU GU\ VDYDQQD FHUUDGRf DQG PRLVW RU GU\ IRUHVW PDWDf

PAGE 23

)LJXUH 0DUDQK£R DQG QHLJKERULQJ VWDWHV ZLWK ORFDWLRQV RI UHVHDUFK VLWHV

PAGE 24

)LJXUH 0DMRU VRLO JURXSV RI 0DUDQK£R DFFRUGLQJ WR WKH %UD]LOLDQ FODVVLILFDWLRQ V\VWHP (,%5$3$ f + +\GURPRUSKLF /DWHULWH 3G '\VWURSKLF 5HG
PAGE 25

0$1*529( $1' %($&+ 9(*(7$7,21 J 6($621$//< :(7 6$9$11$ e :22'/$1'6&58% $1' '5< 6$9$11$ P 02,67 )25(67 NP )LJXUH 3ULQFLSDO HFRORJLFDO ]RQHV RI 0DUDQKDR 6RXUFH .XKOPDQQ f

PAGE 26

)LJXUH &RYHUDJH RI EDEDVVX VWDQGV LQ 0DUDQKDR 6RXUFH $QRQ\PRXV f

PAGE 27

6HDVRQDOO\ ZHW VDYDQQD} $Q H[WHQVLYH DUHD WUDYHUVHG E\ WKH ORZHU UHDFKHV RI 0DUDQKDRnV PDMRU ULYHUV LV IORRGHG GXULQJ WKH 'HFHPEHU0D\ UDLQ\ VHDVRQ DQG EXUQHG GXULQJ WKH GU\ VHDVRQ 7KLV ORZO\LQJ DUHD LV FDOOHG EDL[DGD 7KH SUHGRPLQDQW VRLOV JURXQGZDWHU ODWHULWHV DQG JOH\Vf DUH K\GURPRUSKLF WKH PDMRU YHJHWDWLRQ W\SH LV D ORZ VDYDQQD GRPLQDWHG E\ VHGJHV DQG JUDVVHV DQG WKH SULQFLSDO ODQG XVH LV JUD]LQJ RI FDWWOH DQG EXIIDOR RQ XQLPSURYHG UDQJH 7KURXJKRXW WKLV UHJLRQ IRUHVWV RI EDEDVVX DUH JHQHUDOO\ OLPLWHG WR WRSRJUDSKLFDOO\ HOHYDWHG LVODQGV ZKLFK DUH DOVR WKH SULQFLSDO VLWHV IRU VKLIWLQJ FXOWLYDWLRQ :RRGODQG VFUXE RU GU\ VDYDQQD 7KH WHUP FHUUDGR UHIHUV WR ERWK DQ H[WHQVLYH SK\WRJHRJUDSKLF UHJLRQ RI %UD]LO DQG WKH UHJLRQnV SUHGRPLQDQW YHJHWDWLRQ :LWKLQ 0DUDQK£R FHUUDGR LV FKDUDFWHUL]HG E\ DQQXDO UDLQn IDOO PP D GU\ VHDVRQ RI PR RU ORQJHU DQG ZHOO WR H[FHVVLYHO\ GUDLQHG VRLOV 7KH YHJHWDWLRQ FRQVLVWV RI GHFLGXRXV ZRRGODQG VFUXE DQG RFFDVLRQDO VDYDQQD 7KH SUHGRPLQDQW ODQG XVH LV JUD]LQJ RI FDWWOH JRDWV DQG VKHHS RQ XQLPSURYHG UDQJH 6KLIWLQJ FXOWLYDWLRQ DQG SHUPDn QHQW SODQWDWLRQV RI VXJDUFDQH DQG FRWWRQ DUH JHQHUDOO\ FRQILQHG WR WKH PRLVW ERWWRPODQGV ZKLFK DUH DOVR WKH KDELWDWV RFFXSLHG E\ EDEDVVX IRUHVWV 0RLVW RU GU\ IRUHVW 7KLV ]RQH LV FKDUDFWHUL]HG E\ PRGHUDWH WR KLJK UDLQIDOO DQG D GU\ VHDVRQ RI PR ,W H[WHQGV DFURVV ERWK 7URSLFDO 0RLVW )RUHVW DQG 7URSLFDO 'U\ )RUHVW /LIH =RQHV VHQVX +ROGULGJH f 7KH YHJHWDWLRQ ZDV RULJLQDOO\ WDOO GLYHUVH IRUHVW UDQJLQJ IURP HYHUn JUHHQ WR GHFLGXRXV 'XH WR ZLGHVSUHDG GHIRUHVWDWLRQ GXULQJ WKH SDVW \U WKH RULJLQDO IRUHVW KDV EHHQ YLUWXDOO\ HOLPLQDWHG DQG WRGD\ VWDQGV

PAGE 28

RI EDEDVVX GRPLQDWH WKH ODQGVFDSH 'XULQJ WKH SDVW \U UDSLG LQIOX[ RI VHWWOHUV IURP WKH DULG %UD]LOLDQ QRUWKHDVWf§FRPELQHG ZLWK DQ DPELn WLRXV URDGEXLOGLQJ SURJUDP WKURXJKRXW WKH VWDWH f§ KDYH SXVKHG KXPDQ VHWWOHPHQW GHHS LQWR 7URSLFDO 0RLVW )RUHVW DQG H[WHQVLYH DUHDV DUH SUHVHQWO\ RFFXSLHG E\ \RXQJ VWDQGV RI EDEDVVX 7KH SULQFLSDO ODQG XVHV WKURXJKRXW WKH PRLVW RU GU\ IRUHVW ]RQH DUH VKLIWLQJ FXOWLYDWLRQ DQG FDWWOH UDQFKLQJ ,Q FRQWUDVW WR RWKHU HFRORJLFDO ]RQHV LQ 0DUDQKDR UDQFKLQJ KHUH LV FDUULHG RXW SULPDULO\ RQ LPSURYHG SDVWXUHV 7KH SDOP IRUHVW ,Q WKH PDMRU HFRORJLFDO ]RQHV RI 0DUDQKDR GHVn FULEHG DERYH EDEDVVX VWDQGV DUH JHQHUDOO\ IRXQG RQ WKH VLWHV PRVW LQWHQVLYHO\ XWLOL]HG E\ SHRSOH 2Q VXFK VLWHV FURSV DUH HVWDEOLVKHG DQG SDVWXUHV PDLQWDLQHG E\ SHULRGLF FXWWLQJ DQG EXUQLQJ 7KHVH SUDFn WLFHV DFFRXQW IRU WKH RULJLQ RI WKH EDEDVVX VWDQGV LQ 0DUDQKDR 3ULRU WR WKH DUULYDO RI (XURSHDQV 0DUDQKDR KDG ORQJ EHHQ RFFXSLHG E\ QXPHURXV LQGLJHQRXV JURXSV GHVFULEHG LQ 6WHZDUG f $PRQJ WKH GRPLQDQW VHGHQWDU\ WULEHV HJ 7XSLQDPED 7LPELUDf DJULFXOWXUH ZDV JHQHUDOO\ FRQILQHG WR WKH PRLVW ERWWRPODQGV DQG YLOODJHV ZHUH HVWDEn OLVKHG QHDUE\ 'XULQJ WKH GU\ PRQWKV ERWWRPODQG VWDQGV RI EDEDVVX DQG EXULWL 0DXULWLD IOH[XRVDf SDOPV SURYLGHG DQ LPSRUWDQW GLHWDU\ VXSSOHn PHQW $V ZLWK WKH PRUH QRPDGLF WULEHV WHUULWRULDO ULJKWV ZHUH LGHQWLn ILHG ZLWK DFFHVV WR WKHVH VWDQGV ZKLFK VHUYHG DV D VRXUFH RI ILEHU DQG IXHO DV ZHOO DV D KHGJH DJDLQVW IDPLQH +LJK GHQVLWLHV RI EDEDVVX ZHUH SUREDEO\ OLPLWHG WR WKHVH ERWWRPODQG VLWHV 7KH RULJLQDO QDWXUH RI WKH IDU PRUH H[WHQVLYH XSODQG IRUHVWV FDQ EH GLVFHUQHG LQ WKH UHPQDQWV WKDW

PAGE 29

VWLOO H[LVW EDEDVVX ZDV FRPPRQO\ SUHVHQW EXW LQYDULDEO\ DVVRFLDWHG ZLWK D KLJK GLYHUVLW\ RI KDUGZRRGV :LWK WKH DUULYDO RI (XURSHDQV GXULQJ WKH V WKH ODQGVFDSH RI 0DUDQK£R FKDQJHG ,Q UHVSRQVH WR WKH QHHGV RI WKH FRORQLDO HFRQRP\ SODQWDWLRQV RI ULFH FRWWRQ DQG VXJDUFDQH ZHUH HVWDEOLVKHG RQ ERWWRPn ODQGV RYHU DQ H[WHQVLYH DUHD RI WKH VWDWH
PAGE 30

FLHQW IRU UDLQIHG DJULFXOWXUH RQ XSODQG VLWHV )LJ f 0HDQ PRQWKO\ WHPSHUDWXUH DQG UHODWLYH KXPLGLW\ DUH VXEMHFW WR PRGHUDWH YDULDWLRQV LQ UHVSRQVH WR UDLQIDOO WKH IRUPHU UDQJHV IURP & )HEUXDU\f WR & 1RYHPEHUf DQG WKH ODWWHU IURP b 2FWREHUf WR b -DQXDU\ )HEUXDU\f 7KURXJKRXW WKH UHJLRQ WKH WHUUDLQ LV IODW WR JHQWO\ UROOLQJ DQG LV GLVVHFWHG E\ QXPHURXV VWUHDPV WKDW IHHG LQWR WKH 0HDULP 5LYHU 7KH SUHGRPLQDQW VRLO )LJ f LV WKH (XWURSKLF 5HG
PAGE 31

5$,1)$// FPf )LJXUH 0RQWKO\ UDLQIDOO DW IRXU UHVHDUFK VLWHV 1XPEHUV DUH ORQJn WHUP DQQXDO PHDQV EDUV DUH ORQJWHUP PRQWKO\ PHDQV VKDGHG FLUFOHV DUH PRQWKO\ GDWD IRU DQG RSHQ FLUFOHV DUH PRQWKO\ GDWD IRU 3HULRG RYHU ZKLFK ORQJWHUP PHDQV ZHUH FDOFXODWHG DQG VRXUFH RI XQSXEOLVKHG GDWD IRU HDFK VLWH DUH %HOHP &HQWUR GH 3HVTXLVD $JURSHFXDULD GR 7USLFR 8PLGR &3$78f 3LQGDUH 0LULP 6XSHULQn WHQGHQFLD GH 'HVHQYROYLPLHQWR GR 1RUGHVWH 68'(1(f %DFDEDO ,QVWLWXWR 1DFLRQDO GH 0HWHRURORJD ,10(7f DQG &D[LDV ,10(7

PAGE 32

WKH PRVW DJULFXOWXUDOO\ SURGXFWLYH DUHDV LQ 0DUDQKDR
PAGE 33

/$1' bf .Pf NPf )LJXUH 'LVWULEXWLRQ RI ODQG DQG KRXVHKROGV DFFRUGLQJ WR VL]H RI ODQGKROGLQJ LQ UXUDO DUHDV RI 0DUDQKDR GXULQJ DQG 6L]H RI KROGLQJ GHILQHG E\ HIIHFWLYH RFFXSDWLRQ RI WKH ODQG $PRXQW RI ODQG LQFUHDVHG IURP WR GXH WR IURQWLHU H[SDQVLRQ 6RXUFHV ,%*(

PAGE 34

6KLIWLQJ FXOWLYDWLRQ 7KURXJKRXW 0DUDQKDR VKDUHFURSSLQJ LV WKH PHDQV E\ ZKLFK WKH YDVW PDMRULW\ RI FXOWLYDWRUV JDLQ DFFHVV WR ODQG &URSV DUH UDLVHG SULQFLSDOO\ E\ VKLIWLQJ FXOWLYDWLRQ 7DEOH f 7KH IRUHVW LV IHOOHG E\ D[HV DQG EXUQHG WRZDUG WKH HQG RI WKH GU\ VHDVRQ 3ODQWLQJ XVXDOO\ WDNHV SODFH LQ 'HFHPEHU ZLWK WKH ILUVW ZHW VHDVRQ UDLQV ,Q WKH VWXG\ UHJLRQ WKH SULQFLSDO FURSV DUH XSODQG ULFH FDVVDn YD PDL]H DQG EHDQV $V ULFH LV VHQVLWLYH WR GURXJKW DQG FDVVDYD LV LQWROHUDQW RI SURORQJHG ZDWHUORJJLQJ IDUPHUV RIWHQ SODQW ERWK DQG LQWHUFURSSLQJ LV WKH UXOH &XOWLYDWHG DUHDV DUH IUHTXHQWO\ ZHHGHG XQWLO WKH $SULO-XQH ULFH KDUYHVW FDVVDYD PD\ EH KDUYHVWHG XS WR \U DIWHU SODQWLQJ &XOWLYDWHG SORWV RQ XSODQG VLWHV DUH YLUWXDOO\ QHYHU XVHG IRU WZR FURSV LQ D URZ VR QHZ SORWV DUH UHTXLUHG HDFK \HDU 6KRUW FURSSLQJ SHULRGV PDNH WKLV IRUP RI ODQG XVH H[WUHPHO\ H[WHQVLYH 6KLIWLQJ FXOWLn YDWLRQ RI WKLV VRUW FDQ RQO\ UHPDLQ YLDEOH LQ DUHDV RI ORZ SRSXODWLRQ GHQVLWLHV 2YHU ODUJH SRUWLRQV RI 0DUDQK£R WKLV LV QR ORQJHU WKH FDVH )URP WR ODQG DUHD LQ 0DUDQKDR XWLOL]HG IRU WHPSRUDU\ FURSSLQJ LQFUHDVHG WKUHHIROG ZKLOH WKH UDWLR RI IDOORZHG WR FURSSHG ODQG ZDV FXW LQ KDOI 7DEOH f 7KH LPPHGLDWH LPSOLFDWLRQV RI VKRUWHQHG IDOORZV IRU FURS \LHOGVf§DQG ORQJWHUP LPSOLFDWLRQV IRU VLWH TXDOLW\f§ DUH REYLRXV ,QFUHDVHG SRSXODWLRQ GHQVLWLHV DUH WKH SULQFLSDO FDXVH RI VKRUWHQHG IDOORZV WKURXJKRXW 0DUDQKDR %XW H[FHSWLRQDOO\ VKRUW IDOORZV LQ WKH VWXG\ UHJLRQ DUH D FRQVHTXHQFH RI ZLGHVSUHDG FRQYHUVLRQ WR SDVn WXUH

PAGE 35

7DEOH 8VHV RI RFFXSLHG ODQGV LQ 0DUDQKDR GXULQJ 1XPEHUV UHSUHVHQW KD SHUFHQW RI WRWDO SDUHQWKHVHVf 6RXUFHV ,%*( DQG FODVVLILHG LQ
PAGE 36

*UD]LQJ ,Q FRQWUDVW WR WKH UHVW RI 0DUDQK£R JUD]LQJ LQ WKH VWXG\ UHJLRQ LV FDUULHG RXW SUHGRPLQDQWO\ RQ LPSURYHG SDVWXUHV 7KLV KDV EHHQ D UHFHQW GHYHORSPHQW LQ LPSURYHG SDVWXUHV FRPSULVHG RQO\ b RI 0DUDQKDRnV WRWDO RFFXSLHG DUHD 7DEOH f 7KH IRUPDWLRQ RI LPSURYHG SDVWXUHV KDV HQDEOHG JUD]LQJ WR HQFURDFK LQWR DUHDV RQFH XVHG H[FOXVLYHn O\ IRU VKLIWLQJ FXOWLYDWLRQ &RQYHUVLRQ WR SDVWXUH LV JHQHUDOO\ FDUULHG RXW DV SDUW RI VKLIWLQJ FXOWLYDWLRQ 2Q XSODQG SODQWDWLRQV RI FURSV VXFK DV ULFH DQG PDL]H EXW UDUHO\ FDVVDYD GXH WR LWV UHODWLYHO\ ORQJ FURSSLQJ SHULRGf VHHGV RI WKH $IULFDQ JUDVV +\SDUUKHQLD UXID NQRZQ ORFDOO\ DV ODM HDGRf DUH LQWURGXFHG MXVW SULRU WR KDUYHVW 0DUFK0D\f 8VH RI IHUWLOL]HUV LV UDUH DQG OHJXPHJUDVV DVVRFLDWLRQV DUH DSSDUHQWO\ XQNQRZQ 3DVWXUHV DUH XVXDOO\ EXUQHG GXULQJ WKH GU\ VHDVRQ DQG WKH EXUQ RIWHQ JHWV RXW RI FRQWURO ZLOGILUHV DUH D FRQVSLFXRXV IHDWXUH RI WKH ODQGVFDSH WKURXJKRXW 0DUDQK£R %DEDVVX 7KH VWXG\ UHJLRQ LV FKDUDFWHUL]HG E\ H[FHSWLRQDOO\ KLJK FRYHUDJH RI EDEDVVX 6SRQWDQHRXV RFFXUUHQFH RI WKH SDOP DSSHDUV WR EH OLPLWHG RQO\ RQ VWHHS KLOOWRSV ZKHUH IUXLW GLVSHUVDO PD\ OLPLW FRORQn L]DWLRQf DQG ZDWHUORJJHG VRLOV ZKHUH URRW UHVSLUDWLRQ PD\ EH LQKLEn LWHGf (OVHZKHUH WKH SDOP IRUHVW YLUWXDOO\ EODQNHWV WKH ODQGVFDSH H[FHSW ZKHUH LW LV WKLQQHG RU FOHDUFXW E\ VKLIWLQJ FXOWLYDWRUV DQG UDQFKHUV $OWKRXJK FOHDUFXWWLQJ RI EDEDVVX LV LOOHJDO LQ 0DUDQK£R WKH SUDFWLFH DSSHDUV WR EH LQFUHDVLQJO\ FRPPRQ HVSHFLDOO\ LQ DUHDV ZKHUH SDVWXUH FRQYHUVLRQ LV ZLGHVSUHDG

PAGE 37

6WXG\ 6LWHV /DJR 9HUGH 7KH HFRORJLFDO UHVHDUFK ZDV FHQWHUHG RQ DQ KD IDUP LQ WKH FRXQW\ RI /DJR 9HUGH FD NP 1: RI %DFDEDO )LJ f 7KH IDUP LV IODW DQG GLVVHFWHG E\ VHYHUDO LQWHUPLWWHQW VWUHDPV 6RLOV DUH JHQHUn DOO\ IHUWLOH DQG ZHOOGUDLQHG 7KH SUHGRPLQDQW VRLO W\SH LV (XWURSKLF 5HG
PAGE 38

LQIRUPDQWV UHSRUWHG WKDW WKH VLWH ZDV VWLOO FRYHUHG E\ VWUXFWXUDOO\ FRPSOH[ VSHFLHVULFK IRUHVW ZKHQ WKH\ ILUVW DUULYHG LQ WKH UHJLRQ GXULQJ WKH V 6LQFH WKLV WLPH WKH VLWH KDV EHHQ VXEMHFWHG WR PLQLPDO GLVWXUEDQFH $W WKH WLPH RI P\ VWXG\ HPHUJHQWV DWWDLQHG P DQG WKH FDQRS\ ZDV PRUH RU OHVV FRQWLQXRXV DW P 0RVW RI WKH HPHUJHQW VSHFLHV ZHUH GHFLGXRXV GXULQJ DOO RU SDUW RI WKH GU\ VHDVRQ 9LQHV ZHUH UHODWLYHO\ DEXQGDQW ZKHUHDV HSLSK\WHV ZHUH UDUH 6SHFLHV ULFKQHVV DW VSHFLHV GEK L/ FPf SHU KD ZDV ORZHU WKDQ WKDW RI XSODQG IRUHVWV LQ $PD]RQLD $QGHUVRQ DQG %HQVRQ f :LWK UHODWLYHO\ KLJK DEXQGDQFH bf IUHTXHQF\ bf DQG GRPLQDQFH bf EDEDVVX ZDV WKH PRVW LPSRUWDQW VSHFLHV RQ WKH VLWH 7KH VHFRQGDU\ IRUHVW ZDV RULJLQDOO\ FXW DQG EXUQHG IRU VKLIWLQJ FXOWLYDWLRQ LQ WKH VLWH KDG VXIIHUHG QR PDMRU GLVWXUEDQFH VLQFH (PHUJHQWV DWWDLQHG P DQG WKH FDQRS\ f§ FRPSULVHG ODUJHO\ RI D XQLIRUP VWDQG RI EDEDVVXf§ZDV UHODWLYHO\ FRQWLQXRXV DW P 6SHFLHV ULFKQHVV ZDV VSHFLHV GEK A FPf SHU KD 7KH SHUFHQWDJH RI WRWDO VSHFLHV WKDW RFFXUUHG LQ ERWK WKH SULPDU\ DQG VHFRQGDU\ IRUHVW VLWHV FRHIILFLHQW RI FRPPXQLW\ :KLWWDNHU f ZDV b ,Q DEXQGDQFH bf IUHTXHQF\ bf DQG GRPLQDQFH bf EDEDVVX ZDV IDU DQG DZD\ WKH PRVW LPSRUWDQW VSHFLHV RQ WKH VLWH 7KH SDVWXUH ZDV HVWDEOLVKHG WKURXJK LQWURGXFWLRQ LQ RI WKH H[RWLF JUDVV +\SDUUKHQLD UXID RQ D VLWH WKDW KDG DOUHDG\ XQGHUJRQH VHYHUDO F\FOHV RI VKLIWLQJ FXOWLYDWLRQ 7KH DEXQGDQW XQGHUVWRU\ FRPSHWn LWRUV LQFOXGLQJ /DQWDQD FDPDUD 6WDFK\WDUSKHWD FDMDQHQHVLV 6LGD VSS

PAGE 39

*\ SH UXV VSS DQG MXYHQLOHV RI EDEDVVXf ZHUH DQQXDOO\ FXW DQG EXUQHG 7KH P KLJK FDQRS\ ZDV FRPSULVHG H[FOXVLYHO\ RI EDEDVVX ZKLFK RFFXUUHG DW D GHQVLW\ RI FD VWHPV SHU KD 6DWHOOLWH VLWHV ,Q DGGLWLRQ WR WKH SULQFLSDO ILHOG VLWH DW /DJR 9HUGH VDWHOOLWH VLWHV ZHUH HVWDEOLVKHG DW %HOP 3LQGDU 0LULP DQG &D[LDV )LJ f 7KH %HOHP VLWH ZDV XVHG IRU H[SHULPHQWV LQYROYLQJ HVWDEOLVKPHQW &KDSWHU f} WKH 3LQGDU 0LULP DQG &D[LDV VLWHV ZHUH XVHG IRU SKHQRORJLFDO REVHUYDWLRQV &KDSWHU f 0HWHRURORJLFDO GDWD )LJ f VKRZ D ZLGH UDQJH RI UDLQIDOO DW WKHVH VLWHV ,QIRUPDWLRQ FRQn FHUQLQJ ODQG XVHV DQG H[SHULPHQWDO GHVLJQ DW HDFK VLWH LV SURYLGHG LQ WKH DSSURSULDWH FKDSWHUV

PAGE 40

&+$37(5 7$;2120< $1' 3+<72*(2*5$3+< %DEDVVX EHORQJV WR WKH FRFRVRLG JURXS RQH RI PDMRU JURXSV ZLWKLQ WKH SDOP IDPLO\ 0RRUH Ef &RFRVRLG SDOPV DUH GLVWLQJXLVKHG IURP WKH RWKHU JURXSV E\ WKHLU IUXLWV ZKLFK FRQWDLQ D WKLFN JHQHUDOO\ WKUHHSRUHG HQGRFDUS 7KH SULQFLSDO FHQWHU RI FRFRVRLG GLYHUVLW\ LV 6RXWK $PHULFD ZKHUH RI JHQHUD RFFXU 0RRUH Ef 7ZR ZHOO NQRZQ UHSUHVHQWDWLYHV RI WKH JURXS DUH WKH FRFRQXW &RFRV QXFLIHUDf DQG WKH $IULFDQ RLO SDOP (ODHLV JXLQHHQVLVf 7KH WD[RQRP\ RI EDEDVVX KDV EHHQ D VRXUFH RI FRQIXVLRQ VLQFH WKH SDOP ZDV ILUVW GHVFULEHG ZHOO RYHU D FHQWXU\ DJR 7KLV FRQIXVLRQ EHJLQV RQ WKH JHQHULF OHYHO $OWKRXJK EDEDVVX LV JHQHUDOO\ DVVLJQHG WR WKH JHQXV 2UELJQYD WKH VWDWXV RI WKLV DQG WKH IRXU WR ILYH RWKHU JHQHUD LQ WKH $WWDOHD DOOLDQFH KDV EHHQ TXHVWLRQHG E\ :HVVHOV %RHU f 7KHVH JHQHUD DUH GLVWLQJXLVKHG VROHO\ RQ WKH EDVLV RI VWDPLQDWH IORZHU PRUn SKRORJ\ 1R FRUUHODWLRQ ZLWK RWKHU FKDUDFWHUV KDV EHHQ REVHUYHG DQG VWDPLQDWH IORZHU PRUSKRORJ\ UXQV DFURVV D VHSDUDWLRQ EDVHG RQ RWKHU FKDUDFWHUV VXFK DV HQGRFDUS SRUHV /LQHV EHWZHHQ JHQHUD LQ WKH $WWDOHD DOOLDQFH DUH IXUWKHU EOXUUHG E\ FROOHFWLRQV RI LQWHUPHGLDWH VWDPLQDWH IORZHU W\SHV ZKLFK KDYH SURPSWHG WKH GHVFULSWLRQ RI WZR QHZ JHQHUD RQH RI WKHVH 0DUNOHYD ZDV GHVFULEHG E\ %RQGDU f DV D SRVVLEOH LQWHUn JHQHULF K\EULG LQYROYLQJ EDEDVVX $ FODVVLILFDWLRQ EDVHG VROHO\ RQ VWDPLQDWH IORZHU PRUSKRORJ\ LV RI OLWWOH XVH LQ WKH ILHOG DV PRVW VSHFLHV IORZHU RQO\ IRU D VKRUW WLPH

PAGE 41

7KXV LW LV QR VXUSULVH WKDW EDEDVVX LV RIWHQ FRQIXVHG ZLWK RWKHU VSHFLHV DQG HYHQ JHQHUD QRWDEO\ $WWDOHD 2LO SDOPV FRPPHUFLDOO\ UHIHUUHG WR DV EDEDVVX LQFOXGH 2UELJQYD DJUHVWLV LQ 3DUD $WWDOHD ROHIHUD LQ *RL£V DQG 0LQDV *HUDLV $ JHUDHQVLV LQ 0LQDV *HUDLV DQG $ SLQGREDVVX LQ %DKLD 0DUNOH\ f %RWDQLVWV DUH QRW LPPXQH WR JHQHULF FRQIXVLRQ $ERULJLQDO GLVSHUVDO ZDV HYRNHG WR DFFRXQW IRU WKH GLVMXQFW RFFXUUHQFH RI EDEDVVX LQ WKH VWDWH RI 6DR 3DXOR )HUUL f DOWKRXJK WKH SRSXODWLRQ KDV EHHQ LGHQWLILHG DV SHUWDLQLQJ WR $WWDOHD 7 GH 0 &RVWD SHUV FRPPf 7R UHVROYH LQWHUJHQHULF FRQIXVLRQ LQ WKH $WWDOHD DOOLDQFH :HVVHOV %RHU f UHFRPPHQGHG UHGXFLQJ 2UELJQYD DQG UHODWHG JHQHUD WR WKH VLQJOH JHQXV $WWDOHD EXW KLV VXJJHVWLRQ KDV QRW EHHQ IROORZHG E\ RWKHU FRQWHPSRUDU\ WD[RQRPLVWV 0RRUH E *ODVVPDQ f &RXQWHUYDLOLQJ DUJXPHQWV LQFOXGH WKH ZHLJKW RI WD[RQRPLF WUDGLWLRQ DQG WKH REVHUYDWLRQ WKDW :HVVHOV %RHUnV VFKHPH PHUHO\ UHGXFHV WKH FRQIXVLRQ WR WKH VSHFLILF OHYHO + ( 0RRUH SHUV FRPPf 'XH WR WKH XQFHUWDLQ OLPLWV EHWZHHQ JHQHUD WKH SRVVLEOH RFFXUUHQFH RI LQWHUJHQHULF K\EULGV DQG WKH SUDFWLFDO GLIILFXOW\ RI GLVWLQJXLVKLQJ JHQHUD LQ WKH ILHOG EHOLHYH WKDW WKH FXUUHQW VSOLWWLQJ RI WKH $WWDOHD DOOLDQFH LQWR ILYH JHQHUD 0RRUH Ef LV QRW MXVWLILHG )LQDO UHVROXWLRQ RI WKH FRQIXVHG JHQHULF OLPLWV ZLWKLQ WKH $WWDOHD DOOLDQFH ZLOO UHTXLUH H[KDXVWLYH VWXG\ RI WKH HQWLUH JURXS $V VXFK D VWXG\ LV QRW RQ WKH KRUL]RQ DW SUHVHQW VKDOO IROORZ WD[RQRPLF WUDGLWLRQ DQG FRQVLGHU EDEDVVX ZLWKLQ WKH JHQXV 2UELJQYD

PAGE 42

2Q WKH VSHFLILF OHYHO DW OHDVW QDPHV DVVRFLDWHG ZLWK EDEDVVX KDYH EHHQ SXEOLVKHG ZLWKLQ WKH JHQXV 2UELDQYD 7DEOH f 0RVW RI WKLV SUROLIHUDWLRQ KDV UHVXOWHG IURP VSHFLHV GHVFULSWLRQV EDVHG RQ LQDGHTXDWH ERWDQLFDO VSHFLPHQV D SUREOHP LQKHUHQW WR SDOP WD[RQRP\ LQ JHQHUDO 7RPOLQVRQ f ,Q WKH PRUH UHFHQW WD[RQRPLF OLWHUDWXUH HJ 5L]]LQL %RQGDU *ODVVPDQ f PRVW RI WKH QDPHV OLVWHG LQ 7DEOH DUH FRQVLGHUHG V\QRQ\PRXV DQG LW LV JHQHUDOO\ DJUHHG WKDW WKH EDEDVVX FRPSOH[ FRQVLVWV RI D PD[LPXP RI WKUHH VSHFLHV 'XH WR SRWHQWLDO FRQIXVLRQ DPRQJ WKHVH VSHFLHV HPEDUNHG RQ D WD[RQRPLF VWXG\ RI WKH EDEDVVX FRPSOH[ 0\ REMHFWLYHV ZHUH f WR LGHQWLI\ WKH VSHFLHV WKDW ZDV WKH VXEMHFW RI WKH HFRORJLFDO VWXGLHV GHVFULEHG LQ WKH IROORZLQJ FKDSWHUV DQG f WR GLVWLQJXLVK LW IURP PRUSKRORJLFDOO\ VLPLODU VSHFLHV 0HWKRGV /DFN RI DGHTXDWH VSHFLPHQV KDV EHHQ WKH SULQFLSDO LPSHGLPHQW WR UHVROYLQJ WKH WD[RQRP\ RI WKH EDEDVVX FRPSOH[ 7R UHPHG\ WKLV VLWXn DWLRQ FROOHFWHG UHSUHVHQWDWLYHV RI WKH FRPSOH[ RYHU D VXEVWDQWLDO SRUWLRQ RI LWV SUHVHQWO\ NQRZQ UDQJH )LJ f 'XH WR WKH XQXVXDO VWUXFWXUH DQG EXONLQHVV RI ODUJH SDOPV VXFK DV WKRVH LQ WKH EDEDVVX FRPSOH[ FRPSOHWH ERWDQLFDO FROOHFWLRQV UHTXLUH H[WHQVLYH ILHOG GRFXPHQn WDWLRQ UHFRPPHQGHG SURFHGXUHV DUH GHVFULEHG DW OHQJWK LQ %DOLFN HW DO f 6SHFLPHQV RI DOO FROOHFWLRQV ZHUH GHSRVLWHG DW WKH 0XHX 3DUDHQVH (PLOLR *RHOGL LQ %HOHP %UD]LO GXSOLFDWHV ZHUH VHQW WR RWKHU %UD]LOLDQ LQVWLWXWLRQV &(1$5*(1 ,13$ &3$78f DQG WKH 1HZ
PAGE 43

7DEOH 1RPHQFODWXUD KLVWRU\ RI EDEDVVX DQG LWV VXVSHFWHG K\EULGV VFLHQWLILF 1$0( 38%,, &$7,21 0DUWOXV 'UXGH % 5RGULJXHV % 5RGULJXHV f %XUUHW %XUUHW %XUUHW %RQGDU r} %RQGDU : %RHU &ODVVPDQ $WWDOHD VSHFLRVD 0DUW 'HVFULEHG PDUW ODQD m r EDUERVODQD 2UEOJQ\D O\GODH 2UXGH 'HVFULEHG $ O\GODH 2UEOJQ\D PDUW ODQD %DUE 5RGU 'HVFULEHG PDUW ODQD ROHOIHUD 2UEOJQ\D PDFURVWDFK\D 'UXGH H[ %DUE 5RGU VSHFORVD m GDPPHUODQD FRKXQH 3DUW 0DUWf %DUE 5RGU PDUWODQD EDUERVODQD ROHIHUD 2UEOJQ\D GDPQHU ODQD 'HVFULEHG 3DUW ‘ 3DUW %DWK 5RGU e VSHF ORVD FRKXQH $WWDOHD O\GODH 'HVFULEHG ‘ e O\GODH f EDUERVODQD 72UXGHO %DUE 5RGU 2UEOJQ\D PDFURSHWDO D 'HVFULEHG $ VSHF ORVD EDUERVODQD %XU UHW 2UEOJQ\D KXHEQHU, %XUUHW 'HVFULEHG 8QFHUWDLQ VSHFLHV 2E,JQ\D EDUERV ODQD %XUUHW 3URSRVHG $ VSHFORVD EDUERVODQD 2UE JQ\D ROHOIHUD %XUUHW 2UELJQ\D WH ,[H, UDQD %RQGDU 0DUN ,H\D GDKOTUUQODQD %RQGDU $WWDOHD GDKOJUHQODQD %RQGDUf : %RHU 'HVFULEHG EDUERVODQD 'HVFULEHG 'HVFULEHG $ GDKOJUHQODQD8QFHUWDLQ VSHFLHV 'HVFULEHG 0 GDKOJUHQODQD VSHFORVD +DUWf %DUE FXUUHQWO\ UHGXFHG WR 5RGU ,V 6\QRQ\P\ QRW WR EH FRQIXVHG ZLWK VSHF ORVD %DUE 5RGU XQGHU FRKXQH 0DUWf 'DKLJUQ H[ 6WDQGL D QDPH 3RVVLEOH K\EULG 8! 2

PAGE 44

)LJXUH 'LVWULEXWLRQ RI VSHFLPHQV RI 2UELJQ\D PDUWLDQD DQG LWV VXVSHFWHG K\EULG FRPSOH[HV &LUFOHV UHSUHVHQW f PDUn WLDQD VTXDUHV UHSUHVHQW WKH 2UELJQ\D0D[LPLOLDQD FRPn SOH[ WULDQJOHV UHSUHVHQW WKH -2 PDUWLDQD2 HLFKOHUL FRPSOH[ 6RXUFHV :HVVHOV %RHU f *ODVVPDQ f DQG 0 %DOLFN DQG $ % $QGHUVRQ XQSXEOLVKHGf

PAGE 45

$ ILUVWKDQG IDPLOLDULW\ ZLWK WKH WD[D LQ WKH ILHOG SURYLGHG D VRXQG EDVLV IRU UHYLHZLQJ WKH H[WHQVLYH DQG RIWHQ FRQWUDGLFWRU\ OLWHUDn WXUH RQ QRPHQFODWXUH 5HVXOWV SHUWDLQLQJ WR WKH PRUSKRORJ\ DQG GLVWULn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f ,WV YDOLG SXEOLFDWLRQ E\ %DUERVD 5RGULJXHV f UHSUHVHQWV WKH ILUVW WD[RQRPLF GHVFULSWLRQ WKDW FDQ EH GHILQLWHO\ DWWULEXWHG WR EDEDVVX 3UHYLRXVO\ SXEOLVKHG QDPHV DVVRFLDWHG ZLWK EDEDVVX I$WWDOHD VSHFLRVD 0DUWf %DUE 5RGU DQG 2UELJn QYD OYGLDH 'UXGH@ ZHUH EDVHG RQ LQFRPSOHWH FROOHFWLRQV DQG WKHLU DFFRPn SDQ\LQJ GHVFULSWLRQV ZHUH HLWKHU LQVXIILFLHQW RU LQFRUUHFW QHLWKHU FDQ EH GHILQLWHO\ DWWULEXWHG WR EDEDVVX %DUERVD 5RGULJXHV f VXEVHn TXHQWO\ UHGXFHG PDUWLDQD WR V\QRQ\P\ XQGHU WKH QHZ FRPELQDWLRQ VSHFLRVD 0DUWf %DUE 5RGU $OWKRXJK ZLGHO\ DFFHSWHG LQ %UD]LO WKH ODWWHU QDPH LV LQYDOLG EHFDXVH WKH VDPH FRPELQDWLRQ 4LB VSHFLRVD %DUE

PAGE 46

5RGUf KDG EHHQ SUHYLRXVO\ SXEOLVKHG E\ %DUERVD 5RGULJXHV f LQ UHIHUHQFH WR DQRWKHU VSHFLHV FI $UWLFOH ,QWHUQDWLRQDO &RGH RI %RWDQLFDO 1RPHQFODWXUH 6WDIOHX S f 2UELJQYD PDUWLDQD LV WKHUHIRUH WKH HDUOLHVW YDOLGO\ SXEOLVKHG QDPH GHILQLWHO\ UHIHUDEOH WR WKLV WD[RQ DQG LW PDLQWDLQV SUHFHGHQFH RYHU VXEVHTXHQWO\ SXEOLVKHG QDPHV DVVRFLDWHG ZLWK EDEDVVX 7DEOH f 0\ XVH RI PDUWLDQD LV WHQWDWLYH GXH WR DQ ROGHU QDPH WKDW UHPDLQV WR EH FRQVLGHUHG 0DUWLXV f SXEOLVKHG D UHODWLYHO\ FRPSOHWH GHVFULSWLRQ RI 4cB SKDOHUDWD 0DUW ZKLFK KH GHVLJQDWHG DV WKH W\SH VSHFLHV RI WKH JHQXV 6XEVHTXHQW FROOHFWLRQV RI WKLV VSHFLHV KDYH DSSDUHQWO\ QRW EHHQ REWDLQHG *ODVVPDQ f DQG 0DUWLXVr SXEOLVKHG GHVFULSWLRQ DQG LOOXVWUDWLRQ SURYLGH QR EDVLV IRU GLVWLQJXLVKLQJ LW IURP PDUWLDQD 7KH ODFN RI VSHFLPHQV RI SKDOHUDWD SURPSWHG D UHFHQW -XO\$XJXVW f H[SHGLWLRQ VSRQVRUHG E\ WKH 1HZ
PAGE 47

RXWHU DED[LDOf VXUIDFH RI WKH VKHDWK ODUJH LQIORUHVFHQFHV WKDW DUH HLWKHU H[FOXVLYHO\ VWDPLQDWH RU KHUPDSKURGLWLF HDFK FRQWDLQLQJ D SHUVLVWHQW ZRRG\ EUDFW ZLWK D SURPLQHQW DSH[ VWDPLQDWH LQIORUHVFHQFHV ZLWK QXPHURXV HUHFW EUDQFKHV UDFKLOODHf HDFK EUDQFK EHDULQJ WZR WR IRXU URZV RI VWDPLQDWH IORZHUV RQ LWV ORZHU DED[LDOf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f LV VRPHZKDW EOXUUHG VXVSHFW WKDW WKHVH ODWWHU VSHFLHV K\EULGL]H ZLWK EDEDVVX 0RUSKRORJLFDO FRPSDULVRQV $SSHQGLFHV DQG (f LOOXVWUDWH WKH LQWHUPHGLDWH PRUSKRORJ\ RI WKH SXUSRUWHG K\EULGV 2UELJQYD WHL[HLUDQD DQG 0DUNOHYD GDKOJUHQLDQDf *HUPLQDWLRQ H[SHULPHQWV KDYH VKRZQ WKDW ERWK RI WKH ODWWHU SURGXFH YLDEOH VHHGV DQG WKXV DUH FDSDEOH RI UHJHQHUDWLQJ KRZHYHU WKH\ DUH RQO\ NQRZQ WR RFFXU LQ DVVRFLDWLRQ ZLWK WKHLU DSSDUHQW SURJHQLWRUV 7KH GLVWULEXWLRQ RI EDEDVVX DQG LWV SXUSRUWHG K\EULGV LV VKRZQ LQ )LJXUH $QDWRPLFDO SDO\QRORJLFDO DQG HOHFWURSKRUHWLF VWXGLHV RI WKHVH FRPSOH[HV KDYH EHHQ LQLWLDWHG E\ FROOHDJXHV DW WKH 1HZ
PAGE 48

*DUGHQ LQ WKH 86 DQG WKH &HQWUR 1DFLRQDO GH 5HFXUVRV *HQWLFRV &(1$5*(1f LQ %UD]LO 'LVWULEXWLRQ DQG 2ULJLQ %DEDVVX KDV D ZLGHVSUHDG GLVWULEXWLRQ LQ 6RXWK $PHULFD )LJ f 'LVVHPLQDWLRQ RI WKH ODUJH KHDY\ IUXLWV KDV SUREDEO\ EHHQ H[FOXVLYHO\ E\ PDPPDOV HLWKHU OLYLQJ RU H[WLQFWf RU ZDWHU $PRQJ OLYLQJ PDPPDOV URGHQWV VXFK DV SDFDV $JRXWL SDFDf DQG DJRXWLV 'DVYSURFWD SXQFWDWDf DUH HIIHFWLYH VKRUWUDQJH GLVSHUVDO DJHQWV VHH &KDSWHU f KXPDQV DQG SRVVLEO\ PRQNH\V GLVVHPLQDWH WKH IUXLWV RYHU JUHDWHU GLVWDQFHV &RHYROXWLRQ EHWZHHQ KDUGIUXLWHG SDOPV VXFK DV EDEDVVX DQG ODUJH 3OHLVWRFHQH PDPPDOV VHHPV OLNHO\ -DQ]HQ DQG 0DUWLQ f EXW D SDXFLW\ RI SDOHRQWRORJLFDO HYLGHQFH PDNHV LW LPSRVVLEOH WR DVVHVV WKH SRWHQWLDO FRQWULEXWLRQ RI WKHVH QRZ H[WLQFW DQLPDOV WR WKH FXUUHQW GLVWULEXWLRQ RI EDEDVVX 0DUNOH\ f DUJXHG FRQYLQFLQJO\ WKDW WKH EDEDVVX SDOP RULJLQDWHG RQ WKH %UD]LOLDQ 6KLHOG LQ WKH VWDWH RI *RL£V D PDMRU FHQWHU RI GLYHUVLW\ IRU WKH HQWLUH $WWDOHD DOOLDQFH 6LQFH LWV RULJLQ GXULQJ WKH 3UHFDPEULDQ WKH %UD]LOLDQ 6KLHOG KDV XQGHUJRQH FRQVLGHUDEOH HURVLRQ E\ QXPHURXV ULYHU V\VWHPV :LWK IHZ H[FHSWLRQV GLVFXVVHG EHORZf EDEDVVX DUHDV IURP WKH 3DUDJXD\ 5LYHU WR WKH $PD]RQ %DVLQ DUH FRQQHFWHG E\ ZDWHUVKHGV WR WKH %UD]LOLDQ 6KLHOG LQ *RL£V )LJV DQG f 7KURXJKRXW LWV UDQJH EDEDVVX LV ODUJHO\ FRQILQHG WR IORRGSODLQV ULYHU YDOOH\V DQG IRUPHU GHOWDV ZKLFK OHQGV VXSSRUW WR WKH LGHD RI GLVVHPLQDWLRQ E\ ULYHUV $OWKRXJK WKHLU KLJK VSHFLILF JUDYLW\ FD J IUHVK ZHLJKWFPAf SUHYHQWV WKH IUXLWV IURP IORDWLQJ WKH\ FDQ EH

PAGE 49

FDUULHG FRQVLGHUDEOH GLVWDQFHV E\ UXQRII DQG ULYHUV LQ IORRG /LJKWHU IUXLWV ZRXOG WHQG WR EH FDUULHG IDUWKHU ZKLFK FRXOG H[SODLQ ZK\ EDEDVVX IUXLWV DUH UHODWLYHO\ OLJKW J IUHVK PDVVf RYHU PRVW RI WKH VSHFLHVn UDQJH ZKLOH WKH KHDYLHVW IUXLWV NQRZQ XS WR J IUHVK PDVVf DUH IRXQG LQ LWV SUHVXPHG FHQWHU RI RULJLQ LQ *RLDV )RQVHFD 0 %DOLFN DQG $ % $QGHUVRQ XQSXEOLVKHG GDWDf 7KHUH DUH WKUHH LPSRUWDQW H[FHSWLRQV WR WKH OLQNV EHWZHHQ WKH %UD]LOLDQ 6KLHOG DQG EDEDVVXnV SUHVHQW UDQJH 7KH ILUVW LV LWV GLVMXQFW RFFXUUHQFH LQ 6XULQDPH DQG *X\DQD )LJ f 8QWLO WKH HDUO\ 3OHLVWRn FHQH WKH HDVWHUQ SRUWLRQV RI WKH %UD]LOLDQ DQG *XD\DQD 6KLHOGV ZHUH FRQQHFWHG ZKLFK SUREDEO\ DFFRXQWV IRU WKH FORVH IORULVWLF WLHV EHWZHHQ WKH WZR UHJLRQV 0DJXLUH f 7KH *XLDQDV UHSUHVHQW D VHFRQG FHQWHU RI GLYHUVLW\ RI WKH HQWLUH $WWDOHD DOOLDQFH :HVVHOV %RHU f $ VHFRQG H[FHSWLRQ LV EDEDVVXnV RFFXUUHQFH LQ WKH %UD]LOLDQ VWDWHV RI 0DUDQK£R DQG 3LDXL )LJ f 0DUNOH\ f VXJJHVWHG WKDW WKH 7RFDQWLQV 5LYHU PD\ KDYH RQFH VHUYHG DV D OLQN IRU GLVVHPLQDWLRQ RI WKH SDOP WR WKHVH VWDWHV 7KH 7RFDQWLQV KDV LWV KHDGZDWHUV LQ WKH %UD]LOLDQ 6KLHOG DQG SUHVHQWO\ GUDLQV LQWR WKH $PD]RQ 5LYHU 'XULQJ WKH &UHWDFHRXV KRZHYHU LW PD\ KDYH IORZHG DFURVV WKH QRUWKHUQ SDUW RI 0DUDQK£R DQG EXLOW XS WKH KXJH GHOWD WKDW SUHVHQWO\ FRYHUV PRVW RI WKH QRUWKHUQ WKLUG RI WKH VWDWH ,W VHHPV XQOLNHO\ WKDW WKLV GHOWD FRXOG KDYH EHHQ EXLOW E\ WKH UHODWLYHO\ LQVLJQLILFDQW ULYHUV RI 0DUDQK£R WKDW H[LVW WRGD\ 3LQGDU *UDMDX 0HDULP ,WDSHFXUX DQG 3DUQDLEDf PRVW RI ZKLFK KDYH WKHLU KHDGZDWHUV RQO\ D VKRUW GLVWDQFH IURP WKH GHOWD 8SOLIW RI WKH 6HUUD GR *XUXSL LQ ZHVWHUQ 0DUDQK£R SUREDEO\ GXULQJ WKH

PAGE 50

ORZHU 7HUWLDU\f PD\ KDYH IRUFHG WKH 7RFDQWLQV WR FKDQJH WR LWV SUHVHQW FRXUVH $ ILQDO DQRPDO\ LQ EDEDVVXnV GLVWULEXWLRQ LV LWV GLVMXQFW RFFXUUHQFH LQ WKH %UD]LOLDQ VWDWH RI &HDU£ )LJ f +HUH WKH SDOP LV FRQILQHG WR LVRODWHG DQG JHRORJLFDOO\ YHU\ ROG XSOLIWV DQG WKHUH DUH QR ULYHU V\VWHPV WKDW UXQ EHWZHHQ WKHVH VWUXFWXUHV DQG WKH KLJKODQGV RI *RL£V 7KH XSOLIWHG DUHDV WR ZKLFK EDEDVVX LV FRQILQHG LQ &HDU£ KDYH ORQJ VHUYHG DV IDYRUDEOH VLWHV IRU KXPDQ VHWWOHPHQW DQG DJULFXOWXUH GXH WR WKHLU UHODWLYHO\ FRRO WHPSHUDWXUHV KLJK UDLQIDOO DQG IHUWLOH VRLOV 0DUNOH\ f DUJXHG WKDW EDEDVVXnV RFFXUUHQFH LQ &HDU£ LV GXH WR GLVVHPLQDWLRQ E\ KXPDQV (DUO\ DERULJLQDO JURXSV PD\ KDYH LQWURGXFHG IUXLWV REWDLQHG IURP WKH 3DUQDLED YDOOH\ FD NP GLVWDQW

PAGE 51

&+$37(5 5(352'8&7,9( %,2/2*< ,Q DXWHFRORJLFDO VWXGLHV RI SODQWV WKH YDULRXV FRPSRQHQWV RI UHSURGXFWLRQ HJ SKHQRORJ\ SROOLQDWLRQ GLVSHUVDO SUHGDWLRQ JHUn PLQDWLRQf DUH RIWHQ DQDO\]HG DV VHSDUDWH HQWLWLHV
PAGE 52

0HWKRGV 3KHQRORJ\ FDUULHG RXW PRQWKO\ SKHQRORJLFDO REVHUYDWLRQV RI EDEDVVX IURP -DQXDU\ WR 0DUFK $OO DGXOW SDOPV DW HDFK VLWH ZHUH ODEHOOHG DQG PDSSHG 2Q WKH LQWHQVLYH VWXG\ VLWHV DW /DJR 9HUGH WKH VDPSOH FRQVLVWHG RI DOO DGXOWV LQ D KD DUHD LQ HDFK RI WKH VDWHOOLWH VLWHV 3LQGDUH 0LULP DQG &D[LDVf DGXOWV ZHUH UDQGRPO\ VHOHFWHG IRU REVHUn YDWLRQ 7KH IRUPHU VLWH KDG EHHQ UHFHQWO\ FRQYHUWHG WR SDVWXUH ZKHUHDV WKH ODWWHU ZDV LQ IDOORZ IROORZLQJ LWV XVH IRU VKLIWLQJ FXOWLYDWLRQ LQ 7KHVH VLWHV DUH UHSUHVHQWDWLYH RI FXUUHQW PDQDJHPHQW SUDFWLFHV RI EDEDVVX VWDQGV LQ WKH VHDVRQDOO\ ZHW VDYDQQD 3LQGDU 0LULPf DQG GU\ ZRRGODQG &D[LDVf HFRORJLFDO ]RQHV RI 0DUDQKDR 2Q WKH VLWH DW 3LQGDUH 0LULP OHDYHV RI EDEDVVX KDG EHHQ KDUYHVWHG IRU WKDWFK FD \U SULRU WR LQLWLDWLRQ RI WKH VWXG\ 7R SHUPLW REVHUYDWLRQV RI OHDI SURGXFWLRQ DQG ORVV WKH SDOPV ZHUH FOLPEHG DQG JUHHQ OHDYHV ZHUH ODEHOOHG E\ VSUD\LQJ SDLQW RQ WKH XQGHUn VLGHV RI WKH UDFKLVHV 7R HQKDQFH WKH YLVLELOLW\ RI QHZO\ HPHUJHG OHDYHV FD EDVDO OHDIOHWV SHU VLGH UHSUHVHQWLQJ FD b RI WKH WRWDOf ZHUH UHPRYHG IURP PDUNHG OHDYHV FDOFXODWHG OHDI SURGXFWLRQ SHU PRQWK E\ FRXQWLQJ WKH QHZO\ IRUPHG XQODEHOOHGf OHDYHV WKDW KDG FRPn SOHWHO\ H[SDQGHG LQ WKH FURZQ ORVV SHU PRQWK ZDV GHWHUPLQHG E\ FRXQWLQJ WKH QXPEHU RI ODEHOOHG OHDYHV WKDW OLDG IDOOHQ WR WKH JURXQG )DOOHQ OHDYHV ZHUH FXW LQ KDOI DIWHU EHLQJ FRXQWHG VR DV WR SUHYHQW EHLQJ UHFRXQWHG LQ VXEVHTXHQW PRQWKV DQG WKH OHDYHV RI DGMDFHQW SDOPV ZHUH

PAGE 53

SDLQWHG ZLWK GLIIHUHQW FRORUV DQGRU SDWWHUQV WR DYRLG FRQIXVLQJ WKH VRXUFHV RI IDOOHQ OHDYHV $V LQ RWKHU SDOPV WKH UHSURGXFWLYH VWUXFWXUHV LQIORUHVFHQFHV DQG LQIUXFWHVFHQFHVf LQ EDEDVVX DUH UHODWLYHO\ IHZ ODUJH DQG SHUPDQHQW ,Q DGGLWLRQ WR REWDLQLQJ VHSDUDWH GDWD RQ HDFK VWUXFWXUH FRXOG WKXV GHGXFH LWV SKHQRORJLFDO DFWLYLW\ GXULQJ WKH FD PR LQWHUYDO VLQFH WKH ODVW REVHUYDWLRQ 7R SUHYHQW FRQIXVLRQ DPRQJ VHYHUDO IORZHULQJ DQG IUXLWLQJ VWUXFWXUHV ERUQH VLPXOWDQHRXVO\ RQ D SDOP SKHQRORJLFDO REVHUn YDWLRQV ZHUH DFFRPSDQLHG E\ D PHDVXUHPHQW RI WKH FRPSDVV EHDULQJ RI HDFK VWUXFWXUH LQ UHODWLRQ WR WKH SDOP VWHP ,Q WKH FDVH RI IXQFWLRQDOO\ IHPDOH LQIORUHVFHQFHV FRQWDLQLQJ DW OHDVW RQH SLVWLOODWH IORZHU DQG D YDULDEOH QXPEHU RI VWDPLQDWH IORZHUV WKDW ZHUH XVXDOO\ LQYLDEOHf GLVWLQJXLVKHG WKH IROORZLQJ FDWHJRULHV f LQIORUHVFHQFH LQ IORZHU SLVWLOODWH IORZHUV \HOORZf f LQIUXFWHVFHQFH ZLWK LPPDWXUH IUXLWV IUXLWV WDQ WR EURZQ DQG QRW IDOOLQJf f LQIUXFWHVFHQFH ZLWK PDWXUH IUXLWV IUXLWV IDOOLQJf DQG f LQIUXFWHVFHQFH ZLWK QR IUXLWV DOO IUXLWV IDOOHQf ,Q WKH FDVH RI PDOH LQIORUHVFHQFHV FRQWDLQLQJ H[FOXn VLYHO\ VWDPLQDWH IORZHUVf GLVWLQJXLVKHG WKH IROORZLQJ FDWHJRULHV f LQIORUHVFHQFH LQ IORZHU IORZHUV LQWDFW RU IDOOLQJf f LQIORUHVn FHQFH LQ IORZHU GXULQJ WKH SDVW PRQWK IORZHUV IDOOHQ LQIORUHVFHQFH D[HV \HOORZ WR OLJKW WDQf DQG f LQIORUHVFHQFH LQ IORZHU SULRU WR SDVW PRQWK LQIORUHVFHQFH D[HV WDQf 2Q WKH WKUHH VWXG\ VLWHV DW /DJR 9HUGH IUXLW SURGXFWLRQ SHU PRQWK ZDV PHDVXUHG IURP 6HSWHPEHU WR $SULO 0DWXUH LQIUXFWHVFHQFHV ZHUH KDUYHVWHG DQG WKH QXPEHU RI IUXLWV SURGXFHG SHU LQIUXFWHVFHQFH ZDV

PAGE 54

GHWHUPLQHG E\ FRXQWLQJ IUXLWV DQGRU FDO\FHV LQ WKH FDVH RI DEVFLVHG IUXLWVf $ VXEVDPSOH RI FD IUXLWV IURP HDFK LQIUXFWHVFHQFH ZDV DLU GULHG DQG ZHLJKHG WR WKH QHDUHVW JUDP &RQVLGHUDEOH YDULDELOLW\ LQ WKH UDWLR RI PDOH WR IHPDOH LQIORUHVn FHQFHV VH[ UDWLRf ZDV REVHUYHG DPRQJ WKH SDOPV LQ WKH SKHQRORJLFDO VWXG\ 7R FRPSDUH VH[ UDWLR YDULDELOLW\ RQ WKH SRSXODWLRQ OHYHO PHDVXUHG WKH GLVWULEXWLRQV RI LQIORUHVFHQFH W\SHV LQ VL[ SRSXODWLRQV 'LVWULEXWLRQV ZHUH REWDLQHG E\ GHWHUPLQLQJ WKH SHUFHQWDJH RI HDFK LQn IORUHVFHQFH W\SH SHU SDOP 7KH LQIORUHVFHQFHV WHQG WR UHPDLQ DWWDFKHG IRU VHYHUDO \HDUV IROORZLQJ DQWKHVLV P\ FRXQWV WKXV SURYLGHG D ORQJn WHUP SLFWXUH RI UHSURGXFWLYH EHKDYLRU 'DWD ZHUH REWDLQHG IURP WZR JHRJUDSKLFDOO\ LVRODWHG SRSXODWLRQV ORFDWHG LQ 7RFDQWLQSROLV *RL£V DQG 3LUDSRUD 0LQDV *HUDLV $GGLWLRQDO GDWD ZHUH REWDLQHG IURP IRXU SRSXODWLRQV ORFDWHG ZLWKLQ P\ SKHQRORJLFDO VWXG\ VLWHV DW /DJR 9HUGH 3LQGDU 0LULP DQG &D[LDV DOO RI WKH ODWWHU VLWHV RFFXU ZLWKLQ WKH PRUH RU OHVV FRQWLQXRXV EDQG RI EDEDVVX VWDQGV WKDW VWUHWFKHV DFURVV QRUWK FHQWUDO 0DUDQKDR )LJ f )ORUDO %LRORJ\ 7R GHWHUPLQH GXUDWLRQ RI IORZHULQJ LQ LQGLYLGXDO LQIORUHVFHQFHV FDUULHG RXW GDLO\ SKHQRORJLFDO REVHUYDWLRQV IURP 0DUFK RQ SDOPV RQ WKH VHFRQGDU\ IRUHVW VLWH DQG SDOPV RQ WKH SDVWXUH VLWH 3DOPV ZLWK EUDFWV DERXW WR RSHQ ZHUH VHOHFWHG IRU REVHUYDWLRQ )ORZHUn LQJ ZDV MXGJHG WR EHJLQ XSRQ RSHQLQJ RI WKH EUDFW 3URSRUWLRQ RI IORZHUV WKDW XOWLPDWHO\ EHFDPH IUXLWV ZDV GHWHUPLQHG WKURXJK H[DPLQDWLRQ

PAGE 55

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n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

PAGE 56

RFFXUV LQ EDEDVVX IORZHUV ZHUH OHIW EDJJHG WKURXJKRXW WKH IORZHULQJ SHULRG DQG UHPRYHG WKH SLVWLOV IURP RWKHUV ERWK VHWV RI IORZHUV ZHUH VXEVHTXHQWO\ FKHFNHG IRU IUXLW VHW $EXQGDQFH DQG EHKDYLRU RI LQVHFW YLVLWRUV WR ERWK LQIORUHVFHQFH W\SHV ZHUH GHWHUPLQHG E\ GLUHFW REVHUYDWLRQ DQG FROOHFWLRQ 6WLFN\ WUDSV ZHUH XVHG WR PRQLWRU LQVHFWV EHWZHHQ REVHUYDWLRQV ,QVHFW VSHFLn PHQV ZHUH LGHQWLILHG DW WKH 0XVHX 3DUDHQVH (PLOLR *RHOGL LQ %HOHP XQLGHQWLILHG VSHFLPHQV ZHUH VHQW WR VSHFLDOLVWV LQ %UD]LO DQG WKH 86 7R GHWHUPLQH ZKHWKHU ZLQG SROOLQDWLRQ RFFXUV LQ EDEDVVX VOHHYHV ZHUH GHVLJQHG WR H[FOXGH LQVHFW YLVLWRUV IURP IHPDOH LQIORUHVFHQFHV ZKLOH SHUPLWWLQJ SHQHWUDWLRQ RI ZLQGERUQH SROOHQ 7KH VOHHYHV ZHUH PDGH RXW RI Q\ORQ QHWWLQJ ZLWK D FD PP PHVK VXIILFLHQW WR H[FOXGH DOO REVHUYHG LQVHFW YLVLWRUV 'LVSHUVDO RI SROOHQ E\ ZLQG ZDV PRQLWRUHG WKURXJK WKH XVH RI SROOHQ WUDSV ZKLFK ZHUH GLVWULEXWHG DW UHJXODU LQWHUYDOV RQ ERWK WKH VHFRQGDU\ IRUHVW DQG SDVWXUH VLWHV 7KH WUDS FRQVLVWHG RI D PLFURVFRSH VOLGH FRYHUHG ZLWK D WKLQ FRDW RI YDVHOLQH WKH VOLGH ZDV SODFHG LQVLGH D KRUL]RQWDOO\ RULHQWHG VHFWLRQ RI FP ZLGH 39& SLSH VHFXUHG WR D VWDNH RU VWHP $IWHU D ZN H[SRVXUH WKH VOLGHV ZHUH UHPRYHG FRYHUHG DQG VXEVHTXHQWO\ H[DPLQHG LQ WKH ODERUDn WRU\ )UXLW %LRORJ\ 'LVSHUVDO PRQLWRUHG UHPRYDO RI EDEDVVX IUXLWV LQ WKH SDVWXUH DQG VHFRQGDU\ IRUHVW DV ZHOO DV RQ D QHDUE\ ILHOG 7KLV ODWWHU VLWH KDG EHHQ FXW EXUQHG DQG SODQWHG DW WKH HQG RI 'XULQJ WKH FRXUVH RI WKH VWXG\ WKH VLWH UHPDLQHG YLUWXDOO\ DEDQGRQHG H[FHSW IRU VSRUDGLF

PAGE 57

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f RQ WKH VHFRQGDU\ IRUHVW VLWH 'LVSHUVDO ZDV DUELn WUDULO\ GHILQHG DV PRYHPHQW RI PRUH WKDQ FP 1R VLJQLILFDQW GLIIHUn HQFH S f ZDV IRXQG LQ GLVSHUVDO RI ODEHOOHG DQG XQODEHOOHG IUXLWV DFFRUGLQJ WR WKH UDQGRPL]DWLRQ WHVW IRU PDWFKHG SDLUV 6LHJHO f 3UHGDWLRQ 7R PRQLWRU FRORQL]DWLRQ E\ VHHGSUHGDFHRXV EHHWOHV VL[ IUXLW FOXVWHUV ZHUH VHW XS RQ WKH VHFRQGDU\ IRUHVW VLWH SUHYHQWHG GLVSHUVDO E\ VWDSOLQJ WKH DSLFHV RI WKH IUXLWV WR D Q\ORQ OLQH IUXLWV SHU OLQHf DWWDFKHG WR D VWDNH 7ZHQW\ IUXLWV ZHUH JDWKHUHG IURP HDFK FOXVWHU DW LQWHUYDOV RI DQG G IROORZLQJ LQLWLDWLRQ RI WKH H[SHULPHQW 2FWREHU f 7KH IUXLWV JDWKHUHG IURP HDFK FOXVWHU ZHUH LVRODWHG LQ FORVHG VDFNV WR SUHYHQW IXUWKHU SUHGDWRU FRORQL]DWLRQ $IWHU DLU GU\LQJ WKH IUXLWV ZHUH RSHQHG WR GHWHUPLQH WRWDO QXPEHU RI VHHGV QXPEHU RI DWWDFNHG VHHGV DQG QXPEHU RI EHHWOH ODUYDH SHU IUXLW 6HHG IDWH 7R PRQLWRU WKH IDWH RI VHHGV RYHU WLPH XQGHU YDULRXV HQYLURQPHQWDO FRQGLWLRQV ILYH IUXLW FOXVWHUV ZHUH VHW XS LQ HDFK RI

PAGE 58

WKUHH VLWHV VHFRQGDU\ IRUHVW SDVWXUH DQG ILHOG SUHYHQWHG GLVn SHUVDO E\ VWDSOLQJ WKH IUXLWV WR Q\ORQ OLQHV DWWDFKHG WR D VWDNH 7ZHQn W\ IUXLWV SHU SLOH ZHUH JDWKHUHG DW WKH RXWVHW RI WKH H[SHULPHQWV 2FWREHU f DQG DW PR LQWHUYDOV GXULQJ \U REWDLQHG D ORQJWHUP FD PRf PHDVXUH RI VHHG IDWH IURP IUXLWV XVHG LQ WKH LUULJDWLRQ H[SHULPHQW GHVFULEHG LQ &KDSWHU f} $W WKH HQG RI WKLV H[SHULPHQW IUXLWV ZHUH JDWKHUHG IURP HDFK RI ILYH ZHHGHG VXQf DQG ILYH QRQZHHGHG VKDGHf SORWV WKDW KDG QRW EHHQ LUULJDWHG ,Q ERWK VHHG IDWH H[SHULPHQWV JDWKHUHG IUXLWV ZHUH VWRUHG LQ FORVHG VDFNV DLUGULHG DQG RSHQHG WR GHWHUPLQH QXPEHUV RI HDWHQ GHIHFWLYH DERUWHG RU URWWHGf JHUPLQDWHG DQG TXLHVFHQW YLDEOH EXW QRW JHUPLQDWHGf VHHGV *HUPLQDWLRQ $Q H[SHULPHQW ZDV VHW XS RQ WKH VHFRQGDU\ IRUHVW VLWH WR PRQLWRU JHUPLQDWLRQ RYHU WLPH DQG WR WHVW IRU EHWZHHQSDOP YDULDELOLn W\ LQ IUXLW JHUPLQDWLRQ 5HFHQWO\ IDOOHQ IUXLWV IURP HDFK RI SDOPVf ZHUH GHKXVNHG WR LQKLELW GLVSHUVDO DQG SODQWHG LQ SORWV IUXLWV SHU SORW IRXU SORWV SHU SDOPf $W FD PR LQWHUYDOV PHDVn XUHG JHUPLQDWLRQ E\ FKHFNLQJ WKH EDVH RI HDFK IUXLW IRU HPHUJHQFH RI RQH RU PRUH HPEU\RQLF D[HV LH SOXPXOH FRW\OHGRQDU\ QRGH DQG UDGLFOHf FRQGXFWHG DGGLWLRQDO H[SHULPHQWV WR WHVW WKH HIIHFWV RI ILUH KHUELYRU\ DQG VKDGLQJ RQ JHUPLQDWLRQ RI EDEDVVX IUXLWV 7KHVH H[SHULn PHQWV ZHUH HVWDEOLVKHG RQ D FOHDUHG SHULRGLFDOO\ ZHHGHG VLWH FD P IURP VHFRQGDU\ IRUHVW RQ ZHOOGUDLQHG VRLO 7KH IUXLWV XVHG LQ HDFK RI WKHVH H[SHULPHQWV SHU WUHDWPHQWf ZHUH REWDLQHG IURP D VLQJOH SDOP 8QOHVV RWKHUZLVH VSHFLILHG IUXLWV ZHUH GHKXVNHG DQG SODFHG RQ SORWV VKDGHG ZLWK VXVSHQGHG IUDPHV XSRQ ZKLFK EDEDVVX OHDYHV ZHUH WLHG $W

PAGE 59

ZN LQWHUYDOV WKH IUDPHV ZHUH OLIWHG WR FKHFN IRU JHUPLQDWLRQ 7R WHVW WKH HIIHFWV RI ILUH H[SRVHG IUXLWV WR HDFK RI WKH IROORZLQJ WUHDWn PHQWV f D SURORQJHG KLJKLQWHQVLW\ ILUH VLPXODWLQJ D IRUHVW IDOORZ EXUQ f D UDSLG ORZLQWHQVLW\ ILUH VLPXODWLQJ D SDVWXUH EXUQ DQG f QR ILUH 7R WHVW WKH HIIHFW RI FRQVXPSWLRQ RI WKH IUXLW PHVRFDUS E\ KHUELYRUHV FRPSDUHG JHUPLQDWLRQ LQ IUXLWV WKDW KDG EHHQ GHKXVNHG HSLFDUS DQG PHVRFDUS UHPRYHG VLPXODWLQJ WKH HIIHFW RI KHUELYRU\f DQG IUXLWV WKDW KDG QRW EHHQ GHKXVNHG 7R WHVW WKH HIIHFW RI VKDGLQJ SODQWHG IUXLWV RQ DQ H[SRVHG SORW DQG IUXLWV RQ D SORW VKDGHG DV GHVFULEHG DV DERYH 5HVXOWV 3KHQRORJ\ /HDI WXUQRYHU )LJXUH VKRZV GDWD RQ OHDI WXUQRYHU RI EDEDVVX DW WKH ZHW 3LQGDU 0LULPf LQWHUPHGLDWH /DJR 9HUGHf DQG GU\ &D[LDVf VLWHV VLPLODU GDWD IURP WKH SULPDU\ IRUHVW VHFRQGDU\ IRUHVW DQG SDVWXUH VLWHV DUH VKRZQ LQ )LJXUH $W DOO VLWHV OHDI SURGXFWLRQ SHDNHG GXULQJ WKH WZR UDLQ\ VHDVRQV DQG OHDI ORVV SHDNHG GXULQJ WKH VLQJOH GU\ VHDVRQ RI REVHUYDWLRQV $V LQ SDOPV JHQHUDOO\ OHDI SURGXFn WLRQ LQ EDEDVVX LV V\QRQ\PRXV ZLWK JURZWK WKH UDLQ\ VHDVRQ LV WKHUHn IRUH WKH JURZLQJ VHDVRQ IRU WKH SDOP 2YHU D \U SHULRG DIWHU LQLWLDn WLRQ RI REVHUYDWLRQV QR GLIIHUHQFH ZDV GHWHFWHG DPRQJ OHDI SURGXFWLRQ UDWHV SHU SDOP DW WKH ZHW [ f LQWHUPHGLDWH [ mf DQG GU\ [ f VLWHV %\ FRQWUDVW 'XQFDQnV 0XOWLSOH 5DQJH WHVW UHYHDOHG VLJQLn ILFDQW GLIIHUHQFHV S Af LQ UDWHV RI OHDI SURGXFWLRQ SHU SDOP DW WKH WKUHH /DJR 9HUGH VLWHV WKH RUGHU ZDV SDVWXUH [ f SULPDU\ IRUHVW

PAGE 60

/($) 352'8&7,21 3(5 3$/0 )0$0-$ 621')0 )LJXUH 0HDQ PRQWKO\ JDLQ DQG ORVV RI OHDYHV SHU SDOP RQ WKUHH VLWHV LQ 0DUDQKDR :HW VLWH 3LQGDU 0LULP LQWHUPHGLDWH VLWH /DJR 9HUGH DQG GU\ VLWH &D[LDV 1 QXPEHU RI SDOPV REVHUYHG SHU VLWH 1' QR GDWD

PAGE 61

/($) 352'8&7,21 3(5 3$/0 f§ f§ )0$0-$ 621')0 35,0$5< )25(67 M Q QG Q @ f§ OOOOOLLOL OXX & & 7-L U )LJXUH 0HDQ PRQWKO\ JDLQ DQG ORVV RI OHDYHV SHU SDOP RQ WKUHH VWXG\ VLWHV DW /DJR 9HUGH 1 QXPEHU RI SDOPV REVHUYHG SHU VLWH 1' QR GDWD

PAGE 62

[ f VHFRQGDU\ IRUHVW [ f 7KHVH UHVXOWV FDQ SUREDEO\ EH DWWULEXWHG WR GLIIHUHQWLDO VKDGLQJ RI DGXOW SDOPV RQ WKH WKUHH VLWHV 6KDGLQJ ZDV ORZHVW LQ WKH SDVWXUH 3RWHQWLDOO\ KLJK VKDGLQJ LQ WKH SULPDU\ IRUHVW PD\ KDYH EHHQ UHGXFHG E\ WKH UHODWLYHO\ SURQRXQFHG FDQRS\ URXJKQHVV ZKLFK WKXV SHUPLWWHG JUHDWHU ODWHUDO SHQHWUDWLRQ RI OLJKWf DQG WKH HPHUJHQW VWDWXV RI PDQ\ RI WKH DGXOW SDOPV ,Q WKH VHFRQGDU\ IRUHVW VKDGLQJ ZDV KLJKHVW EHFDXVH WKH GHQVH XQLIRUP EDEDVVX VWDQG IRUPHG D FORVHG LQWHUORFNLQJ FDQRS\ ([FHSW DW WKH ZHW VLWH OHDI SURGXFWLRQ DQG ORVV UDWHV ZHUH DSSUR[n LPDWHO\ HTXDO DIWHU \U RI REVHUYDWLRQV QHW JDLQ IRU WKLV SHULRG DYHUDJHG OHDYHV SHU SDOP ZKLFK LV YLUWXDOO\ VWHDG\ VWDWH $VVXPLQJ VWHDG\ VWDWH FRQGLWLRQV RQH FDQ GLYLGH WKH QXPEHU RI OHDYHV ERUQH SHU SDOP [ Q f E\ DQQXDO SURGXFWLRQ SHU SDOP [ r} Q f WR REWDLQ DQ DYHUDJH OHDI OLIHVSDQ RI \U $W WKH ZHW VLWH SURGXFWLRQ UDWHV ZHUH IDU JUHDWHU WKDQ ORVV UDWHV WKH QHW JDLQ SHU \HDU ZDV r OHDYHV SHU WUHH 7KH ORZ ORVVHV ZHUH SUREDEO\ GXH WR OHDI KDUYHVWLQJ IRU WKDWFK \U SULRU WR LQLWLDWLRQ RI WKH VWXG\
PAGE 63

,1)/25(6&(1&( 352'8&7,21 3(5 3$/0 )LJXUH 0HDQ PRQWKO\ SURGXFWLRQ RI PDOH DQG IXQFWLRQDOO\ IHPDOH LQIORUHVFHQFHV SHU SDOP DW WKUHH VLWHV LQ 0DUDQKDR :HW VLWH 3LQGDU 0LULP LQWHUPHGLDWH VLWH /DJR 9HUGH DQG GU\ VLWH &D[LDV 1 QXPEHU RI SDOPV REVHUYHG SHU VLWH

PAGE 64

,1)/25(6&(1&( 352'8&7,21 3(5 3$/0 G f§ f§ )0$0 -$621' )0 35,0$5< )25(67 1 f f§ f§ f§ 8 8 G 6(&21'$5< )25(67 1 f LW% FURU )LJXUH 0HDQ PRQWKO\ SURGXFWLRQ RI PDOH DQG IXQFWLRQDOO\ IHPDOH LQIORUHVFHQFHV SHU SDOP DW WKUHH VWXG\ VLWHV DW /DJR 9HUGH 1 QXPEHU RI SDOPV REVHUYHG SHU VLWH

PAGE 65

ORJLHV LV SUREDEO\ GXH WR WKH SK\VLFDO DVVRFLDWLRQ RI WKH WZR VWUXFn WXUHV HDFK LQIORUHVFHQFH LV DWWDFKHG WR WKH LQQHU DGD[LDOf VXUIDFH RI D OHDI VKHDWK +RZHYHU D FD \U ODJ RFFXUUHG EHWZHHQ D OHDInV HPHUJn HQFH LQ WKH FURZQ DQG WKH RSHQLQJ RI LWV DVVRFLDWHG LQIORUHVFHQFH (DUOLHU LQIORUHVFHQFH H[SDQVLRQ ZDV DSSDUHQWO\ SURKLELWHG E\ WKH WLJKW SDFNLQJ RI QHZO\ HPHUJHG OHDYHV QHDU WKH WRS RI WKH FURZQ 2QO\ FRQWLQn XHG JURZWK RI WKH SDOP GXULQJ WKH IROORZLQJ UDLQ\ VHDVRQ ORRVHQHG WKLV SDFNLQJ VXIILFLHQWO\ WR SHUPLW LQIORUHVFHQFH H[SDQVLRQ 'LVVHFWLRQ RI FURZQV VKRZHG WKDW HDFK OHDI EDVH FRQWDLQV D VLQJOH LQIORUHVFHQFH 8QGHU VXLWDEOH FRQGLWLRQV DOO LQIORUHVFHQFHV XOWLPDWHO\ GHYHORS WR DQWKHVLV WKXV UHVXOWLQJ LQ DQ LQIORUHVFHQFHWROHDI SURGXFn WLRQ UDWLR RI 3URGXFWLRQ GDWD RYHU WKH HQWLUH SHULRG RI REVHUYDWLRQV UHYHDOHG KLJK UDWLRV IRU WKH GU\ VLWH DW &D[LDV f DQG WKH SDVWXUH DW /DJR 9HUGH f 7KH SRVVLELOLW\ RI REWDLQLQJ LQIORUHVFHQFHWR OHDI SURGXFWLRQ UDWLRV JUHDWHU WKDQ UHVXOWV IURP FRXQWLQJ WKHVH VWUXFn WXUHV VLPXOWDQHRXVO\ $ PRUH DFFXUDWH UDWLR ZRXOG EH REWDLQHG E\ FRPn SDULQJ SURGXFWLRQ RI OHDYHV LQ D JLYHQ \HDU DQG SURGXFWLRQ RI WKHLU FRUUHVSRQGLQJ LQIORUHVFHQFHV GXULQJ WKH IROORZLQJ \HDUf 6KDGLQJ SUREn DEO\ FDXVHG LQWHUPHGLDWH UDWLRV RQ WKH SULPDU\ IRUHVW f DQG VHFRQGn DU\ IRUHVW f VLWHV DW /DJR 9HUGH 7KH ORZ UDWLR DW WKH ZHW VLWH f FDQ EH DWWULEXWHG WR WKH OHDI KDUYHVW \U SUHYLRXVO\ %DEDVVXnV DSSDUHQW UHVSRQVH WR GHIROLDWLRQ ZDV WR LQFUHDVH LQIORUHVFHQFH DERUWLRQ UDWKHU WKDQ WR GHFUHDVH OHDI SURGXFWLRQ 1R GLIIHUHQFH LQ WKH SKHQRORJ\ RI PDOH DQG IHPDOH LQIORUHVFHQFHV ZDV GHWHFWHG $SSUR[LPDWHO\ b RI WRWDO LQIORUHVFHQFHV ZHUH IHPDOH LQ

PAGE 66

DOO SRSXODWLRQV REVHUYHG GHVSLWH KLJK YDULDELOLW\ bf DPRQJ LQGLn YLGXDO SDOPV 9DULDELOLW\ RQ WKH SRSXODWLRQ OHYHO ZDV DQDO\]HG E\ FRPSDULQJ WKH GLVWULEXWLRQV RI IHPDOH LQIORUHVFHQFHV SHU SDOP LQ VL[ SRSXODWLRQV )LJ f 'HVSLWH D ZLGH UDQJH RI HFRORJLFDO FRQGLWLRQV WKH IRXU FRQWLJn XRXV SRSXODWLRQV IRUHVW DQG SDVWXUH DW /DJR 9HUGH ZHW VLWH DW 3LQGDU 0LULP DQG GU\ VLWH DW &D[LDVf VKRZHG QR GLIIHUHQFHV LQ GLVWULEXWLRQV RI IHPDOH LQIORUHVFHQFHV SHU SDOP DFFRUGLQJ WR WKH .ROPRJRURY6PLUQRY 7HVW 6LHJHO f +RZHYHU GLVWULEXWLRQV LQ WKH WZR GLVMXQFW SRSXODWLRQV QRUWKHUO\ VLWH LQ *RL£V DQG VRXWKHUO\ VLWH LQ 0LQDV *HUDLVf ZHUH VLJQLn ILFDQWO\ GLIIHUHQW IURP WKH ILUVW IRXU SRSXODWLRQV DV ZHOO DV IURP HDFK RWKHU S LQ DOO FDVHVf 7KHVH UHVXOWV VXJJHVW WKDW WKH GLVWULEXn WLRQ RI IHPDOH LQIORUHVFHQFHV SHU SDOP LQ SRSXODWLRQV RI EDEDVVX LV JRYHUQHG E\ JHQHWLF IDFWRUV 3RWHQWLDO LPSOLFDWLRQV RI WKLV ILQGLQJ DUH GLVFXVVHG LQ &KDSWHU )UXLW SURGXFWLRQ $W DOO VLWHV IUHTXHQFLHV RI LPPDWXUH LQIUXFWHV FHQFHV SHDNHG DV IORZHULQJ HQGHG LQ $SULO-XQH DQG ZHUH ORZHVW ZKHQ IUXLWLQJ SHDNHG LQ 2FWREHU'HFHPEHU )LJV f 7KLV SDWWHUQ ZDV FRQVLVWHQW GHVSLWH WKH ZLGH UDQJH RI FOLPDWLF FRQGLWLRQV DPRQJ VLWHV )LJ f $FFRUGLQJ WR SKHQRORJLFDO REVHUYDWLRQV WKHUH ZDV D FD PR ODJ EHWZHHQ WKH IORZHULQJ SHDN -DQXDU\0DUFK )LJV f DQG WKH IUXLWLQJ SHDN 2FWREHU'HFHPEHU )LJV f 2EVHUYDWLRQV RQ LQGLn YLGXDO SDOPV FRQILUPHG WKDW IUXLWV UHTXLUH DQ DYHUDJH RI PR WR UHDFK PDWXULW\

PAGE 67

)(0$/( ,1)/25(6&(1&(6 3(5 3$/0 bf )LJXUH 'LVWULEXWLRQV RI IHPDOH LQIORUHVFHQFHV SHU SDOP LQ VL[ SRSXODWLRQV RI EDEDVVX )RUHVW DQG SDVWXUH VLWHV /DJR 9HUGH ZHW VLWH 3LQGDU 0LULP GU\ VLWH &D[LDV QRUWKHUO\ VLWH 7RFDQWLQSROLV *RL£V DQG VRXWKHUO\ VLWH 3LUDSRUD 0LQDV *HUDLV 1 QXPEHU RI SDOPV REn VHUYHG SHU VLWH

PAGE 68

,1)58&7(6&(1&(6 3(5 3$/0 ,00$785( 0$785( ,00$785( 0$785( ,00$785( 0$785( ) 0 $ 0 f§ f§ $ 621')0 Q ,17(50(',$7( 6,7( 1 f )LJXUH 0HDQ PRQWKO\ QXPEHU RI LPPDWXUH DQG PDWXUH LQIORUHVFHQFHV SHU SDOP DW WKUHH VLWHV LQ 0DUDQKDR :HW VLWH 3LQGDU 0LULP LQWHUPHGLDWH VLWH /DJR 9HUGH DQG GU\ VLWH &D[LDV 1 QXPEHU RI SDOPV REVHUYHG SHU VLWH

PAGE 69

,1)58&7(6&(1&(6 3(5 3$/0 ) 0 $ 0 $ f§ 621 ) 0 WX U L + XL e( XL RF XL e( + 8, FF + )LJXUH 0HDQ PRQWKO\ QXPEHU RI LPPDWXUH DQG PDWXUH LQIORUHVFHQFHV SHU SDOP DW WKUHH VWXG\ VLWHV DW /DJR 9HUGH 1 QXPEHU RI SDOPV REVHUYHG SHU VLWH

PAGE 70

)58,7 352'8&7,21 NJKDf )LJXUH 0RQWKO\ IUXLW SURGXFWLRQ SHU KD RQ WKUHH VWXG\ VLWHV DW /DJR 9HUGH 9DOXHV UHSUHVHQW DLUGULHG PDVV

PAGE 71

)58,7 352'8&7,21 3(5 3$/0 NJf )LJXUH 0HDQ PRQWKO\ IUXLW SURGXFWLRQ SHU SDOP RQ WKUHH VWXG\ VLWHV DW /DJR 9HUGH 9DOXHV DUH DLUGULHG PDVV 1 QXPEHU RI SDOPV SHU VLWH

PAGE 72

)UXLW \LHOGV SHU KD DQG SHU SDOP 7DEOH f ZHUH KLJKHVW LQ WKH SDVWXUH SUREDEO\ GXH WR WKH PRGHUDWH GHQVLW\ DQG UHODWLYH RSHQQHVV RI WKH EDEDVVX VWDQG RQ WKLV VLWH 3HUKD \LHOGV ZHUH ORZHVW LQ WKH SULn PDU\ IRUHVW DQG LQWHUPHGLDWH LQ WKH VHFRQGDU\ IRUHVW 7KLV LV SUREDEO\ GXH WR GLIIHUHQFHV LQ SDOP GHQVLWLHV ZKLFK ZHUH ORZ LQ WKH SULPDU\ IRUHVW PDWXUH SDOPVKDf DQG KLJK LQ WKH VHFRQGDU\ IRUHVW PDWXUH SDOPVKDf &RQYHUVHO\ SHUSDOP \LHOGV ZHUH LQWHUPHGLDWH LQ WKH SULPDU\ IRUHVW DQG ORZ LQ WKH VHFRQGDU\ IRUHVW SUREDEO\ GXH WR LQFUHDVHG VKDGn LQJ RI PDWXUH SDOPV RQ WKH ODWWHU VLWH $OO WKUHH VLWHV VKRZHG VXEVWDQWLDO LQFUHDVHV LQ \LHOGV EHWZHHQ WKH DQG IUXLWLQJ VHDVRQV 2Q DOO VLWHV WKH SULQFLSDO VRXUFH RI WKHVH KLJKHU \HLOGV ZDV DQ LQFUHDVH LQ WKH QXPEHU RI SDOPV WKDW ERUH IUXLW LQFUHDVHG QXPEHUV RI LQIUXFWHVFHQFHV SHU IUXLWLQJ SDOP FRQVWLn WXWHG WKH VHFRQG PRVW LPSRUWDQW VRXUFH 7DEOH f )ORUDO %LRORJ\ )UXLW VHW ,Q WKH VHFRQGDU\ IRUHVW b VWDQGDUG GHYLDWLRQ bf RI IORZHUV EHFDPH IUXLWV b bf HLWKHU DERUWHG RU ZHUH QRW IHUWLOL]HG DQG b bf ZHUH DSSDUHQWO\ DWWDFNHG E\ WKH ODUYDH RI DQ XQLGHQWLILHG ZHHYLO ,Q WKH SDVWXUH WKH ILJXUHV ZHUH b bf b bf DQG b bf UHVSHFWLYHO\ 7KHVH QXPEHUV LQGLFDWH KLJK EUHHGLQJ VXFFHVV RI EDEDVVX LQ ERWK KDELWDWV 3DUWKHQRJHQHVLV )UXLW VHW FRQVLVWHQWO\ IDLOHG WR RFFXU LQ IORZHUV WKDW ZHUH QRW H[SRVHG WR SROOHQ VRXUFHV Q RQ WKUHH LQIORUHVn FHQFHVf DV ZHOO DV LQ IORZHUV LQ ZKLFK WKH VWLJPDV KDG EHHQ UHPRYHG

PAGE 73

7DEOH )UXLW \LHOGV RI EDEDVVX SHU KD DQG SHU SDOP GXULQJ DQG RQ WKUHH VLWHV 9DOXHV DUH RYHQGULHG PDVV LQ NJ 6,7( 3(5 +(&7$5( 3(5 3$/0 3ULPDU\ )RUHVW 6HFRQGDU\ )RUHVW 3DVWXUH

PAGE 74

7DEOH 6RXUFHV RI LQFUHDVHG IUXLW \LHOGV RI EDEDVVX SHU KD EHWZHHQ DQG RQ WKUHH VLWHV 1XPEHUV UHSUHVHQW SHUFHQW RI WRWDO LQFUHDVH LQ \LHOGV 6285&( 6,7( 3ULPDU\ )RUHVW 6HFRQGDU\ )RUHVW 3DVWXUH
PAGE 75

Q RQ RQH LQIORUHVFHQFHf 7KHVH UHVXOWV LQGLFDWH WKDW SDUWKHQRJHQn HVLV GRHV QRW RFFXU LQ EDEDVVX 6HOISROOLQDWLRQ ,QIORUHVFHQFHV LQ EDEDVVX DUH JHQHUDOO\ HLWKHU PDOH FRQWDLQLQJ H[FOXVLYHO\ VWDPLQDWH IORZHUVf RU IHPDOH FRQWDLQLQJ DW OHDVW RQH SLVWLOODWH IORZHU DQG D YDULDEOH QXPEHU RI VWDPLQDWH IORZHUV WKDW DUH XVXDOO\ LQYLDEOHf 7KH IRUPHU DUH IDU PRUH QXPHURXV WKDQ WKH ODWWHU 'XULQJ D ZN SHULRG LQ 0DUFK bf RI LQIORUHVn FHQFHV H[DPLQHG FRQWDLQHG SLVWLOODWH IORZHUV 2I WKHVH LQIORUHVn FHQFHV RQO\ WKUHH Af FRQWDLQHG VWDPLQDWH IORZHUV WKDW VKHG SROOHQ DQG WKXV DSSHDUHG WR EH YLDEOH &RQVHTXHQWO\ PRVW LQIORUHVFHQFHV WKDW FRQWDLQ SLVWLOODWH IORZHUV DUH H[FOXVLYHO\ IHPDOH LQ D IXQFWLRQDO VHQVH ,Q D SLVWLOODWH IORZHU EDJJLQJ H[SHULPHQW IUXLW VHW RFFXUUHG RQO\ LQ IORZHUV H[SRVHG K DIWHU RSHQLQJ RI WKH LQIORUHVFHQFH VXEVHn TXHQWO\ H[SRVHG IORZHUV GLG QRW VHW IUXLW 7KXV SLVWLOODWH IORZHUV DSSHDU WR EH UHFHSWLYH IRU D PD[LPXP RI K DIWHU LQIORUHVFHQFHV RSHQ ,Q WKH WKUHH IHPDOH LQIORUHVFHQFHV ZLWK IXQFWLRQDO VWDPLQDWH IORZHUV WKDW REVHUYHG SROOHQ UHOHDVH LQYDULDEO\ EHJDQ PRUH WKDQ K DIWHU RSHQLQJ RI WKH LQIORUHVFHQFH ,Q RQH RI WKHVH LQIORUHVFHQFHV IUXLW VHW GLG QRW RFFXU ZKHQ H[WHUQDO VRXUFHV RI SROOHQ ZHUH H[FOXGHG E\ EDJJLQJ 7KXV SROOHQ WUDQVIHU IURP YLDEOH VWDPLQDWH IORZHUV WR UHFHSWLYH SLVWLOn ODWH IORZHUV RQ WKH VDPH LQIORUHVFHQFH DSSHDUV WR EH SUHYHQWHG E\ SUR WRJ\Q\ 3ROOHQ WUDQVIHU IURP PDOH LQIORUHVFHQFHV WR UHFHSWLYH IHPDOH LQn IORUHVFHQFHV RQ WKH VDPH SDOP OLNHZLVH DSSHDUV WR EH UDUH 2Q D JLYHQ SDOP VLPXOWDQHRXV RSHQLQJ RI WKH WZR LQIORUHVFHQFH W\SHV LV XQOLNHO\

PAGE 76

,Q IHPDOH LQIORUHVFHQFHV REVHUYHG RYHUODS RI UHFHSWLYLW\ K DIWHU EUDFW RSHQLQJf ZLWK SROOHQ UHOHDVH IURP D PDOH LQIORUHVFHQFH RQ WKH VDPH SDOP RFFXUUHG RQO\ RQFH 7KH SRVVLELOLWLHV RI LQGXFLQJ VHOISROOLQDWLRQ LQ EDEDVVX DUH XQn NQRZQ $V LQIORUHVFHQFHV FRQWDLQLQJ SLVWLOODWH IORZHUV DUH HLWKHU IXQFn WLRQDOO\ IHPDOH RU PRUH UDUHO\f SURWRJ\QRXVf§DQG RSSRUWXQLWLHV IRU SROOHQ WUDQVIHU EHWZHHQ LQIORUHVFHQFHV RQ WKH VDPH SDOP DUH UDUH f§VHOI SROOLQDWLRQ LV SUREDEO\ XQXVXDO XQGHU QDWXUDO FRQGLWLRQV ,QVHFW SROOLQDWLRQ 3UHVHQWDWLRQ RI FRPSOHWH GDWD RQ DEXQGDQFH DQG EHKDYLRU RI LQVHFW YLVLWRUV WR LQIORUHVFHQFHV RI EDEDVVX DZDLWV LGHQWLn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

PAGE 77

VHUYHG WR IHHG RQ SROOHQ DQG SUREDEO\ XVH ERWK IORZHU W\SHV DV EUHHGLQJ VLWHV $OWKRXJK WKH\ ZHUH IDU OHVV DEXQGDQW RQ IHPDOH LQIORUHVFHQFHV PRUH WKDQ RQH EHHWOH ZDV FRQVLVWHQWO\ REVHUYHG HQWHULQJ HDFK SLVWLOODWH IORZHU (QWU\ ZDV PDGH HLWKHU DORQJ WKH SHWDO PDUJLQV RU PRUH FRPPRQn O\ EHWZHHQ WKH EDVDOO\ IXVHG VWLJPDV DW WKH WLS RI WKH IORZHU 3ROOHQ RI EDEDVVX ZDV IRXQG LQ LVRODWHG FROOHFWLRQV RI 0\ VWURSV VS REWDLQHG IURP IHPDOH LQIORUHVFHQFHV WKDW FRQWDLQHG QR YLDEOH VWDPLQDWH IORZHUV LQGLFDWLQJ SROOHQ WUDQVSRUW IURP DQRWKHU LQIORUHVFHQFH 7KH EHHWOHV ZHUH REVHUYHG LQ ERWK VHFRQGDU\ IRUHVW DQG SDVWXUH DQG VLPLODU VSHFLn PHQV RI 0\ VWURSV ZHUH FROOHFWHG IURP WKH FURZQ RI D EDEDVVX SDOP LQ 3LUDSRUD 0LQDV *HUDLV LQGLFDWLQJ WKDW WKH DVVRFLDWLRQ LV ZLGHVSUHDG :LQG SROOLQDWLRQ 7UDQVSRUW RI EDEDVVX SROOHQ E\ ZLQG ZDV UHn SHDWHGO\ FRQILUPHG E\ YLVXDO REVHUYDWLRQV LQ ERWK WKH SDVWXUH DQG VHFn RQGDU\ IRUHVW VLWHV 3UHOLPLQDU\ DQDO\VLV RI SROOHQ WUDSSLQJ GDWD LQGLn FDWHV D KLJKHU DEXQGDQFH RI ZLQGERUQH SROOHQ LQ WKH RSHQ SDVWXUH WKDQ LQ WKH FORVHG VHFRQGDU\ IRUHVW 9DULDQFHV LQ TXDQWLWLHV RI SROOHQ WUDSSHG LQ ERWK KDELWDWV ZHUH H[WUHPHO\ KLJK ,QVHFW H[FOXVLRQ IURP RQH IHPDOH LQIORUHVFHQFH LQ HDFK KDELWDW UHVXOWHG LQ b IUXLW VHW LQ WKH VHFRQGDU\ IRUHVW WRWDO IORZHUVf DQG b IUXLW VHW LQ WKH SDVWXUH WRWDO IORZHUVf 7KHVH UHVXOWV LQGLFDWH WKDW ZLQG SROOLQDWLRQ RFFXUV LQ ERWK KDELWDWV EXW PD\ EH PRUH HIIHFWLYH RQ WKH UHODWLYHO\ RSHQ SDVWXUH VLWH )UXLW %LRORJ\ 'LVSHUVDO 7KH SURSRUWLRQ RI EDEDVVX IUXLWV UHPRYHG P ZDV VLJQLILFDQWO\ KLJKHU S RQHZD\ $129$f LQ WKH VHFRQGDU\ IRUHVW

PAGE 78

WKDQ LQ WKH ILHOG RU SDVWXUH )LJ f ,Q WKH VHFRQGDU\ IRUHVW IUXLWV ZHUH SUREDEO\ UHPRYHG E\ SDFDV $JRXWL SDFDf DQG DJRXWLV 'DV\SURFWD SXQFWDWDf ,Q WKH ILHOG DQG SDVWXUH UHPRYDO ZDV HIIHFWHG HLWKHU E\ GRPHVWLFDWHG DQLPDOV RU E\ UXQRII 3UHGDWLRQ 7KH RQO\ QRQKXPDQ VHHG SUHGDWRU RI EDEDVVX REVHUYHG DW P\ VWXG\ VLWHV ZDV D VSHFLHV RI EUXFKLG EHHWOH 3DFK\PHUXV QXFOHRUXP $WWDFN RI IUXLWV DQG VHHGV E\ WKLV SUHGDWRU )LJ f RFFXUUHG RQO\ DIWHU IUXLW DEVFLVVLRQ Gf 7KH UDWH RI FRORQL]DWLRQ LQFUHDVHG VKDUSn O\ DW G DQG OHYHOHG RII DW G $OWKRXJK FD b RI IUXLWV ZHUH FRORQL]HG RQO\ FD b RI VHHGV ZHUH GHVWUR\HG DQ DYHUDJH RI RXW RI D WRWDO RI VHHGV SHU IUXLW HVFDSHG SUHGDWLRQ $PRQJ IUXLWV FROOHFWHG G RU ORQJHU DIWHU DEVFLVVLRQ b VXIIHUHG SUHGDWLRQ RI DOO VHHGV &RORQL]DWLRQ RI WKHVH IUXLWV E\ 3 QXFOHRUXP ZDV DSSUR[LPDWHO\ WKH VDPH UHJDUGOHVV RI WKH QXPEHU RI VHHGV SHU IUXLW 7DEOH f %HFDXVH HDFK ODUYD ZDV OLPLWHG WR RQH VHHG KRZn HYHU ORZHU SURSRUWLRQV RI VHHGV ZHUH HDWHQ DV WKH QXPEHU RI VHHGV SHU IUXLW LQFUHDVHG 7DEOH f 7KHVH UHVXOWV VXJJHVW WKDW VHOHFWLRQ E\ VHHG SUHGDFHRXV EHHWOHV IDYRUV PDQ\VHHGHG IUXLWV LQ EDEDVVX 6HHG IDWH ,Q P\ H[SHULPHQW RQ VHHG IDWH RYHU WLPH )LJ f WKH SHUFHQWDJH RI GHIHFWLYH DERUWHG RU URWWHGf VHHGV GLG QRW YDU\ VLJQLILn FDQWO\ DPRQJ VLWHV $IWHU \U PRVW RI WKH VHHGV RQ WKH VHFRQGDU\ IRUHVW DQG SDVWXUH VLWHV KDG VXFFXPEHG WR SUHGDWLRQ UHPRYDO RI SDOP WUHHV WKDW KDUERU VHHGSUHGDFHRXV EUXFKLG EHHWOHV SUREDEO\ UHVXOWHG LQ VLJQLILFDQWO\ ORZHU OHYHOV RI SUHGDWLRQ LQ WKH ILHOG S RQHZD\ $19$f /DUJHO\ GXH WR GLIIHUHQFHV LQ SUHGDWLRQ SHUFHQWDJH RI TXLHV

PAGE 79

)58,76 5(029(' )LJXUH 5HPRYDO RI EDEDVVX IUXLWV RQ WKUHH VWXG\ VLWHV 9HUGH 6KDGHG SRUWLRQV RI EDUV VKRZ SHUFHQWDJH UHPRYHG P DW /DJR RI IUXLWV

PAGE 80

3(5&(17 $77$&.(' )LJXUH $WWDFN RI EDEDVVX IUXLWV DQG VHHGV E\ WKH EUXFKLG EHHWOH 3DFK\PHUXV QXFOHRUXP RYHU WLPH %DUV UHSUHVHQW s VWDQGDUG HUURU

PAGE 81

7DEOH $WWDFN RI EDEDVVX IUXLWV DQG VHHGV E\ WKH VHHG SUHGDFHRXV EHHWOH 3DFK\PHUXV QXFOHRUXP 6(('6)58,7 )58,76 6(('6 7RWDO 1 &RORQL]HG 7RWDO 1 &RORQL]HG 1 b 1 b

PAGE 82

6(('6 3(5 )58,7 , 3$6785( ),(/' )LJXUH 6HHG IDWH RYHU WLPH LQ WKUHH VWXG\ VLWHV DW /DJR 9HUGH

PAGE 83

FHQW VHHGV ZDV VLJQLILFDQWO\ JUHDWHU LQ WKH ILHOG DIWHU \U S RQHZD\ $129$f 7KH UHODWLYH ODFN RI VKDGLQJ LQ WKH SDVWXUH SUREDEO\ DFFRXQWHG IRU WKH VLJQLILFDQWO\ ORZHU SHUFHQWDJH RI JHUPLQDWHG IUXLWV LQ WKLV KDELWDW S RQHZD\ $129$f ,Q WKH H[SHULPHQW RQ ORQJWHUP PRf VHHG IDWH WKH SHUFHQWDJH RI IUXLWV ZLWK JHUPLQDWHG VHHGVf ZDV VLJQLILFDQWO\ KLJKHU S L &KLVTXDUH WHVWf RQ QRQZHHGHG SORWV bf WKDQ RQ ZHHGHG SORWV bf &RQYHUVHO\ WKH SHUFHQWDJH RI IUXLWV ZLWK TXLHVFHQW VHHGVf ZDV VLJQLILFDQWO\ KLJKHU S &KLVTXDUH WHVWf RQ ZHHGHG SORWV bf WKDQ RQ QRQZHHGHG SORWV bf VXVSHFW WKDW WKHVH ODWWHU VHHGV ZHUH TXLHVFHQW UDWKHU WKDQ GRUPDQW WKHLU JHUPLQDWLRQ ZDV SUREDEO\ ORZ GXH WR LQVXIILFLHQW PRLVWXUH RQ WKH QRQZHHGHG SORWV 7KHVH UHVXOWV VXJJHVW WKDW XQGHU UHODWLYHO\ H[SRVHG FRQGLWLRQV EDEDVVX VHHGV FDQ PDLQWDLQ WKHLU YLDELOLW\ IRU H[WHQGHG SHULRGV *HUPLQDWLRQ )LJXUH VKRZV EDEDVVXnV FXPXODWLYH JHUPLQDWLRQ RYHU WLPH EHJLQQLQJ DW WKH SHDN RI IUXLWLQJ GXULQJ WKH ODWWHU SDUW RI WKH GU\ VHDVRQ 0RVW JHUPLQDWLRQ RFFXUUHG FD PR DIWHU IUXLW IDOO KLJK VRLO PRLVWXUH FRQGLWLRQV SUHYDOHQW GXULQJ WKH UDLQ\ VHDVRQ DUH SUREDEO\ QHFHVVDU\ IRU VXFFHVVIXO JHUPLQDWLRQ $OWKRXJK WKLV VHDVRQDO SDWWHUQ RI JHUPLQDWLRQ DSSHDUV WR EH UHODWLYHO\ FRQVWDQW LQ EDEDVVX JHUPLQDWLRQ VXFFHVV RI IUXLWV REWDLQHG IURP GLIIHUHQW SDOP WUHHV YDULHGFRQVLGHUDEO\ DQG LQ VRPH FDVHV VLJQLILFDQWO\ )LJ f $V DOO IUXLWV ZHUH SODQWHG XQGHU LGHQWLFDO FRQGLWLRQV WKHVH UHVXOWV VXJJHVW WKDW REVHUYHG LQWUD VSHFLILF YDULDELOLW\ LQ EDEDVVXnV JHUPLQDWLRQ PD\ EH XQGHU JHQHWLF FRQWURO

PAGE 84

&808/$7,9( *(50,1$7,21 bf )LJXUH &XPXODWLYH JHUPLQDWLRQ RI EDEDVVX IUXLWV RYHU WLPH ,QWHUn YDOV UHSUHVHQW VWDQGDUG HUURU 6KDGHG EDUV UHSUHVHQW UDLQIDOO 'DWD ZHUH REWDLQHG IURP 1RYHPEHU WR $SULO 5$,1)$// FP SHU GD\Vf

PAGE 85

3$/0 180%(5 , , , *(50,1$7,21 bf )LJXUH *HUPLQDWLRQ LQ IUXLWV REWDLQHG IURP EDEDVVX SDOPV (DFK KDWFKHG EDU UHSUHVHQWV RQH SDOP LQWHUYDOV UHSUHVHQW LO VWDQGDUG HUURU

PAGE 86

7DEOH *HUPLQDWLRQ RI EDEDVVX IUXLWV LQ IRXU H[SHULPHQWV $VWHULVNV LQGLFDWH VLJQLILFDQW GLIIHUHQFHV r IRU S 6 rr IRU S6f (;3(5,0(17 4-20, 1 $ 7 1 IWf 7UHDWPHQW &RQWURO 5HPRYDO RI 0HVRFDUS m ,QWHQVLYH %XUQ r 0RGHUDWH %XUQ 6KDGLQJ r

PAGE 87

7KH DEXQGDQFH RI VHHGOLQJV DQG MXYHQLOHV RQ EXUQHG VLWHV KDV OHG WR WKH VXJJHVWLRQ WKDW ILUH SURPRWHV JHUPLQDWLRQ LQ EDEDVVX 6PLWK fr 5HPRYDO RI WKH PHVRFDUS E\ GLVSHUVDO DJHQWV DQG VKDGLQJ FRXOG DOVR VWLPXODWH JHUPLQDWLRQ LQ WKH SDOP 5HVXOWV RI H[SHULPHQWV GHVLJQHG WR WHVW WKHVH K\SRWKHVHV DUH VKRZQ LQ 7DEOH IRXQG WKDW D PRGHUDWH ILUH VXFK DV RFFXUV DQQXDOO\ LQ SDVWXUHV KDG QR HIIHFW RQ JHUPLQDWLRQ VXFFHVV +RZHYHU IUXLWV H[SRVHG WR DQ LQWHQVLYH ILUH VXFK DV RFFXUV ZKHQ IRUHVW IDOORZ LV EXUQHG LQ VKLIWLQJ FXOWLYDWLRQ H[KLELWHG VLJQLILn FDQWO\ UHGXFHG JHUPLQDWLRQ S e &KLVTXDUH WHVWf 5HPRYDO RI WKH PHVRFDUS VLPXODWLQJ WKH HIIHFWV RI GLVSHUVDO DJHQWV ZDV IRXQG WR SURPRWH JHUPLQDWLRQ RI EDEDVVX S ef /LNHZLVH JHUPLQDWLRQ ZDV VLJQLILFDQWO\ KLJKHU LQ VKDGHG IUXLWV WKDQ LQ IUXLWV OHIW LQ WKH VXQ Sef 7KLV ILQDO UHVSRQVH SUREDEO\ UHVXOWV IURP HQKDQFHPHQW RI JHUPLQDWLRQ E\ KLJKHU VRLO PRLVWXUH OHYHOV LQ WKH VKDGH 'LVFXVVLRQ 3KHQRORJ\ %DEDVVXnV SKHQRORJ\ VKRZHG UHPDUNDEOH FRQVWDQF\ RYHU D ZLGH UDQJH RI HFRORJLFDO FRQGLWLRQV 7KH JHQHUDO SDWWHUQ ZDV D SURQRXQFHG SHDN LQ OHDI SURGXFWLRQ DQG IORZHULQJ GXULQJ WKH ZHW VHDVRQ ZKLOH OHDI ORVV DQG IUXLWLQJ SHDNHG GXULQJ WKH GU\ VHDVRQ 'HVSLWH LWV GU\VHDVRQ SHDN IORZHULQJ ZLWKLQ D JLYHQ SRSXODWLRQ RFFXUUHG LQ PRVW PRQWKV 0DOH LQIORUHVFHQFHV ZHUH JHQHUDOO\ PRUH DEXQGDQW WKDQ IHPDOH LQIORUHVFHQFHV DQG FRXOG GHWHFW QR VHDVRQDO YDULDWLRQ LQ VH[ UDWLRV ZLWKLQ WKH SRSXn ODWLRQV REVHUYHG

PAGE 88

/HDI DQG LQIORUHVFHQFH SURGXFWLRQ LQ EDEDVVX DUH VWURQJO\ FRUn UHODWHG GXH WR WKH FORVH SK\VLFDO DVVRFLDWLRQ RI WKH WZR VWUXFWXUHV 1HYHUWKHOHVV PRVW SKHQRORJLFDO VWXGLHV RI SDOPV LQ WKHLU QDWXUDO KDELn WDWV HJ 0RUD 8USL DQG 6ROLV %XOORFN 0\HUV %HDFK LQ SUHVVf KDYH GHDOW H[FOXVLYHO\ ZLWK UHSURGXFWLYH FRPSRQHQWV 7R P\ NQRZOHGJH VWXGLHV RI ERWK OHDI DQG LQIORUHVFHQFH SURGXFWLRQ LQ SDOPV KDYH EHHQ OLPLWHG WR FXOWLJHQV XQGHU SODQWDWLRQ FRQGLWLRQV HJ &KLOG +DUWOH\ f ,Q WKH $IULFDQ RLO SDOP IRU H[DPSOH OHDI SURGXFn WLRQ RFFXUV WKURXJKRXW WKH \HDU ZKLFK UHVXOWV LQ QHDUO\ FRQWLQXRXV IORZHULQJ &RUOH\ HW DO f LQ EDEDVVX OHDI SURGXFWLRQ DQG IORZHUn LQJ WHQG WR EH VHDVRQDO /LNHZLVH WKH PLQLPXP DQQXDO UDWH RI OHDI SURGXFWLRQ LQ RLO SDOP UHFRUGHG LQ WKH OLWHUDWXUH LV FD OHDYHV &RUOH\ HW DO f FRPSDUHG WR D PHDQ RI FD IRXU OHDYHV LQ EDEDVVX 3UREDEO\ DV D UHVXOW RI LWV KLJK UDWH RI OHDI SURGXFWLRQ PLQLPXP OHDI OLIHVSDQ LQ ZLOG VWDQGV RI RLO SDOP LQ 1LJHULD LV RQO\ \U =HYHQ f LQ EDEDVVX PHDQ OHDI OLIHVSDQ LV r \U EHOLHYH WKDW WKLV ODWWHU ILJXUH KDV LPSRUWDQW LPSOLFDWLRQV IRU PDQDJHPHQW RI EDEDVVX XQGHU VKLIWLQJ FXOWLYDWLRQ D WRSLF H[SORUHG IXUWKHU LQ &KDSWHU $V KDV EHHQ GRFXPHQWHG LQ PDQ\ RWKHU VSHFLHV HJ +DUSHU -DQ]HQ E 3LQHUR DQG 6DUXNKDQ f IUXLW \LHOGV LQ EDEDVVX YDU\ FRQVLGHUDEO\ IURP \HDU WR \HDU $V LV WKH FDVH LQ RLO SDOP +DUWOH\ f VXVSHFW WKDW FOLPDWLF IDFWRUV VXFK DV UDLQIDOO PD\ KDYH D VLJQLILFDQW HIIHFW RQ DQQXDO YDULDELOLW\ LQ EDEDVVXnV IUXLW \LHOGV 7KH LQIOXHQFH LV SUREDEO\ PRVW SURQRXQFHG DW WKH WLPH RI LQIORUHVFHQFH GLIIHUHQWLDWLRQ UDWKHU WKDQ DW IUXLW VHW ([DPLQDWLRQ RI GHYHORSPHQWDO

PAGE 89

VHULHV VKRZV WKDW LQIORUHVFHQFH W\SHV PDOH RU IHPDOHf FDQ EH FOHDUO\ GLVWLQJXLVKHG XS WR OHDYHV SULRU WR WKH QHZHVW HPHUJHG OHDI *LYHQ DQ DYHUDJH SURGXFWLRQ RI FD IRXU OHDYHV SHU \HDU WKLV PHDQV WKDW VH[ GHWHUPLQDWLRQ RFFXUV DW OHDVW \U SULRU WR OHDI H[SDQVLRQ LQ WKH FURZQ \U SULRU WR IORZHULQJ DQG QHDUO\ \U SULRU WR IUXLWLQJ 'HWHFWLQJ SRVVLEOH UHODWLRQVKLSV EHWZHHQ FOLPDWH DQG \LHOG ZLOO WKHUHIRUH UHTXLUH ORQJWHUP VWXGLHV RI EDEDVVXnV SKHQRORJ\ 3ROOLQDWLRQ 0\ UHVXOWV LQGLFDWH WKDW UHSURGXFWLRQ LQ EDEDVVX LV SUHGRPLQDQWO\ RU SHUKDSV H[FOXVLYHO\ YLD RXWFURVVLQJ DQG WKDW SROOLQDWLRQ LV HIIHFWHG E\ QLWLGXOLG EHHWOHV DQG ZLQG 6XFK D FRPELQDWLRQ RI SROOLQDWLRQ V\Vn WHPV KDV EHHQ REVHUYHG LQ RWKHU VSHFLHV RI SDOPV 8KO DQG 0RRUH 0RRUH DQG 8KO f 7KH FRFRQXW &RFRV QXFIHUDf DSSHDUV WR EH SROOLn QDWHG E\ HLWKHU KRQH\ EHHV RU ZLQG LQ +DZDLL 6KRGW DQG 0LWFKHOO f 7KH RLO SDOP ZDV WKRXJKW WR EH H[FOXVLYHO\ ZLQGSROOLQD WHG +DUWOH\ f EXW D UHFHQW VWXG\ 6\HG f KDV LPSOLFDWHG ZHHYLOV DV SROOLQDn WRUV ZLWKLQ WKH SDOPnV QDWXUDO UDQJH LQ :HVW $IULFD DQG WKULSV RQ WKH 0DOD\ SHQLQVXOD /LNHZLVH 0\HUV f VXJJHVWHG WKDW ERWK ZLQG DQG LQVHFWV HIIHFW SROOLQDWLRQ LQ 5DSKLD WDHGLJHUD LQ &RVWD 5LFD 1LWLGXOLG EHHWOHV DUH UDUHO\ FLWHG DV SROOLQDWRUV )DHJUL DQG YDQ GHU 3LMO f EXW WKHLU UROH LQ WKH SROOLQDWLRQ RI SDOPV KDV HLWKHU EHHQ GRFXPHQWHG (VVLJ f RU LPSOLFDWHG 'UDQVILHOG f :LQG SROOLQDWLRQ LQ SDOPV KDV EHHQ UHSRUWHG VLQFH DQWLTXLW\ DFFRUGLQJ WR )DHJUL DQG YDQ GHU 3LMO f

PAGE 90

,Q EDEDVVX LQVHFW SROOLQDWLRQ LV SUREDEO\ SUHGRPLQDQW LQ IRUHVWV ZKLOH ZLQG SROOLQDWLRQ LV OLNHO\ WR EH HQKDQFHG LQ RSHQ KDELWDWV VXFK DV SDVWXUHV 1HLWKHU V\VWHP LV H[FOXVLYH RI WKH RWKHU DQG ERWK PD\ KDYH FRPH LQWR SOD\ GXULQJ EDEDVVXnV HYROXWLRQDU\ KLVWRU\ 3UHGRPLQDQFH RI PDOH LQIORUHVFHQFHV VPDOO \P LQ GLDPHWHUf PRQRVXOFDWH SROOHQ SHUVRQDO REVHUYDWLRQf DQG ODFN RI QHFWDULHV DOO SRLQW WR SROOLQDWLRQ E\ ZLQG 0RRUH DQG 8KO f
PAGE 91

E\ ORFDO LQKDELWDQWV WR IHHG RQ WKH PHVRFDUS EXW QRW WKH VHHGVf RI EDEDVVX IUXLWV LQFOXGH SRUFXSLQHV &RHQG SUHKHQ VL OLV DQG &RHQG VSf VSLQ\ UDWV 3URHFKLP\V ORQJLFDXGDWXV DQG 0HVRP\V KLVSLGXVf DQG SHFn FDULHV 7D\DVVX WD DFX DQG 7 SFDUL ERWK ORFDOO\ H[WLQFWf 1RQH RI WKHVH VSHFLHV DSSHDUV WR KDYH D VLJQLILFDQW UROH LQ GLVSHUVDO RI EDEDVVX IUXLWV 0\ ILHOG REVHUYDWLRQV LQGLFDWH WKDW ZDWHUPHGLDWHG GLVSHUVDO LV D VLJQLILFDQW IDFWRU RQ UHODWLYHO\ RSHQ VLWHV VXFK DV SDVWXUHV +HUH VRLO FRPSDFWLRQ UHVXOWV LQ LQFUHDVHG VXUIDFH UXQRII E\ ZKLFK IUXLWV ZHUH REVHUYHG WR EH FDUULHG GLVWDQFHV RI FD P 7KH SRWHQWLDO UROH RI ZDWHUPHGLDWHG GLVSHUVDO LV IDU JUHDWHU 'LVVHPLQDWLRQ E\ ULYHUV ODUJHn O\ DFFRXQWV IRU WKH SUHVHQW GLVWULEXWLRQ RI EDEDVVX RYHU QHDUO\ KDOI D FRQWLQHQW VHH &KDSWHU f $OWKRXJK LW UHPDLQV XQGRFXPHQWHG WKH UROH RI KXPDQV LQ WKH GLVSHUn VDO RI EDEDVVX LV SRWHQWLDOO\ VLJQLILFDQW $V GLVFXVVHG LQ &KDSWHU WKH SUHVHQW RFFXUUHQFH RI EDEDVVX LQ WKH %UD]LOLDQ VWDWH RI &HDU£ PD\ EH GXH WR GLVVHPLQDWLRQ E\ LQGLJHQRXV SHRSOHV 7RGD\ VHWWOHUV WKURXJKRXW WKH $PD]RQ UHJLRQ FRPPRQO\ WUDQVSODQW EDEDVVX WR DUHDV ZKHUH LW GLG QRW IRUPHUO\ RFFXU 2Q D ORFDO VFDOH LQ 0DUDQK£R JDWKHULQJ RI EDEDVVX IUXLWV LV D ZLGHVSUHDG DFWLYLW\ SDUWLFXODUO\ DPRQJ ZRPHQ ,QWHQVLW\ RI IUXLW JDWKHULQJ LV HVSHFLDOO\ KLJK RQ UHODWLYHO\ RSHQ DFFHVVLEOH VLWHV VXFK DV SDVWXUHV $OWKRXJK WKH VHHGV IURP DOPRVW DOO JDWKHUHG IUXLWV DUH XOWLPDWHO\ FRQVXPHG VRPH DFFLGHQWDO GLVSHUVDO RFFXUV GXULQJ JDWKHUn LQJ DQG WUDQVSRUW 0RUH VLJQLILFDQWO\ SHUKDSV LV WKH UROH RI KXPDQ JDWKHUHUV RI EDEDVVX IUXLWV DV DJHQWV RI VHOHFWLRQ 3HRSOH SUHIHU

PAGE 92

HQWLDOO\ JDWKHU ODUJH PDQ\VHHGHG IUXLWV IRU H[WUDFWLRQ RI WKH RLOULFK VHHGV VPDOOHU IHZVHHGHG IUXLWV DUH OHIW WR UHJHQHUDWH $OWKRXJK URGHQWV UXQRII DQG SHRSOH SOD\ VLJQLILFDQW UROHV DV GLVSHUVDO DJHQWV EDEDVVXnV PDVVLYH IUXLWV XVXDOO\ Jf DUH UHODn WLYHO\ LPPRELOH 7KLV LPPRELOLW\ FRQVWLWXWHV D FUXFLDO OLPLWLQJ IDFWRU LQ WKH SDOPnV HFRORJ\ DQG ODUJHO\ DFFRXQWV IRU LWV FXUUHQW GLVWULEXWLRQ VHH &KDSWHU f 5HGXFHG GLVSHUVDO LV WKH SULFH EDEDVVX SD\V IRU HQn KDQFHG SURWHFWLRQ IURP VHHG SUHGDWRUV 3UHGDWLRQ 2I LQVHFW VSHFLHV NQRZQ WR IHHG RQ EDEDVVX 7DEOH f QLQH DUH SRWHQWLDO VHHG SUHGDWRUV 0RVW RI WKH ODWWHU IHHG RQ VWRUHG VHHGV WKH RQO\ VHHGSUHGDFHRXV LQVHFWV XQGHU QDWXUDO FRQGLWLRQV DUH WZR VSHFLHV RI EUXFKLG EHHWOH 3DFK\PHUXV QXFOHRUXP DQG &DU\EUXFKXV OLSLVPDWXV 2I WKHVH RQO\ WKH IRUPHU ZDV SUHVHQW DW DOO RI P\ VWXG\ VLWHV LQ 0DUDQKDR 3RSXODUO\ NQRZQ LQ %UD]LO DV ELFKR GR FRFR WKH ODUYDH RI WKLV EHHWOH SUH\ RQ WKH VHHGV RI D ZLGH YDULHW\ RI SDOPV LQFOXGLQJ HFRQRPLF VSHFLHV VXFK DV FD UQDXED &RSHUQLFLD SUXQLIHUDf OL FXUL 6\D JUXV FRU RQ DWDf SLDVVDYD $WWDOHD IXQLIHUDf WXFXP $VWURFDU\XP YXOJDUHf FRFRQXW &RFRV QXFIHUDf DQG $IULFDQ RLO SDOP (ODHLV JXLQHHQVLVf 7KH SHQHWUDWLRQ RI EDEDVVX IUXLWV E\ WKLV EHHWOH ZDV GHVFULEHG E\ %RQGDU f 2YLn SRVLWLQJ EHHWOHV UHPDLQ LQ WKH FURZQ RI EDEDVVX GXULQJ WKH GD\ DW QLJKW WKH\ GHSRVLW HJJV RQ WKH VXUIDFHV RI IDOOHQ IUXLWV 7KH RGRU RI WKH IHUPHQWLQJ IUXLW PHVRFDUS SUREDEO\ DWWUDFWV RYLSRVLWLQJ EHHWOHV VHH -DQ]HQ f 6XFK DWWUDFWDQWV DSSHDU WR SHUVLVW IRU D PD[LPXP RI G DIWHU IUXLW DEVFLVVLRQ DIWHU ZKLFK FRORQLVDWLRQ RI EDEDVVX IUXLWV E\

PAGE 93

7DEOH 3UHOLPLQDU\ FKHFNOLVW RI LQVHFWV WKDW IHHG RQ EDEDVVX $VWHULVNV LQGLFDWH HFRQRPLFDOO\ LPSRUWDQW SHVWV 6RXUFH &RVWD /LPD f /&8////P 6WRUHG 7$;21 6WHPV /HDYHV ,QIORUHVFHQFHV )UXLWV 6HHGV +RPRSWHUD $VWHUROHFDQLLGDH $VWHUROHFDQLXP VS 'LDVSLGLGDH $VSLGLRWXV GHVWUXFWRU /HSLGRSWHUD 3K\FLWLGDH &DGUD FDXWHOOD 3ORGLD LQWHUSXQFWHOOD &ROHSWHUD %UXFKLGDH &DUYREUXFKXV OLSLVPDWXVr &BW SHUJDQGHL 3DFKYPHUXV QXFOHRUXPr eMB ROHDULXVr %RVWULFKLGDH 'LQRGHUXV PLQXWXV &HUDPE\FLGDH 0DFURGRQWLD FHUYLFRUQLV &KU\VRPHOLGDH &RUDOLRPHOD EUXQQHD &OHULGDH 1HFURELD UXILSHVr 'HUPHVWLGDH f 'HUPHVWHV DWHUr

PAGE 94

7DEOH FRQWnG 2&&855(1&( 7$;21 6WHPV /HDYHV ,QIORUHVFHQFHV )UXLWV 6WRUHG 6HHGV &ROHSWHUD FRQWnGf &XFXMLGDH 2UY]DH'KLOXV VXULQDPHQVLVr &XUFXOLRQLGDH %DULGLQDH 7RQHVLD EDEDVVX &KROLQDH +RUDDOLQRWXV FRULDFHXVr % YDOLGXV 3HWDORFKLOLQDH &HOHVWHV EDEDVVX &I PRVHVL & RUELHQYDH & URVHQWDOL & WHL[HLUDOHLWHL 3\URSLQDH %HORSRHXV VS % RUELHQYDH 5KLQRVWRPXV EDUELURVWULV 5KYQFKR'KRUXV 'DOPDUXPr 6'KHQR'KRUXV VGG 1LWLGXOLGDH &DU'2'KLOXV VGGr

PAGE 95

EHHWOH ODUYDH DSSDUHQWO\ FHDVHG )LJ f $SSUR[LPDWHO\ G IROn ORZLQJ RYLSRVLWLRQ WLQ\ [ PPf ODUYDH HPHUJH DQG SURFHHG WR WKH EDVH RI WKH IUXLWV 3HQHWUDWLRQ WDNHV SODFH WKURXJK WKH JHUPLQDWLRQ SRUHV ZKLFK RIIHU PLQLPDO UHVLVWDQFH 3ULRU WR DEVFLVVLRQ IUXLWV DUH DWWDFKHG DW WKH EDVHV DQG WKHLU JHUPLQDWLRQ SRUHV DUH WKXV SURWHFWHG ZKLFK H[SODLQV ZK\ IUXLWV DUH RQO\ DWWDFNHG DIWHU WKH\ IDOO $SSUR[Ln PDWHO\ G DIWHU SHQHWUDWLRQ EHJLQV WKH ODUYDH UHDFK WKH VHHGV 2EVHUYDWLRQV RI IUXLW FRORQL]DWLRQ UHYHDOHG D PD[LPXP RI RQH ODUYD SHU VHHG FKDPEHU ORFXOHf VXJJHVWLQJ WKDW ODWH DUULYDOV DUH SUREDEO\ FDQQLn EDOL]HG VHH -DQ]HQ f PRYHPHQW EHWZHHQ VHHG FKDPEHUV LV SURKLELWHG E\ WKH WKLFNQHVV RI WKH LQQHU HQGRFDUS ZDOOV 2QH VHHG SURYLGHV VXIILFn LHQW IRRG IRU D ODUYDnV FRPSOHWH GHYHORSPHQW $GXOWV HPHUJH FD PR DIWHU RYLSRVLWLRQ E\ EXUURZLQJ D FD PPZLGH H[LW KROH WKURXJK WKH HQGRFDUS ,Q D FORVHO\ UHODWHG SDOP 6FKHHOHD URVWUDWDf LQ &RVWD 5LFD %UDGn IRUG DQG 6PLWK f IRXQG WKDW FD b RI WKH IUXLWV ORVW DOO WKHLU VHHGV WR D SUHGDFHRXV EUXFKLG EHHWOH &DU\REUXFKXV EXVFNLf RU WR URn GHQWV LQ EDEDVVX WKH WRWDO ILJXUH ZDV FD b H[FOXGLQJ SRWHQWLDO ORVVHV WR KXPDQVf $ NH\ FRPSRQHQW RI WKLV KLJK SUHGDWRU HVFDSH LV WKH SUHGRPLQDQFH RI PDQ\VHHGHG IUXLWV LQ EDEDVVX 7DEOH f LQ FRQWUDVW PRVW FD bf IUXLWV LQ WKH &RVWD 5LFDQ SRSXODWLRQ RI 6K URVWUDWD ZHUH RQHVHHGHG %UDGIRUG DQG 6PLWK f ,Q ERWK VSHFLHV WKH SURSRUWLRQ RI VHHGV HDWHQ E\ SUHGDWRUV GHFOLQHG DV WKH QXPEHU RI VHHGV SHU IUXLW LQFUHDVHG 7KXV LQ 6B URVWUDWD PHDQ VHHG ORVVHV WR SUHGDWRUV LQ RQH WZR DQG WKUHHVHHGHG IUXLWV ZHUH b b DQG b UHVSHFWLYHO\

PAGE 96

%UDGIRUG DQG 6PLWK f WKH FRUUHVSRQGLQJ ILJXUHV IRU EDEDVVXf§b b DQG b 7DEOH ff§DUH UHPDUNDEO\ VLPLODU 7KH WUHQG VXJJHVWV WKDW LQ ERWK VSHFLHV VHOHFWLRQ E\ VHHG SUHGDWRUV IDYRUV PDQ\VHHGHG IUXLWV +RZHYHU VXFK VHOHFWLRQ SUREDEO\ RSHUDWHV ZLWKLQ VWULFW FRQVWUDLQWV %UDGIRUG DQG 6PLWK f IRXQG WKDW LQ URVWUDWD LQGLYLGXDO VHHGV LQ RQHVHHGHG IUXLWV FRQWDLQHG b DQG b KLJKHU HQHUJ\ FRQWHQW WKDQ WKRVH LQ WZR DQG WKUHHVHHGHG IUXLWV UHVSHFWLYHO\ 6HHGOLQJV IURP RQHVHHGHG IUXLWV VKRXOG WKHUHIRUH KDYH D FRPSHWLWLYH DGYDQWDJH ,Q DGGLWLRQ VHHGOLQJV HPHUJLQJ IURP WKH VDPH PDQ\VHHGHG IUXLW FRPSHWH LQWHQVLYHO\ ZLWK HDFK RWKHU IXUWKHU LQFUHDVLQJ WKH FRPSHWLWLYH DGYDQn WDJH RI VHHGOLQJV IURP RQHVHHGHG IUXLWV 7KH QXPEHU RI VHHGV SHU IUXLW WKXV UHSUHVHQWV D FRPSURPLVH EHWZHHQ RSSRVLQJ VHOHFWLYH IRUFHV 7KH HQGRFDUS RI EDEDVVX LV H[FHSWLRQDOO\ WKLFN [ PP VH f DQG UHVLVWDQW WR FUXVKLQJ )LJ f FRPSDUHG WR RWKHU VSHFLHV RI SDOPV ,W ZRXOG WKXV DSSHDU WR SURYLGH HIIHFWLYH GHWHUUHQFH DJDLQVW VHHG SUHGDWRUV
PAGE 97

-HVVHQLD EDWDXD f $VWURFRU\XP FI PDFURFDO\[ f 6RFURWHD H[KRUUKL]D f ,ULDUWHR YHQWULFRVD f 0DXULWLD IOH[XRVD f 6FKHHOHD VS f 3K\IHOHSKDV PLFURFDUSD f 2UELTQ\D PDUWLDQD f ,, O O f§Of§, n, O _f§ )25&( 72 %5($. Wf L L L U )LJXUH )RUFH UHTXLUHG WR EUHDN SDOP IUXLWV ,QWHUYDOV UHSUHVHQW W VWDQGDUG GHYLDWLRQ YDOXHV LQ SDUHQWKHVHV DUH VDPSOH VL]HV 'DWD IRU 2UELJQ\D PDUWLDQD IURP )RQVHFD f GDWD IRU DOO RWKHU VSHFLHV IURP .LOWLH f

PAGE 98

$RWXV WULNLQJDWXV DFFRUGLQJ WR ORFDO LQIRUPDQWVf VXFFHVVIXOO\ SUH\ RQ EDEDVVX E\ SXOOLQJ RII LPPDWXUH IUXLWV DQG FRQVXPLQJ WKH OLTXLG HQGRn VSHUP ZLWKLQ VHH ,]DZD DQG 0L]XQR f 7KXV VHHG SUHGDWRUV HLWKHU DUH FRPSOHWHO\ GHWHUUHG RU VXFFHVVIXOO\ E\SDVV EDEDVVXnV WKLFN HQGRFDUS VXVSHFW WKDW WKHUH LV QR SUHGDWRU FXUUHQWO\ H[HUWLQJ VXIILFLHQW VHOHFn WLYH SUHVVXUH WR DFFRXQW IRU WKH HYROXWLRQ RI WKLV H[WUDRUGLQDU\ VWUXFn WXUH &RHYROXWLRQ ZLWK ODUJH 3OHLVWRFHQH PDPPDOV -DQ]HQ DQG 0DUWLQ f WKXV SURYLGHV WKH PRVW VDWLVIDFWRU\ H[SODQDWLRQ IRU WKH RULJLQ RI WKH WKLFN PDVVLYH IUXLWV FKDUDFWHULVWLF RI EDEDVVX *HUPLQDWLRQ 7KH UHVXOWV RI P\ H[SHULPHQWV LQGLFDWH WKDW JHUPLQDWLRQ LQ EDEDVVX LV OHDVW VXFFHVVIXO RQ GLVWXUEHG VLWHV VXFK DV SDVWXUHV DQG VKLIWLQJ FXOWLYDWLRQ SORWV ZKHUH VKDGLQJ LV PLQLPDO ILUHV HLWKHU IUHTXHQW RU LQWHQVLYH DQG PHVRFDUSUHPRYLQJ GLVSHUVDO DJHQWV UDUH :KLOH JHUPLQDn WLRQ VXFFHVV RI EDEDVVX LV WKXV UHGXFHG LWV SHFXOLDU PRGH RI JHUPLQDn WLRQ FRQIHUV XSRQ WKH SDOP DQ H[WUDRUGLQDU\ FDSDFLW\ WR UHFRYHU IURP WKH VWUHVVHV LQKHUHQW WR WKHVH VLWHV *HUPLQDWLRQ LQ EDEDVVX RFFXUV ZKHQ RQH RU PRUH HPEU\RQLF D[HV HPHUJH IURP WKH JHUPLQDWLRQ SRUHV DW WKH EDVH RI WKH IUXLW (DFK D[LV HQWHUV WKH VRLO DQG SHQHWUDWHV WR D YDULDEOH GHSWK ZKHUH LW VXEVHn TXHQWO\ GLIIHUHQWLDWHV LQWR URRWV DQG OHDYHV )LJ f 7KLV VRFDOOHG FU\SWFJHDO PRGH RI JHUPLQDWLRQ LV IRXQG WKURXJKRXW WKH $WWDOHD DOOLDQFH DQG LQ GLVWDQWO\ UHODWHG SDOP JHQHUD VXFK DV $FDQWKRFRFRV DQG 6DEDO &RUQHU f DV ZHOO DV LQ PDQ\ RWKHU SODQW IDPLOLHV -DFNVRQ f ,Q SODQWV WKDW H[KLELW FU\SWRJHDO JHUPLQDWLRQ WKH VKRRW DULVHV IURP

PAGE 99

L!XUH &U\SWRJHDO JHUPLQDWLRQ LQ EDEDVVX IURP OHIW WR ULJKW VHHGOLQJV ZHUH DQG ZN ROG 6FDOH FP

PAGE 100

EHORZ JURXQG HYHQ WKRXJK WKH VHHG JHUPLQDWHV RQ WKH VXUIDFH ,Q WKH FDVH RI EDEDVVX WKH XQGHUJURXQG PHULVWHP LQ VHHGOLQJV DQG VWHPOHVV MXYHQLOHV HQDEOHV WKHP WR UHJHQHUDWH RQ VLWHV WKDW DUH VXEMHFW WR FXWn WLQJ DQG EXUQLQJ 0DQ\ VSHFLHV H[KLELWLQJ FU\SWRJHDO JHUPLQDWLRQ RFFXU LQ ILUHSURQH KDELWDWV DQG -DFNVRQ f SURSRVHG WKDW LW UHSUHVHQWV DQ HYROXWLRQDU\ UHVSRQVH WR ILUH +RZHYHU VXVSHFW WKDW EDEDVVXf§DV ZHOO DV PDQ\ RI LWV UHODWLYHV LQ WKH $WWDOHD DOOLDQFHf§HYROYHG LQ PRLVW IRUHVWHG KDELWDWV ZKHUH QDWXUDO ILUHV ZHUH LQIUHTXHQW VHH &KDSWHU f 7KH SUHVHQFH RI FU\SWRJHDO JHUPLQDWLRQ LQ WKHVH VSHFLHV LV PRUH OLNHO\ GXH WR WKH JHQHUDO ODFN RI VHFRQGDU\ JURZWK WKURXJKRXW WKH IDPLO\ /DWHUDO H[SDQVLRQ RI WKH SDOP VWHP LV DFFRPSOLVKHG E\ VRFDOOHG HVWDEn OLVKPHQW JURZWK 7RPOLQVRQ DQG =LPPHUPDQQ f GXULQJ ZKLFK WKH VWHP WDNHV RQ D KLJKO\ XQVWDEOH REFRQLFDO VKDSH 2QH RI WKH PHDQV E\ ZKLFK SDOPV SURYLGH VXSSRUW IRU WKLV VWUXFWXUH LV WR EXU\ WKHLU VKRRW D[HV XQGHUJURXQG 5HJDUGOHVV RI WKH VHOHFWLYH SUHVVXUHV WKDW EURXJKW DERXW LWV HYROXn WLRQ WKHUH LV QR GRXEW WKDW FU\SWRJHDO JHUPLQDWLRQ SOD\V DQ LPSRUWDQW UROH LQ WKH FXUUHQW SUROLIHUDWLRQ RI EDEDVVX RQ VLWHV VXEMHFWHG WR KXPDQ GLVWXUEDQFHV

PAGE 101

&+$37(5 (67$%/,6+0(17 (VWDEOLVKPHQW IURP VHHG FRQVLVWV RI VHHGOLQJ UHFUXLWPHQW DQG VXEVHTXHQW JURZWK 7KLV FRQVWLWXWHV D SDUWLFXODUO\ YXOQHUDEOH SKDVH RI D SODQWnV OLIH F\FOH +DUSHU f )ROORZLQJ VXFFHVVIXO GLVSHUVDO WR D PLFURVLWH VXLWDEOH IRU JHUPLQDWLRQ PRVW VHHGOLQJV IDFH GHDWK GXH WR VXFK FDXVHV DV VKDGLQJ GDPDJH E\ IDOOLQJ OLWWHU KHUELYRU\ GLVHDVH RU GURXJKW &RPELQDWLRQV RI WKHVH FDXVHV PD\ RFFXU IRU H[DPSOH VKDGHG VHHGOLQJV DUH PRUH VXVFHSWLEOH WR GLVHDVH WKDQ QRQVKDGHG VHHGOLQJV 7DKHU DQG &RRNH f 'XH WR WKH VXFFHVV RI EDEDVVX LQ DWWDLQLQJ FRPPXQLW\ GRPLQDQFH WKH SDOPnV SHUIRUPDQFH GXULQJ WKH YXOQHUDEOH HVWDEOLVKPHQW SKDVH LV RI VSHFLDO LQWHUHVW ,Q WKLV FKDSWHU GHVFULEH H[SHULPHQWV GHVLJQHG WR WHVW EDEDVVXnV SHUIRUPDQFH GXULQJ HVWDEOLVKPHQW 3HUIRUPDQFH LQFOXGHV ERWK DGDSWLYHQHVV DQG FRPSHWLWLYHQHVV 7KH IRUPHU LV GHOLPLWHG E\ SK\VLRORJLFDO OLPLWV RI WROHUDQFH WKH ODWWHU LV FRQWUROOHG E\ WKH LQWHQVLW\ RI FRPSHWLWLRQ +XWFKLQVRQ f $GDSWLYHQHVV DQG FRPSHWLn WLYHQHVV RI EDEDVVX ZHUH UHVSHFWLYHO\ PRQLWRUHG RQ ZHHGHG DQG QRQZHHGHG SORWV 7KH H[SHULPHQWDO WUHDWPHQWV FRQVLVWHG RI WKH SUHVHQFH RU DEVHQFH RI VXEVLGLHV LQ WKH IRUP RI f LUULJDWLRQ f IHUWLOL]DWLRQ f LQVHFWLFLGH DSSOLFDWLRQV DQG f FOLPDWH 7KH K\SRWKHVL]HG RXWFRPHV 7DEOH f SUHGLFWHG WKDW DGDSWLYHQHVV ZRXOG EH KLJKHU XQGHU VXEVLGL]HG FRQGLWLRQV ZKHUHDV FRPSHWLWLYHQHVV ZRXOG EH KLJKHU XQGHU QRQVXEVLGL]HG FRQGLWLRQV 3UHFHGHQW H[LVWV IRU WKLV ODWWHU SUHGLFWLRQ 6WUHVVHV VXFK

PAGE 102

7DEOH +\SRWKHVL]HG VXFFHVV RI EDEDVVX LQ IRXU H[SHULPHQWV 6XFFHVV UHVXOWV IURP WKH LQWHUDFWLRQ RI DGDSWLYHQHVV WR WKH VWUHVV IDFWRU ZHHGHG SORWVf DQG FRPSHWLWLYHQHVV QRQZHHGHG SORWVf 3UHGLFWHG UHVXOWV DUH LQGLFDWHG E\ WKH GLJLWV PRVW VXFFHVVIXO OHDVW VXFFHVVIXO (;3(5,0(17 675(66 )$&725 :(('(' 121:(('(' ,QVHFWLFLGHV ,QVHFWLFLGH 1R ,QVHFWLFLGHV )HUWLOL]HUV )HUWLOL]HU 1R )HUWLOL]HUV ,UULJDWLRQ ,UULJDWLRQ 1R ,UULJDWLRQ :HW &OLPDWH &OLPDWH 'U\ &OLPDWH

PAGE 103

DV QXWULHQWSRRU VRLOV &RQQHOO f IORRGLQJ +RRSHV f LQWHQVLYH KHUELYRU\ &KXUFKLOO HW DO f DQG LQFUHDVLQJ GU\QHVV *HQWU\ f KDYH UHSHDWHGO\ EHHQ LPSOLFDWHG DV LPSRUWDQW FDXVDO DJHQWV RI UHGXFHG GLYHUVLW\f§ZLWK FRQFRPLWDQW LQFUHDVHV LQ GRPLQDQFH E\ RQH RU D IHZ VSHFLHVf§LQ SODQW FRPPXQLWLHV 0HWKRGV ([SHULPHQWV LQYROYLQJ LUULJDWLRQ IHUWLOL]DWLRQ DQG LQVHFWLFLGH DSSOLFDWLRQV ZHUH VHW XS LQ /DJR 9HUGH 0DUDQKVR WKH FOLPDWH H[SHULPHQW ZDV FDUULHG RXW DW /DJR 9HUGH GU\ VLWHf DQG %HOP 3DUD ZHW VLWHf 7KH /DJR 9HUGH VLWH FRQVLVWHG RI FD \UROG VHFRQGDU\ IRUHVW GRPLQDWHG E\ MXYHQLOH EDEDVVX $Q DUHD RI FD P ZDV FXW RQ 6HSWHPEHU DQG EXUQHG RQ 2FWREHU 7R SURPRWH KRPRJHQHLW\ UHPDLQLQJ XQEXUQHG PDWHULDO ZDV VXEVHTXHQWO\ JDWKHUHG LQWR SLOHV DQG EXUQHG DJDLQ 7R SUHYHQW IXWXUH FRQIXVLRQ EHWZHHQ SODQWHG DQG SUHYLn RXVO\ HVWDEOLVKHG VHHGOLQJV DQG MXYHQLOHV WKH ODWWHU ZHUH NLOOHG HLWKHU E\ XSURRWLQJ LQ WKH FDVH RI VHHGOLQJVf RU E\ GULYLQJ VWDNHV LQWR WKH DSLFDO PHULVWHPV IROORZLQJ GHIROLDWLRQ LQ WKH FDVH RI MXYHQLOHVf 7KLV ODWWHU WHFKQLTXH RIWHQ KDG WR EH UHSHDWHG WR DFKLHYH WKH GHVLUHG HIIHFW 2Q 1RYHPEHU HLJKW\ [ P SORWV ZHUH VHW XS DQG EDEDVVX IUXLWV SODQWHG LQ HDFK :LWK WKH H[FHSWLRQ RI WKH FOLPDWH H[SHULPHQW GHVFULEHG EHORZf IUXLWV ZHUH REWDLQHG ORFDOO\ DQG YLUWXDOO\ DOO KDG ULSHQHG LH DEVFLVHGf ZLWKLQ WKH SDVW PRQWK DFFRUGLQJ WR ORFDO LQIRUn PDQWV 7KH KXVNV ZHUH UHPRYHG WR SURPRWH JHUPLQDWLRQ DQG LQKLELW GLVSHUVDO E\ DQLPDOV )UXLWV ZHUH SDUWLDOO\ EXULHG WR SUHYHQW GLVSHUVDO E\ UXQRII

PAGE 104

7KH %HOHP VLWH FRQVLVWHG RI FD \UROG VHFRQGDU\ IRUHVW LQ ZKLFK EDEDVVX ZDV DEVHQW 7KH VRLO ZDV D '\VWURSKLF
PAGE 105

)LJ f HQWLUH OHDYHV RI YDU\LQJ OHQJWKV ZHUH KDUYHVWHG IURP EDEDVVX VHHGOLQJV WKH OHDYHV ZHUH WKHQ WUDFHG DQG WKH FXWRXW WUDFLQJV SDVVHG WKURXJK DQ DUHD PHWHU /L&RU 0RGHO f 7KH /$, RI EDEDVVX ZDV GHWHUPLQHG E\ PHDVXULQJ DOO OHDI OHQJWKV FRQYHUWLQJ WR DUHDV DQG GLYLGLQJ WKH VXP E\ WKH WRWDO DUHD RI WKH SORW 7R SHUPLW DFFXUDWH PHDVXUHV RI LQGLYLGXDO JURZWK ZLWKRXW ELDV GXH WR QHZO\ JHUPLQDWHG VHHGOLQJV FDOFXODWHG OHDI DUHD SHU SDOP XVLQJ RQO\ WKRVH VHHGOLQJV SUHVHQW VLQFH WKH ILUVW PHDVXUHPHQW FD PR DIWHU LQLWLDWLRQ RI WKH H[SHULPHQWVf $Q DGGLWLRQDO PHDVXUH RI JURZWK ZDV REWDLQHG DIWHU WKH H[SHULPHQWV KDG UXQ FD PR 0D\ f E\ KDUYHVWLQJ DERYHJURXQG ELRPDVV RI EDEDVVX ZKLFK FRQVLVWHG HQWLUHO\ RI OHDYHV DV VWHPV KDG QRW HPHUJHGf ZLWKLQ WKH FHQWUDO [ P DUHD RI HDFK SORW $ FP ERUGHU ZDV OHIW WR UHGXFH HGJH HIIHFWVf 7KH KDUYHVWHG PDWHULDO ZDV RYHQGULHG DW & DQG ZHLJKHG WR QHDUHVW JUDP $ PHDVXUH RI FRPSHWLWLYHQHVV ZDV REWDLQHG E\ FRPSDULQJ WKH ELRPDVV RI EDEDVVX ZLWK WKDW RI RWKHU VSHFLHV RQ WKH QRQZHHGHG SORWV %LRPDVV RI VSHFLHV RWKHU WKDQ EDEDVVX ZDV VHSDUDWHG LQWR WKH IROORZLQJ FRPSRn QHQWV f IORZHUV DQG IUXLWV f JUHHQ OHDYHV JUHHQ EUDQFKHV DQG JUHHQ VWHPV f EURZQ EUDQFKHV DQG EURZQ VWHPV DQG f VWDQGLQJ GHDG &RPSHWLWLYHQHVV ZDV IXUWKHU PHDVXUHG E\ SHULRGLFDOO\ GHWHUPLQLQJ WKH /$, RI EDEDVVX UHODWLYH WR LWV FRPSHWLWRUV 7KH /$, ZDV PHDVXUHG E\ ORZHULQJ D SOXPE ERE RQWR SRLQWV GLVWULEXWHG WKURXJKRXW HDFK [ P SORW H[FOXGLQJ PDUJLQV DW HDFK OHDI FRQWDFW WKH VSHFLHV DQG FP KHLJKW LQWHUYDO ZHUH UHFRUGHG 7KLV PHWKRG XQGHUHVWLPDWHV WUXH /$, E\

PAGE 106

/($) $5($ FP YM )LJXUH /($) /(1*7+ FPf /HDI DUHD DV D IXQFWLRQ RI OHDI OHQJWK LQ VHHGOLQJV RI EDEDVVX 7KH EHVW ILW ZDV OHDI DUHD FPnf OHDI OHQJWK FPf f OHDI OHQJWK FPf

PAGE 107

DQ DPRXQW GHSHQGHQW RQ WKH OHDInV GHYLDWLRQ IURP WKH KRUL]RQWDO (ZHO HW DO f ,UULJDWLRQ ([SHULPHQW 7KLV H[SHULPHQW IROORZHG D [ FRPSOHWHO\ UDQGRPL]HG GHVLJQ LQYROYLQJ SORWV 2Q DOWHUQDWH GD\V EHJLQQLQJ -DQXDU\ HDFK RI SORWV ZDV LUULJDWHG ZLWK / RI ZDWHU ZKLFK ZDV HTXLYDOHQW WR DQ LUULJDWLRQ RI FP GHSWK )HUWLOL]HU ([SHULPHQW 7KLV H[SHULPHQW IROORZHG D [ FRPSOHWHO\ UDQGRPL]HG GHVLJQ LQYROYLQJ SORWV )ROORZLQJ DGYLFH RI ORFDO DJURQRPLVWV IHUWLOL]HU GRVDJHV HTXLYDOHQW WR NJ 1 DV 1+Af NJ 3 DV 3Af DQG NJ DV .&f SHU KD ZHUH DSSOLHG WR SORWV RQ 'HFHPEHU ,GHQWLFDO GRVDJHV RI 1 ZHUH DSSOLHG DW PR LQWHUYDOV DQG RI 3 DQG DW PR LQWHUYDOV &RPSRVLWH VRLO VDPSOHV IURP DOO SORWV FRPSULVHG RI VXEVDPSOHV SHU SORW ZHUH REWDLQHG DW WKH FRQFOXVLRQ RI WKH H[SHULPHQW 0D\ f ,QVHFWLFLGH ([SHULPHQW 7KLV H[SHULPHQW IROORZHG D [ UDQGRPL]HG FRPSOHWH EORFN GHVLJQ LQYROYLQJ SORWV $SSOLFDWLRQ RI LQVHFWLFLGH DQG QHPDWLFLGH ZDV EORFNHG WR PLQLPL]H EHWZHHQSORW FRQWDPLQDWLRQ $W ZHHNO\ LQWHUYDOV EHJLQQLQJ RQ -DQXDU\ SORWV ZHUH VSUD\HG ZLWK D FF SHU / VROXWLRQ RI WKH QRQV\VWHPLF LQVHFWLFLGH V\QWKHWLF S\UHWKUXP NQRZQ FRPPHUFLDOO\ DV 'HFLVf $W PRQWKO\ LQWHUYDOV EHJLQQLQJ RQ 'HFHPEHU JP RI WKH V\VWHPLF QHPDWLFLGH )XUDGDQ )0& ZHUH DSSOLHG

PAGE 108

&OLPDWH ([SHULPHQW 7KLV H[SHULPHQW IROORZHG D [ FRPSOHWHO\ UDQGRPL]HG GHVLJQ LQ WZR EORFNV GU\ DQG ZHW VLWHVf WKXV LQYROYLQJ D WRWDO RI SORWV 7R WHVW IRU SRVVLEOH HFRW\SLF GLIIHUHQFHV EHWZHHQ ORFDOO\ DGDSWHG SRSXODWLRQV IUXLW RULJLQ ZDV VXSHULPSRVHG DV D WUHDWPHQW ,Q DGGLWLRQ WR /DJR 9HUGH IUXLWV ZHUH REWDLQHG IURP D SRSXODWLRQ RI EDEDVVX QHDU WKH YLOODJH RI 7UDFXDWHXD FD NP ( RI %HOHP )LJ f 7KH YDULDQFHV RI WKH GDWD RQ EDEDVVXnV SHUIRUPDQFH 7DEOHV WR f ZHUH QRW KRPRJHQHRXV DFFRUGLQJ WR %DUWOHWWnV WHVW 6QHGHFRU DQG &RFKUDQ f EXW ZHUH JHQHUDOO\ SURSRUWLRQDO WR WKH VTXDUH RI WKH PHDQV 7KH YDOXHV ZHUH WKHUHIRUH ORJDULWKPLFDOO\ WUDQVIRUPHG 6WDWLVWLFDO DQDO\VHV ZHUH SHUIRUPHG RQ WKH WUDQVIRUPHG GDWD PHDQV DUH UHSRUWHG IRU XQWUDQVIRUPHG GDWD 6WDWLVWLFDO DQDO\VHV LQYROYHG DSSOLFDWLRQ RI WKH JHQHUDO OLQHDU PRGHO )UHXQG DQG /LWWHOO f IROORZHG E\ 'XQFDQnV 0XOWLSOH 5DQJH WHVW ZKHQ VLJQLILFDQW GLIIHUHQFHV ZHUH GHWHFWHG 5HVXOWV $ FRQVLVWHQW RXWFRPH RI WKH LUULJDWLRQ )LJ f IHUWLOL]HU )LJ f LQVHFWLFLGH )LJ f DQG FOLPDWH )LJ f H[SHULPHQWV ZDV KLJKHU VHHGOLQJ UHFUXLWPHQW DV PHDVXUHG E\ GHQVLW\f RQ WKH QRQZHHGHG SORWV &RQYHUVHO\ VHHGOLQJ JURZWK DV PHDVXUHG E\ OHDI DUHD SHU SDOPf ZDV FRQVLVWHQWO\ KLJKHU RQ ZHHGHG SORWV +RZHYHU WKH VHFRQG HIIHFW FRQVLVWHQWO\ IDLOHG WR FRPSHQVDWH IRU WKH ILUVW HIIHFW DQG EDEDVVXnV RYHUDOO /$, ZDV FRQVHTXHQWO\ KLJKHU RQ WKH QRQZHHGHG SORWV LQ DOO H[SHULPHQWV

PAGE 109

m Rf R )LJXUH 3DOPV SHU P Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX RYHU WLPH LQ LUULJDWLRQ H[SHULPHQW %DUV UHSUHVHQW s VWDQGDUG HUURU

PAGE 110

)LJXUH 3DOPV SHU Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX RYHU WLPH LQ IHUWLOL]HU H[SHULPHQW %DUV UHSUHVHQW VWDQGDUG HUURU

PAGE 111

3DOPV SHU Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX RYHU WLPH LQ LQVHFWLFLGH H[SHULPHQW %DUV UHSUHVHQW s VWDQGDUG HUURU )LJXUH

PAGE 112

)LJXUH 3DOPV SHU QUn Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX RYHU WLPH LQ FOLPDWH H[SHULPHQW %DUV UHSUHVHQW s VWDQGDUG HUURU

PAGE 113

,UULJDWLRQ H[SHULPHQW ,Q DGGLWLRQ WR WKH SUHVHQFH RI RWKHU SODQWV KHQFHIRUWK UHIHUUHG WR DV FRPSHWLWRUVf LUULJDWLRQ KDG D SURQRXQFHG HIIHFW RQ JHUPLQDWLRQ LQ EDEDVVX DV LQGLFDWHG E\ WKH KLJK GHQVLW\ RI SDOPV RQ LUULJDWHG SORWV 7DEOH Df :HHGHG SORWV WKDW ZHUH QRW LUULJDWHG FRQWDLQHG D VLJQLILn FDQWO\ ORZHU S BAf SDOP GHQVLW\ WKDQ DQ\ RWKHU WUHDWPHQW FRPELQDWLRQ %XW LUULJDWLRQ ZDV D OHVV HIIHFWLYH PHDQV RI SURPRWLQJ JHUPLQDWLRQ WKDQ FRPSHWLWRU SUHVHQFH DV LQGLFDWHG E\ WKH VLJQLILFDQWO\ ORZHU Ssf GHQVLW\ RI SDOPV RQ ZHHGHG LUULJDWHG SORWV WKDQ HLWKHU WUHDWPHQW FRPELn QDWLRQ RQ QRQZHHGHG SORWV ,UULJDWLRQ GLG QRW LQIOXHQFH JURZWK 7DEOH Ef DQG EDEDVVXnV VXEVWDQWLDOO\ KLJKHU /$, 7DEOH Ff RQ LUULJDWHG SORWV LV FRQVHTXHQWO\ GXH WR LQFUHDVHG GHQVLW\ RI WKH SDOP XQGHU LUULJDn WLRQ (IIHFWV RI LUULJDWLRQ RQ EDEDVVXnV FRPSHWLWRUV DSSHDU WR EH PLQLPDO &RPSHWLWRU /$, ZDV VXEVWDQWLDOO\ KLJKHU RQ QRQLUULJDWHG SORWV 7DEOH f +RZHYHU GU\ PDVV RI FRPSHWLWRU OHDYHV ZDV PDUJLQDOO\ JUHDWHU RQ LUULJDWHG SORWV 7DEOH f 7KH KLJKHU RYHUDOO PDVV RI FRPSHWLWRUV RQ QRQLUULJDWHG SORWV LV DOPRVW HQWLUHO\ DWWULEXWDEOH WR VWDQGLQJ GHDG )HUWLOL]HU H[SHULPHQW $SSOLFDWLRQ RI VXEVWDQWLDO OHYHOV RI 1 3 DQG KDG QR HIIHFW RQ WKH SHUIRUPDQFH RI HLWKHU EDEDVVX RU LWV FRPSHWLWRUV 7DEOHV DQG f 1R GLIIHUHQFHV LQ WKH UHODWLYHO\ PRELOH 1 DQG ZHUH GHWHFWHG EHWZHHQ WUHDWPHQWV FD PR DIWHU WKH ILQDO DSSOLFDWLRQ RI IHUWLOL]HUV

PAGE 114

7DEOH f RQO\ WKH UHODWLYHO\ LPPRELOH 3 LQ WKH IRUP RI 32f ZDV VLJQLILFDQWO\ KLJKHU S Bf RQ WKH IHUWLOL]HG SORWV ,QVHFWLFLGH H[SHULPHQW $SSOLFDWLRQ RI LQVHFWLFLGHV SURGXFHG QR HIIHFW RQ GHQVLW\ 7DEOH Df D VLJQLILFDQW UHGXFWLRQ LQ EDEDVVXnV JURZWK S eB 7DEOH Ef DQG D VXEVWDQWLDO UHGXFWLRQ LQ LWV /$, S 7DEOH Ff $V WKH ODWWHU WZR HIIHFWV ZHUH FRQVLVWHQW RQ ZHHGHG DV ZHOO DV QRQZHHGHG SORWV FRQFOXGH WKDW RQH RU ERWK RI WKH LQVHFWLFLGHV XVHG LQ WKH H[SHULPHQW PD\ KDYH D GHELOLWDWLQJ HIIHFW RQ EDEDVVXnV JURZWK :LWK FRPSHWLWRUV SUHVHQW UHODWLYH /$, 7DEOH f DQG UHODWLYH ELRPDVV 7DEOH f RI EDEDVVX ZHUH VXEVWDQWLDOO\ KLJKHU ZKHQ LQVHFWLn FLGHV ZHUH QRW DSSOLHG 7KHVH UHVXOWV DSSHDU WR EH GXH WR DQ DSSDUHQWO\ QHJDWLYH HIIHFW RI WKH LQVHFWLFLGHV RQ EDEDVVX FRPELQHG ZLWK WKHLU DSSDUHQWO\ SRVLWLYH HIIHFW RQ WKH YHJHWDWLRQ DV D ZKROH $OO FRPSRQHQWV RI FRPSHWLWRU GU\ PDVV ZHUH KLJKHU 7DEOH ff§DQG /$, RI FRPSHWLWRUV RWKHU WKDQ JUDVVHV VLJQLILFDQWO\ KLJKHU S e 7DEOH ff§ZKHQ LQVHFWLFLGHV ZHUH DSSOLHG +RZHYHU SORWV ZLWKRXW LQVHFWLFLGHV ZHUH FKDUDFWHUL]HG E\ D VLJQLILFDQWO\ KLJKHU /$, RI YLQHV S e 7DEOH f $SSOLFDWLRQ RI LQVHFWLFLGHV WKXV UHVXOWHG LQ UHGXFHG GRPLQDQFH E\ YLQHV KLJKHU RYHUDOO FRPSHWLWRU YLJRU DQG DEVROXWH DQG UHODWLYH GHFUHDVHV LQ EDEDVVXnV SHUIRUPDQFH &OLPDWH H[SHULPHQW ,Q DGGLWLRQ WR WKH SUHVHQFH RI FRPSHWLWRUV FOLPDWH KDG D SURn QRXQFHG HIIHFW RQ JHUPLQDWLRQ LQ EDEDVVX DV LQGLFDWHG E\ WKH KLJK GHQVLW\ RI SDOPV RQ ZHW FOLPDWH SORWV 7DEOHV D DQG Df :HHGHG

PAGE 115

2 7DEOH 3DOPV SHU P Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI LUULJDWLRQ H[SHULPHQW $VWHULVNV LQGLFDWH VLJQLILFDQW GLIIHUHQFHV r IRU S e f EHWZHHQ PDLQ HIIHFW PHDQV :KHUH WKHUH LV LQWHUDFWLRQ EHWZHHQ PDLQ HIIHFWV VXSHUn VFULSWV DUH XVHG WR FRPSDUH WUHDWPHQW PHDQV 9DOXHV WKDW GR QRW VKDUH WKH VDPH VXSHUVFULSW Df GLIIHU VLJQLILFDQWO\ S e f 3DOPV 'HU P 1R ,UULJDWLRQ ,UULJDWLRQ 0HDQ 1RW :HHGHG D D :HHGHG E OE 0HDQ Ef /HDI $UHD 'HU 3DOP FPf 1R ,UULJDWLRQ ,UULJDWLRQ 0HDQ 1RW :HHGHG e :HHGHG 0HDQ Ff /HDI $UHD ,QGH[ 1R ,UULJDWLRQ ,UULJDWLRQ 0HDQ 1RW :HHGHG r :HHGHG 0HDQ

PAGE 116

7DEOH 3DOPV SHU P Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI IHUWLOL]HU H[SHULPHQW $VWHULVNV LQGLFDWH VLJQLILFDQW GLIIHUHQFHV rr IRU S sf EHWZHHQ PDLQ HIIHFW PHDQV Df S 3DOPV GHU QU 1R )HUWLOL]HUV )HUWLOL]HUV 0HDQ 1RW :HHGHG rf :HHGHG 0HDQ Ef /HDI $UHD SHU 3DOP FP f 1R )HUWLOL]HUV )HUWLOL]HUV 0HDQ 1RW :HHGHG }r :HHGHG 0HDQ Ff /HDI $UHD ,QGH[ 1R )HUWLOL]HUV )HUWLOL]HUV 0HDQ 1RW :HHGHG :HHGHG 0HDQ

PAGE 117

7DEOH 3DOPV SHU P Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI LQVHFWLFLGH H[SHULPHQW $VWHULVNV LQGLFDWH Df VLJQLILFDQW GLIIHUHQFHV rr IRU S s EHWZHHQ PDLQ HIIHFW PHDQV 3DOPV 'HU P f 1R ,QVHFWLFLGHV ,QVHFWLFLGHV 0HDQ 1RW :HHGHG rr :HHGHG 0HDQ Ef /HDI $UHD 'HU 3DOP FPf 1R ,QVHFWLFLGHV ,QVHFWLFLGHV 0HDQ 1RW :HHGHG }} :HHGHG 0HDQ rr Ff /HDI $UHD ,QGH[ 1R ,QVHFWLFLGHV ,QVHFWLFLGHV 0HDQ 1RW :HHGHG ‘ U RQ f }r :HHGHG 0HDQ

PAGE 118

7DEOH 3DOPV SHU P Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI FOLPDWH H[SHULPHQW ZLWK FOLPDWH DQG ZHHGLQJ DV PDLQ HIIHFWV $VWHULVNV LQGLFDWH VLJQLILFDQW GLIIHUHQFHV rr IRU S ef EHWZHHQ PDLQ HIIHFW PHDQV :KHUH WKHUH LV LQWHUDFWLRQ EHWZHHQ PDLQ HIIHFWV VXSHUVFULSWV DUH XVHG WR FRPSDUH WUHDWPHQW PHDQV 9DOXHV Df WKDW GR QRW VKDUH WKH VDPH VXSHUVFULSW GLIIHU VLJQLILFDQWO\ S ef S 3DOPV 'HU QU 'UY &OLPDWH :HW &OLPDWH 0HDQ 1RW :HHGHG D D :HHGHG r E 0HDQ Ef /HDI $UHD SHU 3DOP FPf 'UY &OLPDWH :HW &OLPDWH 0HDQ 1RW :HHGHG E E :HHGHG D E 0HDQ Ff /HDI $UHD ,QGH[ 'UY &OLPDWH :HW &OLPDWH 0HDQ 1RW :HHGHG r} :HHGHG 0HDQ rr

PAGE 119

7DEOH 3DOPV SHU P Df OHDI DUHD SHU SDOP Ef DQG OHDI DUHD LQGH[ Ff RI EDEDVVX DW HQG RI FOLPDWH H[SHULPHQW ZLWK FOLPDWH DQG HFRW\SH DV PDLQ HIIHFWV $VWHULVNV LQGLFDWH VLJQLILFDQW GLIIHUHQFHV rr IRU S e f EHWZHHQ PDLQ HIIHFW PHDQV :KHUH WKHUH LV LQWHUDFWLRQ EHWZHHQ PDLQ HIIHFWV VXSHUVFULSWV DUH XVHG WR FRPSDUH WUHDWPHQW PHDQV 9DOXHV WKDW GR QRW VKDUH WKH VDPH VXSHUVFULSW VLJQLILFDQWO\ S ef GLIIHU Df 2 3DOPV 'HU P 'UY &OLPDWH :HW &OLPDWH 0HDQ 'U\ (FRW\SH :HW (FRW\SH 0HDQ rr Ef /HDI $UHD 'HU 3DOP FPf 'UY &OLPDWH :HW &OLPDWH 0HDQ 'U\ (FRW\SH D r :HW (FRW\SH EnF DE 0HDQ Ff /HDI $UHD ,QGH[ 'UY &OLPDWH :HW &OLPDWH 0HDQ 'U\ (FRW\SH E E RQ &?:HW (FRW\SH r D 0HDQ

PAGE 120

7DEOH /HDI DUHD LQGH[ /$,f RI EDEDVVX DQG FRPSHWLWRUV LQ IRXU H[SHULPHQWV 0HDVXUHPHQWV ZHUH REWDLQHG GXULQJ 0D\ LQ FD PRROG YHJHWDWLRQ 9DOXHV DUH H[SUHVVHG DV /$, PHDQ SHUFHQW RI WRWDO /$, LQ SDUHQWKHVHV 6LJQLILFDQW GLIIHUHQFHV DUH VKRZQ DV r S I f RU S f ,UULJDWLRQ )HUWLOL]HUV ,QVHFWLFLGHV &OLPDWH 1R
PAGE 121

7DEOH 2YHQ GULHG PDVV RI EDEDVVX DQG FRPSHWLWRUV LQ IRXU H[SHULPHQWV 9HJHWDWLRQ ZDV FD PR ROG DW WLPH RI KDUYHVW 0D\ f 9DOXHV DUH J SHU PA PHDQ SHUFHQW RI WRWDO GU\ PDVV LQ SDUHQWKHVHV 6LJQLILFDQW GLIIHUHQFHV DUH VKRZQ DV r S ef ,UULJDWLRQ )HUWLOL]HUV ,QVHFWLFLGHV &OLPDWH 1R
PAGE 122

7DEOH 6RLO FKHPLVWU\ RI IRXU WUHDWPHQWV DW HQG RI QXWULHQW H[SHULPHQW 9DOXHV DUH PHDQ VWDQGDUG GHYLDWLRQVf RI ILYH VDPSOHV IURP FP HDFK D FRPSRVLWH RI FRUHV )(57,/,=(' 127 )(57,/,=(' :HHGHG 1RQZHHGHG :HHGHG 1RQZHHGHG & bf f f f f 2UJDQLF PDWWHU bf f f f f 1 bf f f f f &1 3 SSPf f f f f S+ LQ +2 S+ LQ .& &D 0J PHTO22Jf f f f f 1D PHTO22Jf f f f f PHTJf f f f f + PHTJf f f f f $ PHJJf f f f f 7RWDO %DVHV PHTJf f f f f %DVH 6DWXUDWLRQ bf f f f f

PAGE 123

7DEOH 6RLO FKHPLVWU\ RI H[SHULPHQWDO VLWHV DW /DJR 9HUGH DQG %HOHP 6DPSOHV ZHUH REWDLQHG DIWHU FOHDULQJ DQG EHIRUH EXUQLQJ WKH IDOORZ 9DOXHV DUH PHDQV VWDQGDUG GHYLDWLRQVf RI IRXU VDPSOHV IURP FP HDFK D FRPSRVLWH RI FRUHV /DJR 9HUGH 'UY &OLPDWHf %HOHP :HW &OLPDWHf & bf f f 2UJDQLF 0DWWHU bf f f 1 bf f f &1 3 SSPf f f S+ &D 0J PHT Jf f f 1D PHT Jf f f PHT Jf f f + PHT Jf f f $ PHT Jf f f 7RWDO %DVHV PHT Jf f f %DVH 6DWXUDWLRQ f f f

PAGE 124

,OO SORWV XQGHU WKH GU\ FOLPDWH FRQWDLQHG D VLJQLILFDQWO\ ORZHU SDOP GHQVLW\ WKDQ DQ\ RWKHU WUHDWPHQW FRPELQDWLRQ S e 7DEOH Df /LNH LUULJDWLRQ KRZHYHU WKH ZHW FOLPDWH ZDV OHVV HIIHFWLYH LQ SURPRWLQJ JHUPLQDWLRQ WKDQ FRPSHWLWRU SUHVHQFH DV LQGLFDWHG E\ WKH VLJQLILFDQWO\ ORZHU S ef GHQVLW\ RI SDOPV RQ ZHHGHG ZHW FOLPDWH SORWV WKDQ RQ QRQZHHGHG SORWV XQGHU HLWKHU WKH ZHW RU GU\ FOLPDWH 5HVSRQVHV LQ LQGLYLGXDO JURZWK ZHUH FRPSOH[ /HDI DUHD SHU SDOP ZDV VLJQLILFDQWO\ KLJKHU RQ ZHHGHG GU\ FOLPDWH SORWV WKDQ XQGHU DQ\ RWKHU WUHDWPHQW FRPELQDWLRQf§LQFOXGLQJ ZHHGHG ZHW FOLPDWH SORWV S en 7DEOH Ef 7KLV VXJJHVWV WKDW FRQGLWLRQV XQGHU WKH GU\ FOLPDWH ZHUH PRUH IDYRUDEOH IRU EDEDVVXnV JURZWK SRVVLEO\ GXH WR KLJKHU VRLO IHUWLOLW\ 7DEOH f +RZHYHU WKHUH ZHUH DOVR VLJQLILFDQW GLIIHUHQFHV LQ HFRW\SLF UHVSRQVHV 7DEOH Ef 8QGHU ERWK FOLPDWHV LQGLYLGXDO OHDI DUHD ZDV VLJQLILFDQWO\ KLJKHU S e f LQ SDOPV SURGXFHG IURP ORFDO IUXLWV %XW WKH GDWD RQ OHDI DUHD SHU SDOP LQGLFDWH WKDW GU\ HFRW\SH SDOPV SHUIRUPHG OHVV ZHOO LQ WKH ZHW FOLPDWH WK£Q ZHW HFRW\SH SDOPV LQ WKH GU\ FOLPDWH 7KLV VXJJHVWV WKDW VRPH FRPSHQVDWRU\ IDFWRU VXFK DV VRLO IHUWLOLW\ 7DEOH f ZDV RSHUDWLQJ DW WKH GU\ FOLPDWH VLWH /HDI DUHD LQGH[ ZDV VLJQLILFDQWO\ KLJKHU DW WKH ZHW FOLPDWH VLWH S e 7DEOH Ff ZKLFK LV DWWULEXWDEOH WR LQFUHDVHG JHUPLQDWLRQ XQGHU WKH ZHW FOLPDWH ,Q ERWK FOLPDWHV /$, ZDV VLJQLILFDQWO\ KLJKHU LQ SDOPV SURGXFHG IURP ORFDO IUXLWV S e 7DEOH Ff :LWK FRPSHWLWRUV SUHVHQW UHODWLYH /$, 7DEOH f DQG ELRPDVV 7DEOH f RI EDEDVVX ZHUH VLJQLILFDQWO\ KLJKHU XQGHU WKH ZHW FOLPDWH

PAGE 125

S e DQG S e UHVSHFWLYHO\f 7KHVH UHVSRQVHV DUH ODUJHO\ DWWULEXWDEOH WR GHFUHDVHG YLJRU RI FRPSHWLWRUV 8QGHU WKH GU\ FOLPDWH VWHP DQG WRWDO ELRPDVV RI FRPSHWLWRUV ZHUH VXEVWDQWLDOO\ KLJKHU 7DEOH ff§DQG FRPSHWLWRU /$, VLJQLILFDQWO\ KLJKHU S e 7DEOH f f§ WKDQ XQGHU WKH ZHW FOLPDWH 7KH ODWWHU VLWH KRZHYHU ZDV FKDUDFWHUL]HG E\ D VLJQLILFDQWO\ KLJKHU /$, RI JUDVVHV S e 7DEOH f /RZHU VRLO IHUWLOLW\ 7DEOH f DW WKH ZHW FOLPDWH VLWH PD\ DFFRXQW IRU WKH ORZHU RYHUDOO FRPSHWLWLYH YLJRU DQG D VKLIW WR JUHDWHU GRPLQDQFH E\ JUDVVHV ZLWK FRQFRPLWDQW LQFUHDVHV LQ EDEDVVXnV UHODWLYH SHUIRUPDQFH 'LVFXVVLRQ 6HHGOLQJ *URZWK ,Q WKH H[SHULPHQWV GHVFULEHG DERYH EDEDVVXnV SHUIRUPDQFH ZDV VHSDUDWHG LQWR WZR FRPSRQHQWV JURZWK OHDI DUHD SHU SDOPf DQG GHQVLW\ SDOPV SHU Pf $V LQ RWKHU SDOPV HJ 5HHV f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

PAGE 126

JURZWK RI HDUO\ VXFFHVVLRQDO YHJHWDWLRQ ZDV OLNHZLVH QRW HQKDQFHG E\ DSSOLFDWLRQ RI IHUWLOL]HUV +DUFRPEH f SUHGLFW WKDW JURZWK RI EDEDVVX RQ FRPSHWLWLRQIUHH SORWVf DQG LWV FRPSHWLWRUV RQ QRQZHHGHG SORWVf ZRXOG EH HQKDQFHG E\ DSSOLFDWLRQ RI IHUWLOL]HUV 7KH UHVXOWV RI WKH FOLPDWH H[SHULPHQW SURYLGH LQGLUHFW VXSSRUW IRU WKLV SUHGLFWLRQ *URZWK RI EDEDVVX RQ FRPSHWLWLRQIUHH SORWVf DQG LWV FRPSHWLWRUV RQ QRQZHHGHG SORWVf ZDV VLJQLILFDQWO\ HQKDQFHG XQGHU ZKDW ZHUH SUHVXPHG WR EH OHVV IDYRUDEOH FOLPDWLF FRQGLWLRQV DW WKH GU\ VLWH /DJR 9HUGHf 7KH UHODWLYHO\ KLJK VRLO IHUWLOLW\ DW WKLV VLWH FRPSDUHG WR WKH ZHW VLWH %HOHPf VHHPV WR EH WKH PRVW OLNHO\ FDXVH RI WKLV UHVSRQVH 7DEOH f 6HHGOLQJ 5HFUXLWPHQW $V PRUWDOLW\ RI EDEDVVX ZDV H[WUHPHO\ ORZ LQ WKHVH H[SHULPHQWV VKDOO FRQVLGHU GHQVLW\ WR EH HVVHQWLDOO\ V\QRQ\PRXV ZLWK VHHGOLQJ UHFUXLWPHQW 7KH ODWWHU WHUP LV PRUH XVHIXO LQ WKH GLVFXVVLRQ WKDW IROORZV ,Q WKH H[SHULPHQWV GHVFULEHG DERYH UHVSRQVHV LQ UHFUXLWPHQW ZHUH PXFK PRUH SURQRXQFHG WKDQ UHVSRQVHV LQ JURZWK 5HFUXLWPHQW ZDV FRQVLVWn HQWO\ HQKDQFHG RQ SORWV WKDW ZHUH QRW ZHHGHG LUULJDWHG RU ORFDWHG LQ WKH ZHW FOLPDWH 7KH UHVSRQVH ZDV GXH WR LQFUHDVHG JHUPLQDWLRQ ZKLFK SUREDEO\ UHVXOWHG IURP KLJKHU VRLO PRLVWXUH OHYHOV RQ WKHVH SORWV &RPSHWLWRU SUHVHQFH KDG D FRQVLVWHQWO\ JUHDWHU HIIHFW RQ UHFUXLWPHQW WKDQ HLWKHU LUULJDWLRQ RU WKH ZHW FOLPDWH /LNHZLVH UHFUXLWPHQW FRQWULEXWHG PRUH VXEVWDQWLDOO\ WR EDEDVVXnV RYHUDOO SHUIRUPDQFH DV PHDVXUHG E\ /$,f WKDQ GLG JURZWK 7KXV FRPSHWLWRU SUHVHQFH ZDV WKH NH\ IDFWRU JRYHUQLQJ EDEDVVXnV SHUIRUPDQFH GXULQJ WKH PR H[SHULPHQWV

PAGE 127

7KLV LV D FOHDU FDVH RI IDFLOLWDWLRQ LQ ZKLFK WKH SUHVHQFH RI DQ HDUOLHU SODQW FRPPXQLW\ SURPRWHV WKH UHFUXLWPHQW RI D VXEVHTXHQW VSHFLHV 'HVSLWH XQUHDOLVWLFDOO\ KLJK GHQVLWLHV RI SODQWHG IUXLWV S OPAf VHHGOLQJ UHFUXLWPHQW RI EDEDVVX ZDV H[WUHPHO\ ORZ RQ ZHHGHG SORWV 7KLV VXJJHVWV WKDW XQGHU QDWXUDO FRQGLWLRQV LQLWLDO UHFUXLWPHQW RI EDEDVVX UHTXLUHV WKH SUHVHQFH RI SUHYLRXVO\ HVWDEOLVKHG FRPSHWLWRUV )DFLOLWDWLRQ RI UHFUXLWPHQW GRHV QRW DOZD\V WUDQVODWH LQWR QHW IDFLOLWDWLRQ EHFDXVH LQKLELWLRQ ODWHU LQ WKH OLIH F\FOH PD\ EH VWURQJHU WKDQ WKH LQLWLDO IDFLOLWDWLRQ 7XUQHU f 7KH H[SHULPHQWV GHVFULEHG LQ WKLV FKDSWHU VKRZ WKDW ZKHUHDV FRPSHWLWRUV IDFLOLWDWH UHFUXLWPHQW LQ EDEDVVX WKH\ DOVR LQKLELW VXEVHTXHQW JURZWK ,Q WKH IROORZLQJ FKDSWHU VKDOO H[DPLQH JURZWK RI EDEDVVX XQGHU FRQGLWLRQV RI KLJK FRPSHWLWRU DEXQGDQFH

PAGE 128

&+$37(5 *52:7+ $1' 352'8&7,9,7< 7KLV FKDSWHU FRPSOHWHV WKH OLIH F\FOH RI EDEDVVX ZLWK DQ DQDO\VLV RI WKH SDOPnV JURZWK DQG SURGXFWLYLW\ *URZWK UHIHUV WR WKH DQQXDO LQFUHPHQW RI D PRUSKRORJLFDO SDUDPHWHU VWHP KHLJKW RU OHDIOHWV SHU OHDIf LQ LQGLYLGXDO SDOPV 3URGXFWLYLW\ LV KHUH GHILQHG DV WKH DERYHn JURXQG PDVV SURGXFHG SHU XQLW WLPH HLWKHU E\ D SRSXODWLRQ RU E\ LQGLn YLGXDOV %RWK JURZWK DQG SURGXFWLYLW\ DUH UHODWLYHO\ HDV\ WR PHDVXUH LQ SDOPV $V LQ RWKHU PRQRFRW\OHGRQV VHFRQGDU\ JURZWK GRHV QRW RFFXU LQ SDOPV GXH WR DQ DEVHQFH RI D ODWHUDO FDPELXP DQG SULPDU\ JURZWK DERYH WKH JURXQG LV JHQHUDOO\ OLPLWHG WR D VLQJOH VKRRW DSH[ )RU SUDFWLFDO SXUSRVHV JURZWK RI WKH VWHP FDQ WKXV EH PHDVXUHG VLPSO\ DV FKDQJHV LQ KHLJKW RYHU WLPH 7RPOLQVRQ f /LNHZLVH EHFDXVH OHDYHV DQG UHSURn GXFWLYH VWUXFWXUHV WHQG WR EH IHZ DQG ODUJH GLUHFW PHDVXUHV RI SURGXFn WLYLW\ DUH UHODWLYHO\ HDV\ LQ SDOPV 7KH\ DUH WKHUHIRUH LGHDO VXEMHFWV IRU VWXGLHV RI ERWK JURZWK DQG SURGXFWLYLW\ ,Q WKLV FKDSWHU H[DPLQH JURZWK DQG SURGXFWLYLW\ LQ LQGLYLGXDO SDOPV RI GLIIHUHQW DJHV DV ZHOO DV SURGXFWLYLW\ LQ SRSXODWLRQV JURZLQJ XQGHU GLIIHULQJ HQYLURQPHQWDO FRQGLWLRQV 7KH REMHFWLYH LV WR GHWHUPLQH KRZ UHVRXUFHV DUH DOORFDWHG DPRQJ WKH YDULRXV FRPSRQHQWV RI WKH SDOP RU SRSXODWLRQ WKH JRDO LV D GHHSHU XQGHUVWDQGLQJ RI KRZ JURZWK DQG UHSURn GXFWLRQ DUH FRQWUROOHG LQ EDEDVVX

PAGE 129

0HWKRGV *URZWK 7R PHDVXUH JURZWK LQ EDEDVVX GHILQHG OLIH VWDJHV EDVHG RQ GLVWLQFWLYH PRUSKRORJLFDO FKDUDFWHUV 7DEOH f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bf RI WKHVH SDOPV H[KLELWHG QHJDWLYH RU QHJOLJLEOH JURZWK UDWHV ,I DOO LQGLYLGXDOV PHDVXUHG ZHUH XVHG WR GHULYH PHDQ JURZWK UDWHV FDOFXODWH WKDW WKH DYHUDJH WLPH IRU D SDOP WR SDVV WKURXJK VWDJHV DQG ZRXOG EH \U LQ WKH VHFRQGDU\ IRUHVW DQG \U LQ WKH SULPDU\ IRUHVW $V WKHVH QXPEHUV DUH FOHDUO\ XQWHQDEOH DUELWUDULO\ DVVXPHG WKDW LQGLYLGXDOV H[KLELWLQJ QHJDWLYH RU ]HUR JURZWK

PAGE 130

7DEOH 'HILQLWLRQ RI 67$*( OLIH VWDJHV RI EDEDVVX XVHG LQ VWXG\ %(*,11,1* 'HVFUL'WLRQ 1XPEHU 6HHGV $EVFLVVLRQ RI IUXLW 6HHGOLQJV $SSHDUDQFH RI ILUVW OHDI IROORZLQJ JHUPLQDWLRQ 3UHHVWDEOLVKHG MXYHQLOHV ,QLWLDO GLYLVLRQ RI OHDI EODGH (WDEOLVKHG MXYHQLOHV /HDIOHWV SHU VLGH +HLJKW FP +HLJKW FP 0DWXUH SDOPV r +HLJKW B! FP +HLJKW FP +HLJKW FP +HLJKW FP +HLJKW FP +HLJKW FP f6WDJH LQFOXGHV SDOPV WKDW KDG QRW IORZHUHG DQG WKXV ZHUH QRW UHSURGXFWLYHO\ PDWXUH

PAGE 131

UDWHV ZHUH GRRPHG DQG GLG QRW XVH WKHP WR FDOFXODWH PHDQ JURZWK UDWHV RU UHVLGHQFH WLPHV ,QGHSHQGHQW FRUUHODWLRQ RI DJH HVWLPDWHV ZLWK D KLVn WRULFDOO\ NQRZQ HYHQW VHH &KDSWHU f LQGLFDWHV WKDW WKLV DVVXPSWLRQ LV SUREDEO\ UHDVRQDEOH *URZWK UDWHV LQ VWDJHV ZHUH FDOFXODWHG DV FKDQJH LQ KHLJKW SHU \HDU +HLJKW RI WKH SDOP ZDV GHILQHG DV WKH GLVWDQFH IURP WKH JURXQG WR WKH VKHDWK DSH[ RI WKH QHZHVW H[SDQGHG OHDI WKLV ZDV LQLWLDOO\ PHDVXUHG WR QHDUHVW FP HLWKHU GLUHFWO\ VWDJHV f RU ZLWK DQ LQFOLQRPHWHU VWDJHV f $IWHU \U WKH KHLJKW RI SDOPV LQ VWDJHV ZDV UHPHDVXUHG GLUHFWO\ )RU WDOOHU SDOPV PHDQ LQFUHPHQW SHU OHDI ZDV FDOFXODWHG DV WKH OHQJWK RI VWHP FRYHUHG ZLWK OHDI EDVHV GLYLGHG E\ WKH WRWDO QXPEHU RI OHDI EDVHV *URZWK UDWHV ZHUH WKHQ FDOFXODWHG DV WKH SURGXFW RI WKH PHDQ LQFUHPHQW SHU OHDI PXOWLSOLHG E\ WKH QXPEHU RI OHDYHV SURGXFHG SHU \HDU 3URGXFWLYLW\ /DFN RI EUDQFKLQJ DQG DERYHJURXQG VWHP WKLFNHQLQJ JUHDWO\ IDFLOLn WDWHG WKH PHDVXUHPHQW RI SURGXFWLYLW\ LQ EDEDVVX :KHUHDV PRVW VWXGLHV HVWLPDWH DQQXDO SURGXFWLRQ RI OHDYHV DQG UHSURGXFWLYH VWUXFWXUHV LQGLn UHFWO\ WKURXJK VDPSOLQJ RI OLWWHU IDOO LQ EDEDVVX LW ZDV SRVVLEOH WR PHDVXUH WKHVH FRPSRQHQWV GLUHFWO\ 0\ PHDVXUHPHQWV FRQVHTXHQWO\ DYRLGHG SRWHQWLDO XQGHUHVWLPDWHV GXH WR UHDEVRUSWLRQ GHFRPSRVLWLRQ DQG OHDFKn LQJ SULRU WR DEVFLVVLRQ /RVVHV GXH WR KHUELYRU\ ZHUH QRW PHDVXUHG EXW ZHUH SUREDEO\ QHJOLJLEOH 3RSXODWLRQ SURGXFWLYLW\ PHDVXUHG SURGXFWLRQ RI DERYHJURXQG FRPSRQHQWV LQ DOO OLIH VWDJHV RI EDEDVVX SUHVHQW RQ WKH SULPDU\ DQG

PAGE 132

VHFRQGDU\ IRUHVW VLWHV RYHU D \U SHULRG $SULO 0DUFK f 'XH WR JUD]LQJ FXWWLQJ DQG EXUQLQJ SURGXFWLYLW\ PHDVXUHPHQWV LQ WKH SDVWXUH ZHUH OLPLWHG WR D KD VDPSOH RI PDWXUH SDOPV 7R DFFRXQW IRU SRSXODWLRQ WXUQRYHU GXULQJ WKH \U SHULRG REn WDLQHG GDWD RQ SRSXODWLRQZLGH PRUWDOLW\ DQG UHFUXLWPHQW GXULQJ FHQVXVHV FDUULHG RXW DW PR LQWHUYDOV DW HDFK VLWH VHH &KDSWHU f 3URGXFWLYLn W\ RI SDOPV WKDW GLHG GXULQJ WKH ILUVW LQWHUYDO ZDV QRW FDOFXODWHG SURGXFWLYLW\ RI SDOPV WKDW GLHG GXULQJ WKH VHFRQG LQWHUYDO ZDV FDOFXn ODWHG IRU WKH ILUVW LQWHUYDO ,Q WKH FDVH RI QHZO\ UHFUXLWHG VHHGOLQJV SURGXFWLYLW\ ZDV FDOFXODWHG DV HVWLPDWHG JURZWK E\ WKH ILQDO FHQVXV $QQXDO KHLJKW LQFUHPHQWV RI SDOP VWHPV ZHUH PHDVXUHG LQ WKH DERYH GHILQHG VDPSOLQJ DUHDV RI HDFK SRSXODWLRQ 7R REWDLQ D UHJUHVVLRQ EHWZHHQ VWHP ELRPDVV DQG KHLJKW UDQGRPO\ VHOHFWHG VHYHQ VWHPV IRU IHOOLQJ DQG PHDVXUHPHQW RI GLPHQVLRQV WKUHH VXEVDPSOHV SHU VWHP FD FP KLJK FURVV VHFWLRQV IURP WKH EDVDO PLGGOH DQG DSLFDO SRUWLRQV RI HDFK VWHPf ZHUH RYHQGULHG DW & DQG ZHLJKHG ,Q SDOPV P WDOO WKH VWHP LH DSLFDO PHULVWHPf KDG QRW HPHUJHG IURP XQGHUJURXQG 7KH UHJUHVVLRQ REWDLQHG IRU SDOPV 0 P WDOO ZDV R VWHP ELRPDVV NJf VWHP KHLJKW Pf U 8VLQJ KHLJKW LQFUHPHQW GDWD DQG WKLV UHJUHVVLRQ DQQXDO VWHP SURGXFWLRQ ZDV FDOFXODWHG IRU SDOPV LQ WKH VDPSOHG DUHDV DQG FRUUHFWHG WR D SHUKD EDVLV 7KH PHDQ QXPEHU RI OHDYHV SURGXFHG SHU \HDU LQ HDFK OLIH VWDJH ZDV GHWHUPLQHG E\ PRQLWRULQJ OHDI SURGXFWLRQ LQ WKH DERYHGHILQHG DUHDV RI HDFK SRSXODWLRQ 7R REWDLQ GDWD RQ OHDI PDVV RQH QHZO\ H[SDQGHG OHDI

PAGE 133

IURP HDFK RI UDQGRPO\ VHOHFWHG SDOPV ZDV KDUYHVWHG RYHQGULHG DQG ZHLJKHG OHDIOHWV DQG OHDI D[HV VKHDWK SHWLROH DQG UDFKLVf ZHUH ZHLJKHG VHSDUDWHO\ )RU HDFK RI WKH SDOPV PXOWLSOLHG WKH PDVV RI LWV OHDI E\ WKH PHDQ QXPEHU RI OHDYHV SURGXFHG SHU \HDU LQ WKH FRUUHVn SRQGLQJ OLIH VWDJH 7KH HVWLPDWHG OHDI SURGXFWLYLW\ RI HDFK SDOP ZDV WKHQ SORWWHG DJDLQVW LWV DJH HVWLPDWHG VHSDUDWHO\ IRU HDFK VLWHf WR SURGXFH WKH IROORZLQJ UHJUHVVLRQV IRU WKH SULPDU\ DQG VHFRQGDU\ IRUHVWV UHVSHFWLYHO\ ORJWOHDI SURGXFWLRQ NJ\Uf@ >DJH \Uf@ DJH \Uf U ORJ>OHDI SURGXFWLRQ NJ\Uf@ >DJH \Uf@ DJH \Uf U 8VLQJ WKHVH UHJUHVVLRQV DQQXDO OHDI SURGXFWLRQ ZDV FDOFXODWHG IRU SDOPV LQ WKH VDPSOHG DUHD DQG FRUUHFWHG WR D SHUKD EDVLV ,QIORUHVFHQFH DQG IUXLW SURGXFWLRQ LQ PDWXUH SDOPV ZDV PRQLWRUHG RYHU WKH \U SHULRG VHH &KDSWHU f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

PAGE 134

,QGLYLGXDO SURGXFWLYLW\ 3URGXFWLYLW\ RI LQGLYLGXDO SDOPV DV D IXQFWLRQ RI DJH ZDV FDOFXODWHG XVLQJ WKH SULPDU\ IRUHVW SRSXODWLRQ ZKHUH ORQJOLYHG SDOPV ZHUH UHSUHVHQWHG FDOFXODWHG WKH SURGXFWLYLW\ RI HDFK DERYHJURXQG FRPSRQHQW VHSDUDWHO\ )RU OHDYHV XVHG WKH UHJUHVVLRQ GHVFULELQJ OHDI SURGXFWLYLW\ DV D IXQFWLRQ RI DJH SURYLGHG DERYH )RU VWHPV ILUVW FRQYHUWHG WKH PHDQ DQQXDO KHLJKW LQFUHPHQW RI HDFK VWDJH 7DEOH f WR PDVV XVLQJ WKH PDVVKHLJKW UHJUHVVLRQ SURYLGHG DERYH WKHQ SORWWHG VWHP ELRPDVV LQFUHPHQW DJDLQVW WKH PHDQ DJH RI HDFK VWDJH )RU UHSURGXFWLYH FRPSRQHQWV OLNHZLVH SORWWHG HDFK VWDJHnV PHDQ DQQXDO SURGXFWLRQ DJDLQVW LWV PHDQ DJH /LIHWLPH SURGXFWLYLW\ UHIHUV WR DQ LQGLYLGXDO SDOPn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f LQGLFDWH WKDW EDEDVVXnV LQLWLDO JURZWK ZDV H[WUHPHO\ VORZ $FFRUGLQJ WR WKH PHDQ JURZWK UDWH PHDVXUHPHQWV DQ DYHUDJH RI FD \U LQ WKH SULPDU\ IRUHVW DQG FD \U LQ WKH VHFRQGDU\ IRUHVW HODSVHG

PAGE 135

7DEOH (VWLPDWHG JURZWK UDWHV DQG DJHV RI EDEDVVX SDOPV LQ WKH SULPDU\ IRUHVW VLWH DW /DJR 9HUGH 6WDJHV DUH GHILQHG LQ 7DEOH 1 QXPEHU RI SDOPV PHDVXUHG 1 /($)/(76 3(5 6,'( +(,*+7 FPf 7,0( ,1 $*( $7 (1' 5DQJH LQ 6WDJH SHU
PAGE 136

7DEOH (VWLPDWHG JURZWK UDWHV DQG DJHV RI EDEDVVX SDOPV LQ WKH VHFRQGDU\ IRUHVW QXPEHU VLWH DW /DJR 9HUGH RI SDOPV PHDVXUHG 6WDJHV DUH GHILQHG LQ 7DEOH 1 67$*( 1 /($)/(76 3(5 6,'( +(,*+7 FPf 7,0( ,1 $*( $7 (1' 5DQJH LQ 6WDJH 'HU
PAGE 137

EHWZHHQ JHUPLQDWLRQ DQG LQLWLDWLRQ RI YHUWLFDO VWHP JURZWK 6ORZHU JURZWK UDWHV LQ WKH SULPDU\ IRUHVW DUH SUREDEO\ LQGLFDWLYH RI JUHDWHU VXSSUHVVLRQ RI VHHGOLQJV DQG SUHHVWDEOLVKHG MXYHQLOHV LQ WKLV KDELWDW 3HUFHQWDJH RI SDOPV LQ WKHVH VWDJHV H[KLELWLQJ ]HUR RU QHJDWLYH JURZWK UDWHV ZDV KLJKHU LQ WKH SULPDU\ IRUHVW bf WKDQ LQ WKH VHFRQGDU\ IRUHVW bf 7KH KLJK YDULDQFHV LQ HDUO\ OLIH VWDJHV 7DEOHV DQG f VXJJHVW WKDW PHDQ JURZWK UDWHV DUH QRW D JRRG PHDVXUH RI VXFFHVV 3DOPV WKDW XOWLPDWHO\ EHFDPH HVWDEOLVKHG ZHUH SUREDEO\ VLWXDWHG LQ OLJKW JDSV ZKHUH JURZWK UDWHV ZHUH HQKDQFHG 7KLV PD\ KDYH EHHQ HVSHFLDOO\ WUXH LQ WKH SULPDU\ IRUHVW ZKHUH HVWDEOLVKHG SDOPV VWDJHV f DOPRVW LQYDULn DEO\ JUHZ IDVWHU WKDQ LQ WKH VHFRQGDU\ IRUHVW ,Q ERWK KDELWDWV YHUWLn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f RQ WKH SULPDU\ IRUHVW VLWH KDG DQ HVWLPDWHG DJH RI \U 2YHU WKH HQWLUH OLIHWLPH RI WKLV SDOP D ODUJH

PAGE 138

SURSRUWLRQ RI WRWDO DERYHJURXQG SURGXFWLRQ ZDV DOORFDWHG WR UHSURGXFn WLYH VWUXFWXUHV bf DQG OHDYHV bf ZKLOH RQO\ b ZHQW WR SHUPDQHQWO\ DFFUXLQJ ELRPDVV LQ WKH VWHP 7DEOH f 3DUWLWLRQLQJ RI DERYHJURXQG SURGXFWLYLW\ YDULHG FRQVLGHUDEO\ DFFRUGLQJ WR DJH )LJV f /LNHZLVH D FRPSDULVRQ RI PDWXUH SDOPV RQ WKH WKUHH VWXG\ VLWHV DW /DJR 9HUGH 7DEOH f VKRZV WKDW SURGXFWLYLW\ YDULHG FRQVLGHUDEO\ DFFRUGLQJ WR VLWH 2Q WKH IRUHVWHG VLWHV DEVROXWH SURGXFWLYLW\ SHU SDOPf§DV ZHOO DV UHODWLYH DOORFDWLRQ WR WKH SDOPnV YDULRXV FRPSRQHQWVf§ ZHUH DOPRVW LGHQWLFDO +RZHYHU SURGXFWLYLW\ RI DOO FRPSRQHQWV LQn FUHDVHG GUDPDWLFDOO\ LQ WKH SDVWXUH $OORFDWLRQ WR YHJHWDWLYH FRPn SRQHQWV LH VWHP DQG OHDYHVf LQFUHDVHG E\ FD b DQG DOORFDWLRQ WR UHSURGXFWLYH VWUXFWXUHV E\ DOPRVW b $PRQJ WKH HQWLUH SRSXODWLRQV RI EDEDVVX RQ WKH SULPDU\ DQG VHFRQGn DU\ IRUHVW VLWHV SDUWLWLRQLQJ RI SURGXFWLYLW\ WR VWHPV OHDYHV DQG UHSURGXFWLYH VWUXFWXUHV ZDV VWULNLQJO\ VLPLOLDU 7DEOH f 7KH FRQVLVn WHQWO\ KLJK DOORFDWLRQ WR OHDYHV FD bf LV DWWULEXWDEOH WR KLJK GHQVLWLHV RI VHHGOLQJV DQG MXYHQLOHV LQ ZKLFK DERYHJURXQG SURGXFWLYLW\ LV DOORFDWHG H[FOXVLYHO\ RU SULPDULO\ WR OHDYHV ,Q IDFW PRVW RI WKH OHDI SURGXFWLYLW\ LQ WKH WZR SRSXODWLRQV ZDV GXH WR VHHGOLQJV DQG MXYHn QLOHV RI WRWDO VWDQG SURGXFWLYLW\ WKH OHDYHV RI WKHVH SDOPV FRPSULVHG b LQ WKH SULPDU\ IRUHVW DQG b LQ WKH VHFRQGDU\ IRUHVW

PAGE 139

7DEOH 3DUWLWLRQLQJ RI DERYHJURXQG GU\ PDWWHU SURGXFn WLRQ RI EDEDVVX RYHU DQ HVWLPDWHG OLIH VSDQ RI \U MeB B,B /($9(6 /HDIOHWV /HDI $[HV 6XEWRWDO /HDYHV 67(06 5(352'8&7,9( 6758&785(6 0DOH ,QIORUHVFHQFHV )HPDOH ,QIORUHVFHQFH $[HV )HPDOH ,QIORUHVFHQFH %UDFWV 6XEWRWDO $QFLOODU\ 6WUXFWXUHV D )UXLW :DOOV 6HHGV 6XEWRWDO )UXLWV 6XEWRWDO 5HSURGXFWLYH 6WUXFWXUHV 727$/

PAGE 140

7DEOH 6,7( 3ULPDU\ )RUHVW 6HFRQGDU\ )RUHVW $QQXDO DERYH JURXQG GU\ PDWWHU SURGXFWLRQ RI PDWXUH EDEDVVX SDOPV RQ WKUHH VWXG\ VLWHV DW /DJR 9HUGH 352'8&7,9,7< 3(5 3$/0 NJ\Uf '(16,7< 352'8&7,9,7< WrKDA r\Un 6WHP /HDYHV 5HSURGXFWLYH 6WUXFWXUHV 7RWDO SDOPVKDf 3DVWXUH

PAGE 141

7DEOH $QQXDO DERYHJURXQG GU\ PDWWHU SURGXFWLRQ RI DOO EDEDVVX SDOPV RQ WKH SULPDU\ DQG VHFRQGDU\ IRUHVW VLWHV DW /DJR 9HUGH 9DOXHV DUH NJrKDfnn\Un SHUFHQWDJHV LQ SDUHQn WKHVHVf 35,0$5< )25(67 6(&21'$5< )25(67 /($9(6 /HDIOHWV f f /HDI $[HV f f 6XEWRWDO /HDYHV f f 67(06 5(352'8&7,9( 6758&785(6 f f $QFLOODU\ 6WUXFWXUHV f f )UXLWV f f 6XEWRWDO 5HSURGXFWLYH 6WUXFWXUHV f f 727$/ f f

PAGE 142

'LVFXVVLRQ *URZWK DQG 3URGXFWLYLW\ 3DUWLWLRQLQJ RI GU\ PDWWHU SURGXFWLRQ FKDQJHV FRQVLGHUDEO\ GXULQJ WKH SRWHQWLDO OLIH RI D EDEDVVX SDOP JURZLQJ RQ WKH SULPDU\ IRUHVW VLWH )LJV f $OWKRXJK URRW SURGXFWLYLW\ ZDV QRW PHDVXUHG H[FDYDWLRQV UHYHDOHG WKDW PRVW GU\ PDWWHU SURGXFWLRQ LV LQLWLDOO\ DOORFDWHG WR URRWV )ROORZLQJ JHUPLQDWLRQ D URRW V\VWHP IRUPV SULRU WR H[SDQVLRQ RI WKH ILUVW OHDI )LJ f 7KH V\VWHP FRQWLQXHV WR HQODUJH GXULQJ WKH SHULRG RI VRFDOOHG HVWDEOLVKPHQW JURZWK 7RPOLQVRQ DQG =LPPHUPDQQ f ZKHQ WKH XQGHUJURXQG VWHP JUDGXDOO\ LQFUHDVHV LQ JLUWK 7KH EDVH RI WKH VWHP IRUPV DQ LQYHUWHG FRQH IURP ZKLFK DGYHQWLWLRXV URRWV DULVH 7KHVH URRWV SURYLGH FUXFLDO VXSSRUW IRU WKH PHFKDQLFDOO\ XQVWDEOH VWHP GXULQJ HVWDEOLVKPHQW JURZWK +ROWWXP =LPPHUPDQQ f DV GLVn FXVVHG LQ &KDSWHU WKH XQGHUJURXQG SODFHPHQW RI WKH DSLFDO PHULVWHP YLD FU\SWRJHDO JHUPLQDWLRQ SURYLGHV DGGLWLRQDO VXSSRUW LQ EDEDVVX (VWDEOLVKPHQW JURZWK FRQWLQXHV XQWLO WKH VWHP DSSURDFKHV PD[LPXP JLUWK ZKHUHXSRQ VWHP JURZWK EHFRPHV SUHGRPLQDQWO\ YHUWLFDO :DWHUKRXVH DQG 4XLQQ f $W WKH EHJLQQLQJ RI WKH VHHGOLQJ VWDJH PRVW DERYHJURXQG SURGXFn WLRQ FD bf LV DOORFDWHG WR SKRWR V\QWKHWLF WLVVXHV OHDIOHWVf $V WKH SDOP JURZV KRZHYHU LQFUHDVHG DOORFDWLRQ JRHV WR VXSSRUW LQ WKH IRUP RI OHDI D[HV DQG XOWLPDWHO\ WKH VWHP 3ULRU WR FRPPHQFHPHQW RI HPHUJHQFH RI WKH VWHP IURP WKH VRLO DOORFDWLRQ WR OHDI D[HV FRPSULVHV RYHU b RI WRWDO SURGXFWLYLW\ 7KHVH D[HV SURYLGH FUXFLDO VXSSRUW DV WKH OHDYHV DW WKLV WLPH DSSURDFK WKHLU PD[LPXP OHQJWK DQG DUH RULHQWHG

PAGE 143

352'8&7,9,7< NJ\Uf /(1*7+ 2) 67(0 Pf 2 , 22+ + 5(352'8&7,9( 6758&785(6 + + f§A $*( \Uf )LJXUH $EVROXWH DOORFDWLRQ RI DQQXDO DERYHJURXQG GU\ PDWWHU SURGXFWLRQ SHU EDEDVVX SDOP DV D IXQFWLRQ RI DJH

PAGE 144

352'8&7,9,7< bf /(1*7+ 2) 67(0 Pf 2 L22 $*( \Uf )LJXUH 5HODWLYH DOORFDWLRQ RI SURGXFWLRQ SHU EDEDVVX DQQXDO DERYHJURXQG GU\ PDWWHU SDOP DV D IXQFWLRQ RI DJH

PAGE 145

YHUWLFDOO\ 7KXV D VWHPOHVV SDOP LV DEOH WR SODFH SKRWRV\QWKHWLF WLVVXH FD P DERYH WKH JURXQG WKH OHDI D[HV DUH DOVR JUHHQ DQG FRQVHTXHQWO\ SURYLGH DGGLWLRQDO SKRWRV\QWKHVLV ([FOXVLYH DOORFDWLRQ RI DERYHJURXQG SURGXFWLYLW\ WR OHDYHV GXULQJ WKH VWHPOHVV SKDVH PD\ SURYLGH DQ LPSRUWDQW FRPSHWLWLYH HGJH WR EDEDVn VX DV ZHOO DV WR UDLQIRUHVW SDOPV LQ JHQHUDO 7KHVH OHDYHV DFW HVVHQn WLDOO\ DV WKURZDZD\ EUDQFKHV WKDW SHUPLW VWHPOHVV SDOPV WR RFFXS\ ODUJH YROXPHV DW D FRPSDUDWLYHO\ ORZ PHWDEROLF FRVW *LYQLVK f &KDQQHOOLQJ RI DERYHJURXQG SURGXFWLYLW\ WR SKRWRV\QWKHWLF WLVVXHV PD\ WKXV HQDEOH SDOPV WR SHUVLVW IRU H[WHQVLYH SHULRGV LQ WKH XQGHUVWRU\ XQWLO HYHQWV VXFK DV JDS RSHQLQJV SURYLGH RSSRUWXQLWLHV IRU YHUWLFDO JURZWK %HFDXVH WKH DSLFDO PHULVWHP RI SUHHVWDEOLVKHG MXYHQLOHV LV EXULHG EDEDVVXnV VWHP EHJLQV WR JURZ ORQJ EHIRUH LW HPHUJHV IURP WKH VRLO $OORFDWLRQ RI SURGXFWLYLW\ WR WKH VWHP DFWXDOO\ EHJLQV DW JHUPLQDWLRQ 9HUWLFDO JURZWK FRPPHQFHV DW FD \U LQ WKH SULPDU\ IRUHVW VLWH 7KH VWHP GRHV QRW HPHUJH IURP WKH VRLO XQWLO WKH SDOP LV FD \U ROG E\ ZKLFK WLPH WKH VKRRW LV FD FP ORQJ 8QWLO WKH RQVHW RI UHSURGXFn WLRQ DOORFDWLRQ RI DERYHJURXQG SURGXFWLYLW\ WR WKH VWHP UDSLGO\ LQn FUHDVHV )LJV f DQG VWHP JURZWK DWWDLQV LWV KLJKHVW UDWHV 7DEOH f 5HSURGXFWLRQ EHJLQV ZKHQ WKH SDOP LV VWLOO EHORZ WKH IRUHVW FDQRS\ $V WKH SDOP UHDFKHV WKH FDQRS\ DW FD \U RI DJH SDUWLWLRQLQJ RI SURGXFWLYLW\ WR UHSURGXFWLYH VWUXFWXUHV LQFUHDVHV UDSLGO\ $W WKH VDPH fWLPH WRWDO SURGXFWLYLW\ SHDNV DQG VXEVHTXHQWO\ XQGHUJRHV D VWHHS GH

PAGE 146

FOLQH )LJ f SUREDEO\ GXH WR WKH LQFUHDVHG PHWDEROLF FRVWV RI PDLQn WDLQLQJ D ORQJ VWHP =LPPHUPDQQ f 7KLV GHFOLQH LV DEVRUEHG SULQFLn SDOO\ E\ WKH SDOPnV YHJHWDWLYH VWUXFWXUHV HVSHFLDOO\ WKH OHDYHV SURn GXFWLYLW\ RI UHSURGXFWLYH VWUXFWXUHV GHFUHDVHV PRGHUDWHO\ EXW FRQWLQXHV WR LQFUHDVH RQ D UHODWLYH EDVLV $W WKH RXWVHW RI UHSURGXFWLYH DFWLYLW\ PRVW LQIORUHVFHQFHV SURn GXFHG DUH PDOH )LJ f $V WKH SDOP DSSURDFKHV WKH FDQRS\ SURGXFn WLYLW\ RI IHPDOH LQIORUHVFHQFHV DQG VHHGV SHU SDOP LQFUHDVHV UDSLGO\ SHDNLQJ DW \U 3URGXFWLYLW\ RI IHPDOH LQIORUHVFHQFHV )LJ f DQG VHHGV SHU SDOP )LJ f VXEVHTXHQWO\ XQGHUJRHV PRGHUDWH GHFOLQH ZKLOH SURGXFWLYLW\ RI PDOH LQIORUHVFHQFHV LQFUHDVHV $V ZLQG SROOLQDWLRQ RFFXUV LQ EDEDVVX VHH &KDSWHU f SROOHQ GLVSHUVDO IURP HPHUJHQW SDOPV LV WKXV HQKDQFHG ZKLFK SUREDEO\ UHVXOWV LQ LQFUHDVHG RXWFURVVLQJ $W KLJK GHQVLWLHV SRSXODWLRQV RI EDEDVVX PD\ DWWDLQ KLJK SURGXFn WLYLW\ 7KH SRSXODWLRQ RQ WKH VHFRQGDU\ IRUHVW VLWH IRU H[DPSOH SURGXFHG FD W RI GU\ PDWWHU SHU \U 7DEOH f $OWKRXJK WKH HVWLn PDWHG SURGXFWLYLW\ RI WKH EDEDVVX VWDQG LV OHVV WKDQ WKDW RI WKH HQWLUH VHFRQGDU\ IRUHVW LW LV QRQHWKHOHVV RQH RI WKH KLJKHVW ILJXUHV UHFRUGHG IRU D IRUHVWHG VLWH ZKHWKHU QDWLYH RU SODQWHG .LUD &DQQHOO f $PRQJ WKH IHZ IRUHVWHG VLWHV WKDW KDYH IRXQG ZLWK KLJKHU DERYHJURXQG GU\ PDWWHU SURGXFWLRQ WrKDf r\Uf f LV D SODQWDWLRQ RI $IULFDQ RLO SDOP LQ 0DOD\VLD 1J HW DO f $ FRPPRQ IHDWXUH RI ERWK EDEDVVX VWDQGV DQG RLO SDOP SODQWDWLRQV LV WKDW PRVW RI WKHLU SURGXFn WLYLW\ LV DOORFDWHG WR OHDYHV ZKLFK DUH QRW SHUPDQHQWO\ DFFUXLQJ DQG WKXV PD\ EH UHODWLYHO\ LQH[SHQVLYH WR SURGXFH *LYQLVK f ,Q IDFW

PAGE 147

,1)/25(6&(1&(6 3(5 3$/0 K 727$/ ,1)/25(6&(1&(6 2 )(0$/( ,1)/25(6&(1&(6 DLLLLLLLL,U 0($1 $*( \Uf , )LJXUH $QQXDO QXPEHU RI LQIORUHVFHQFHV SURGXFHG SHU SDOP DV D IXQFWLRQ RI PHDQ DJH SHU OLIH VWDJH $UURZ LQGLFDWHV DJH RI \RXQJHVW PDWXUH SDOP RQ WKH SULn PDU\ IRUHVW VLWH ZKHUH GDWD ZHUH REWDLQHG

PAGE 148

6(('6 3(5 3$/0 )LJXUH $QQXDO QXPEHU RI VHHGV SURGXFHG SHU SDOP DV D IXQFWLRQ RI PHDQ DJH SHU OLIH VWDJH %DUV UHSUHVHQW VWDQGDUG HUURU $UURZ LQGLFDWHV DJH RI \RXQJHVW PDWXUH SDOP RQ SULPDU\ IRUHVW VLWH ZKHUH GDWD ZHUH REWDLQHG

PAGE 149

E\SDVVLQJ WKH SURGXFWLRQ RI SHUPDQHQWO\ DFFUXLQJ VWUXFWXUHV VXFK DV VWHPV LV D KLJKO\ VXFFHVVIXO VWUDWHJ\ LQ PDQ\ JUDVVGRPLQDWHG DJULFXOn WXUDO HFRV\VWHPV ZKLFK DUH WKH PRVW SURGXFWLYH RI WHUUHVWLDO VLWHV /RRPLV DQG *HUDNLV f $V KDV EHHQ VKRZQ IRU EDEDVVX SDOPV PD\ EH PRUH VLPLODU WR JUDVVHV WKDQ WR RWKHU WUHHV LQ WHUPV RI WKHLU DOORFDWLRQ RI GU\ PDWWHU SURGXFWLRQ 7KH SURGXFWLYLW\ RI ZLOG VWDQGV RU SODQWDn WLRQV RI SDOPV FDQ WKXV EH H[WUHPHO\ KLJK \HW QRW DV FRQVSLFXRXV DV WKDW RI SODQWDWLRQV FRPSULVHG RI IDVWJURZLQJ WUHHV VXFK DV (XFDO\SWXV VSS DQG *PHOLQD DUEUHD LQ ZKLFK GU\ PDWWHU SURGXFWLRQ LV DOORFDWHG SULn PDULO\ WR VWHPV 5HSURGXFWLRQ 6WXGLHV RI UHVRXUFH DOORFDWLRQ LQ SDOPV LQFOXGH HVWLPDWHV RI GU\ PDWWHU SURGXFWLRQ RYHU WKH SRWHQWLDO OLIH RI &RU\SKD HO DW D 7RPOLQVRQ DQG 6RGHUKROP f -HVVHQLD ED WDXD %DOLFN f DQG $V WURFDU \XP PH[LFDQXP 6DUXNKDQ f $SSUR[LPDWHO\ b RI WKH PDVV RI DERYHn JURXQG FRPSRQHQWV SURGXFHG RYHU D \U OLIHWLPH LQ EDEDVVX ZDV DOORn FDWHG WR UHSURGXFWLYH VWUXFWXUHV 7DEOH f 7KLV LV KLJKHU WKDQ VLPLn ODU PHDVXUHV RI DOORFDWLRQ WR UHSURGXFWLYH VWUXFWXUHV LQ $V WURFDU\XP PH[LFDQXP bf -HVVHQLD EDWDXD bf DQG &RU\SKD HODWD bf :LWK WKH H[FHSWLRQ RI &RU\SKD HODWD WKH ILJXUHV FLWHG DERYH JHQn HUDOO\ H[FHHG WKRVH UHSRUWHG IRU SHUHQQLDO KHUEV JURZLQJ LQ WHPSHUDWH DUHDV HJ 6WUXLN $EUDKDPVRQ f 7KH ODWWHU DUH VXEMHFW WR ZLQWHU GLHEDFN ZKLFK XVXDOO\ UHTXLUHV FRQVLGHUDEOH DOORFDWLRQ RI UHn VRXUFHV WR YHJHWDWLYH VWRUDJH RUJDQV 6XFK DOORFDWLRQ LV QRW QHFHVVDU\ DPRQJ DQQXDOV JURZLQJ LQ WHPSHUDWH DUHDV DQG WKLV KDV DSSDUHQWO\ OHG WR

PAGE 150

WKH JHQHUDO REVHUYDWLRQ WKDW DQQXDOV DOORFDWH SURSRUWLRQDWHO\ PRUH SURn GXFWLYLW\ WR UHSURGXFWLYH VWUXFWXUHV WKDQ GR SHUHQQLDOV +DUSHU DQG :KLWH f 7KH JHQHUDOL]DWLRQ REYLRXVO\ GRHV QRW H[WHQG WR WKH KXPLG WURSLFV ZKHUH HYHUJUHHQ SHUHQQLDOV DUH QRW VXEMHFW WR GLH EDFN HDFK \HDU ,W LV QR VXUSULVH WKDW WKH VHPHOSDURXV LH PRQRFDUSLFf &RU\SKD HODWD DOORFDWHV D PXFK ORZHU SURSRUWLRQ RI OLIHWLPH SURGXFWLYLW\ WR UHSURGXFWLYH VWUXFWXUHV WKDQ LWHURSDURXV LH SRO\FDUSLFf SDOPV VXFK DV EDEDVVX LQ ZKLFK UHSURGXFWLRQ RFFXUV DQQXDOO\ RYHU FD \U 7KH VKHHU VL]H RI WKH VLQJOH UHSURGXFWLYH HYHQW LQ &K HODWD LV QRQHWKHOHVV LPSUHVVLYH ,Q DEVROXWH WHUPV LW LV HTXLYDOHQW WR FD \U RI UHSURn GXFWLRQ LQ DQ DYHUDJH PDWXUH SDOP RI EDEDVVX &RPSDUHG WR RWKHU LWHURSDURXV SDOPV VXFK DV -HVVHQLD EDWDXD DQG $VWURFDU\XP PH[LFDQXP EDEDVVXnV DOORFDWLRQ WR UHSURGXFWLYH VWUXFWXUHV LV KLJK 7KLV LV SUREDEO\ GXH WR WKH H[FHSWLRQDOO\ ORQJ SHULRG RI VHQHVFHQFH LQ EDEDVVX GXULQJ ZKLFK YHJHWDWLYH JURZWK GHFUHDVHV VKDUSO\ ZKLOH DOORFDWLRQ WR UHSURGXFWLYH VWUXFWXUHV XQGHUJRHV RQO\ D PRGHUDWH GHFOLQH )LJV f %DEDVVXnV SDWWHUQ RI LQFUHDVLQJ VHQHVFHQFH ZLWK DJH FRQIRUPV WR WKDW SUHGLFWHG IRU WUHHV E\ +DUSHU DQG :KLWH fm +RZHYHU 6DUXNKDQ f IRXQG QR HYLGHQFH IRU VHQHVFHQFH LQ D QXPEHU RI WURSLFDO WUHHV LQFOXGLQJ $VWURFDU\XP PH[LFDQXP 3LQHUR HW DO f $OWKRXJK LW KDV EHHQ DUJXHG WKDW VHQHVFHQFH LQ SODQWV LV LQHYLWDEOH +DPLOWRQ f PDQ\ VSHFLHV SUREDEO\ IDLO WR PDQLIHVW VHQHVFHQFH XQGHU QDWXUDO FRQGLWLRQV VXVSHFW WKDW WKH SUHVHQFH RU DEVHQFH RI VHQHVn FHQFH LQ VSHFLHV RI SDOPV PD\ EH SULPDULO\ D IXQFWLRQ RI WKHLU PD[LPXP

PAGE 151

VWHP OHQJWK $V VHFRQGDU\ FRQGXFWLYH WLVVXH LV DEVHQW WKH SKORHP LV HVSHFLDOO\ VXVFHSWLEOH WR PHFKDQLFDO VWUHVV ,Q WDOOHU SDOPV VXFK DV EDEDVVX HYHQWXDO GHWHULRUDWLRQ RI ORQJGLVWDQFH SKORHP WUDQVSRUW LV SUREDEO\ WKH SULQFLSDO FDXVH RI VHQHVFHQFH =LPPHUPDQQ f 7KH DSSDUHQW ODFN RI VHQHVFHQFH LQ $ PH[LFDQXP ZKLFK DWWDLQV D PD[LPXP VWHP OHQJWK RI P 6DUXNK£Q f PD\ EH GXH WR UHGXFHG PHFKDQLFDO VWUHVV RQ WKH SKORHP ,Q WKH WUHHVL]HG EDWDXD VHQHVFHQFH DSSDUn HQWO\ GRHV RFFXU %DOLFN f EXW LWV GXUDWLRQ DQG HIIHFWV RQ SDUWLn WLRQLQJ RI SURGXFWLYLW\ ZLWKLQ WKH SDOP DUH XQNQRZQ 7KH HFRORJLFDO DQG HYROXWLRQDU\ FRQVHTXHQFHV RI EDEDVVXnV UHODWLYHn O\ KLJK DOORFDWLRQ RI SURGXFWLYLW\ WR UHSURGXFWLYH VWUXFWXUHV UHPDLQ WR EH H[DPLQHG +LJK DOORFDWLRQ WR VXFK VWUXFWXUHV GRHV QRW QHFHVVDULO\ WUDQVODWH LQWR LQFUHDVHG SURGXFWLRQ RI VHHGV )RU H[DPSOH FDOFXODWH WKDW EDEDVVX SURGXFHV DQ DYHUDJH RI FD VHHGV RYHU LWV OLIHWLPH WKH RQFH IORZHULQJ &RU\SKD HODWD ZLWK D IDU ORZHU DEVROXWH DV ZHOO DV UHODWLYH DOORFDWLRQ WR UHSURGXFWLYH VWUXFWXUHV SURGXFHV DQ HVWLPDWHG PLQLPXP RI VHHGV 7RPOLQVRQ DQG 6RGHUKROP f 2I WKH WRWDO GU\ PDVV RI UHSURGXFWLYH VWUXFWXUHV SURGXFHG RYHU WKH OLIHWLPH RI D EDEDVVX SDOP RQO\ b LV GLUHFWO\ DOORFDWHG WR UHSURGXFWLRQ LQ WKH IRUP RI VHHGV 0RUH WKDQ WZLFH WKLV DPRXQW bf LV LQYHVWHG LQ PDOH LQIORUHVFHQFHV 3URGXFWLRQ RI PDOH DQG IHPDOH IORZHUV RQ VHSDUDWH LQIORUHVFHQFHV HQKDQFHV WKH SUREDELOLW\ RI RXWFURVVLQJ ZKLFK LQ WXUQ OHDGV WR LQFUHDVHG JHQHWLF YDULDELOLW\ ZLWKLQ SRSXODWLRQV RI EDEDVVX VHH &KDSWHU f 6XFK JHQHWLF YDULDELOLW\ SHUPLWV WKH RFFXUUHQFH RI ORFDOO\ DGDSWHG HFRW\SHV VHH &KDSWHU f 0RVW RI WKH UHPDLQLQJ

PAGE 152

DOORFDWLRQ WR UHSURGXFWLYH VWUXFWXUHV LV GLUHFWHG WR SURWHFWLRQ HLWKHU RI VHHGV E\ IUXLW ZDOOV bf RU RI GHYHORSLQJ IHPDOH IORZHUV E\ EUDFWV bf 0RUSKRORJLFDO SURWHFWLRQ RI UHSURGXFWLYH FRPSRQHQWV LV D FKDUDFWHULVWLF IHDWXUH RI SDOPV 8KO DQG 0RRUH f 6HOHFWLYH SUHVn VXUHV WKDW PD\ KDYH UHVXOWHG LQ EDEDVVXnV SDUWLFXODUO\ KLJK LQYHVWPHQW ZHUH GLVFXVVHG LQ &KDSWHU +LJK LQYHVWPHQW LQ SURWHFWLRQ RI UHSURGXFn WLYH FRPSRQHQWV PD\ XOWLPDWHO\ OHDG WR KLJK VHHGOLQJ UHFUXLWPHQW D WRSLF WKDW VKDOO H[DPLQH IXUWKHU LQ WKH IROORZLQJ FKDSWHU

PAGE 153

&+$37(5 3238/$7,21 6758&785( $1' '<1$0,&6 3RSXODWLRQ ELRORJ\ RI SODQWV ZDV WKH H[FOXVLYH GRPDLQ RI IRUHVWHUV DQG DJURQRPLVWV XQWLO UHFHQWO\ ZKHQ HFRORJLVWV EHFDPH LQWHUHVWHG LQ WKH VXEMHFW 3ODQW HFRORJLVWV KDYH WUDGLWLRQDOO\ FRQFHQWUDWHG RQ WKH GHn VFULSWLRQ FODVVLILFDWLRQ DQG SK\VLRORJ\ RI YHJHWDWLRQ ZKLOH ODUJHO\ LJQRULQJ WKH G\QDPLF DVSHFWV RI SRSXODWLRQ SKHQRPHQD $V D UHVXOW SODQW SRSXODWLRQ ELRORJ\ KDV ERUURZHG PXFK RI LWV WKHRUHWLFDO EDVLV IURP ]RRORJLVWV +DUSHU DQG :KLWH f} +RZHYHU LQ D V\QWKHVLV RI WKH HPHUJLQJ ILHOG +DUSHU f GUHZ LPSRUWDQW GLVWLQFWLRQV EHWZHHQ SRSXn ODWLRQV RI DQLPDOV DQG SODQWV 7KH PRVW VLJQLILFDQW GLVWLQFWLRQ LV WKH PRGXODU FRQVWUXFWLRQ RI SODQWV ZKLFK FRQIHUV H[WUHPH PRUSKRORJLFDO SODVWLFLW\ ,Q FRQWUDVW WR DQLPDOV LQGLYLGXDO SODQWV FDQ UHVSRQG WR VWUHVVHV E\ YDU\LQJ WKH QXPEHU RI WKHLU FRPSRQHQW SDUWV $V D UHVXOW SODQW SRSXODWLRQV PD\ EH YLHZHG DV FRQVLVWLQJ RI ERWK LQGLYLGXDOV DQG WKHLU FRPSRQHQW SDUWV 7R GDWH PRVW HFRORJLFDO VWXGLHV LQYROYLQJ WKH VWUXFWXUH DQG G\QDn PLFV RI SODQW SRSXODWLRQV KDYH EHHQ FDUULHG RXW RQ WHPSHUDWH KHUEV 7R P\ NQRZOHGJH ZLWK WKH H[FHSWLRQ RI WZR GLFRW\OHGRQRXV WUHHV VWXGLHG E\ +DUWVKRUQ f DOO VXFK VWXGLHV LQ WKH WURSLFV KDYH LQYROYHG SDOPV 9DQ 9DOHQ EDVHG RQ GDWD IURP %DQQLVWHU 6DUXNKDQ DQG %XOORFN f SUREDEO\ EHFDXVH WKH DJH RI SDOPV LV UHODWLYHO\ HDV\ WR HVWLPDWH VHH &KDSWHU f :KHWKHU LQ WHPSHUDWH RU WURSLFDO DUHDV IHZ VWXGLHV KDYH DWWHPSWHG WR H[DPLQH WKH SRSXODWLRQ ELRORJ\ RI D VSHFLHV

PAGE 154

XQGHU YDU\LQJ GHJUHHV RI GLVWXUEDQFH HJ 0DUNV &RRN DQG /\RQV f ,Q WKLV FKDSWHU H[DPLQH KRZ VWDQGV RU SRSXODWLRQV RI EDEDVVX UHVSRQG WR GLIIHUHQW GLVWXUEDQFHV 0\ REMHFWLYH LV WR FRPSDUH WKH VWUXFWXUH DQG G\QDPLFV RI WKUHH SRSXODWLRQV RQH LQ SULPDU\ IRUHVW RQH LQ VHFRQGDU\ IRUHVW DQG RQH LQ SDVWXUH 7KH VWUXFWXUH RI WKHVH SRSXODn WLRQV UHIOHFWV SDVW DV ZHOO DV SUHVHQW SDWWHUQV RI ODQG XVH DQG WKXV SURYLGHV FOXHV DV WR KRZ KLJKGHQVLW\ VWDQGV RI EDEDVVX RULJLQDWH /LNHZLVH WKH G\QDPLFV RI WKHVH SRSXODWLRQV LQGLFDWH WKHLU SUHVHQW FDSDFLW\ WR UHJHQHUDWH $ FRPSDULVRQ RI WKH VWUXFWXUH DQG G\QDPLFV RI SRSXODWLRQV VXEMHFWHG WR GLIIHUHQW ODQG XVHV WKXV JLYHV LQVLJKW DV WR WKH RULJLQ DQG SRVVLEOH IXWXUH RI WKH SDOP IRUHVWV 0HWKRGV 3RSXODWLRQ VWUXFWXUH RI EDEDVVX ZDV GHWHUPLQHG E\ FHQVXVLQJ VDPSOHV RI HDFK OLIH VWDJH VHH 7DEOH f SUHVHQW RQ WKH SULPDU\ IRUHVW VHFRQGn DU\ IRUHVW DQG SDVWXUH VLWHV 7KH DUHDV DOORFDWHG IRU FHQVXVLQJ ZHUH KD IRU VWDJHV KD IRU VWDJH KD IRU VWDJHV KD IRU VWDJH DQG KD IRU VWDJHV 7KH SDOPV ZLWKLQ WKHVH DUHDV ZHUH ODEHOOHG DQG PDSSHG )ROORZLQJ DQ LQLWLDO FHQVXV LQ $SULO WKH SRSXODWLRQV ZHUH UHFHQVXVHG WZLFH DW PR LQWHUYDOV GXULQJ ZKLFK HDFK VWDJHnV PRUWDOLW\ DQG UHFUXLWPHQW ZHUH UHFRUGHG $ VWDWLF RU WLPHVSHFLILF OLIH WDEOH ZDV FRQVWUXFWHG IRU WKH SULn PDU\ IRUHVW SRSXODWLRQ 7DEOH f 7R FRQVWUXFW WKH OLIH WDEOH WKH SRSXODWLRQ ZDV DVVXPHG WR EH QXPHULFDOO\ VWDEOH ZKLFK LPSOLHV WKDW WKH DJH GLVWULEXWLRQ DQG VXUYLYRUVKLS FXUYHV ZHUH LGHQWLFDO 6XUYLYRUVKLS

PAGE 155

O[f FRXOG WKXV EH GHWHUPLQHG VLPSO\ E\ FRXQWLQJ WKH QXPEHU RI LQGLn YLGXDOV SUHVHQW DW D JLYHQ VWDJH RU DJH ,QLWLDO DJHV LQ WKH OLIH WDEOH ZHUH FDOFXODWHG IURP RYXOH IHUWLOLn ]DWLRQ UDWKHU WKDQ VHHG JHUPLQDWLRQ 7KH WLPH EHWZHHQ WKHVH HYHQWV ZDV HVWLPDWHG WR EH \U DVVXPLQJ DQ DYHUDJH RI PR IRU IUXLW PDWXUDWLRQ DQG PR IRU JHUPLQDWLRQ VHH &KDSWHU f XVHG VHHGV SURGXFHG SHU \U LQVWHDG RI VHHGV RQ WKH JURXQG VWDJH f LQ WKH ILUVW URZ RI WKH OLIH WDEOH 7KH ODWWHU UHIOHFWV VHHG QXPEHUV DW RQH SRLQW LQ WLPH ZKHUHDV WKH IRUPHU LQWHJUDWHV VHHG G\QDPLFV RYHU D \U SHULRG GXH WR WKH UDSLGLW\ RI VHHG G\QDPLFV FRQVLGHUHG WKH DQQXDO UDWH RI VHHG SURGXFn WLRQ WR EH D EHWWHU LQGLFDWRU RI SRSXODWLRQ VWUXFWXUH 7KH YDOXHV XQGHU VXUYLYRUVKLS LQ WKH ILUVW DQG VXEVHTXHQW URZV DUH \U PHDQV 7RWDO VHHGV SURGXFHG SHU \U f WKXV UHSUHVHQW WKH PHDQ RI WKH f DQG f KDUYHVWV /LNHZLVH WRWDOV LQ VXEVHTXHQW URZV UHSUHVHQW PHDQV RI WKH ILUVW $SULO f DQG ILQDO $SULO f SRSXODWLRQ FHQVXVHV VHH 7DEOHV f ZKLFK ZHUH FDUULHG RXW DW WKH HQG RI EDEDVVXnV IUXLWLQJ VHDVRQ 0RUWDOLW\ G[f ZDV FDOFXODWHG DV WKH GLIIHUHQFH EHWZHHQ WRWDO VXUYLYRUVKLS LQ D JLYHQ URZ DQG WKDW LQ WKH IROORZLQJ URZ 0RUWDOLW\ GLYLGHG E\ WKH WLPH LQ \U VSHQW LQ HDFK URZ JDYH PRUWDOLW\ SHU \U T[f 7KH WLPH VSHQW LQ WKH ILQDO URZ VWDJH f ZDV HVWLPDWHG DW \U VHH 7DEOH f 7R FDOFXODWH YDOXHV IRU WKH ILQDO WKUHH FROXPQV LQ VWDJH WKH VROH UHSUHVHQWDWLYH ZDV DVVXPHG WR GLH DW WKH FRQFOXVLRQ RI WKH VWXG\ /LIH H[SHFWDQF\ H[f ZDV FDOFXODWHG IURP WKH LQLWLDO DJH LQ HDFK URZ 'HWDLOV RQ KRZ OLIH H[SHFWDQF\ LV FDOFXODWHG DUH SURYLGHG E\ %UHZHU DQG 0F&DQQ f

PAGE 156

$ VWDJH SURMHFWLRQ PDWUL[ /HINRYLWFK f ZDV QRW FRQVWUXFWHG IRU WKLV SRSXODWLRQ 6XFK D PDWUL[ UHTXLUHV WUDQVIHU FRHIILFLHQWV WKDW DUH ERWK FRQVWDQW 6LOYHUWRZQ f DQG SRVLWLYH EHWZHHQ DOO VWDJHV RI WKH SRSXODWLRQ 'XULQJ D \U SHULRG VL[ RI WKH OLIH VWDJHV UHFHLYHG QR UHFUXLWPHQW 7DEOH f $ ORQJHU SHULRG RI REVHUYDWLRQV ZRXOG WKXV EH UHTXLUHG WR FRQVWUXFW D SURMHFWLRQ PDWUL[ IRU WKLV SRSXODWLRQ 5HVXOWV 3ULPDU\ )RUHVW ,Q WKH SULPDU\ IRUHVW SRSXODWLRQ RI EDEDVVX DQQXDO PRUWDOLW\ UDWHV DV FDOFXODWHG LQ WKH OLIH WDEOH 7DEOH f FORVHO\ SDUDOOHOHG WKRVH REWDLQHG GXULQJ WKH \U SHULRG RI ILHOG REVHUYDWLRQV 7DEOH f 7KLV OHQGV VXSSRUW WR WKH LQLWLDO DVVXPSWLRQ RI VWDELOLW\ LQ WKH SRSXODWLRQ %DEDVVXfV VXUYLYRUVKLS DV D IXQFWLRQ RI DJH LQ WKH SULPDU\ IRUHVW 7DEOH f VKRZHG D UHYHUVH -VKDSHG FXUYH FKDUDFWHULVWLF RI PDQ\ PDWXUH IRUHVW WUHH SRSXODWLRQV DQG LQGLFDWLYH RI D VWDEOH DJH SRSXODWLRQ 6LOYHUWRZQ f 3DOP LQ ROGHU OLIH VWDJHV DOPRVW FRQVLVWHQWO\ GHFOLQHG LQ IUHn TXHQF\ \HW YLUWXDOO\ DOO VWDJHV ZHUH UHSUHVHQWHG )LJ f} RQH ZRXOG H[SHFW VXFK D SDWWHUQ LQ D VWDEOH SRSXODWLRQ $JH GLVWULEXWLRQ LQ WKH SULPDU\ IRUHVW OLNHZLVH IROORZHG D UHYHUVH -VKDSHG FXUYH )LJ f 6HHG G\QDPLFV VKRZHG D VLPLODU SDWWHUQ LQ ERWK WKH SULPDU\ DQG VHFRQGDU\ IRUHVWV 7DEOHV f $SSUR[LPDWHO\ KDOI WKH DQQXDO VHHG SURGXFWLRQ VXFFXPEHG WR PRUWDOLW\ D WHQWK JHUPLQDWHG DQG WKH UHVW ZHQW WR VRLO VWRUDJH 2Q ERWK VLWHV VHHGOLQJ UHFUXLWPHQW ZDV PRUH WKDQ VXIILFLHQW WR PDLQWDLQ VHHGOLQJ SRSXODWLRQV

PAGE 157

c7DEOH /LIH WDEOH RI WKH SRSXODWLRQ RI EDEDVVX /LIH VWDJHV DUH GHILQHG LQ 7DEOH RQ WKH KD SULPDU\ IRUHVW VLWH DW /DJR 9HUGH /,)( 67$*( $*( \Uf 6859,9256+,3 0257$/,7< /,)( 1XPEHU DQG 'HVFULSWLRQ ,QLWLDO 0HDQ 1 b 1 SHU \U (;3(&7$1&< \Uf 6HHGV SURGXFHG SHU \U 6HHGOLQJV 3UHHVWDEOLVKHG MXYHQLOHV (VWDEOLVKHG MXYHQLOHV 0DWXUH SDOPV

PAGE 158

7DEOH $QQXDO FKDQJHV LQ QXPEHU RI SDOPV SHU OLIH VWDJH RQ WKH KD SULPDU\ IRUHVW VLWH DW /DJR 9HUGH /LIH VWDJHV DUH GHILQHG LQ 7DEOH /,)( 67$*( ,1,7,$/ 1 *$,16 /266(6 1(7 &+$1*( ),1$/ 1 1XPEHU DQG 'HVFULSWLRQ 5HFUXLWPHQW IURP /DVW 6WDJH 5HFUXLWPHQW WR 1H[W 6WDJH 0RUWDOLW\ 6HHGV RQ JURXQG 6HHGOLQJV 3UHHVWDEOLVKHG MXYHQLOHV (VWDEOLVKHG MXYHQLOHV 0DWXUH SDOPV 67

PAGE 159

7DEOH $QQXDO FKDQJHV LQ QXPEHU RI SDOPV SHU OLIH VWDJH RQ WKH KD VHFRQGDU\ IRUHVW VLWH DW /DJR 9HUGH /LIH VWDJHV DUH GHILQHG LQ 7DEOH /,)( 67$*( ,1,7,$/ 1 *$,16 /266(6 1(7 &+$1*( ),1$/ 1 1XPEHU DQG 'HVFULSWLRQ 5HFUXLWPHQW IURP /DVW 6WDJH 5HFUXLWPHQW WR 1H[W 6WDJH 0RUWDOLW\ 6HHGV RQ JURXQG 6HHGOLQJV 3UHHVWDEOLVKHG MXYHQLOHV (VWDEOLVKHG MXYHQLOHV 0DWXUH SDOPV 97

PAGE 160

7DEOH $QQXDO FKDQJHV LQ QXPEHU RI SDOPV SHU OLIH VWDJH RQ WKH KD SDVWXUH VLWH DW /DJR 9HUGH /LIH VWDJHV DUH GHILQHG LQ 7DEOH /,)( 67$*( ,1,7,$/ 1 *$,16 /266(6 1(7 &+$1*( ),1$/ 1 1XPEHU DQG 'HVFULSWLRQ 5HFUXLWPHQW IURP /DVW 6WDJH 5HFUXLWPHQW WR 1H[W 6WDJH 0RUWDOLW\ 6HHGV RQ JURXQG 6HHGOLQJV 3UHHVWDEOLVKHG MXYHQLOHV (VWDEOLVKHG MXYHQLOHV 0DWXUH SDOPV

PAGE 161

3$/06 QRKDf , , , , P 6(('/,1*6 + 35((67$%/,6+(' A -89(1,/(6 + (67$%/,6+(' -89(1,/(6 Â’ 0$785( 3$/06 )LJXUH 6WDJH GLVWULEXWLRQV RI EDEDVVX RQ WKH SULPDU\ IRUHVW VHFRQGDU\ IRUHVW DQG SDVWXUH VLWHV DW /DJR 9HUGH

PAGE 162

3$/06 QRKDf )LJXUH $JH GLVWULEXWLRQ RI EDEDVVX RQ WKH SULPDU\ IRUHVW VLWH DW /DJR 9HUGH

PAGE 163

3$/06QRKDf )LJXUH $JH GLVWULEXWLRQ RI EDEDVVX RQ WKH VHFRQGDU\ IRUHVW VLWH DW /DJR 9HUGH

PAGE 164

35,0$5< )25(67 A YOn nOn 6(&21'$5< )25(67 ?nOn nO 3$6785( A A A 6(('/,1*6 35((67$%/,6+(' (67$%/,6+(' -89(1,/(6 -89(1,/(6 0$785( 3$/06 )LJXUH 3RSXODWLRQ G\QDPLFV RI EDEDVVX RQ WKH KD SULPDU\ IRUHVW VHFRQGDU\ IRUHVW DQG SDVWXUH VLWHV DW /DJR 9HUGH 1XPEHUV LQ ER[HV DUH 1 DW W QXPEHUV RQ KRUL]RQWDO DUURZV UHn SUHVHQW UHFUXLWPHQW RYHU \U DQG QXPEHUV RQ YHUWLFDO DUn URZV UHSUHVHQW PRUWDOLW\ RYHU \U

PAGE 165

5HFUXLWPHQW RI SUHHVWDEOLVKHG DQG HVWDEOLVKHG MXYHQLOHV GLG QRW RFFXU LQ WKH SULPDU\ IRUHVW )LJ f} ZKLFK ZRXOG VXJJHVW WKDW WKH SRSXODWLRQ ZDV QRW UHJHQHUDWLQJ
PAGE 166

:KHQ WKH IRUHVW ZDV RULJLQDOO\ FOHDUHG WKHVH SDOPV ZRXOG KDYH EHHQ FD \U ROG RU OHVV WKDQ P WDOO DFFRUGLQJ WR P\ JURZWK UDWH HVWLPDWHV IRU SULPDU\ IRUHVW VHH 7DEOH f 7KH DSLFDO PHULVWHP ZRXOG VWLOO EH EHQHDWK WKH VRLO VXUIDFH LQ WKHVH SDOPV LWV HPHUJHQFH LQ EDEDVVX RQO\ RFFXUV ZKHQ WKH VKRRW LV FD P WDOO VHH &KDSWHU f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f LQGLFDWH WKDW UHFUXLWPHQW RI VHHGOLQJV MXYHQLOHV DQG DGXOWV ZDV PRUH WKDQ VXIILFLHQW WR RIIVHW ORVVHV GXH WR PRUWDOLW\ 7KXV DW \U DIWHU IRUHVW FOHDULQJ WKH EDEDVVX SRSXODWLRQ VWLOO DSSHDUHG WR EH JURZLQJ 3DVWXUH ,Q FRPSDULVRQ WR WKH VHFRQGDU\ IRUHVW WKH SDVWXUH SRSXODWLRQ VKRZHG DQ HYHQ PRUH SURQRXQFHG EXOJH RI VWDJH SDOPV DSSDUHQWO\ HPSKDn VL]HG E\ UHODWLYHO\ ORZ IUHTXHQFLHV RI VWDJHV )LJ ff 7KH UHVXOW ZDV D GLVWLQFWO\ WZRWLHUHG VWDQG FKDUDFWHULVWLF RI EDEDVVX LQ SDVWXUHV D GHQVH XQGHUVWRU\ RI MXYHQLOHV P WDOO ZLWK DQ RYHUVWRU\ RI PDWXUH SDOPV WKDW ZDV PRUH RU OHVV XQLIRUP LQ KHLJKW 7KH RYHUVWRU\ SDOPV ZHUH SUREDEO\ UHOHDVHG VLPXOWDQHRXVO\ DIWHU WKH RULJLQDO IRUHVW FOHDULQJ

PAGE 167

$QQXDO FXWWLQJ DQG EXUQLQJ ZHUH LQLWLDWHG DIWHU WKH SDVWXUH KDG EHHQ HVWDEOLVKHG 8QGHU WKLV UHJLPHQ H[SRVXUH RI DSLFDO PHULVWHPV DW RU QHDU WKH VRLO VXUIDFH UHVXOWHG LQ KLJK PRUWDOLW\ DPRQJ VWDJHV WKXV DFFRXQWLQJ IRU WKHLU UHODWLYHO\ ORZ IUHTXHQFLHV RQ WKH VLWH 6HHG G\QDPLFV LQ WKH SDVWXUH GLIIHUHG IURP WKRVH LQ WKH IRUHVWHG KDELWDWV 7DEOH f ,Q WKH SDVWXUH RYHU b RI WKH DQQXDO VHHG SURn GXFWLRQ HLWKHU GLHG RU ZDV UHPRYHG IURP WKH VLWH E\ IUXLWJDWKHULQJ KXPDQV &RQVHTXHQWO\ RQO\ D VPDOO SHUFHQWDJH FD bf ZDV DYDLODEOH IRU VHHGOLQJ UHFUXLWPHQW $PRQJ VHHGOLQJV DQG MXYHQLOHV UHFUXLWPHQW GXULQJ WKH \U SHULRG ZDV LQVXIILFLHQW WR FRYHU PRUWDOLW\ ORVVHV )LJ fr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n PRUH f ,Q WKH DEVHQFH RI PDMRU GLVWXUEDQFHV VRPH VSHFLHV UHO\ RQ WKH IRUPDWLRQ RI VPDOO JDSV LQ WKH IRUHVW FDQRS\ +DUWVKRUQ f :KHQ VXEMHFWHG WR PDMRU GLVWXUEDQFHV VXFK DV D F\FORQH :HEE f RU ILUH 0DLVVXURZ f} WKH VLWH LV W\SLFDOO\ RFFXSLHG E\ D GLIIHUHQW VHW RI

PAGE 168

VSHFLHV FKDUDFWHULVWLF RI HDUO\ VXFFHVVLRQDO FRPPXQLWLHV 9HU\ IHZ IRUHVW VSHFLHV DUH FKDUDFWHULVWLF RI ERWK H[WUHPHV PRVW VHHP WR IDOO VRPHZKHUH LQ EHWZHHQ DQG UHJHQHUDWH IROORZLQJ PRGHUDWH GLVWXUEDQFHV 0DUNV &RRN DQG /\RQV f %DEDVVX LV H[WUDRUGLQDU\ LQ LWV FDSDFLW\ WR UHJHQHUDWH DQG SHUVLVW RYHU DQ H[WUHPHO\ ZLGH JUDGLHQW RI GLVWXUEDQFHV UDQJLQJ IURP RFFDVLRQDO JDSV LQ PDWXUH IRUHVW WR SHULRGLF FOHDULQJ DQG EXUQLQJ RI WKH IRUHVW E\ VKLIWLQJ FXOWLYDWRUV %DEDVVXnV FDSDFLW\ WR PDLQWDLQ SRSXODWLRQV LQ SULPDU\ IRUHVW UHn VXOWV IURP f KLJK VHHGOLQJ UHFUXLWPHQW DQG f SHUVLVWHQFH RI ERWK VHHGOLQJV DQG MXYHQLOHV LQ WKH XQGHUVWRU\ +LJK VHHGOLQJ UHFUXLWPHQW LV SUREDEO\ GXH WR WKH SDOPnV UHODWLYH VXFFHVV LQ HVFDSLQJ VHHG SUHGDWLRQ DQG LWV HQKDQFHG JHUPLQDWLRQ XQGHU VKDGH VHH &KDSWHU f 3HUVLVWHQFH RI VHHGOLQJV DQG MXYHQLOHV UHVXOWV IURP WKHLU FDSDFLW\ WR VXUYLYH XQGHU FRPSHWLWLYHO\ FURZGHG FRQGLWLRQV )RU H[DPSOH WKH PHDQ OLIH H[SHFWDQF\ RI VHHGOLQJV LQ WKH SULPDU\ IRUHVW VLWH ZDV FDOFXODWHG DW \U 7DEOH f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

PAGE 169

GXULQJ ZKLFK WKH ULVN RI GHDWK E\ FXWWLQJ DQG EXUQLQJ LV JUHDWHVW 7KH IDOORZ DOVR SURYLGHV VKDGH IRU HQKDQFHG JHUPLQDWLRQ DQG VHHGOLQJ UHn FUXLWPHQW 3HULRGLF DOWHUQDWLRQV RI IRUHVW FOHDULQJ DQG UHJURZWK WKXV OHDG WR WKH IRUPDWLRQ RI H[FHSWLRQDOO\ GHQVH VWDQGV RI EDEDVVX )ROORZLQJ WKH LQLWLDO FOHDULQJ RI SULPDU\ IRUHVW D SRSXODWLRQ RI EDEDVVX ZLOO FRQWLQXH WR IORXULVK HYHQ LI VXEVHTXHQW FOHDULQJ LV GHOD\HG FRQVLGHUDEO\ DV LQGLFDWHG E\ FRQWLQXHG UHJHQHUDWLRQ RI WKH VHFRQGDU\ IRUHVW SRSXODWLRQ DIWHU D IDOORZ RI \U )LJ fr 7KH WLPH UHTXLUHG WR UHWXUQ WR WKH RULJLQDO IRUHVW FRPSRVLWLRQ SULRU WR GLVWXUEDQFH PD\ EH H[WUHPHO\ SURORQJHG $V IDOORZV \U DUH H[FHSWLRQDO LQ ]RQHV FXUn UHQWO\ GRPLQDWHG E\ EDEDVVX RQH FDQ VDIHO\ SUHGLFW WKDW RQFH D SRSXODn WLRQ FDSWXUHV D VLWH LW FDQ PDLQWDLQ LWVHOI LQGHILQLWHO\ 7KH VWDQG ZLOO RQO\ EH UHSODFHG E\ LQWHQWLRQDO HUDGLFDWLRQ RU E\ DQQXDO FXWWLQJ DQG EXUQLQJ RQ D ORQJWHUP EDVLV :KHQ VXEMHFWHG WR VSRUDGLF GLVWXUEDQFHV SRSXODWLRQV RI EDEDVVX FRQVHTXHQWO\ WKULYH +RZHYHU ZKHQ D VWDQG LV VXEMHFWHG WR PRUH IUHn TXHQW GLVWXUEDQFHV VXFK DV DQQXDO SDVWXUH ILUHV UHJHQHUDWLRQ LV QR ORQJHU VXIILFLHQW IRU ORQJWHUP PDLQWDQFH RI WKH SRSXODWLRQ 7KLV UHVXOWV IURP D FRPELQDWLRQ RI ORZ VHHGOLQJ UHFUXLWPHQW GXH WR IUXLW JDWKHULQJ KXPDQV DQG KLJK PRUWDOLW\ RI MXYHQLOHV GXH WR UHSHDWHG H[SRn VXUH RI WKHLU DSLFDO PHULVWHUQV WR ILUH %HFDXVH RI EDEDVVXnV ORQJHYLW\ WKH XOWLPDWH GHPLVH RI SDVWXUH VWDQGV ZLOO QRW RFFXU LQ WKH QHDU IXWXUH 1HYHUWKHOHVV ZLGHVSUHDG FRQYHUVLRQ RI IRUPHUO\ FXOWLYDWHG ODQGV WR SDVWXUHV WKURXJKRXW 0DUDQK£R KDV LPSRUWDQW LPSOLFDWLRQV IRU WKH IXWXUH RI WKH SDOP IRUHVWV

PAGE 170

&+$37(5 ,03/,&$7,216 )25 0$1$*(0(17 3HRSOH DQG WKH 3DOP )RUHVW 7KH DGDSWDELOLW\ RI EDEDVVX WR D ZLGH UDQJH RI HFRORJLFDO FRQGLn WLRQV KDV EHHQ D FRPPRQ WKHPH WKURXJKRXW WKLV VWXG\ +LJK JHQHWLF YDULDELOLW\ LV PDQLIHVW LQ EDEDVVXnV SURSHQVLW\ WR K\EULGL]H DQG LWV FDSDFLW\ WR IRUP ORFDOO\ DGDSWHG HFRW\SHV %URDGO\ V\QFKURQL]HG IORZHUn LQJ RYHU ZLGH JHRJUDSKLF DUHDVf§FRPELQHG ZLWK EUHHGLQJ YLD RXWFURVVLQJf§ SURYLGH DPSOH RSSRUWXQLWLHV IRU JHQHWLF UHFRPELQDWLRQ %DEDVVXnV SROn OLQDWLRQ E\ ZLQG DV ZHOO DV EHHWOHV PD\ KDYH IDFLOLWDWHG LWV VXFFHVVIXO WUDQVLWLRQ IURP FORVHG IRUHVWV WR PRUH RSHQ VLWHV $OWKRXJK PDVVLYH WKH IUXLWV RI WKH SDOP KDYH EHHQ GLVSHUVHG RYHU QHDUO\ KDOI D FRQWLQHQW E\ IORRGLQJ ULYHUV DQG UXQRII DQG RQ D PRUH ORFDO VFDOH E\ KXPDQV DQG URGHQWV /DUJH QRZH[WLQFW PDPPDOV VXFK DV JLDQW JURXQG VORWKV DQG JRPSKRWKHUHV PD\ DOVR KDYH KDG D UROH DV DJHQWV RI GLVSHUVDO DV ZHOO DV DJHQWV RI VHOHFWLRQ LQ WKH HYROXWLRQ RI EDEDVVXnV H[WUDRUGLQDULO\ WKLFN HQGRFDUS 2WKHU WKDQ D[HZLHOGLQJ KXPDQV WKHUH DUH QR PDPPDOV FXUUHQWn O\ RQ WKH VFHQH FDSDEOH RI SHQHWUDWLQJ WKLV IRUPLGDEOH VWUXFWXUH %Dn EDVVXnV RWKHU LPSRUWDQW VHHG SUHGDWRU D EUXFKLG EHHWOH VNLUWV WKH HQGRFDUS E\ HQWHULQJ DV WLQ\ ODUYDH WKURXJK WKH JHUPLQDWLRQ SRUHV +RZHYHU EHFDXVH WKHUH DUH XVXDOO\ VHYHUDO VHHGV SHU IUXLWVf§DQG RQH LV VXIILFLHQW WR VDWLVI\ HDFK ODUYDf§VRPH DOPRVW DOZD\V HVFDSH ,Q FORVHG IRUHVWV JHUPLQDWLRQ LQ EDEDVVX W\SLFDOO\ WDNHV SODFH PR DIWHU IUXLW IDOO 'XH WR KLJK VKDGH WROHUDQFH VHHGOLQJV DQG VWHP

PAGE 171

OHVV MXYHQLOHV FDQ SHUVLVW LQ WKH XQGHUVWRU\ IRU \HDUV EXLOGLQJ XS KLJK GHQVLWLHV ,Q SULPDU\ IRUHVW IRU H[DPSOH D SDOP UHPDLQV DSSDUHQWO\ VWHPOHVV IRU DQ DYHUDJH RI FD \U 'XULQJ WKLV SHULRG WKH VWHP JURZV EHORZ JURXQG GXH WR EDEDVVXnV VRFDOOHG FU\SWRJHDO PRGH RI JHUPLn QDWLRQ LQ ZKLFK WKH DSLFDO PHULVWHP LV SODFHG EHQHDWK WKH VRLO 3UREn DEO\ EHFDXVH LW JURZV VR VORZO\ WKH SDOP LV SRWHQWLDOO\ ORQJOLYHG DWWDLQLQJ D PD[LPXP OLIHVSDQ RI DW OHDVW \U (VFDSH IURP VHHG SUHGDWRUV VKDGH WROHUDQFH DQG ORQJHYLW\ DUH WKH NH\ IDFWRUV DFFRXQWLQJ IRU EDEDVVXnV KLJK LPSRUWDQFH LQ SULPDU\ IRUHVW 3ODFHPHQW RI WKHLU DSLFDO PHULVWHPV EHQHDWK WKH VRLO HQDEOHV VHHGn OLQJV DQG VWHPOHVV MXYHQLOHV WR VXUYLYH ZKHQ IRUHVWV DUH FXW DQG EXUQHG E\ VKLIWLQJ FXOWLYDWRUV :KHQ WKH VLWHV DUH DEDQGRQHG WKHVH SDOPV DUH UHOHDVHG DQG IRUP KLJKGHQVLW\ YLUWXDOO\ PRQRVSHFLILF VWDQGV GXULQJ WKH VXEVHTXHQW IDOORZ %DEDVVXn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b RI VDPSOHG KRXVHKROGV LQ WKH YLFLQLW\ RI WKH SDOP IRUHVWV $QGHUn VRQ DQG $QGHUVRQ f 7KXV EDEDVVX LV DOUHDG\ WLJKWO\ LQWHJUDWHG LQWR PDUNHW DQG VXEVLVWHQFH HFRQRPLHV XSRQ ZKLFK UXUDO SRRU GHSHQG

PAGE 172

6KLIWLQJ FXOWLYDWRUV WUDGLWLRQDOO\ OHDYH WKH SDOP IRUHVWV LQWDFW RU FDUU\ RXW RQO\ PRGHUDWH WKLQQLQJ )LJ f $ KLJK SURSRUWLRQ FD bf RI DERYHJURXQG GU\ PDWWHU SURGXFWLRQ LQ EDEDVVX VWDQGV LV DOORFDWHG WR OHDYHV ZKLFK SUREDEO\ DFFRXQWV IRU WKH SUHVHUYDWLRQ RI WKH SDOP IRUHVWV XQGHU VKLIWLQJ FXOWLYDWLRQ $FFRUGLQJ WR VKLIWLQJ FXOWLYDWRUV RQ EDEDVn VXGRPLQDWHG VLWHV LQ 0DUDQKDR D PLQLPXP IDOORZ SHULRG RI \U UHn VXOWV LQ VXIILFLHQW DFFXPXODWLRQ RI IXHO LQ WKH IRUP RI OHDYHV WR SURn GXFH WKH KRW ILUH QHFHVVDU\ IRU QXWULHQW UHF\FOLQJ DQG ZHHG FRQWRO 1\H m DQG *UHHQODQG f 7KH PLQLPXP IDOORZ SHULRG FRLQFLGHV ZLWK WKH DYHUDJH OHDI OLIHVSDQ LQ EDEDVVX \Uf EHFDXVH OHDI SURGXFWLRQ DSSHDUV WR UHPDLQ FRQVWDQW DIWHU GHIROLDWLRQ VHH &KDSWHU f SDOPV UHTXLUH WKLV SHULRG WR UHJDLQ WKHLU IXOO FRPSOHPHQW RI OHDYHV ,Q VKRUW EDEDVVX LV H[WUHPHO\ ZHOO DGDSWHG WR WUDGLWLRQDO V\VWHPV RI VKLIWLQJ FXOWLYDWLRQ 7KH SDOP IRUHVWV IRUP VSRQWDQHRXVO\ UHTXLUH OLWWOH RU QR PDLQWHQDQFH H[KLELW UHODWLYHO\ IHZ SHVW SUREOHPV VHH %RQGDU Ef FRPSDUHG WR SODQWDWLRQ PRQRFXOWXUHV VXFK DV RLO SDOP +DUWOH\ f DQG FRFRQXW &KLOG f} SURPRWH VLWH UHFRYHU\ GXULQJ WKH IDOORZ VHH )XUOH\ f DQG DUH DOUHDG\ WLJKWO\ LQWHJUDWHG LQWR PDUNHW DQG VXEVLVWHQFH HFRQRPLHV /LPLWLQJ )DFWRUV ,Q UHFHQW \HDUV FRQYHUVLRQ RI IRUPHUO\ FXOWLYDWHG ODQGV WR FDWWOH SDVWXUHV KDV RFFXUUHG DW DQ DFFHOHUDWLQJ UDWH LQ IURQWLHU DUHDV RI %UD]LO :LWK LQFUHDVHG ODQG SUHVVXUHV RQ VKLIWLQJ FXOWLYDWRUV DYHUDJH IDOORZ SHULRGV LQ WKH VWDWH RI 0DUDQKDR GHFOLQHG IURP \U LQ WR \U LQ LQ WKH VWXG\ UHJLRQ WKH FHQWUDO 0HDULP 9DOOH\f WKH

PAGE 173

)LJXUH 6KLIWLQJ FXOWLYDWLRQ VLWH MXVW SULRU WR EXUQ DW /DJR 9HUGH )LJXUH &OHDUFXW RI SDOP IRUHVW SULRU WR SDVWXUH HVWDEOLVKPHQW QHDU %DFDEDO 0DUDQK£R

PAGE 174

ILJXUH DW WKH ODWWHU GDWH ZDV \U ,%*( f 7KH UHGXFHG IXHO DFFXPXODWHG GXULQJ VKRUWHQHG IDOORZV LV LQGXFLQJ VKLIWLQJ FXOWLYDWRUV WR WKLQ WKH VWDQGV PRUH LQWHQVLYHO\ LQ RUGHU WR JHQHUDWH VXIILFLHQW IXHO IRU KRW ILUHV :LWK GHJUDGDWLRQ RI VLWHV XVHG IRU VKLIWLQJ FXOWLYDWLRQ WKH UXUDO SRRU DUH EHFRPLQJ LQFUHDVLQJO\ GHSHQGHQW RQ SDVWXUH VWDQGV RI EDEDVVX DV D FUXFLDO VRXUFH RI LQFRPH DQG VXEVLVWHQFH SURGXFWV 5DQFKHUV LQ WXUQ LQFUHDVLQJO\ SHUFHLYH WKH SHRSOH ZKR JDWKHU EDEDVVX IUXLWV DV LQWHUn IHULQJ ZLWK SDVWXUH PDQDJHPHQW )UXLW JDWKHUHUV DUH EODPHG IRU VWDUWLQJ ZLOGILUHV FXWWLQJ IHQFHV DQG OHDYLQJ EHKLQG IUXLW KXVNV WKDW GDPDJH WKH KRRYHV RI FDWWOH $QGHUVRQ DQG $QGHUVRQ f 3UREDEO\ DV D FRQVHn TXHQFH FOHDUFXWWLQJ RI EDEDVVX IRUHVWV E\ UDQFKHUV LV EHFRPLQJ LQn FUHDVLQJO\ FRPPRQ )LJ f 7KH FXUUHQW EUHDNGRZQ LQ ODQGXVH V\VWHPV DVVRFLDWHG ZLWK EDEDVVX LV RQO\ SDUW RI D ORQJ KLVWRU\ RI GLVDSSRLQWPHQW 7KH VKHHU H[WHQW RI WKH VWDQGV KDV XQGHUVWDQGDEO\ WHQGHG WR LQVSLUH RSWLPLVWLF HVWLPDWHV RI WKHLU SRWHQWLDO FRQWULEXWLRQ WR WKH QDWLRQDO DQG JOREDO HFRQRP\ 7R FLWH MXVW RQH RI PDQ\ H[DPSOHV LW ZDV UHFHQWO\ FDOFXODWHG WKDW IXOO LQGXVWULDO XWLOL]DWLRQ RI EDEDVVX IUXLWV ZRXOG UHVXOW LQ WKH DQQXDO SURGXFWLRQ RI [ A / RI HWKDQRO [ A W RI FKDUFRDO [ A W RI YHJHWDEOH RLO DQG [ A PA RI FRPEXVWLEOH JDVHV DOO WKLV ZRXOG EH HTXLYDOHQW WR WKH FRQVWDQW JHQHUDWLRQ RI [ A : RYHU D \U SHULRG 67, f 7KHVH HVWLPDWHV DSSHDU WR EH WHFKQRORJLFDOO\ IHDVLEOH GXH WR UHFHQW GHYHORSPHQW RI PDFKLQHU\ FDSDEOH RI SURFHVVLQJ WKH H[WUHPHO\ KDUG IUXLWV RI WKH SDOP 67, f

PAGE 175

7KH NH\ ERWWOHQHFN LQ WKH EDEDVVX HFRQRP\ LV DQG DOZD\V KDV EHHQ VXSSO\ *RQVDOYHV 9DOYHUGH %UDJD DQG 'LDV 67, f ,QFRUSRUDWLQJ WKH SDOPnV IXOO HQHUJHWLF SRWHQWLDO LQWR WKH PDUNHW HFRQRn P\ UHTXLUHV WKDW FHQWUDOL]HG LQGXVWULHV EH VXSSOLHG ZLWK D VWHDG\ IORZ RI UDZ PDWHULDOV LQ WKH IRUP RI IUXLWV 7KH UHTXLUHG FRQWLQXLW\ RI VXSSO\ LV LOOXVRU\ GXH WR WKH VHDVRQDO QDWXUH RI WKH KDUYHVW %HFDXVH RI UHODWLYHO\ ORZ \LHOGV SHU XQLW DUHDf§DV ZHOO DV WKH ORZ YDOXH SHU XQLW SURGXFWf§FRVWV RI WUDQVSRUWLQJ HQWLUH IUXLWV WR D FHQWUDOL]HG IDFLOLW\ TXLFNO\ EHFRPH SURKLELWLYH 7KXV WKH UHTXLUHG YROXPH RI VXSSO\ LV OLNHZLVH LOOXVRU\ )XUWKHUPRUH SHRSOH ZKR KDUYHVW EDEDVVX IUXLWV JHQHUDOO\ H[WUDFW WKH NHUQHOV PRVW RI ZKLFK DUH XOWLPDWHO\ VROG WR ORFDO YHJHWDEOH RLO LQGXVWULHV WKH IUXLW KXVNV DUH WKHQ XVHG WR PDNH FKDUFRDO 7KHVH SHRSOH DUH XQGHUVWDQGDEO\ UHOXFWDQW WR SDUW ZLWK ZKDW LV XVXDOO\ WKHLU H[FOXVLYH VRXUFH RI IXHO 7KXV WKHUH LV FRQVLGHUDEOH VRFLDO UHVLVWDQFH WR LQFRUSRUDWLQJ EDEDVVXnV IXOO HQHUJHWLF SRWHQWLDO LQWR WKH PDUNHW HFRQRP\ $QGHUVRQ DQG $QGHUVRQ f (YHQ WUDGLWLRQDO LQGXVWULHV EDVHG RQ YHJHWDEOH RLO DUH FKURQLFDOO\ SODJXHG E\ LQVXIILn FLHQW VXSSO\ ,Q DGGLWLRQ WKH KLJK FRVWV RI UDZ PDWHULDOf§ODUJHO\ GXH WR ODERULQWHQVLYH PDQXDO H[WUDFWLRQ RI WKH NHUQHOVf§UHGXFH EDEDVVXnV FRPSHWLWLYHQHVV LQ UHODWLRQ WR RWKHU YHJHWDEOH RLOV $V D UHVXOW PRVW RI WKH LQGXVWULHV GHSHQGHQW RQ EDEDVVX DUH FXUUHQWO\ LQ FULVLV DQG WKH FRQWLQXHG YLDELOLW\ RI WKH SDOP DV D VRXUFH RI PDUNHW SURGXFWV LV LQ GRXEW

PAGE 176

3RVVLEOH 6ROXWLRQV ,Q WKH JRYHUQPHQW RI 0DUDQK£R HVWDEOLVKHG D UHVHDUFK LQVWLn WXWH WKH ,QVWLWXWR (VWDGXDO GR %DEDSX ,1(%f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f VXFK VWUDLQV ZLOO XOWLPDWHO\ EH XWLOL]HG LQ LQWHQVLYHO\ FXOWLYDWHG SODQWDWLRQV ZKLFK ZLOO WKHQ IRUP WKH EDVLV IRU DQ HIILFLHQW SURGXFWLRQ V\VWHP FDSDEOH RI SURYLGLQJ FHQWUDOL]HG LQGXVWULHV ZLWK D VWHDG\ IORZ RI UDZ PDWHULDOV %DEDVVXnV KLJK JHQHWLF YDULDELOLW\ LQGLFDWHV WKDW WKH SRWHQWLDO IRU GHYHORSLQJ LPSURYHG VWUDLQV RI WKH SDOP DUH H[FHOOHQW 3UHVHQW DQQXDO IUXLW \LHOGV RI EDEDVVX VWDQGV LQ 0DUDQK£R DYHUDJH WKD $QRQ\PRXV f ZKLFK LV VLPLODU WR WKRVH RI ZLOG VWDQGV RI RLO SDOP LQ 1LJHULD =HYHQ f VHOHFWLRQ LQ WKH ODWWHUf§DV ZHOO DV PDQDJHPHQW VXEVLGLHVf§ KDYH LQFUHDVHG \LHOGV E\ D IDFWRU RI 3XUVHJORYH f /LNH RLO SDOP VH[ UDWLRV LQ EDEDVVX DSSHDU WR EH XQGHU JHQHWLF FRQWURO DQG WKXV PD\ EH DPHQDEOH WR VHOHFWLRQ /LNHZLVH JURZWK UDWHV RI EDEDVVX DUH H[WUHPHO\ VORZ LQ FORVHG IRUHVWV DOWKRXJK FRQVLGHUDEO\ KLJKHU RQ PRUH RSHQ VLWHV SDOPV PD\ VWLOO UHTXLUH PRUH WKDQ D GHFDGH WR EHJLQ IUXLW SURGXFWLRQ DFFRUGLQJ WR ORFDO LQIRUPDQWV 7KLV WLPH SHULRG FDQ SUREDE

PAGE 177

O\ EH UHGXFHG :LWKLQ SRSXODWLRQV RI EDEDVVX KDYH RFFDVLRQDOO\ REVHUYHG UHSURGXFWLYH VKRUWVWDWXUHG SDOP Pf WKDW ZHUH DSSDUHQWO\ SUHFRFLRXV ,QFUHDVHG SUHFRFLW\ ZDV VXFFHVVIXOO\ LQFRUSRUDWHG LQWR RLO SDOP E\ SURJUDPV RI VHOHFWLRQ DQG EUHHGLQJ +DUWOH\ f 2WKHU SRWHQWLDOO\ GHVLUDEOH WUDLWV VXFK DV LQFUHDVHG JHUPLQDWLRQ VXFFHVV RU HQKDQFHG UHVLVWDQFH WR GLVHDVHV PD\ DOVR EH DPHQDEOH WR JHQHWLF LPSURYHn PHQW )URP D EURDGHU SHUVSHFWLYH KRZHYHU GHYHORSPHQW RI EDEDVVX IRU LQWHQVLYH FXOWLYDWLRQ UHTXLUHV FORVHU VFUXWLQ\ KDYH WKUHH UHVHUYDn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ƒ

PAGE 178

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n DJHPHQW RI WKH VWDQGV WKDW DOUHDG\ H[LVW $ NH\ FRPSRQHQW RI VXFK PDQDJHPHQW LV PDLQWHQDQFH RI SDOP GHQVLWLHV VXLWDEOH IRU JURZWK RI RWKHU FURSV 0DWXUH SDOPV ZLWK D ORZ SURSRUWLRQ RI IHPDOH LQIORUHVFHQFHV VKRXOG EH SUHIHUHQWLDOO\ WKLQQHG &RPSHWLWLRQ IURP KLJK GHQVLWLHV RI MXYHQLOH SDOPV LV UHGXFHG E\ SHULRGLF FXWWLQJ RQ VLWHV XVHG IRU SHUPDn QHQW FURSSLQJ RU JUD]LQJ MXYHQLOHV FDQ EH HOLPLQDWHG E\ UHSHDWHG FXWn WLQJ RYHU VHYHUDO \HDUV RU E\ DSSOLFDWLRQ RI V\VWHPLF KHUELFLGHV 6HHGn OLQJ UHFUXLWPHQW FDQ EH FRQWUROOHG WKURXJK PRUH HIILFLHQW IUXLW KDUYHVWn LQJ SURFHGXUHV 5DWKHU WKDQ JDWKHU IDOOHQ IUXLWV IURP WKH JURXQG PDWXUH IUXLW EXQFKHV VKRXOG EH KDUYHVWHG IURP WKH FURZQV ZLWK SUXQLQJ SROHV WKLV SUDFWLFH UHGXFHV WKH LQFLGHQFH RI ORVW IUXLWV DV ZHOO DV LQIHVWDWLRQ E\ EUXFKLG EHHWOHV ZKLFK RQO\ DWWDFN IUXLWV RQ WKH JURXQG %HFDXVH RI LWV ORZ PDLQWHQDQFH UHTXLUHPHQWV EDEDVVX KDV H[FHOOHQW SRWHQWLDO IRU LQWURGXFWLRQ LQWR H[WHQVLYH ODQGXVH V\VWHPV ([SHULPHQn WDO UHVXOWV GHVFULEHG LQ &KDSWHU LQGLFDWH WKDW EDEDVVX FRPSHWHV HIIHFn WLYHO\ XQGHU D ZLGH UDQJH RI FOLPDWLF FRQGLWLRQV DQG WKXV FDQ EH VXFn FHVVIXOO\ LQWURGXFHG LQ DUHDV RXWVLGH RI LWV FXUUHQW UDQJH 3ODQWLQJ RI

PAGE 179

GHVLUDEOH VWUDLQV RI EDEDVVX PD\ VRRQ EH IHDVLEOH 7KH SRWHQWLDO H[LVWV IRU LQGXFLQJ VHOISROOLQDWLRQ LQ SDOPV ZLWK SURPLVLQJ FKDUDFWHULVWLFV VXFK DV D KLJK SURSRUWLRQ RI IHPDOH LQIORUHVFHQFHV RU SUHFRFLW\ :KHWKHU RU QRW VHOILQJ FDQ EH LQGXFHG LQ EDEDVVX LV QRW SUHVHQWO\ NQRZQ EXW LQFRPSDWLELOLW\ GRHV QRW RFFXU LQ RWKHU FXOWLJHQV VXFK DV FRFRQXW DQG RLO SDOP DQG WKH WHFKQLTXHV IRU DUWLILFLDO VHOILQJ LQ WKHVH SDOPV DUH VLPSOH &KLOG } +DUWOH\ f 'HVLUDEOH VWUDLQV RI EDEDVVX FRXOG EH LQWURGXFHG HLWKHU XQGHU VWDQGV WKDW SUHVHQWO\ H[LVW RU RQ GHIRUHVWHG VLWHV :LWK SURSHU ILUH FRQWURO XQGHUSODQWLQJ ZRXOG EH HVSHFLDOO\ DSSURSULDWH LQ SDVWXUHV IROn ORZLQJ HOLPLQDWLRQ RI VHHGOLQJV DQG MXYHQLOHV 6XFK FRQWUROOHG UHJHQn HUDWLRQ ZRXOG HQDEOH UDQFKHUV WR LPSURYH WKHLU VWDQGV JHQHWLFDOO\ ZKLOH PDLQWDLQLQJ FRQVWDQW IRUHVW FRYHU FRQVLGHUDEOH SUHFHGHQW H[LVWV IRU FRPELQLQJ SODQWDWLRQV RI SDOPV DQG SDVWXUHV HJ 3OXFNQHWW f $OWHUQDWLYHO\ GHVLUDEOH VWUDLQV RI EDEDVVX FRXOG EH XVHG IRU UHIRUHVWDn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n FLHQWO\ SURGXFH WKHLU RZQ VXEVLVWHQFH SURGXFWV VXFK DV RLO IXHO DQG

PAGE 180

IHHGFDNH WKH H[FHVV FRXOG WKHQ EH VROG DV VHPLSURFHVVHG PDWHULDOV JUHDWO\ UHGXFLQJ WKH FXUUHQW FRVWV RI WUDQVSRUWLQJ EXON\ IUXLWV 7KH SRWHQWLDO UROH RI EDEDVVX LQ UXUDO FRPPXQLWLHV LV LOOXVWUDWHG E\ WKH LQWHUGHSHQGHQFH RI SHRSOH DQG ZLOG VWDQGV RI WKH ORQWDU SDOP %RUDVVXV VXQGDLFXVf LQ LVODQGV RI ,QGRQHVLD )R[ f 'HVSLWH D GU\ FOLPDWH DQG QXWULHQWSRRU VRLOV KLJK SRSXODWLRQ GHQVLWLHV DUH PDLQn WDLQHG RQ WKHVH LVODQGV E\ SURGXFWLYH V\VWHPV RI JDWKHULQJ VHPLSHUPDn QHQW DJULFXOWXUH DQG DQLPDO KXVEDQGU\ HLWKHU GLUHFWO\ RU LQGLUHFWO\ EDVHG RQ WKH SDOP /LNH WKH ORQWDU EDEDVVX LV D VRXUFH RI HVVHQWLDO VXEVLVWHQFH SURGXFWV DQG LV WKH EDVLV IRU D ZLGHVSUHDG FRWWDJH LQGXVWU\ /LNHZLVH EDEDVVX PDLQWDLQV G\QDPLFf§DOWKRXJK FXUUHQWO\ OHVV H[SORLWHGf§ LQWHUDWLRQV ZLWK V\VWHPV RI DJULFXOWXUH DQG DQLPDO KXVEDQGU\ )XOO XWLOL]DWLRQ RI WKH SDOP E\ UXUDO FRPPXQLWLHV FRXOG RSHQ WKH ZD\ WR VHOI VXIILFLHQF\ DQG HFRORJLFDO VWDELOLW\ ZKLOH DW WKH VDPH WLPH LQWHJUDWLQJ WKHVH FRPPXQLWLHV LQWR WKH ODUJHU PDUNHW HFRQRP\

PAGE 181

5()(5(1&(6 &,7(' $EUDKDP VRQ : 3DWWHUQV RI UHVRXUFH DOORFDWLRQ LQ ZLOGI ORZHU SRSXODWLRQV RI ILHOGV DQG ZRRGV $PHULFDQ -RXUQDO RI %RWDQ\ $EUHX 6 ) &RFR EDEDSX H R SUREOHPD GR FRPEXVWLYHO 6HFRQG HGLWLRQ ,QVWLWXWR 1DFLRQDO GH 7HFQRORJD 5LR GH -DQHLUR %UD]LO $QGHUVRQ $ % DQG ( 6 $QGHUVRQ 3HRSOH DQG WKH SDOP IRUHVW ELRORJ\ DQG XWLOL]DWLRQ RI EDEDVVX IRUHVWV LQ 0DUDQKDR %UD]LO )LQDO UHSRUW WR 86'$ )RUHVW 6HUYLFH &RQVRUWLXP IRU WKH 6WXG\ RI 0DQnV 5HODWLRQVKLS ZLWK WKH *OREDO (QYLURQPHQW :DVKLQJWRQ 'LVWULFW RI &ROXPELD 86$ DQG : : %HQVRQ 2Q WKH QXPEHU RI WUHH VSHFLHV LQ $PD]RQLDQ IRUHVWV %LRWURSLFD $QRQ\PRXV 0DSHDPHQWR GDV RFRUUQFLDV H SURVSHFSDR GR SRWHQFLDO DWXDO GR EDEDSX QR (VWDGR GR 0DUDQKDR &RPSDQKLD GH 3HVTXLVD H $SURYHLWDUQHQWR GH 5HFXUVRV 1DWXU£LV &23(1$7f DQG )XQGDSDR ,QVWLn WXWR (VWDGXDO GR %DEDSX ,1(%f 6DR /XLV %UD]LO %DOLFN 0 7KH ELRORJ\ DQG HFRQRPLFV RI WKH 2HQRFDUSXV -HVVHQLD 3DOPDHf FRPSOH[ 'LVVHUWDWLRQ +DUYDUG 8QLYHUVLW\ &DPEULGJH 0DVVDFKXVHWWV 86$ $ % $QGHUVRQ DQG 0 ) GD 6LOYD 3DOP WD[RQRP\ LQ %UD]LOLDQ $PD]RQLD WKH VWDWH RI V\VWHPDWLF FROOHFWLRQV LQ UHJLRQDO KHUEDULD %ULWWRQLD r %DQQLVWHU % $ (FRORJLFDO OLIH F\FOH RI (XWHUSH JORERVD *DHUWQ 3DJHV %% LQ + 7 2GXP DQG 5 ) 3LGJHRQ HGLWRUV $ WURSLFDO UDLQ IRUHVW $WRPLF (QHUJ\ &RPPLVVLRQ :DVKLQJWRQ 'LVWULFW RI &ROXPELD 86$ %DUERVD 5RGULJXHV 3ODQWDV QRYDV FXOWLYDGDV QR -DUGLP %RW£QLFR GR 5LR GH -DQHLUR 5LR GH -DQHLUR %UD]LOf 3DOPDH 0DWWRJURVVHQVHV QRYDH YHO PLQXV FRJQLWDH 5LR GH -DQHLUR %UD]LO 6HUWXP SDOPDUXP %UDVLOLHQVLXP %UXVVHOV %HOJLXP

PAGE 182

%HDFK + ,Q SUHVV 7KH UHSURGXFWLYH ELRORJ\ RI WKH SHMLED\H SDOP %DFWULV JDVLSDHVf DQG %cB SRUVFKLDQD LQ WKH $WODQWLF ORZODQGV RI &RVWD 5LFD 3ULQFLSHV %HDUG 6 } &OLPD[ YHJHWDWLRQ LQ WURSLFDO $PHULFD (FRORJ\ %ODFN & $ 0HWKRGV RI VRLO DQDO\VLV 3DUW &KHPLFDO DQG PLFURELRORJLFDO SURSHUWLHV $PHULFDQ 6RFLHW\ RI $JURQRP\ 0DGLVRQ :LVFRQVLQ 86$ %RQGDU 1RWDV ELROJLFDV VREUH %UXFKGHRV REVHUYDGRV QR %UDVLO $UTXLYRV GR ,QVWLWXWR %LROJLFR GH 9HJHWDS£R 5LR GH -DQHLUR %UD]LOf r D 1RYD HVSHFLH GH 2UELJQ\D SURGXWRUD GR OHR GH EDEDFX $UTXLYRV GR -DUGLP %RW£QLFR 5LR GH -DQHLUR %UD]LOf E $ HQWRPRORJD GH EDEDVVX %ROHWLP GR ,QVWLWXWR GH %LRORJLD 6DOYDGRU %UD]LOf f 1RYR JQHUR H QRYD HVSHFLH GH SDOPHLUDV GD WULER $WWDOHLQH $UTXLYRV GR -DUGLP %RW£QLFR 5LR GH -DQHLUR %UD]LOf 3DOPHLUDV GR %UDVLO ,QVWLWXWR GH %RW£QLFD GH 6DR 3DXOR 6DR 3DXOR %UD]LO %UDGIRUG ) DQG & & 6PLWK 6HHG SUHGDWLRQ DQG VHHG QXPEHU LQ 6FKHHOHD SDOP IUXLWV (FRORJ\ %UDJD + & DQG & 'LDV $VSHFWRV VRFLRHFRQPLFRV GR EDEDVn VX ,QVWLWXWR GH OHRV 0LQLVWULR GH $JULFXOWXUD 5LR GH -DQHLUR %UD]LO %UHZHU 5 DQG 0 7 0F&DQQ /DERUDWRU\ DQG ILHOG PDQXDO RI HFRORJ\ 6DXQGHUV &ROOHJH 3XEOLVKLQJ 3KLODGHOSKLD 3HQQV\OYDQLD 86$ %XOORFN 6 + 'HPRJUDSK\ RI DQ XQGHUJURZWK SDOP LQ OLWWRUDO &DPHURRQ %LRWURSLFD 1RWHV RQ WKH SKHQRORJ\ RI LQIORUHVFHQFHV DQG SROn OLQDWLRQ LQ VRPH UDLQ IRUHVW SDOPV LQ &RVWD 5LFD 3ULQFLSHV %XUUHW 0 'LH 3DOPHQJDWWXQJHQ 2UELJQ\D $WWDOHD 6FKHHOHD XQG 0D[LPLOLDQD 1RWL]EODWW R

PAGE 183

$WWDOHD FRKXQH 0DUW ZLUNOLFK HLQH 4UELJQ\D 1RWL] EODWW 3DOPDH %UDVLOLHQVLV 1RWL]EODWW %XUWW % $ UHFRUG RI IUXLWV DQG VHHGV GLVSHUVHG E\ PDPPDOV DQG ELUGV IURP WKH 6LQJLGD 'LVWULFW RI 7DQJDQ\LND 7HUULWRULW\ -RXUQDO RI (FRORJ\ &DQQHOO 0 5 :RUOG IRUHVW ELRPDVV DQG SULPDU\ SURGXFWLRQ GDWD $FDGHPLF 3UHVV /RQGRQ (QJODQG &KLOG 5 &RFRQXWV 6HFRQG HGLWLRQ /RQJPDQ /RQGRQ (QJODQG &KXUFKLOO % + + -RKQ 3 'XQFDQ DQG $ & +RGVRQ /RQJ WHUP HIIHFWV RI GHIROLDWLRQ RI DVSHQ E\ WKH IRUHVW WHQW FDWHUn SLOODU (FRORJ\ &RQQHOO + 'LYHUVLW\ LQ WURSLFDO UDLQ IRUHVWV DQG FRUDO UHHIV 6FLHQFH &RRN 5 ( DQG ( ( /\RQV 7KH ELRORJ\ RI 9L ROD ILPEULDWXOD LQ D QDWXUDO GLVWXUEDQFH (FRORJ\ &RUOH\ 5 + 9 +DUGRQ DQG % :RRG 2LO SDOP UHn VHDUFK (OVHYLHU 6FLHQWLILF 3XEOLVKLQJ &RPSDQ\ $PVWHUGDP 7KH 1HWKHUODQGV &RUQHU ( + 7KH QDWXUDO KLVWRU\ RI SDOPV 8QLYHUVLW\ RI &DOLIRUQLD 3UHVV %HUNHOH\ &DOLIRUQLD 86$ &RVWD /LPD $ 0 4XDUWR FDW£ORJR GRV LQVHWRV TXH YLYHP DV SODQWDV GR %UDVLO VHXV SDUDVLWDV H SUHGDGRUHV 'HSDUWDPHQWR GH 'HIHVD H ,QVSHS£R $JURSHFXDULD 0LQLVWHULR GD $JULFXOWXUD 5LR GH -DQHLUR %UD]LO 'UDQVILHOG 3LQDQJD GHLV WDQ WKD D QHZ VSHFLHV ZLWK KLGGHQ IORZHUV 3ULQFLSHV 'UXGH 3DOPDH ,Q & ) 3 YRQ 0DUWLXV HGLWRU )ORUD %UDVLOLHQVLV 0XQLFK *HUPDQ\ (0%5$3$ 0DSD HVTXHP£WLFR GRV VRORV GDV UHJLRHV QRUWH PHLR QRUWH H FHQWURRHVWH GR %UDVLO WH[WR H[SOLFDWLYR 0LQLVWHULR GD $JULFXOWXUD %ROHWLP 7FQLFR 5LR GH -DQHLUR %UD]LO 0DSD GH VRORV GR %UDVLO 6HUYLSR 1DFLRQDO GH /HYDQWD PHQWR H &RQVHUYDS£R GH 6RORV 5LR GH -DQHLUR %UD]LO

PAGE 184

(VFROD 7FQLFD )HGHUDO GR 0DUDQKDR %DEDSX LQGXVWULDQ]DS£R WRWDO 6DR /XLV %UD]LO (VVLJ ) % 2EVHUYDWLRQV RI SROOLQDWLRQ LQ %DFWULV 3ULQFLSHV (ZHO 6 *OLHVVPDQ 0 $PDGRU ) %HQHGLFW & %HULVK 5 %HUPXGH] % %URZQ $ 0DUWLQH] 5 0LUDQGD DQG 1 3ULFH /HDI DUHD OLJKW WUDQVPLVVLRQ URRWV DQG OHDI GDPDJH LQ QLQH WURSLFDO SODQW FRPPXQLWLHV $JUR(FRV\VWHPV )DHJUL DQG / YDQ GHU 3LMO 7KH SULQFLSOHV RI SROOLQDWLRQ HFRORJ\ 7KLUG HGLWLRQ 3HUJDPRQ 3UHVV 2[IRUG (QJODQG )HUUL 0 (FRORJLD WHPDV H SUREOHPDV EUDVLOHLURV (GLWRUD 8QLYHUVLGDGH GH 6DR 3DXOR 6DR 3DXOR %UD]LO 9HJHWDS£R EUDVLOHLUD (GLWRUD 8QLYHUVLGDGH GH 6DR 3DXOR 6DR 3DXOR %UD]LO )RQVHFD ( 7 EDEDVVX $WWDOHD VSHFLRVD 0DUW 2UELJQ\D PDUWLDQD %DUE 5RGUf 0LQLVWULR GD $JULFXOWXUD ,QGVWULD H &RPHUFLR 5LR GH -DQHLUR %UD]LO )R[ +DUYHVW RI WKH SDOP +DUYDUG 8QLYHUVLW\ 3UHVV &DPEULGJH 0DVVDFKXVHWWV 86$ )UHXQG 5 DQG 5 & /LWWHOO 6$6 IRU OLQHDU PRGHOV $ JXLGH WR WKH $129$ DQG */0 SURFHGXUHV 6$6 ,QVWLWXWH ,QFRUSRUDWHG *DU\ 1RUWK &DUROLQD 86$ )XUOH\ 3 $ 7KH VLJQLILFDQFH RI WKH FRKXQH SDOP 2UELJQ\D FRKXQH 0DUWf 'DKOJUHQ RQ WKH QDWXUH DQG GHYHORSPHQW RI WKH VRLO SURILOH %LRWURSLFD *HQWU\ $ + 3DWWHUQV RI QHRWURSLFDO SODQW GLYHUVLW\ (YRn OXWLRQDU\ %LRORJ\ *LYQLVK 7 2Q WKH DGDSWLYH VLJQLILFDQFH RI FRPSRXQG OHDYHV ZLWK SDUWLFXODU UHIHUHQFH WR WURSLFDO WUHHV 3DJHV LQ 3 % 7RPOLQVRQ DQG 0 + =LPPHUPDQQ HGLWRUV 7URSLFDO WUHHV DV OLYLQJ V\VWHPV &DPEULGJH 8QLYHUVLW\ 3UHVV &DPEULGJH (QJODQG *ODVVPDQ 6 ) 3UHOLPLQDU\ WD[RQRPLF VWXGLHV LQ WKH SDOP JHQXV 2UELJQ\D 0DUW 3K\WRORJLD &RUUHFWLRQV DQG FKDQJHV LQ UHFHQW SDOP DUWLFOHV SXEOLVKHG LQ 3K\WRORJLD 3K\WRORJLD

PAGE 185

*RQVDOYHV $ EDEDSX FRQVLGHUDSHV FLHQWILFDV WFQLFDV H HFRQPLFDV 6HUYLFR GH ,QIRUPDS£R $JUFROD 0LQLVWHULR GH $JULn FXOWXUD 5LR GH -DQHLUR %UD]LO +DPLOWRQ : 7KH PRXOGLQJ RI VHQHVFHQFH E\ QDWXUDO VHOHFWLRQ -RXUQDO RI 7KHRUHWLFDO %LRORJ\ +DUFRPEH 3 $ 7KH LQIOXHQFH RI IHUWLOL]DWLRQ RQ VRPH DVSHFWV RI VXFFHVVLRQ LQ D KXPLG WURSLFDO IRUHVW (FRORJ\ +DUSHU / 3RSXODWLRQ ELRORJ\ RI SODQWV $FDGHPLF 3UHVV 1HZ
PAGE 186

E 6LQRSVH SUHOLPLQDU GR FHQVR GHPRJU£ILFR 0DUDQKDR )XQ GDS£R ,QVWLWXWR %UDVLOHLUR GH *HRJUDILD H (VWDWLVWLFD 5LR GH -DQHLUR %UD]LO ,37 $Q£OLVH WHFQROJLFD HFRQPLFD H VRFLDO GR DSURYHLWDPHQWR LQWHJUDO GR FRFR GH EDEDFX 7ZR YROXPHV ,QVWLWXWR GH 3HVTXLVDV 7HFQROJLFDV GR (VWDGR GH 6DR 3DXOR 6DR 3DXOR %UD]LO ,]DZD DQG $ 0L]XQR 3DOPIUXLW FUDFNLQJ EHKDYLRU RI ZLOG EODFNFDSSHG FDSXFKLQ &HEXV DSHOODf 3ULPDWHV -DFNVRQ } &U\SWRJHDO JHUPLQDWLRQ DQG RWKHU VHHGOLQJ DGDSWDn WLRQV WR WKH EXUQLQJ RI YHJHWDWLRQ LQ VDYDQQD UHJLRQV 7KH RULJLQ RI WKH S\URSK\WLF KDELW 1HZ 3K\WRORJLVW V -DQ]HQ + 7KH IDWH RI 6FKHHOHD URVWUDWD IUXLWV EHQHDWK WKH SDUHQW WUHH SUHGLVSHUVDO DWWDFN E\ %UXFKLGV 3ULQFLSHV f D 5HGXFWLRQ RI VHHG SUHGDWLRQ RQ %DXKLQLD SDXOHWLD /HJXPLQRVDHf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cB SFDULf %LRWURSLFD .LUD 7 3ULPDU\ SURGXFWLRQ RI IRUHVWV 3DJHV LQ 3 &RRSHU HGLWRU 3KRWRV\QWKHVLV DQG SURGXFWLYLW\ LQ GLIIHUHQW HQYLURQPHQWV &DPEULGJH 8QLYHUVLW\ 3UHVV &DPEULGJH (QJODQG .XKOPDQQ ( 9HJHWDF£R 3DJHV LQ ,%*( *HRJUDILD GR %UDVLO 5HJL£R QRUGHVWH )XQGDS£R ,QVWLWXWR %UDVLOHLUR GH *HRn JUDILD H (VWDWLVWLFD 5LR GH -DQHLUR %UD]LO /HINRYLWFK / 3 7KH VWXG\ RI SRSXODWLRQ JURZWK LQ RUJDQLVPV JURXSHG E\ VWDJHV %LRPHWULFV

PAGE 187

/RRPLV 5 6 DQG 3 $ *HUDNLV 3URGXFWLYLW\ RI DJULFXOWXUDO HFRV\VWHPV 3DJHV LQ 3 &RRSHU HGLWRU 3KRWRV\QWKHVLV DQG SURGXFWLYLW\ LQ GLIIHUHQW HQYLURQPHQWV &DPEULGJH 8QLYHUVLW\ 3UHVV &DPEULGJH (QJODQG 0DJXLUH % 2Q WKH IORUD RI WKH *XD\DQD +LJKODQG %LRWURSLFD 0DLVVXURZ ( 7KH UROH RI ILUH LQ WKH SHUSHWXDWLRQ RI YLUJLQ IRUHVWV LQ QRUWKHUQ :LVFRQVLQ -RXUQDO RI )RUHVWU\ 0DUNOH\ 6 7KH EDEDVVX RLO SDOP RI %UD]LO (FRQRPLF %RWDQ\ 0DUNV 3 / 7KH UROH RI SLQ FKHUU\ 3UXQXV SHQQV\OYDQLFD /f LQ WKH PDLQWHQDQFH RI VWDELOLW\ LQ QRUWKHUQ KDUGZRRG HFRV\VWHPV (FRn ORJLFDO 0RQRJUDSKV 0DUWLXV & ) 3 YRQ +LVWRULD QDWXUDOLV SDOPHUXP r 0XQLFK *HUPDQ\ 3DOPHWXP 2UELJQLDQXP ,Q $ Gn2UELJQ\ HGLWRU 9R\DJH GDQV ,f$PHULTXH PHULGLRQDOH f 3DULV )UDQFH 0F.H\ 7KH HFRORJ\ RI FRHYROYHG VHHG GLVSHUVDO V\VWHPV 3DJHV LQ / ( *LOEHUW DQG 3 + 5DYHQ HGLWRUV &RHYRn OXWLRQ RI DQLPDOV DQG SODQWV 8QLYHUVLW\ RI 7H[DV 3UHVV $XVWLQ 7H[DV 86$ 0HQGHV $ 0 GH & DQG % &DULRFD %DEDSX 9ROXPH LQ (VWXGR LQWHJUDGR GR XVR SRWHQFLDO GH ELRPDVVD SDUD ILQV HQHUJWLFRV QR %UDVLO )RUWDOH]D %UD]LO 0LPHRJUDSKHG 0RRUH + ( D 3DOPV LQ WKH WURSLFDO IRUHVW HFRV\VWHPV RI $IULFD DQG 6RXWK $PHULFD 3DJHV LQ % 0HJJHUV ( 6 $\HQVX DQG : 'XFNZRUWK HGLWRUV 7URSLFDO IRUHVW HFRV\VWHPV RI $IULFD DQG 6RXWK $PHULFD D FRPSDUDWLYH UHYLHZ 6PLWKVRQLDQ ,QVWLWXWLRQ 3UHVV :DVKLQJWRQ 'LVWULFW RI &ROXPELD 86$ E 7KH PDMRU JURXSV RI SDOPV DQG WKHLU GLVWULEXWLRQ *HQWHV +DUEDULXP (QGDQJHUPHQW DW WKH VSHFLILF DQG JHQHULF OHYHO LQ SDOPV 3DJHV LQ 7 3UDQFH DQG 7 6 (OLDV HGLWRUV ([WLQFWLRQ LV IRUHYHU WKH VWDWXV RI WKUHDWHQHG DQG HQGDQJHUHG SODQWV RI WKH $PHULFDV 7KH 1HZ
PAGE 188

DQG 1 : 8KO 0DMRU WUHQGV RI HYROXWLRQ LQ SDOPV %RWDQLFDO 5HYLHZ 0RUD 8USL DQG ( 0 6ROLV 3ROLQL]DFLQ HQ %DFWULV JDVLSDHV + % 3DOPDHf 5HYLVWD %LRORJLD 7URSLFDO 0\HUV 5 / 7KH HFRORJ\ RI ORZ GLYHUVLW\ SDOP VZDPSV QHDU 7RUWXJXHUR &RVWD 5LFD 'LVVHUWDWLRQ 8QLYHUVLW\ RI )ORULGD *DLQHVYLOOH )ORULGD 86$ 1J 6 6 7KDPERR DQG 3 GH 6RX]D 1XWULHQW FRQWHQWV RI RLO SDOPV LQ 0DOD\D ,, 1XWULHQWV LQ YHJHWDWLYH WLVVXHV 0DOD\VLDQ $JULFXOWXUDO -RXUQDO 1\H 3 + DQG *UHHQODQG 7KH VRLO XQGHU VKLIWLQJ FXOWLn YDWLRQ &RPPRQZHDOWK $JULFXOWXUDO %XUHDX[ )DUQKDP 5R\DO %XFNV (QJODQG 3LQHUR DQG 6DUXNK£Q 5HSURGXFWLYH EHKDYLRXU DQG LWV LQGLYLGXDO YDULDELOLW\ LQ D WURSLFDO SDOP $VWURFDU\XP PH[LFDQXP -RXUQDO RI (FRORJ\ 6DUXNKDQ DQG 3 $OEHUGL 7KH FRVWV RI UHSURn GXFWLRQ LQ D WURSLFDO SDOP $VWURFDU\XP PH[LFDQXP -RXUQDO RI (FRORJ\ 3OXFNQHWW / 0DQDJLQJ SDVWXUHV DQG FDWWOH XQGHU FRFRQXWV :HVWYLHZ 3UHVV %RXOGHU &RORUDGR 86$ 3XUVHJORYH g 7URSLFDO FURSV 0RQRFRW\OHGRQV /RQJPDQ /RQGRQ (QJODQG 3\NH 0 7HFKQRORJLFDO HDWLQJ RU ZKHUH GRHV WKH ILVKILQJHU SRLQW" -RKQ 0XUUD\ /RQGRQ (QJODQG 5DWKFNH % ,Q SUHVV 3DWWHUQV RI IORZHULQJ SKHQRORJLHV WHVWn DELOLW\ DQG FDXVDO LQIHUHQFH XVLQJ D UDQGRP PRGHO ,Q 5 6WURQJ 6LPEHUORII DQG / $EHOH HGLWRUV (FRORJLFDO FRPn PXQLWLHV FRQFHSWXDO LVVXHV DQG WKH HYLGHQFH 3ULQFHWRQ 8QLYHUVLW\ 3UHVV 3ULQFHWRQ 1HZ -HUVH\ 86$ 5HHV $ 5 $Q DQDO\VLV RI JURZWK RI RLO SDOP VHHGOLQJV LQ IXOO GD\OLJKW DQG LQ VKDGH $QQDOV RI %RWDQ\ 5LFKDUGV 3 : 7KH WURSLFDO UDLQ IRUHVW &DPEULGJH 8QLYHUVLW\ 3UHVV &DPEULGJH (QJODQG 5L]]LQL & 7 6REUH D GLVWLQS£R H GLVWULEXLSDR GH GXDV HVSHFLHV GH EDEDSX 2UELJQ\Df 5HYLVWD %UDVLOHLUD GH *HRJUDILD

PAGE 189

6DUXNK£Q 6WXGLHV RQ WKH GHPRJUDSK\ RI WURSLFDO WUHHV 3DJHV LQ 3 % 7RPOLQVRQ DQG 0 + =LPPHUPDQQ HGLWRUV 7URSLn FDO WUHHV DV OLYLQJ V\VWHPV &DPEULGJH 8QLYHUVLW\ 3UHVV &DPn EULGJH (QJODQG 'HPRJUDSKLF SUREOHPV LQ WURSLFDO V\VWHPV 3DJHV LQ 7 6ROEULJ HGLWRU 'HPRJUDSK\ DQG HYROXWLRQ LQ SODQW SRSXODWLRQV 8QLYHUVLW\ RI &DOLIRUQLD 3UHVV %HUNHOH\ &DOLIRUQLD 86$ 6KRGW / / DQG : $ 0LWFKHOO 7KH SROOLQDWLRQ RI &RFRV QXFIHUD / LQ +DZDLL 7URSLFDO $JULFXOWXUH 7ULQLGDGf V f§ 6LHJHO 6 1RQSDUDPHWULF VWDWLVWLFV IRU WKH EHKDYLRUDO VFLHQFHV 0F*UDZ+LOO 1HZ
PAGE 190

7DKHU 0 0 DQG 5 & &RRNH 6KDGHLQGXFHG GDPSLQJRII LQ FRQLIHU VHHGOLQJV (IIHFWV RI UHGXFHG OLJKW LQWHQVLW\ RQ LQIHFn WLRQ E\ QHFURWURSKLF IXQJL 1HZ 3K\WRORJLVW 7RPOLQVRQ 3 % 0HDVXULQJ JURZWK UDWHV LQ SDOPV 3ULQFLSHV 6\VWHPDWLFV DQG HFRORJ\ RI WKH 3DOPDH $QQXDO 5HYLHZ RI (FRORJ\ DQG 6\VWHPDWLFV DQG 3 6RGHUKROP 7KH IORZHULQJ DQG IUXLWLQJ RI &RU\SKD HODWD LQ 6RXWK )ORULGD 3ULQFLSHV DQG 0 + =LPPHUPDQQ $QDWRP\ RI WKH SDOP 5KDSLV H[FHOVD ,,, -XYHQLOH SKDVH -RXUQDO RI WKH $UQROG $UERUn HWXP 7XUQHU 7 )DFLOLWDWLRQ DV D VXFFHVVLRQDO PHFKDQLVP LQ D URFN\ LQWHUWLGDO FRPPXQLW\ $PHULFDQ 1DWXUDOLVW 8KO 1 : DQG + ( 0RRUH -U 7KH SURWHFWLRQ RI SROOHQ DQG RYXOHV LQ SDOPV 3ULQFLSHV DQG + ( 0RRUH -U &RUUHODWLRQV RI LQIORUHVFHQFH IORZHU VWUXFWXUH DQG IORUDO DQDWRP\ ZLWK SROOLQDWLRQ LQ VRPH SDOPV %LRWURSLFD 9DOYHUGH *HRJUDILD HFRQPLFD H VRFLDO GR EDEDSX QR PHLR QRUWH 5HYLVWD %UDVLOHLUD GH *HRJUDILD YDQ GHU 3LMO / 3ULQFLSOHV RI GLVSHUVDO LQ KLJKHU SODQWV 6HFRQG HGLWLRQ 6SULQJHU9HUODJ 1HZ
PAGE 191

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n DQLVP %ODFNZHOO 6FLHQWLILF 3XEOLFDWLRQV 2[IRUG (QJODQG :KLWWDNHU 5 + &RPPXQLWLHV DQG HFRV\VWHPV 6HFRQG HGLWLRQ 0DFPLOODQ 3XEOLVKLQJ &RPSDQ\ 1HZ
PAGE 192

$33(1',; $ 63(&,(6 $%81'$1&( )5(48(1&< '20,1$1&( $1' ,03257$1&( ,1 +$ 2) 35,0$5< )25(67 /$*2 9(5'( 0$5$1+$2 S ( & ( 6 $%81'$14% VWHPVKDb )5(48(1&< b '20,1$1&( QUKD ,03257$1&( b 2UELHQYD PDUWLDQD %DUE 5RGUr (XFHQLD VG 6'RQGLDV PRPELUD 8UEr *XVWDYLD DXFXVWD /r 6ZDUW]OD '2OY'KYOOD $'& +LHURQLPD DOFKRUQLRLGHV )ULHUHf $OOHPr 3URWLXP KH'WD'KYOOXP $XEOf 0DUFK &YQRPHWUD PDUFLQDWD %HQWK &URWRQ RI GLDVLL 3LUHV 6D'LXP FXUX'LWD +XE &RUGLD VGr 5LFKHULD FUDQGLV 9DKO %RPED[ VG 7ULFKLOLD DXDGUL LXHD +%.

PAGE 193

$SSHQGL[ $ FRQWLQXHG 63(&,(6 $%81'$1&( VWHPVKD  )5(48(1&; b '20,1$1&( PKD  ,03257$1&( $VWURFDUYXP YXOFDUH 0DUWr 2P'KDOHD GLDQGULD / (VFKZHLOHUD RGRUD 3RH''f 0LHUV 3DOLFRXUHD VG 8QNQRZQ WUHH 1R 0DQLONDUD FI VXULQDUDHQVLV 0LDf 'XFNH -DFDUDWLD 6'LQRVD $'&r 'LDOLXP HXLDQHQVLV $XEO 6DQGZ 'XHXHWLD VG 3VYFKRWULD VG /XHKHD 'DQLFXODWD 0DUWr 6WHUFXOLD VGr 5LFKDUGHOOD VG &HGUHOD RGRUDWD / +LUWHOOD UDFHLQRVD /DPr

PAGE 194

$SSHQGL[ $ FRQWLQXHG 6( -/-//(&DVHDULD VYOYHVWULV 6Zr 6HJXLHULD PDFURSKYOOD %HQWK 2FRWHD VS &RUGLD VFDEULIROLD $'& 3URWLXP LQVLJQH 7U H[ 3,f (QJO 6FOHURORELXP SDQLFXODWXP 9RJ /LFDQLD FI DSWDOD ( 0H\HUf )ULWVFKr &RFRORED ODWLIROLD /DPr &DOYFRSKYOOXP VSUXFHDQXP %HQWK 6FKL]RORELXP DPD]RQLFXP +XE H[ 'XFNH &RXUDWDUL FI PDFURVSHUPD $& 6PLWK $VWURQLXP FI IUD[LQLIROLXP 6FKRWW &KULVWLDQD DIULFDQD '& &LWKDUH[YOXP FI PDFURSKYOOXP 3RHE 0RXULUL FI JXLDQHQVLV $XEO $%81'$1&( )5(48(1&< '20,1$1&( ,03257$1&( VWHPVKDb PI9KL b RR

PAGE 195

$SSHQGL[ $ FRQWLQXHG 63(&,(6 $%81'$1&( VWHPJKDb )5(49(1&;  '20,1$1&( QUKD b ,03257$1&( L $PDLRXD HXLDQHQVLV $XEO 8QNQRZQ 9LQH 1R +LPDWDQWKXV VXFXXED 6'UXFHf :RRG 0YUFLD FI IDOD[ 5LFKf '& &DVHDULD (ULVPD FI XQFLQDWXP :DUP 0DQLONDUD VGr 8QNQRZQ 9LQH 1R 3URWLXP 'DOOLGXP &XDWU 8QNQRZQ 9LQH 1R 7ULFKLOLD PLFUDQWKD %HQWK 3RXWHULD VG 6LPDED FHGUQ 3ODQFK 8QNQRZQ 9LQH 1R $QDFDUGLXP FI PLFURFDPXP 'XFNH

PAGE 196

$SSHQGL[ $ FRQWLQXHG 63(&,(6 $%81'$1&( VWHPVKDb )5(48(1&< b '20,1$1&( P9KDb ,03257$1&( L %URVLPXP XOHDUXP 0LOGEU r 'LRV'YURV FI HXLDQHQVLV $XEOf *XUNHr 9LWH[ VG $P'KLUUKR[ VXULQDPHQVLV (LFK 727$/ 1RWH $VWHULVNV LQGLFDWH VSHFLHV WKDW DOVR RFFXU RQ KD RI VHFRQGDU\ IRUHVW QHDUE\

PAGE 197

$33(1',; % 63(&,(6 $%81'$1&( )5(48(1&< '20,1$1&( $1' ,03257$1&( ,1 +$ 2) 6(&21'$5< )25(67 /$*2 9(5'( 0$5$1+$2 V S ( & ( 6 $%81'$1&( VWHPVKD  )5(48(1&< b '20,1$1&( P9KD ,03257$1&( b 2UELFQYD PDUWLDQD %DUE 5RGUr 6'RQGLDV PRPELP 8UEr &RUGLD VHOORZLDQD &KDPE &HFUR'LD FRQFRORU :LOOG $VWURFDUYXP YXOFDUH 0DUWr *XDWHULD FKUYVR'HWDOD 6WHQGf 0LD +LHURQLPD DOFKRUQLRLGHV )UHLUHf $OOHPr $RHLED WLERXUERU $XEO 6WHUFXOLD VGr 6D'LXP ODQFHRODWXP +XE &RFRORED ODWLIROLD /DPr -DFDUDWLD 6'LQRVD $'&r 9LWH[ FI IODYHQV +%. &DVHDULD VYOYHVWULV 6Zr

PAGE 198

$SSHQGL[ % FRQWLQXHG i S ( & ( 6 $P'$-&J VWHPVKD  )5(28(1&< L '20,1$1&( QUKD b ,03257$1&( b (QWHURORELXP FRQWRUWLVLOLDXDP 9HLOf 0RURQJ *XVWDYLD DX5XVWD /r /LFDQLD D'HWDOD ( 0HYHUf )ULWVFKr *HQL'D DPHULFDQD / 0DQLONDUD VGr &RFRORED 'DQLFXODWD 0HLVVQ ,Q5D HGXOLV 0DUW /XHKHD 'DQLFXODWD 0DUWr 0YUFLD UXIL'LOD 0F9DXFK &RUGLD VGr %DXKLQLD 5XLDQHQVLV $XEOf 0-%DOLT 7DEHEXLD FI EDUEDWD ( 0HYHUf 6DQGZ &URWRQ FDLXFDUD %HQWK +LUWHOOD UDFHPRVD /DPr 8QNQRZQ 9LQH 1R

PAGE 199

$SSHQGL[ % FRQWLQXHG $%81'$1&( )5(48(1&< '20,1$1&( ,03257$1&( 3 ( & IL iVWHPVLL RUKD ,QJD DOED :LOOG 'LRVSYURV JXLDQHQVLV $XEOf *XUNHr 727$/ 1RWH $VWHULVNV LQGLFDWH VSHFLHV WKDW DOVR RFFXU RQ KD RI SULPDU\ IRUHVW QHDUE\

PAGE 200

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f VXUIDFH ZKLWH WR \HOORZOHSLGRWH DW DQWKHVLV EHFRPLQJ WDQ WR UXVW EURZQ RYHU WLPH DOO D[HV RI LQIORUHVFHQFHV ZKLWH WR \HOORZOHSLGRWH DW DQWKHVLV SHGXQFOH FP ORQJ UDFKLV FP ORQJ VWDPLQDWH LQIORUHVFHQFHV EHDULQJ XS WR UDFKLOODH WKH ODWWHU FD FP ORQJ HUHFW VXEWHQGHG E\ D FD PP ORQJ EUDFWHROH EHDULQJ

PAGE 201

VWDPLQDWH IORZHUV DUUDQJHG LQ f ORQJLWXGLQDO URZV RQ DED[LDO VLGH RQO\ KHUPDSKURGLWLF EXW IXQFWLRQDOO\ SLVWLOODWHf LQIORUHVFHQFHV EHDULQJ XS WR UDFKLOODH WKH ODWWHU VXEWHQGHG E\ D FD PP ORQJ EUDFWHROH JHQHUDOO\ EHDULQJ f§f§f SLVWLOODWH IORZHUV DW EDVH WR PLGGOH DQG VHYHUDO VWDPLQDWH IORZHUV DW DSH[ RU PRUH UDUHO\ UDFKLOODH EHDULQJ VWDPLQDWH IORZHUV RQO\ 6WDPLQDWH IORZHUV VOLJKWO\ LPPHUVHG VXEWHQGHG E\ YHU\ VPDOO EUDFWHROHV VHSDOV ODQFHRODWH DFXWH FD PP ORQJ DQG PP ZLGH SHWDOV f FRULDFHRXV VWURQJO\ WR ZHDNO\ LQFXUYHG HOOLSWLF WR RERYDWH GHQWDWH DW DSH[ FD PP ORQJ DQG PP ZLGH VWDPHQV f f ILODPHQWV VOHQGHU FD PP ORQJ WKHFDH VHSDUDWH LUUHJXODUO\ FRLOHG DQG WZLVWHG SLVWLOORGH SUHVHQW 3LVWLOODWH IORZHUV VOLJKWO\ FXSXODWH RYRLGREORQJ UXVWWRPHVWRVH VXEWHQGHG E\ EUDFWHROHV VHSDOV DUUDQJHG LQ ZKRUOV RI VHSDOV HDFK LQQHU VHSDOV SURJUHVVLYHO\ ODUJHU SHWDOV WULDQJXODU HTXDO LQ VL]H WR FD FP ORQJ HQWLUH WR GHQWDWH DW PDUJLQV DQG DSH[ VWLJPDV W\SLFDOO\ HUHFW DSLFDO )UXLWV EURDGO\ HOOLSWLF WR REORQJ PP ORQJ PP ZLGH J GU\ ZHLJKW OHSLGRWH WDQ ZKHQ \RXQJ EHFRPLQJ EURZQ RU UXVW FRORUHG DW PDWXULW\ JUD\ WR ZKLWH DW DSH[ VWLJPDWLF UHVLGXH SHUVLVWHQW VWDPLQRGLDO ULQJ ZHDNO\ WR VWURQJO\ GHILQHG HSLFDUS ILEURXV PP WKLFN PHVRFDUS PHDO\ GU\ PP WKLFN HQGRFDUS ZRRG\ PP WKLFN VHHGV ff RYDWH WR HOOLSWLF FD FP ORQJ HQGRVSHUP ZKLWH KRPRJHQHRXV

PAGE 202

6SHFLPHQV ([DPLQHG %5$=,/ &HDU£ 0XQ RI 8EDMDUD /DW rn6 /RQJ r: 'HF 0 %DOLFN HW DO 1< 0* &(1$5*(1 ,$1 ,13$f 0XQ RI ,S /DW r} 6 /RQJ rn : 'HF 9/ L %DOLFN HW DOA 1< 0* &(1$5*(1 ,$1 ,13$f *RL£V 0XQ RI 7RFDQWLQSROLV /DW rn 6 /RQJ rn : 1RY 0 %DOLFN HW DO 1< 0* &(1$5*(1 ,$1 ,13$f 0DUDQK£R 0XQ RI &RGR /DW rn 6 /RQJ R : $XJXVW $ % $QGHUVRQ DQG -/ %MB $QGHUVRQ 1< 0*f 0XQ 6DR )HOL[ GH %DOVDV /DW rf 6 /RQJ rn : 'HF 0 %DOLFN HW DO 1< 0* &(1$5*(1 ,$1 ,13$f 0DWR *URVVR 0XQ RI $ULSXDQ£ /DW r n 6 /RQJ rn : 0DUFK $ % $QGHUVRQ 0LQDV *HUDLV 0XQ RI 3LUDSRUD /DW rn 6 /RQJ Rn: 0D\ $ % $QGHUVRQ 1< 0*f 3DU£ 0XQ RI 6DR )HOL[ GR ;LQJX /DW rf 6 /RQJ rn : -XO\ $ % $QGHUVRQ 1< 0*f 0XQ RI %UDJDQSD /DW rn 6 /RQJ rn : 1RY 0 %DOLFN HW DO 1< 0* &(1$5*(1 ,$1 ,13$f 0XQ RI ,WXSLUDQJD /DW rn 6 /RQJ rn : 1RY e/B %DOLFN HW DO 1< 0* &(1$5*(1 ,$1 ,13$f 3LDXL 0XQ RI 7HUHVLQD /DW r 6 /RQJ rn : 'HF 0 %DOLFN HW DO 1< 0* &(1$5*(1 ,$1 ,13$f

PAGE 203

$33(1',; 6(/(&7,9( &203$5,621 2) *5266 0253+2/2*,&$/ &+$5$&7(56 ,1 2UELJQYD PDUWLDQD WHL[HLUDQD $1' &/ HLFKOHUL &+$5$&7(5 PDUWLDQD WHL[HLUDQD SRVVLEOH K\EULGf 67(0 /HQJWK LQ DGXOWV FPf '%+ FPf /($) 1XPEHU SHU SDOP 6KHDWK OHQJWK LQ DGXOWV FPf VWULDWLRQV RQ DED[LDO VXUIDFH JHQHUDOO\ SUHVHQW SUHVHQW RU DEVHQW 3HWLROH OHQJWK LQ DGXOWV FPf 5DFKLV OHQJWK LQ DGXOWV FPf FURVV VHFWLRQ s VLGHG s VLGHG DED[LDO VXUIDFH ZHDNO\ WR VWURQJO\ OHSLGRWH ZHDNO\ OHSLGRWH 3LQQDH QXPEHU SHU VLGH RULHQWDWLRQ LQ VDPH SODQH LQ VDPH SODQH WR FULVSDWH GLVSRVLWLRQ UHJXODU FOXVWHUHG WR UHJXODU DGD[LDO VXUIDFH GXOO WR OXVWURXV GXOO DED[LDO VXUIDFH ZHDNO\ JODFRXV ZHDNO\ JODFRXV WR VPRRWK HLFKOHUL DEVHQW WULDQJXODU VPRRWK FULVSDWH FOXVWHUHG GXOO PRVWO\ VPRRWK

PAGE 204

$SSHQGL[ FRQWLQXHG &+$5$&7(5 PDUWLDQD ,1)/25(6&(1&( %UDFWV OHQJWK RI SURSK\OO FPf OHQJWK RI DFXPHQ RI SHGXQFXODU EUDFW FPf DED[LDO VXUIDFH RI SHGXQFXODU EUDFW 3HGXQFOH OHQJWK FPf 5DFKLV OHQJWK FPf FXUYDWXUH WDQ WR UXVW EURZQ QRW UHFXUYHG 5DFKLOODH DWWDFKPHQW WR UDFKLV WKURXJKRXW OHQJWK RI VXEWHQGLQJ EUDFWHROH 6WDPLQDWH IORZHUV SHWDOV SHU IORZHU IXVLRQ RI SHWDOV FD PP f VHSDUDWH VWDPHQV SHU IORZHU ff )58,7 6WDPLQRLGDO ULQJ ZHDNO\ WR VWURQJO\ GHILQHG HLFKOHUL WHL[HLUDQD SRVVLEOH K\EULGf HLFKOHUL EORRG UHG EORRG UHG ZHDNO\ WR VWURQJO\ UHFXUYHG VWURQJO\ UHFXUYHG RQ DED[LDO VLGH RQO\ FD PP RQ DED[LDO VLGH RQO\ FD FP VHSDUDWH ff VHSDUDWH WR SDUWLDOO\ FRQQDWH ZHDNO\ WR PRGHUDWHO\ GHILQHG ZHDNO\ GHILQHG

PAGE 205

2UELJQYD PDUWLDQD 0DUNOHYD GDKOJUHQLDQD $1' 0D[LPOOLDQD PDULSD 2UELJQYD PDUWLDQD 0DUNOHYD GDKOJUHQLDQD 0D[LPLOLDQD PDULSD &+$5$&7(5 SRVVLEOH K\EULGf /($) $UUDQJHPHQW RI OHDI EDVHV LQ FURZQ VSLUDOOHG VSLUDOOHG LQ s YHUWLFDO URZV VSLUDOOHG LQ YHUWLFDO URZV 6KHDWK OHQJWK LQ DGXOWV FPf 3HWLROH OHQJWK LQ DGXOWV FPf 5DFKLV OHQJWK LQ DGXOWV FPf 3LQQDH QXPEHU SHU VLGH RULHQWDWLRQ LQ VDPH SODQH LQ VDPH SODQH WR VOLJKWO\ FULVSDWH VWURQJO\ FULVSDWH GLVSRVLWLRQ UHJXODU FOXVWHUHG LQ JURXSV RI FOXVWHUHG LQ JURXSV RI ,1)/25(6&(1&( 3HGXQFXODU EUDFW FXUYDWXUH VOLJKW PRGHUDWH SURQRXQFHG OHQJWK RI DFXPHQ FPf OHQJWK H[FOXGLQJ DFXPHQ FPf ZLGWK

PAGE 206

$SSHQGL[ ( FRQWLQXHG &+$5$&7(5 2UELHQYD PDUWLDQD 0DUNOHYD GDKOFUHQLDQD SRVVLEOH K\EULGf 0D[LPLOLDQD PDUL'D ,1)/25(6&(1&( FRQWf 3HGXQFOH OHQJWK FPf 5DFKLV OHQJWK FPf 5DFKLOODH OHQJWK RI VXEWHQGLQJ EUDFWHROH PPf 6WDUDLQDWH IORZHUV OHQJWK RI SHWDOV QXPEHU RI VWDPHQV ff f§ff§f RULHQWDWLRQ RI WKHFDH LUUHJXODUO\ FRLOHG DQG WZLVWHG WZLVWHG RU KDPDWH VWUDLJKW IXVLRQ RI WKHFDH VHSDUDWH SDUWLDOO\ XQLWHG XQLWHG 3LVWLOODWH IORZHUV QXPEHU SHU UDFKLOOD f§f§f VL]H RI VHSDOV XQHTXDO XQHTXDO HTXDO )58,7 6KDSH EURDGO\ HOOLSWLF WR REORQJ REORQJ RYRLGREORQJ /HQJWK PPf :LGWK PPf 7KLFNQHVV RI HQGRFDUS PPf

PAGE 207

%,2*5$3+,&$/ 6.(7&+ $QWKRQ\ %HQQHWW $QGHUVRQ ZDV ERP RQ -XO\ LQ 2DN %OXIIV 0DVVDFKXVHWWV ,Q } KH UHFHLYHG WKH GHJUHH RI %DFKHORU RI $UWV LQ QDWXUDO VFLHQFH IURP $PKHUVW &ROOHJH $PKHUVW 0DVVDFKXVHWWV $IWHU VSHQGLQJ PRQWKV DV D 7KRPDV :DWVRQ )HOORZ LQ /DWLQ $PHULFD KH ZDV HPSOR\HG DV D UHVHDUFKHU LQ WKH %RWDQ\ 'HSDUWPHQW RI WKH ,QVWLWXWR 1DFLRQDO GH 3HVTXLVDV GD $PD]RQLD ,13$f LQ 0DQDXV %UD]LO ,Q KH UHFHLYHG WKH GHJUHH RI 0DVWHU RI 6FLHQFH LQ WURSLFDO ERWDQ\ IURP ,13$ +LV WKHVLV ZDV RQ WKH HFRORJ\ RI ZKLWHVDQG YHJHWDWLRQ LQ $PD]RQLD $IWHU FRPSOHWLQJ KLV GRFWRUDWH KH ZLOO DVVXPH D SRVLWLRQ RQ WKH VWDII RI WKH 0XVHX 3DUDHQVH (PLOLR *RHOGL LQ %HOHP %UD]LO &XUUHQW UHVHDUFK LQWHUHVWV LQFOXGH UHSURGXFWLYH HFRORJ\ RI HFRQRPLF SODQWV DQG ODQG XVH SUDFWLFHV LQ WKH KXPLG WURSLFV

PAGE 208

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

PAGE 209

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


13
Figure 12. Coverage of babassu stands in Maranhao. Source:
Anonymous (1981).


53
all populations observed, despite high variability (0-74%) among indi
vidual palms.
Variability on the population level was analyzed by comparing the
distributions of female inflorescences per palm in six populations (Fig.
20). Despite a wide range of ecological conditions, the four contig
uous populations (forest and pasture at Lago Verde, wet site at Pindar
Mirim, and dry site at Caxias) showed no differences in distributions of
female inflorescences per palm, according to the Kolmogorov-Smirnov Test
(Siegel 1956). However, distributions in the two disjunct populations
(northerly site in Gois and southerly site in Minas Gerais) were signi
ficantly different from the first four populations, as well as from each
other (p <.01 in all cases). These results suggest that the distribu
tion of female inflorescences per palm in populations of babassu is
governed by genetic factors. Potential implications of this finding are
discussed in Chapter 7.
Fruit production. At all sites, frequencies of immature infructes-
cences peaked as flowering ended in April-June and were lowest when
fruiting peaked in October-December (Figs. 21-22). This pattern was
consistent despite the wide range of climatic conditions among sites
(Fig. 13). According to phenological observations, there was a ca. 9-mo
lag between the flowering peak (January-March, Figs. 18-19) and the
fruiting peak (October-December, Figs. 23-24). Observations on indi
vidual palms confirmed that fruits require an average of 9 mo to reach
maturity.


11
Figure 10. Major soil groups of Maranho, according to the Brazilian
classification system (E1IBRAPA 1976). H = Hydromorphic
Laterite, Pd = Dystrophic Red-Yellow Podzolic soils, Pe =
Eutrophic Red-Yellow Podzolic soils, Ppd = Pl'inthic Pod
zolic soils, L = Yellow Latosol, Lr = Red-Yellow Latosol,
and Q = Quartz sands. Unlabelled areas = other soil types.
Source: EMBRAPA (1981).


188
staminate flowers arranged in 2(-4) longitudinal rows on abaxial side
only; hermaphroditic (but functionally pistillate) inflorescences
bearing up to 475 rachillae, the latter subtended by a ca. 2 mm long
bracteole, generally bearing 1 2(3) pistillate flowers at base to
middle and 1-several staminate flowers at apex, or more rarely rachillae
bearing staminate flowers only.
Staminate flowers slightly immersed, subtended by 2 very small
bracteoles; sepals 3, lanceolate, acute, ca. 1 mm long and 1 mm wide;
petals 2(-3), coriaceous, strongly to weakly incurved, elliptic to
obovate, dentate at apex, ca. 14 mm long and 8 mm wide; stamens (22-)24-
26(-30), filaments slender, ca. 2 mm long, thecae separate, irregularly
coiled and twisted; pistillode present.
Pistillate flowers slightly cupulate, ovoid-oblong, rust-tomestose,
subtended by 1-2 bracteoles; sepals 3-6, arranged in 1-2 whorls of 3
sepals each, inner sepals progressively larger; petals 3, triangular,
equal in size, to ca. 5 cm long, entire to dentate at margins and apex;
stigmas typically 3-6, erect, apical.
Fruits broadly elliptic to oblong, 74-125 mm long, 49-99 mm wide,
80-480 g dry weight, lepidote, tan when young, becoming brown or rust-
colored at maturity, gray to white at apex; stigmatic residue
persistent; staminodial ring weakly to strongly defined; epicarp
fibrous, 1-4 mm thick; mesocarp mealy dry, 2-12 mm thick; endocarp
woody, 35-76 mm thick; seeds (1-)3-6(-11), ovate to elliptic, ca. 3-6 cm
long; endosperm white, homogeneous.


160
Figure 45. Shifting cultivation site just prior to burn at Lago Verde.
Figure 46. Clearcut of palm forest prior to pasture establishment, near
Bacabal, Maranho.


165
fie stands of babassu form spontaneously, require little maintenance,
and are already utilized by people. Rather than embark on long-term
development of the palm for intensive agriculture, babassu seems most
amenable for immediate use as an extensive, moderate-yielding, low-
maintenance crop that is easily integrated into the predominant forms of
land use currently practiced over widespread areas of Brazil: shifting
cultivation and cattle ranching.
Potential use of babassu as a crop could begin with rational man
agement of the stands that already exist. A key component of such
management is maintenance of palm densities suitable for growth of other
crops. Mature palms with a low proportion of female inflorescences
should be preferentially thinned. Competition from high densities of
juvenile palms is reduced by periodic cutting; on sites used for perma
nent cropping or grazing, juveniles can be eliminated by repeated cut
ting over several years or by application of systemic herbicides. Seed
ling recruitment can be controlled through more efficient fruit harvest
ing procedures. Rather than gather fallen fruits from the ground,
mature fruit bunches should be harvested from the crowns with pruning
poles; this practice reduces the incidence of lost fruits as well as
infestation by bruchid beetles, which only attack fruits on the ground.
Because of its low maintenance requirements, babassu has excellent
potential for introduction into extensive land-use systems. Experimen
tal results described in Chapter 4 indicate that babassu competes effec
tively under a wide range of climatic conditions and thus can be suc
cessfully introduced in areas outside of its current range. Planting of


29
On the specific level, at least 10 names associated with babassu
have been published within the genus Orbianva (Table 2). Most of this
proliferation has resulted from species descriptions based on inadequate
botanical specimens, a problem inherent to palm taxonomy in general
(Tomlinson 1979). In the more recent taxonomic literature (e.g. Rizzini
1963; Bondar 1964; Glassman 1977, 1978), most of the names listed in
Table 2 are considered synonymous, and it is generally agreed that the
babassu complex consists of a maximum of three species. Due to
potential confusion among these species, I embarked on a taxonomic study
of the babassu complex. My objectives were (1) to identify the species
that was the subject of the ecological studies described in the
following chapters and (2) to distinguish it from morphologically
similar species.
Methods
Lack of adequate specimens has been the principal impediment to
resolving the taxonomy of the babassu complex. To remedy this situ
ation, I collected representatives of the complex over a substantial
portion of its presently known range (Fig. 15). Due to the unusual
structure and bulkiness of large palms such as those in the babassu
complex, complete botanical collections require extensive field documen
tation; recommended procedures are described at length in Balick et al.
(1982). Specimens of all collections were deposited at the Mu3eu
Paraense Emilio Goeldi in Belem, Brazil; duplicates were sent to other
Brazilian institutions (CENARGEN, INPA, CPATU) and the New York Botan
ical Garden, Bronx, N.Y., U.S.A.


16
still exist; babassu was commonly present but invariably associated with
a high diversity of hardwoods.
With the arrival of Europeans during the 1600s, the landscape of
Maranho changed. In response to the needs of the colonial economy,
plantations of rice, cotton, and sugarcane were established on bottom
lands over an extensive area of the state. Yet more than two centuries
elapsed before the first historical reference was made to the babassu
forests of Maranho (Prazeres 1820, cited in Abreu 1940: 11), which
suggests that these forests were not as extensive as they are today.
With the abolition of slavery in 1888, the plantation economy collapsed
and vast areas of cultivated land were abandoned. These sites were
recolonized by high-density stands of babassu. Like the Indians before
them, the freed slaves and their descendants were dependent on the palm
for food, fuel, and fiber. As their populations increased, agriculture
in the form of shifting cultivation moved out of the bottomlands into
the mixed upland forests, and stands of babassu gradually spread across
the landscape. The size of these stands today (Fig, 12) indicates the
extent and nature of human impact on the landscape of Maranho.
The Study Region
Climate and soils. The ecological research was located primarily
in the central Mearim Valley of north-central Maranho (Fig. 9), which I
shall henceforth refer to as the study region. Climate and soils in
this region are extremely favorable for agriculture. Meteorological
data from the region's principal town, Bacabal, show a mean annual
precipitation of 157.5 cm; the 6-mo wet season (December-May) is suffi-


I certifv 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 duality, as a dissertation for the
degree of Doctor of Philosophy.
This dissertation was submitted to the Graduate Faculty of the
Department of Botany 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.
December 1983
Dean for Graduate Studies
and Research


41
determined by counting fruits and/or calyces (in the case of abscised
fruits). A subsample of ca. 30 fruits from each infructescence was air-
dried and weighed to the nearest gram.
Considerable variability in the ratio of male to female inflores
cences (sex ratio) was observed among the palms in the phenological
study. To compare sex ratio variability on the population level, I
measured the distributions of inflorescence types in six populations.
Distributions were obtained by determining the percentage of each in
florescence type per palm. The inflorescences tend to remain attached
for several years following anthesis; my counts thus provided a long
term picture of reproductive behavior. Data were obtained from two
geographically isolated populations located in Tocantinpolis, Gois,
and Pirapora, Minas Gerais. Additional data were obtained from four
populations located within my phenological study sites at Lago Verde,
Pindar Mirim, and Caxias; all of the latter sites occur within the more
or less continuous band of babassu stands that stretches across north-
central Maranhao (Fig. 12).
Floral Biology
To determine duration of flowering in individual inflorescences, I
carried out daily phenological observations from 1-15 March 1982 on 38
palms on the secondary forest site and 15 palms on the pasture site.
Palms with bracts about to open were selected for observation. Flower
ing was judged to begin upon opening of the bract. Proportion of
flowers that ultimately became fruits was determined through examination


167
feedcake; the excess could then be sold as semi-processed materials,
greatly reducing the current costs of transporting bulky fruits.
The potential role of babassu in rural communities is illustrated
by the interdependence of people and wild stands of the lontar palm
(Borassus sundaicus) in islands of Indonesia (Fox 1977). Despite a dry
climate and nutrient-poor soils, high population densities are main
tained on these islands by productive systems of gathering, semi-perma
nent agriculture, and animal husbandry either directly or indirectly
based on the palm. Like the lontar, babassu is a source of essential
subsistence products and is the basis for a widespread cottage industry.
Likewise, babassu maintains dynamicalthough currently less exploited
interations with systems of agriculture and animal husbandry. Full
utilization of the palm by rural communities could open the way to self-
sufficiency and ecological stability, while at the same time integrating
these communities into the larger market economy.


171
Escola Tcnica Federal do Maranhao. 1976. Babapu: industrianzapo
total. Sao Luis, Brazil.
Essig, F. B. 1971. Observations of pollination in Bactris. Principes
15: 20-24.
Ewel, J., S. Gliessman, M. Amador, F. Benedict, C. Berish, R. Bermudez,
B. Brown, A. Martinez, R. Miranda, and N. Price. 1982. Leaf area,
light transmission, roots and leaf damage in nine tropical plant
communities. Agro-Ecosystems 7: 305-326.
Faegri, K., and L. van der Pijl. 1979. The principles of pollination
ecology. Third edition. Pergamon Press, Oxford, England.
Ferri, M. G. 1974. Ecologia: temas e problemas brasileiros. Editora
Universidade de Sao Paulo, Sao Paulo, Brazil.
. 1980. Vegetapo brasileira. Editora Universidade de Sao
Paulo, Sao Paulo, Brazil.
Fonseca, E. T. 1924. 0 babassu (Attalea speciosa Mart., Orbignya
martiana Barb. Rodr.). Ministrio da Agricultura, Indstria e
Comercio, Rio de Janeiro, Brazil.
Fox, J. J. 1977. Harvest of the palm. Harvard University Press,
Cambridge, Massachusetts, USA.
Freund, R. J., and R. C. Littell. 1981. SAS for linear models: A guide
to the ANOVA and GLM procedures. SAS Institute Incorporated, Gary,
North Carolina, USA.
Furley, P. A. 1975. The significance of the cohune palm, Orbignya
cohune (Mart.) Dahlgren, on the nature and development of the soil
profile. Biotropica 7: 32-36.
Gentry, A. H. 1982. Patterns of neotropical plant diversity. Evo
lutionary Biology 15: 1-84.
Givnish, T. J. 1978. On the adaptive significance of compound leaves,
with particular reference to tropical trees. Pages 351-380 in P.
B. Tomlinson and M. H. Zimmermann, editors. Tropical trees as
living systems. Cambridge University Press, Cambridge, England.
Glassman, S. F. 1977. Preliminary taxonomic studies in the palm genus
Orbignya Mart. Phytologia 36: 89-115.
. 1978. Corrections and changes in recent palm articles
published in Phytologia. Phytologia 40: 313-315.


The excellent drawings were by George Fuller, and my battery of
dedicated typists included Bob Epting, Donna Epting, and Leslie Rigg.
Less tangible but none theles s meaningful assistance was provided by
Helda Lenz Cesar, Warwick Kerr, Mario Leal, Peter May, and Ghillean
Prance.
I would like to express my sincere appreciation to the people who
made my dedication to the research possible: Clemente Rodrigues and his
extended family; Isa and Nelson Carvalho; my parents, Martha and Seneca
Anderson; and especially my wife, Suely Anderson.
Finally, very special thanks go to my supervisory committee, Drs.
Peter Feinsinger, Walter Judd, Hugh Popenoe, Francis Putz, and especial
ly my chairman, John Ewel, who visited the field sites and provided
patient guidance in all phases of the research.
IV


12
MANGROVE AND BEACH
VEGETATION
g SEASONALLY WET
SAVANNA
£3 WOODLAND,SCRUB,
AND DRY SAVANNA
m MOIST FOREST
200 km
Figure 11. Principal ecological zones of Maranhao.
Source: Kuhlmann (1977).


164
ly be reduced. Within populations of babassu, I have occasionally
observed reproductive, short-statured palm ( <5 m) that were apparently
precocious. Increased precocity was successfully incorporated into oil
palm by programs of selection and breeding (Hartley 1977). Other,
potentially desirable traits such as increased germination success or
enhanced resistance to diseases may also be amenable to genetic improve
ment.
From a broader perspective, however, development of babassu for
intensive cultivation requires closer scrutiny. I have three reserva
tions concerning this approach. First, it can only be accomplished over
an extremely long term, while current issues concerning the continued
viability of the babassu economy require more immediate solutions.
Second, although its potential for genetic improvement is high, babassu
will never approach the African oil palm as a producer of vegetable oil.
Whereas the oil in babassu is limited to the kernels, in oil palm it
occurs in both the kernels and the mesocarp. Likewise, as a producer of
biomass for fuel, babassu will never rival fast-growing trees such as
Eucalyptus spp. and Gmelina arbrea, in which most dry matter production
is allocated to stems. The desirability of babassu as a crop is not its
potential to out-produce already established crops, but its current
capacity to provide a high diversity of products that are useful in both
market and subsistence economies.
My final reservation is that development of babassu for intensive
cultivation ignores the fundamental characteristics that make it unique
as a crop. As emphasized elsewhere in this study, virtually monospeci-



158
less juveniles can persist in the understory for years, building up high
densities. In primary forest, for example, a palm remains apparently
stemless for an average of ca. 50 yr. During this period, the stem
grows below ground due to babassu's so-called cryptogeal mode of germi
nation, in which the apical meristem is placed beneath the soil. Prob
ably because it grows so slowly, the palm is potentially long-lived,
attaining a maximum lifespan of at least 184 yr. Escape from seed
predators, shade tolerance, and longevity are the key factors accounting
for babassu's high importance in primary forest.
Placement of their apical meristems beneath the soil enables seed
lings and stemless juveniles to survive when forests are cut and burned
by shifting cultivators. When the sites are abandoned, these palms are
released and form high-density, virtually monospecific stands during the
subsequent fallow. Babassu's current dominance over extensive areas of
Brazil is a testimony of its adaptability to shifting cultivation.
People who carry out shifting cultivation are also those most
dependent on the palm forests as a source of roof thatch and building
materials; income through extraction and sale of the oil-rich kernals;
food through domestic consumption of vegetable oil, animal ration, and
the heart of the palm; and fuel in the form of charcoal made from the
fruit husks. This charcoal was found to be the exclusive source of fuel
in 96% of sampled households in the vicinity of the palm forests (Ander
son and Anderson 1983). Thus, babassu is already tightly integrated
into market and subsistence economies upon which rural poor depend.


Table 25. Annual changes in number of palms per life stage on the 1-ha primary forest site
at Lago Verde. Life stages are defined in Table 18.
LIFE STAGE
INITIAL N
GAINS
LOSSES
NET CHANGE
FINAL N
Number and Description
Recruitment
from Last Stage
Recruitment
to Next Stage
Mortality
1 Seeds on ground
6,600
23,560
2,400
13,850
+7,310
13,910
2 Seedlings
5,240
2,400
0
1,620
+780
6,020
3 Pre-established juveniles
710
0
0
20
-20
690
4 Established juveniles
225
0
10
0
-10
215
5
60
10
0
0
+10
70
6
34
0
6
0
+6
28
7 Mature palms
36
6
6
0
0
36
8
24
6
0
2
+4
28
9
4
2
2
0
0
4
10
6
2
0
1
+1
7
11
0
0
0
0
0
0
12
1
0
0
0
0
1
S7T


84
Jessenia bataua (20)
Astrocoryum cf. macrocalyx (34)
Socrotea exhorrhiza (35)
Iriarteo ventricosa (32)
Mauritia flexuosa (7)
Scheelea sp. (19)
Phyfelephas microcarpa (17)
Orbiqnya martiana (6)
II l l |
0.05 0.1
1 1 1lI 'I l |
0.5 1.0
FORCE TO BREAK (t)
i I i i r
5.0
10.0
Figure 30. Force required to break palm fruits. Intervals represent "t
1 standard deviation; values in parentheses are sample sizes.
Data for Orbignya martiana from Fonseca (1924); data for all
other species from Kiltie (1982).


ACKNOWLEDGEMENTS
My field research and data analysis were financed by Grant No. 51-
07-09-07 to John Ewel from the USDA Forest Service, Consortium for the
Study of Man's Relationship with the Global Environment. Additional
support was provided by a fellowship from the Superintendencia de Desen
volvi ento Cientfico e Tecnolgico (CNPq). I am grateful to the Museu
Paraense Emilio Goeldi and its Director, Dr. Jos Seixas Lourenpo, for
invaluable logistic as well as moral support during the course of my
research.
Special thanks are due to Bernardo da Silva and Mario Costa, who
gave unfailing assistance in the field, and to Michael Balick and
William Overal, who collaborated with me on various aspects of the
research. I am also grateful to the following people and their asso
ciated institutions: Jos Mrio Frazo and Claudio Pinheiro of the
Instituto Estadual do Babapu (INEB) for their collaboration in studies
on germination, phenology, and taxonomy of babassu; Herclito Aquino and
Walbert Carvalho of the Empresa Maranhense de Pesquisa Agropecuria
(EMAPA) for providing field workers; Sergio Alves, Mrio Dantas,
Elizabeth de Oliveira, Raimundo Reg, and Waldemar Ferreira of the
Centro de Pesquisa Agropecuaria do Trpico Umido (CPATU) for invaluable
logistic support; and Raimundo Bittencourt, Waldemar Franca, Nelson
Rosa, and Sr. Milton of the Museu Goeldi for their assistance in the
field.
iii


22
Table 1.
Uses of occupied lands in Maranhao during
Numbers represent ha (percent of total
parentheses). Sources: IBGE 1956, 1979.
1950 and 1975.
classified in
Year
Land Use
mo
ms
Grazing
Range
Improved Pasture
3,454,400
40,800
(40.2%)
( 0.5$)
2,590,600
1,218,200
(21.7$)
(10.2$)
Cropping
Temporary
Permanent
314,500
14,200
( 3.6$)
( 0.2$)
1,014,100
41,900
( 8.5$)
( 0.4$)
Fallow
2,407,800
(28.0$)
4,058,100
(34.0$)
Permanent Forest
Natural
Planted
2,361,700
4,000
(27.5$)
( 0.0$)
3,014,300
400
(25.2$)
( 0.0$)
Total Classified
8,597,400
(100.0$)
11,937,600
(100.0$)
Not Classified
941,500
471,500
Total Occupied
9,538,900
12,409,100


189
Specimens Examined
BRAZIL
Cear: Mun. of Ubajara, Lat 350'S, Long 4055W, Dec 1981, M. J.
Balick et al. 1353 (NY, MG, CENARGEN, IAN. INPA); Mun. of Ip, Lat 420
S, Long 4045' W, 12 Dec 1981, VL. i. Balick et al^ 1354 (NY, MG,
CENARGEN, IAN, INPA).
Gois: Mun. of Tocantinpolis, Lat 620' S, Long 4728' W, 28 Nov
1981, M. J. Balick et al. 1309 (NY, MG, CENARGEN, IAN, INPA).
Maranho: Mun. of Codo, Lat 440' S, Long 43o40, W, 20 August 1980,
A. B. Anderson 394 and JL. Bj_ Anderson 397 (NY, MG); Mun. Sao Felix de
Balsas, Lat 735 S, Long 4605' W, 4 Dec 1981, M. J. Balick et al. 1342
(NY, MG, CENARGEN, IAN, INPA).
Mato Grosso: Mun. of Aripuan, Lat 9 10' S, Long 6040' W, 17 March
1977, A. B. Anderson 288.
Minas Gerais: Mun. of Pirapora, Lat 1720' S, Long 44o50'W, 14 May
1981, A. B. Anderson 398 (NY, MG).
Par: Mun. of Sao Felix do Xingu, Lat 645 S, Long 5145' W, 7 July
1980, A. B. Anderson 391 (NY, MG); Mun. of Braganpa, Lat 105' S, Long
4655' W, Nov 1981, M. J. Balick et al. 1301 (NY, MG, CENARGEN, IAN,
INPA); Mun. of Itupiranga, Lat 505' S, Long 4920' W, 23 Nov 1981, £L_
J. Balick et al. 1304 (NY. MG, CENARGEN, IAN, INPA).
Piaui": Mun. of Teresina, Lat 5221 S, Long 4250' W, Dec 1981, M. J.
Balick et al. 1351 (NY. MG, CENARGEN, IAN, INPA).


174
Loomis, R. S., and P. A. Gerakis. 1975. Productivity of agricultural
ecosystems. Pages 145-172 in J. P. Cooper, editor. Photosynthesis
and productivity in different environments. Cambridge University
Press, Cambridge, England.
Maguire, B. 1970. On the flora of the Guayana Highland. Biotropica 2:
85-100.
Maissurow, D. E. 1941. The role of fire in the perpetuation of virgin
forests in northern Wisconsin. Journal of Forestry 39: 201-207.
Markley, K. S. 1971. The babassu oil palm of Brazil. Economic Botany
25: 267-304.
Marks, P. L. 1974. The role of pin cherry (Prunus pennsylvanica L.) in
the maintenance of stability in northern hardwood ecosystems. Eco
logical Monographs 44: 73-88.
Martius, C. F. P. von. 1826. Historia naturalis palmerum 2: 91-144*
Munich, Germany.
. 1844. Palmetum Orbignianum. In A. d'Orbigny,
editor. Voyage dans IAmerique meridionale 7(3): 1-140. Paris,
France.
McKey, D. 1975. The ecology of coevolved seed dispersal systems.
Pages 157-191 in L. E. Gilbert and P. H. Raven, editors. Coevo
lution of animals and plants. University of Texas Press, Austin,
Texas, USA.
Mendes, A. M. de C., and J. 0. B. Carioca. 1981. Babapu. Volume 3 in
Estudo integrado do uso potencial de biomassa para fins energticos
no Brasil. Fortaleza, Brazil. Mimeographed.
Moore, H. E. 1973a. Palms in the tropical forest ecosystems of Africa
and South America. Pages 63 -88 in B. J. Meggers, E. S. Ayensu, and
W. D. Duckworth, editors. Tropical forest ecosystems of Africa and
South America: a comparative review. Smithsonian Institution
Press, Washington, District of Columbia, USA.
. 1973b. The major groups of palms and their distribution.
Gentes Harbarium 11: 27-114.
. 1977. Endangerment at the specific and generic level in
palms. Pages 267-282 in G. T. Prance and T. S. Elias, editors.
Extinction is forever: the status of threatened and endangered
plants of the Americas. The New York Botanical Garden, Bronx, New
York, USA.


125
proportion of total above-ground production was allocated to reproduc
tive structures (44.8%) and leaves (39.4%), while only 15.8% went to
permanently accruing biomass in the stem (Table 21). Partitioning of
above-ground productivity varied considerably according to age (Figs.
37-40).
Likewise, a comparison of mature palms on the three study sites at
Lago Verde (Table 22) shows that productivity varied considerably
according to site. On the forested sites, absolute productivity per
palmas well as relative allocation to the palm's various components
were almost identical. However, productivity of all components in
creased dramatically in the pasture. Allocation to vegetative com
ponents (i.e., stem and leaves) increased by ca. 40% and allocation to
reproductive structures by almost 100%.
Among the entire populations of babassu on the primary and second
ary forest sites, partitioning of productivity to stems, leaves, and
reproductive structures was strikingly similiar (Table 23). The consis
tently high allocation to leaves (ca. 70%) is attributable to high
densities of seedlings and juveniles, in which above-ground productivity
is allocated exclusively or primarily to leaves. In fact, most of the
leaf productivity in the two populations was due to seedlings and juve
niles; of total stand productivity, the leaves of these palms comprised
43.1% in the primary forest and 37.9% in the secondary forest.


128
Table 23. Annual above-ground dry matter production of all babassu
palms on the primary and secondary forest sites at Lago
Verde. Values are kg*ha-''yr-"' (percentages in paren
theses).
PRIMARY FOREST SECONDARY FOREST
LEAVES
Leaflets
1,873 (21.6)
4,726 (19.1)
Leaf Axes
4,208 (48.4)
12,053 (48.7)
Subtotal Leaves
6,081 (70.0)
16,779 (67.8)
STEMS
REPRODUCTIVE STRUCTURES
1.545 (17.8)
4.804 (19.4)
Ancillary Structures
323 ( 3.7)
1,502 ( 6.1)
Fruits
732 ( 8.4)
1,650 ( 6.7)
Subtotal Reproductive Structures
1.055 (12.2)
3.152 (12.8)
TOTAL
8,681 (100.0)
24.735 (100.0)


89
Table 8. Hypothesized success of babassu in four experiments.
Success results from the interaction of adaptiveness
to the stress factor (weeded plots) and
competitiveness (non-weeded plots). Predicted
results are indicated by the digits: 1 = most
successful, . 4 = least successful.
EXPERIMENT
STRESS FACTOR
WEEDED
NON-WEEDED
Insecticides
1
4
Insecticide
No Insecticides
2
3
Fertilizers
2
3
Fertilizer
No Fertilizers
2
3
Irrigation
1
4
Irrigation
-
No Irrigation
2
3
Wet Climate
1
4
Climate
Dry Climate
2
3


TABLE OF CONTENTS
ACKNOWLEDGEMENTS iii
LIST OF TABLES vii
LIST OF FIGURES ix
ABSTRACT xii
CHAPTER
1 INTRODUCTION 1
2 TAXONOMY AND PHITOGEOGRAPHY 27
Methods 29
Results and Discussion 32
3 REPRODUCTIVE BIOLOGY 38
Methods 39
Results 46
Discussion 74
4 ESTABLISHMENT 88
Methods 90
Results 95
Discussion 112
5 GROWTH AND PRODUCTIVITY 115
Methods 118
Results 121
Discussion 129
6 POPULATION STRUCTURE AND DYNAMICS 140
Methods 141
Results 143
Discussion 154
7 IMPLICATIONS FOR MANAGEMENT 157
People and the Palm Forest 157
Limiting Factors 159
Possible Solutions 165
v


CUMULATIVE GERMINATION (%)
71
Figure 28. Cumulative germination of babassu fruits over time. Inter
vals represent 1 standard error. Shaded bars represent
rainfall. Data were obtained from 14 November 1981 to 20
April 1982.
RAINFALL (cm per 10 days)


INFRUCTESCENCES PER PALM
IMMATURE MATURE IMMATURE MATURE IMMATURE MATURE
55
1981
F M A M J J
1982
A SONDJ FM
0.5 n
0
0.5 -
1.0 -1
INTERMEDIATE SITE
(N=143)
Figure 21. Mean monthly number of immature and mature inflorescences
per palm at three sites in Maranhao. Wet site = Pindar
Mirim, intermediate site = Lago Verde, and dry site =
Caxias. N = number of palms observed per site.


7
Figure 8. Flowchart of products derivable from babassu fruits via
current (1983) technology. Diamonds represent "either-or"
options. Numbers are percentages of fruit weight. Adapted
from Abreu (1940), Viveiros (1943), Escola Tcnica Federal
do Maranhao (1976), IPT (1979), and Mendes and Carioca
(1981).


68
Table 5. Attack of babassu fruits and seeds by the seed-
predaceous beetle, Pachymerus nucleorum.
SEEDS/FRUIT FRUITS SEEDS
Total N Colonized Total N Colonized
N % N %
16
10
62.5
16
10
62.5
63
50
79.4
126
67
53.2
124
86
69.4
372
157
42.2
113
85
75.2
452
198
43.8
23
19
82.6
115
37
32.2
4
3
75.0
24
6
25.0
6


34
35
36
37
38
39
40
41
42
43
44
45
46
97
Palms per m^ (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in fertilizer experiment . .
Palms per m^ (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in insecticide experiment. . 98
Palms per m^ (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in climate experiment 99
Absolute allocation of annual above-ground dry matter
production per babassu palm as a function of age 130
Relative allocation of annual above-ground dry matter
production per babassu palm as a function of age 131
Mean annual number of inflorescences produced per palm as a
function of mean age per life stage 134
Mean annual number of seeds produced per palm as a function
of mean age per life stage 135
Stage distributions of babassu on the primary forest,
secondary forest, and pasture sites at Lago Verde 148
Age distribution of babassu on the primary forest site at
Lago Verde 149
Age distribution of babassu on the secondary forest site
at Lago Verde 150
Population dynamics of babassu on the 1-ha primary forest,
secondary forest, and pasture sites at Lago Verde 151
Shifting cultivation site just prior to burn at Lago Verde. .160
Clearcut of palm forest prior to pasture establishment,
near Bacabal, Maranho 160
xi


SEEDS PER FRUIT
69
I 00
80
60
40
20
0
I 00
80
60
40
20
0
100
80
60
40
20
0
PASTURE
FIELD
Figure 27. Seed fate over time in three study sites at Lago Verde.


26
Gy pe rus spp., and juveniles of babassu) were annually cut and burned.
The 10-15 m high canopy was comprised exclusively of babassu, which
occurred at a density of ca. 100 stems per ha.
Satellite sites. In addition to the principal field site at Lago
Verde, satellite sites were established at Belm, Pindar Mirim, and
Caxias (Fig. 7). The Belem site was used for experiments involving
establishment (Chapter 4) the Pindar Mirim and Caxias sites were used
for phenological observations (Chapter 3). Meteorological data (Fig.
11) show a wide range of rainfall at these sites. Information con
cerning land uses and experimental design at each site is provided in
the appropriate chapters.


INFLORESCENCE PRODUCTION PER PALM
50
Figure 18. Mean monthly production of male and functionally female
inflorescences per palm at three sites in Maranhao. Wet
site = Pindar Mirim, intermediate site = Lago Verde,
and dry site = Caxias. N = number of palms observed per
site.


152
Recruitment of pre-established and established juveniles did not
occur in the primary forest (Fig. 44) which would suggest that the
population was not regenerating. Yet its relatively stable stage (Fig.
41) and age (Fig. 42) structure strongly indicate that the population
was self-maintaining. As babassu is long-lived, observations of re
cruitment over a 1-yr period are probably insufficient to discern popu
lation trends in as complex an environment as the primary forest. Ran
dom events such as formation of small gaps are undoubtedly necessary for
maintenance of growth in seedlings and juveniles. Less frequent events
such as large tree falls probably result in dramatic pulses of recruit
ment, which could account for the sharp peaks in frequencies of ca. 32-
and 43-yr-old palms (Fig. 42). However, babassus apparently stable
structure (Figs. 41-42) leads me to suspect that small forest gaps are
sufficient for maintenance of the population. Given the limited degree
of recent disturbance on this site (see Chapter 1), I conclude that the
primary forest population of babassu is maintained without major dis
turbances.
Secondary Forest
The stage distribution of babassu on the secondary forest site
(Fig. 41) showed a pronounced bulge of mature palms in stages 7-8 and a
comparative absence of older stages. In its age distribution (Fig. 43),
there were two pronounced peaks at 33 and 73 yr. The site was clearcut
for shifting cultivation 33 yr prior to this study. The 33-yr peak thus
represented release of seedlings that had germinated just before forest
clearing. The 73-yr peak consisted of a dense cohort of mature palms.


106
Table 13. Palms per m2 (a), leaf area per palm (b), and
leaf area index (c) of babassu at end of
climate experiment, with climate and ecotype
as main effects. Asterisks indicate
significant differences (** for p £ .01)
between main effect means. Where there is
interaction between main effects, superscripts
are used to compare treatment means. Values
that do not share the same superscript
significantly (p £.05).
differ
(a)
O
Palms Der m
Drv Climate
Wet Climate
Mean
Dry Ecotype
14.9
21.0
18.0
Wet Ecotype
14.0
19.4
16.7
Mean
14.4
** 20.2
(b)
Leaf Area Der Palm (cm2)
Drv Climate
Wet Climate
Mean
Dry Ecotype
371.9a
231.0
301.4
Wet Ecotype
292.3b'c
301.9a,b
297.1
Mean
332.1
266.4
(c)
Leaf Area Index
Drv Climate
Wet Climate
Mean
Dry Ecotype
.23b
.24b
on
C\J
Wet Ecotype
.16
.34a
.25
Mean
.19
.29


CHAPTER 6
POPULATION STRUCTURE AND DYNAMICS
Population biology of plants was the exclusive domain of foresters
and agronomists until recently, when ecologists became interested in the
subject. Plant ecologists have traditionally concentrated on the de
scription, classification, and physiology of vegetation, while largely
ignoring the dynamic aspects of population phenomena. As a result,
plant population biology has borrowed much of its theoretical basis from
zoologists (Harper and White 1974) However, in a synthesis of the
emerging field, Harper (1977) drew important distinctions between popu
lations of animals and plants. The most significant distinction is the
modular construction of plants, which confers extreme morphological
plasticity. In contrast to animals, individual plants can respond to
stresses by varying the number of their component parts. As a result,
plant populations may be viewed as consisting of both individuals and
their component parts.
To date, most ecological studies involving the structure and dyna
mics of plant populations have been carried out on temperate herbs. To
my knowledge, with the exception of two dicotyledonous trees studied by
Hartshorn (1972), all such studies in the tropics have involved palms
(Van Valen 1975, based on data from Bannister 1970; Sarukhan 1978; and
Bullock 1980), probably because the age of palms is relatively easy to
estimate (see Chapter 5). Whether in temperate or tropical areas, few
studies have attempted to examine the population biology of a species
140


RAINFALL (cm)
18
Figure 13. Monthly rainfall at four research sites. Numbers are long
term annual means, bars are long-term monthly means, shaded
circles are monthly data for 1981, and open circles are
monthly data for 1982. Period over which long-term means
were calculated and source of unpublished data for each
site are: Belem, 1931-60, Centro de Pesquisa Agropecuaria
do Trpico Umido (CPATU); Pindare Mirim, 1966-80, Superin
tendencia de Desenvolvimiento do Nordeste (SUDENE); Bacabal,
1972-80, Instituto Nacional de Meteorologa (INMET); and
Caxias, 1966-80, INMET.


REFERENCES CITED
168
APPENDIX
A SPECIES ABUNDANCE, FREQUENCY, DOMINANCE, AND
IMPORTANCE IN 1 HA OF PRIMARY FOREST, LAGO VERDE,
MARANHAO 179
B SPECIES ABUNDANCE, FREQUENCY, DOMINANCE, AND
IMPORTANCE IN 1 HA OF SECONDARY FOREST, LAGO VERDE,
MARANHAO 184
C MORPHOLOGICAL DESCRIPTION OF Orbignya martiana
Barb. Rodr 187
D SELECTIVE COMPARISON OF GROSS MORPHOLOGICAL
CHARACTERS IN Orbignya martiana, "O.
teixeirana", and 0^ eichleri 190
E SELECTIVE COMPARISON OF GROSS MORPHOLOGICAL
CHARACTERS IN Orbignya martiana, "Markeleya
dahlgreniana", and Maximiliana maripa 192
BIOGRAPHICAL SKETCH 194
vi


9
babassu. The study concludes with a discussion of the implications for
management (Chapter 7).
The Setting
Maranho
The Brazilian state of Maranho (Fig. 9) encompasses a wide range
of ecological conditions. To the south lies a geologically old (Meso
zoic), elevated peneplain, broken by abrupt plateaus and canyons, and by
low, eroded mountain ranges. The predominant soils in this region are
deeply weathered and nutrient poor, and annual rainfall rarely exceeds
1,000 mm. Human population densities are consequently low (<5 inhabi
tants per km^), and the principal land use is grazing on unimproved
range.
In central and northern Maranho, elevations decline and the ter
rain gently undulates as one enters a broad plain comprised of geologi
cally young (Tertiary) sediments, dissected by many of the state's major
rivers. Soils are extremely variable (Fig. 10), with an abundance of
nutrient-rich alluvium, and annual rainfall ranges from 1,000 to over
2,000 mm. Agriculture (most commonly shifting cultivation) and grazing
constitute the principal land uses. Outside the capital (So Luis),
O
population densities in this region average >10 inhabitants per km .
The babassu forests of Maranho are largely confined to the central
and northern portions of the state, where they are associated with three
broadly defined ecological zones (Fig. 11): seasonally wet savanna
(baixada); woodland, scrub, or dry savanna (cerrado); and moist or dry
forest (mata).


44
harvesting of cassava, and dense regrowth became established. At each
of the three sites, 10 one-layered clusters of 25 fruits each were set
up, with a stake marking the center of each custer. Fruit movement was
measured at 1- to 2-wk intervals. The search for dispersed fruits was
conducted to a distance of 4 m from the stake. To label fruits, a 14-mm
staple was driven into the apex of each fruit. To eliminate human odor,
fruits were handled with gloves that had been thoroughly rubbed with
leaves and soil. I tested potential effects of the labelling technique
on dispersal agents by setting up four pairs of fruit clusters (25
fruits per cluster) on the secondary forest site. Dispersal was arbi
trarily defined as movement of more than 50 cm. No significant differ
ence (p = .19) was found in dispersal of labelled and unlabelled fruits,
according to the randomization test for matched pairs (Siegel 1956).
Predation. To monitor colonization by seed-predaceous beetles, six
fruit clusters were set up on the secondary forest site. I prevented
dispersal by stapling the apices of the fruits to a nylon line (10
fruits per line) attached to a stake. Twenty fruits were gathered from
each cluster at intervals of 0, 10, 20, 40, 80, and 160 d following
initiation of the experiment (27 October 1981). The fruits gathered
from each cluster were isolated in closed sacks to prevent further
predator colonization. After air drying, the fruits were opened to
determine total number of seeds, number of attacked seeds, and number of
beetle larvae per fruit.
Seed fate. To monitor the fate of seeds over time under various
environmental conditions, five fruit clusters were set up in each of


Table 14. Leaf area index (LAI) of babassu and competitors in four
experiments. Measurements were obtained during May 1982 in
ca. 18-mo-old vegetation. Values are expressed as LAI; mean
percent of total LAI in parentheses. Significant
differences are shown as (p f. .05) or ## (p .01).
Irrigation Fertilizers Insecticides Climate
No
Yes
No
Yes
No
Yes
Drv
Wet
COMPETITOR LAI
Grasses
1.2
1.0
0.7
0.8
1.0
1.2
0.8

1.7
Vines
0.6
0.3
0.2
0.1
1.2
* 0.2
0.2
0.3
Others
3.2
2.3
2.1
2.0
2.5
** 3.9
3.8
**
1.3
Total Competitors
5.0
3-6
3.0
2.9
4.7
5.3
4.8
**
3.3
BABASSU LAI
0.18
0.24
0.18
0.20
0.25
0.20
0.19
**
0.29
(3.5)
(6.3)
(5.8)
(6.4)
(5.1)
(3.6)
(3.8)
**
(8.1)
TOTAL LAI
5.2
3.8
3.1
3.1
4.9
5.5
5.0
*
176


117
Table 18. Definition of 12
STAGE
life stages
of babassu used in study.
BEGINNING
DescriDtion
Number
Seeds
1
Abscission of fruit
Seedlings
2
Appearance of first
leaf following germination
Pre-established juveniles
3
Initial division of leaf
blade
Etablished juveniles
4
Leaflets > 100 per side
5
Height 100 cm
6
Height > 200 cm
Mature palms
7*
Height _> 500 cm
8
Height > 1000 cm
9
Height > 1500 cm
10
Height > 2000 cm
11
Height > 2500 cm
12
Height > 3000 cm
Stage includes palms that had not flowered and thus were not
reproductively mature.


45
three sites: secondary forest, pasture, and field. I prevented dis
persal by stapling the fruits to nylon lines attached to a stake. Twen
ty fruits per pile were gathered at the outset of the experiments (27
October 1981) and at 3-mo intervals during 1 yr.
I obtained a long-term (ca. 18 mo) measure of seed fate from fruits
used in the irrigation experiment (described in Chapter 4) At the end
of this experiment, 20 fruits were gathered from each of five weeded
(sun) and five non-weeded (shade) plots that had not been irrigated. In
both seed fate experiments, gathered fruits were stored in closed sacks,
air-dried, and opened to determine numbers of eaten, defective (aborted
or rotted), germinated, and quiescent (viable but not germinated) seeds.
Germination. An experiment was set up on the secondary forest site
to monitor germination over time and to test for between-palm variabili
ty in fruit germination. Recently fallen fruits (100 from each of 10
palms) were dehusked to inhibit dispersal and planted in 40 plots (25
fruits per plot, four plots per palm). At ca. 1-mo intervals, I meas
ured germination by checking the base of each fruit for emergence of one
or more embryonic axes (i.e., plumule, cotyledonary node, and radicle).
I conducted additional experiments to test the effects of fire,
herbivory, and shading on germination of babassu fruits. These experi
ments were established on a cleared, periodically weeded site ca. 50 m
from secondary forest on well-drained soil. The fruits used in each of
these experiments (100 per treatment) were obtained from a single palm.
Unless otherwise specified, fruits were dehusked and placed on plots
shaded with suspended frames upon which babassu leaves were tied. At 1-


138
stem length. As secondary conductive tissue is absent, the phloem is
especially susceptible to mechanical stress. In taller palms such as
babassu, eventual deterioration of long-distance phloem transport is
probably the principal cause of senescence (Zimmermann 1973). The
apparent lack of senescence in A. mexicanum, which attains a maximum
stem length of 7 m (Sarukhn 1980), may be due to reduced mechanical
stress on the phloem. In the tree-sized J. bataua, senescence appar
ently does occur (Balick 1980), but its duration and effects on parti
tioning of productivity within the palm are unknown.
The ecological and evolutionary consequences of babassu's relative
ly high allocation of productivity to reproductive structures remain to
be examined. High allocation to such structures does not necessarily
translate into increased production of seeds. For example, I calculate
that babassu produces an average of ca. 99,000 seeds over its lifetime;
the once flowering Corypha elata, with a far lower absolute as well as
relative allocation to reproductive structures, produces an estimated
minimum of 240,000 seeds (Tomlinson and Soderholm 1975). Of the total
dry mass of reproductive structures produced over the lifetime of a
babassu palm, only 5.0% is directly allocated to reproduction in the
form of seeds. More than twice this amount (11.0%) is invested in male
inflorescences. Production of male and female flowers on separate
inflorescences enhances the probability of outcrossing, which in turn
leads to increased genetic variability within populations of babassu
(see Chapter 3). Such genetic variability permits the occurrence of
locally adapted ecotypes (see Chapter 4). Most of the remaining


74
The abundance of seedlings and juveniles on burned sites has led to
the suggestion that fire promotes germination in babassu (Smith 1974)*
Removal of the mesocarp by dispersal agents and shading could also
stimulate germination in the palm. Results of experiments designed to
test these hypotheses are shown in Table 6. I found that a moderate
fire, such as occurs annually in pastures, had no effect on germination
success. However, fruits exposed to an intensive fire, such as occurs
when forest fallow is burned in shifting cultivation, exhibited signifi
cantly reduced germination (p £.01, Chi-square test). Removal of the
mesocarp, simulating the effects of dispersal agents, was found to
promote germination of babassu (p £.05). Likewise, germination was
significantly higher in shaded fruits than in fruits left in the sun
(p£.01). This final response probably results from enhancement of
germination by higher soil moisture levels in the shade.
Discussion
Phenology
Babassu's phenology showed remarkable constancy over a wide range
of ecological conditions. The general pattern was a pronounced peak in
leaf production and flowering during the wet season, while leaf loss and
fruiting peaked during the dry season. Despite its dry-season peak,
flowering within a given population occurred in most months. Male
inflorescences were generally more abundant than female inflorescences,
and I could detect no seasonal variation in sex ratios within the popu
lations observed.


Table 22.
SITE
Primary Forest
Secondary Forest
Annual above ground dry matter production of mature babassu
palms
on three study sites
at Lago Verde.
PRODUCTIVITY PER PALM (kg/yr)
DENSITY
PRODUCTIVITY :
(t*ha-^ *yr'
Stem
Leaves
Reproductive
Structures
Total
(palms/ha)
18.1
48.8
22.0
88.9
48
4.3
18.4
48.7
20.7
87.8
152
13.3
25.2
66.7
40.2
132.1
88
11.6
Pasture


161
figure at the latter date was 3.6 yr (IBGE 1979). The reduced fuel
accumulated during shortened fallows is inducing shifting cultivators to
thin the stands more intensively in order to generate sufficient fuel
for hot fires.
With degradation of sites used for shifting cultivation, the rural
poor are becoming increasingly dependent on pasture stands of babassu as
a crucial source of income and subsistence products. Ranchers, in turn,
increasingly perceive the people who gather babassu fruits as inter
fering with pasture management. Fruit gatherers are blamed for starting
wildfires, cutting fences, and leaving behind fruit husks that damage
the hooves of cattle (Anderson and Anderson 1983). Probably as a conse
quence, clearcutting of babassu forests by ranchers is becoming in
creasingly common (Fig. 46).
The current breakdown in land-use systems associated with babassu
is only part of a long history of disappointment. The sheer extent of
the stands has understandably tended to inspire optimistic estimates of
their potential contribution to the national and global economy. To
cite just one of many examples, it was recently calculated that full
industrial utilization of babassu fruits would result in the annual
production of 1 x 10^ L of ethanol, 2 x 10^ t of charcoal, 0.5 x 10^ t
of vegetable oil, and 2 x 10^ m-^ of combustible gases; all this would be
equivalent to the constant generation of 5 x 10^ W over a 1-yr period
(STI 1979). These estimates appear to be technologically feasible, due
to recent development of machinery capable of processing the extremely
hard fruits of the palm (STI 1979).


Figure 9. Maranho and neighboring states, with locations of
research sites.


87
below ground even though the seed germinates on the surface. In the
case of babassu, the underground meristem in seedlings and stemless
juveniles enables them to regenerate on sites that are subject to cut
ting and burning. Many species exhibiting cryptogeal germination occur
in fire-prone habitats, and Jackson (1974) proposed that it represents
an evolutionary response to fire. However, I suspect that babassuas
well as many of its relatives in the Attalea allianceevolved in moist,
forested habitats where natural fires were infrequent (see Chapter 2).
The presence of cryptogeal germination in these species is more likely
due to the general lack of secondary growth throughout the family.
Lateral expansion of the palm stem is accomplished by so-called estab
lishment growth (Tomlinson and Zimmermann 1966), during which the stem
takes on a highly unstable, obconical shape. One of the means by which
palms provide support for this structure is to bury their shoot axes
underground.
Regardless of the selective pressures that brought about its evolu
tion, there is no doubt that cryptogeal germination plays an important
role in the current proliferation of babassu on sites subjected to human
disturbances.


8
ized by industries for production of vegetable oil, soap, and feedcake.
These industries represent the largest single source of tax revenues in
the Maranhao economy. Over the past 60 yr, kernel production has stead
ily increased (Fig. 6), and as of 1979 babassu provided 89% of all
vegetable oil obtained from non-domesticated sources in Brazil (IBGE
1981a). Recently introduced machinery capable of mechanically breaking
the hard fruits has enabled additional babassu-derived products (e.g.,
charcoal, coke) to enter national markets (Fig. 7), and technology
presently exists to derive a host of additional products from the fruits
(Fig 8).
Despite its economic importance, babassu is not currently domes
ticated, and formation of the stands remains a spontaneous phenomenon.
In the past, stands of babassu were generally preserved and formed an
integral part of shifting cultivation systems. In recent years, how
ever, population pressures and changing land use pabterns are inducing
both shifting cultivators and ranchers to clearcut the stands (Anderson
and Anderson 1983). Sound management of the palm forest requires basic
information concerning the biology of babassu. The purpose of this
study is to provide such information.
The study covers babassu's taxonomy (Chapter 2), life history
(Chapters 3-5), and population biology (Chapter 6); levels of inquiry
change respectively from the species to individual palms to stands of


52
logies is probably due to the physical association of the two struc
tures; each inflorescence is attached to the inner (adaxial) surface of
a leaf sheath. However, a ca. 1-yr lag occurred between a leaf's emerg
ence in the crown and the opening of its associated inflorescence.
Earlier inflorescence expansion was apparently prohibited by the tight
packing of newly emerged leaves near the top of the crown. Only contin
ued growth of the palm during the following rainy season loosened this
packing sufficiently to permit inflorescence expansion.
Dissection of crowns showed that each leaf base contains a single
inflorescence. Under suitable conditions, all inflorescences ultimately
develop to anthesis, thus resulting in an inflorescence-to-leaf produc
tion ratio of 1. Production data over the entire period of observations
revealed high ratios for the dry site at Caxias (1.02) and the pasture
at Lago Verde (1.04). (The possibility of obtaining inflorescence-to-
leaf production ratios greater than 1 results from counting these struc
tures simultaneously. A more accurate ratio would be obtained by com
paring production of leaves in a given year and production of their
corresponding inflorescences during the following year.) Shading prob
ably caused intermediate ratios on the primary forest (.92) and second
ary forest (.89) sites at Lago Verde. The low ratio at the wet site
(.69) can be attributed to the leaf harvest 2 yr previously. Babassu's
apparent response to defoliation was to increase inflorescence abortion
rather than to decrease leaf production.
No difference in the phenology of male and female inflorescences
was detected. Approximately 20% of total inflorescences were female in


153
When the forest was originally cleared, these palms would have been ca.
40 yr old, or less than 1 m tall according to my growth rate estimates
for primary forest (see Table 19). The apical meristem would still be
beneath the soil surface in these palms; its emergence in babassu only
occurs when the shoot is ca. 1.8 m tall (see Chapter 5). Such palms
would consequently survive cutting and burning of the site, while
slightly taller palms would be prone to high mortality due to exposure
of their apical meristems above the soil surface. This appears to
account for the high frequency of the 73-yr-old cohort, as well as the
low frequency of palms older than 73 yr on the site. The consistency
between purported age structure of the population and the precisely
dated disturbance of the site provides independent verification of the
age estimates reported in Chapter 5.
Data on population dynamics in the secondary forest (Fig. 44)
indicate that recruitment of seedlings, juveniles, and adults was more
than sufficient to offset losses due to mortality. Thus, at 33 yr after
forest clearing, the babassu population still appeared to be growing.
Pasture
In comparison to the secondary forest, the pasture population
showed an even more pronounced bulge of stage 7 palms, apparently empha
sized by relatively low frequencies of stages 5-6 (Fig. 41 ) The result
was a distinctly two-tiered stand characteristic of babassu in pastures:
a dense understory of juveniles < 1 m tall with an overstory of mature
palms that was more or less uniform in height. The overstory palms were
probably released simultaneously after the original forest clearing.


\
2
o
10
I 5
20
25
20
Figure 1. Approximate distribution of high-density stands of
babassu (Orbignya martiana) in Brazil. States con
taining stands are labelled. Adapted from Mendes
and Carioca (1981).


17
cient for rain-fed agriculture on upland sites (Fig. 13). Mean monthly-
temperature and relative humidity are subject to moderate variations in
response to rainfall; the former ranges from 25.7C (February) to 28.4C
(November), and the latter from 67.8% (October) to 85.2% (January-
February). Throughout the region, the terrain is flat to gently rolling
and is dissected by numerous streams that feed into the Mearim River.
The predominant soil (Fig. 10) is the Eutrophic Red-Yellow Podzolic type
(Brazilian classification system), roughly equivalent to a Ferric Luvi-
sol (FAO legend), or an Ustalf or Udalf (U.S.D.A. Soil Taxonomy). This
soil is characterized by high base saturation ( 50%) and relatively high
fertility compared to other regional soil types; soil structure is
generally excellent (EMBRAPA 1975).
Settlement. Even after the first wave of non-indigenous settlers
arrived during the 1920s, the region remained thinly populated. At that
time, Babacal was a small village, named after an adjacent stand of
bacaba palms (Oenocarpus bacaba) that has long since disappeared. Hunt
ing, fishing, gathering, and shifting cultivation continued to be the
principal economic activities. During the 1950s and 1960s, droughts in
the Brazilian northeast and roadbuilding throughout the state brought
waves of migrants into the region. Populations soared, and by 1980 the
town of Bacabal contained 43,229 inhabitants (IBGE 1981b) and was the
fifth largest urban center in Maranhao. A roadbuilding program begun
during the 1960s linked the major towns of the Mearim Valley with a
well-developed system of roads. Influx of settlers, enhanced access,
ample rainfall, and excellent soils transformed the region into one of


25
informants reported that the site was still covered by structurally
complex, species-rich forest when they first arrived in the region
during the 1930s. Since this time, the site has been subjected to
minimal disturbance. At the time of my study, emergents attained 30-35
m, and the canopy was more or less continuous at 15-25 m. Most of the
emergent species were deciduous during all or part of the dry season.
Vines were relatively abundant, whereas epiphytes were rare. Species
richness, at 63 species (dbh iL10 cm) per ha, was lower than that of
upland forests in Amazonia (Anderson and Benson 1980). With relatively
high abundance (20.2%), frequency (9.2%), and dominance (27.4%), babassu
was the most important species on the site.
The secondary forest was originally cut and burned for shifting
cultivation in 1948; the site had suffered no major disturbance since.
Emergents attained 20-25 m, and the canopy comprised largely of a
uniform stand of babassuwas relatively continuous at 10-15 m. Species
richness was 31 species (dbh ^.10 cm) per ha. The percentage of total
species that occurred in both the primary and secondary forest sites
(coefficient of community: Whittaker 1975) was 31.9%. In abundance
(66.7%), frequency (21.7%), and dominance (84.1%), babassu was far and
away the most important species on the site.
The pasture was established through introduction in 1974 of the
exotic grass, Hyparrhenia rufa, on a site that had already undergone
several cycles of shifting cultivation. The abundant understory compet
itors (including Lantana camara, Stachytarpheta cajanenesis. Sida spp.,


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
LIST OF TABLES
Uses of occupied lands in Maranhao during 1950 and 1975 ... 22
Nomencatural history of babassu and its suspected hybrids. 30
Fruit yields of babassu per ha and per palm during 1980-81
and 1981-82 at three sites 60
Sources of increased fruit yields per ha between 1980-81
and 1981-82 at three sites 61
Attack of babassu fruits and seeds by the seed-predaceous
beetle, Pachymerus ncleo rum 68
Germination of babassu in four experiments 73
Preliminary checklist of insects that feed on babassu ... 80
Hypothesized success of babassu in four experiments 89
Palms per m^ (a), leaf area per palm (b), and leaf area
index (c) of babassu at end of irrigation experiment 102
Palms per m^ (a), leaf area per palm (b), and leaf area
index (c) of babassu at end of fertilizer experiment 103
Palms per m^ (a), leaf area per palm (b), and leaf area
index (c) of babassu at end of insecticide experiment . .104
Palms per m^ (a), leaf area per palm (b), and leaf area
index (c) of babassu at end of climate experiment, with
climate and weeding as main effects 105
Palms per m^ (a), leaf area per palm (b), and leaf area
index (c) of babassu at end of climate experiment, with
climate and ecotype as main effects 106
Leaf area index (LAI) of babassu and competitors in four
experiments .107
Dry mass of babassu and competitors in four experiments . .108
vii


101
(Table 16); only the relatively immobile P (in the form of P2O5) was
significantly higher (p <_.05) on the fertilized plots.
Insecticide experiment
Application of insecticides produced no effect on density (Table
11a), a significant reduction in babassu's growth (p £_ 0.01, Table 11b),
and a substantial reduction in its LAI (p .07, Table 11c). As the
latter two effects were consistent on weeded as well as non-weeded
plots, I conclude that one or both of the insecticides used in the
experiment may have a debilitating effect on babassu's growth.
With competitors present, relative LAI (Table 14) and relative
biomass (Table 15) of babassu were substantially higher when insecti
cides were not applied. These results appear to be due to an apparently
negative effect of the insecticides on babassu combined with their
apparently positive effect on the vegetation as a whole. All components
of competitor dry mass were higher (Table 15)and LAI of competitors
other than grasses significantly higher (p £..01, Table 14)when
insecticides were applied. However, plots without insecticides were
characterized by a significantly higher LAI of vines (p £.05, Table
14). Application of insecticides thus resulted in reduced dominance by
vines, higher overall competitor vigor, and absolute and relative
decreases in babassu's performance.
Climate experiment
In addition to the presence of competitors, climate had a pro
nounced effect on germination in babassu, as indicated by the high
density of palms on wet climate plots (Tables 12a and 13a). Weeded


97
Figure 34. Palms per (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in fertilizer experiment.
Bars represent 1 standard error.


178
Waterhouse, J. T., and C. Quinn. 1978. Growth patterns in the stem of
the palm Archontopheonix cunninghamiana. Botanical Journal of the
Linnean Society 77: 73-93.
Watt, A. S. 1947. Pattern and process in the plant community. Journal
of Ecology 35: 1-22.
Webb, L. J. 1958. Cyclones as an ecological factor in tropical lowland
rain forest, north Queensland. Australian Journal of Botany 6:
220-228.
Wessels Boer, J. G. 1965. The indigenous palms of Suriname. In J.
Lanjouw, editor. Flora of Suriname. E. J. Brill, Leiden, The
Netherlands.
Whitmore, T. C. 1982. On pattern and process in forests. Pages 45-59
in E. I. Newan, editor. The plant community as a working mech
anism. Blackwell Scientific Publications, Oxford, England.
Whittaker, R. H. 1975. Communities and ecosystems. Second edition.
Macmillan Publishing Company, New York, New York, USA.
Willson, M. F. 1983. Plant reproductive ecology. John Wiley and Sons,
New York, New York, USA.
Zeven, A. C. 1967. The semi-wild oil palm and its industry in Africa.
Dissertation. University of Wageningen, Wageningen, The Nether
lands.
Zimmermann, M. H. 1973. The monocotyledons: their evolution and com
parative biology. IV. Transport problems in arborescent mono
cotyledons. Quarterly Review of Biology 48: 314-321.


Ill
plots under the dry climate contained a significantly lower palm density
than any other treatment combination (p £.01, Table 12a). Like
irrigation, however, the wet climate was less effective in promoting
germination than competitor presence, as indicated by the significantly
lower (p £.05) density of palms on weeded, wet climate plots than on
non-weeded plots under either the wet or dry climate.
Responses in individual growth were complex. Leaf area per palm
was significantly higher on weeded, dry climate plots than under any
other treatment combinationincluding weeded, wet climate plots
(p £..01,' Table 12b). This suggests that conditions under the dry
climate were more favorable for babassu's growth, possibly due to higher
soil fertility (Table 17). However, there were also significant
differences in ecotypic responses (Table 13b). Under both climates,
individual leaf area was significantly higher (p £ .05) in palms
produced from local fruits. But the data on leaf area per palm indicate
that dry ecotype palms performed less well in the wet climate thn wet
ecotype palms in the dry climate. This suggests that some compensatory
factor such as soil fertility (Table 17) was operating at the dry
climate site.
Leaf area index was significantly higher at the wet climate site
(p £ .01, Table 12c), which is attributable to increased germination
under the wet climate. In both climates, LAI was significantly higher
in palms produced from local fruits (p £ .05, Table 13c).
With competitors present, relative LAI (Table 14) and biomass
(Table 15) of babassu were significantly higher under the wet climate


REFERENCES CITED
Abraham son, W. G. 1979. Patterns of resource allocation in wildf lower
populations of fields and woods. American Journal of Botany 66:
71-79.
Abreu, S. F. 1940. Coco babapu e o problema do combustivel. Second
edition. Instituto Nacional de Tecnologa, Rio de Janeiro, Brazil.
Anderson, A. B., and E. S. Anderson. 1983. People and the palm forest:
biology and utilization of babassu forests in Maranhao, Brazil.
Final report to USDA Forest Service, Consortium for the Study of
Man's Relationship with the Global Environment, Washington,
District of Columbia, USA.
, and W. W. Benson. 1980. On the number of tree species
in Amazonian forests. Biotropica 12: 235-237.
Anonymous. 1981. Mapeamento das ocorrncias e prospecpao do potencial
atual do babapu no Estado do Maranhao. Companhia de Pesquisa e
Aproveitarnento de Recursos Naturis (COPENAT) and Fundapao Insti
tuto Estadual do Babapu (INEB), Sao Luis, Brazil.
Balick, M. J. 1980. The biology and economics of the Oenocarpus-
Jessenia (Palmae) complex. Dissertation. Harvard University,
Cambridge, Massachusetts, USA.
, A. B. Anderson, and M. F. da Silva. 1982. Palm taxonomy
in Brazilian Amazonia: the state of systematic collections in
regional herbaria. Brittonia 34* 463-477.
Bannister, B. A. 1970. Ecological life cycle of Euterpe globosa
Gaertn. Pages B299-B314 in H. T. Odum and R. F. Pidgeon, editors.
A tropical rain forest. Atomic Energy Commission, Washington,
District of Columbia, USA.
Barbosa Rodrigues, J. 1891. Plantas novas cultivadas no Jardim Botnico
do Rio de Janeiro (Rio de Janeiro, Brazil) 1:32.
. 1898. Palmae Mattogrossenses novae vel minus
cognitae. Rio de Janeiro, Brazil.
. 1903. Sertum palmarum Brasiliensium. Brussels,
Belgium.
168


119
secondary forest sites over a 1-yr period (1 April 1981 31 March
1982). Due to grazing, cutting, and burning, productivity measurements
in the pasture were limited to a 0.5-ha sample of mature palms.
To account for population turnover during the 1-yr period, I ob
tained data on population-wide mortality and recruitment during censuses
carried out at 6-mo intervals at each site (see Chapter 6). Productivi
ty of palms that died during the first interval was not calculated;
productivity of palms that died during the second interval was calcu
lated for the first interval. In the case of newly recruited seedlings,
productivity was calculated as estimated growth by the final census.
Annual height increments of palm stems were measured in the above-
defined sampling areas of each population. To obtain a regression
between stem biomass and height, I randomly selected seven stems for
felling and measurement of dimensions; three subsamples per stem (ca.
25-cm high cross sections from the basal, middle, and apical portions of
each stem) were oven-dried at 70C and weighed. In palms <1.78 m tall,
the stem (i.e., apical meristem) had not emerged from underground. The
regression obtained for palms M.78 m tall was:
o
stem biomass (kg) = 46.1 stem height (m) 82.1, r = 0.99.
Using height increment data and this regression, annual stem production
was calculated for palms in the sampled areas and corrected to a per-ha
basis.
The mean number of leaves produced per year in each life stage was
determined by monitoring leaf production in the above-defined areas of
each population. To obtain data on leaf mass, one newly expanded leaf


170
. 1932. Attalea cohune Mart, wirklich eine Qrbignya. Notiz-
blatt 11: 688-690.
. 1938. Palmae Brasiliensis. Notizblatt 14 : 231-260.
Burtt, B. D. 1929. A record of fruits and seeds dispersed by mammals
and birds from the Singida District of Tanganyika Territority.
Journal of Ecology 17: 351-355.
Cannell, M. G. R. 1982. World forest biomass and primary production
data. Academic Press, London, England.
Child, R. 1974. Coconuts. Second edition. Longman, London, England.
Churchill, G. B., H. H. John, D. P. Duncan, and A. C. Hodson. 1964.
Long term effects of defoliation of aspen by the forest tent cater
pillar. Ecology 45: 630-633.
Connell, J. H. 1978. Diversity in tropical rain forests and coral reefs.
Science 199: 1302-1310.
Cook, R. E., and E. E. Lyons. 1983. The biology of Vi ola fimbriatula
in a natural disturbance. Ecology 64: 654-660.
Corley, R. H. V., J. J. Hardon, and B. J. Wood. 1976. Oil palm re
search. Elsevier Scientific Publishing Company, Amsterdam, The
Netherlands.
Corner, E. J. H. 1966. The natural history of palms. University of
California Press, Berkeley, California, USA.
Costa Lima, A. M. 1967-1968. Quarto catlogo dos insetos que vivem as
plantas do Brasil, seus parasitas e predadores. Departamento de
Defesa e Inspepo Agropecuaria, Ministerio da Agricultura, Rio de
Janeiro, Brazil.
Dransfield, J. 1982. Pinanga deis tan tha, a new species with hidden
flowers. Principes 26: 126-129.
Drude, 0. 1881. Palmae. In C. F. P. von Martius, editor. Flora
Brasiliensis 3: 253-284. Munich, Germany.
EMBRAPA. 1975. Mapa esquemtico dos solos das regioes norte, meio-
norte e centro-oeste do Brasil: texto explicativo. Ministerio da
Agricultura, Boletim Tcnico 17, Rio de Janeiro, Brazil.
. 1981. Mapa de solos do Brasil. Servipo Nacional de Levanta-
mento e Conservapo de Solos, Rio de Janeiro, Brazil.


CHAPTER 2
TAXONOMY AND PHYTOGEOGRAPHY
Babassu belongs to the cocosoid group, one of 15 major groups
within the palm family (Moore 1973b). Cocosoid palms are distinguished
from the other groups by their fruits, which contain a thick, generally
three-pored endocarp. The principal center of cocosoid diversity is
South America, where 24 of 28 genera occur (Moore 1973b). Two well-
known representatives of the group are the coconut (Cocos nucifera) and
the African oil palm (Elaeis guineensis).
The taxonomy of babassu has been a source of confusion since the
palm was first described well over a century ago. This confusion begins
on the generic level. Although babassu is generally assigned to the
genus Orbignva. the status of this and the four to five other genera in
the Attalea alliance has been questioned by Wessels Boer (1965). These
genera are distinguished solely on the basis of staminate flower mor
phology. No correlation with other characters has been observed, and
staminate flower morphology runs across a separation based on other
characters such as endocarp pores. Lines between genera in the Attalea
alliance are further blurred by collections of intermediate staminate
flower types, which have prompted the description of two new genera; one
of these, Markleva. was described by Bondar (1957) as a possible inter
generic hybrid involving babassu.
A classification based solely on staminate flower morphology is of
little use in the field, as most species flower only for a short time.
27


43
occurs in babassu, flowers were left bagged throughout the flowering
period, and I removed the pistils from others; both sets of flowers were
subsequently checked for fruit set.
Abundance and behavior of insect visitors to both inflorescence
types were determined by direct observation and collection. Sticky
traps were used to monitor insects between observations. Insect speci
mens were identified at the Museu Paraense Emilio Goeldi in Belem;
unidentified specimens were sent to specialists in Brazil and the U.S.
To determine whether wind pollination occurs in babassu, sleeves
were designed to exclude insect visitors from female inflorescences,
while permitting penetration of wind-borne pollen. The sleeves were
made out of nylon netting with a ca. 0.5-mm mesh, sufficient to exclude
all observed insect visitors. Dispersal of pollen by wind was monitored
through the use of pollen traps, which were distributed at regular
intervals on both the secondary forest and pasture sites. The trap
consisted of a microscope slide covered with a thin coat of vaseline;
the slide was placed inside a horizontally oriented section of 15-cm-
wide PVC pipe secured to a stake or stem. After a 1-wk exposure, the
slides were removed, covered, and subsequently examined in the labora
tory.
Fruit Biology
Dispersal. I monitored removal of babassu fruits in the pasture
and secondary forest, as well as on a nearby field. This latter site had
been cut, burned, and planted at the end of 1980. During the course of
the study, the site remained virtually abandoned except for sporadic


155
species characteristic of early successional communities. Very few
forest species are characteristic of both extremes; most seem to fall
somewhere in between and regenerate following moderate disturbances
(Marks 1974, Cook and Lyons 1983). Babassu is extraordinary in its
capacity to regenerate and persist over an extremely wide gradient of
disturbances, ranging from occasional gaps in mature forest to periodic
clearing and burning of the forest by shifting cultivators.
Babassu's capacity to maintain populations in primary forest re
sults from: (1) high seedling recruitment, and (2) persistence of both
seedlings and juveniles in the understory. High seedling recruitment is
probably due to the palm's relative success in escaping seed predation
and its enhanced germination under shade (see Chapter 3). Persistence
of seedlings and juveniles results from their capacity to survive under
competitively crowded conditions. For example, the mean life expectancy
of seedlings in the primary forest site was calculated at 5.0 yr (Table
24), which indicates that seedlings are able to persist long after their
seed endosperm reserves have run out. As a result of high recruitment
and persistence, considerable densities of seedlings and juveniles are
available to exploit opportunities such as light gaps or release from
root competition.
The ideal conditions for growth and maintenance of babassu occur
under shifting cultivation. Initial cutting and burning of primary
forest release a cohort of juvenile palms from competitive suppression.
Respite during subsequent forest regrowth provides time for the cohort
to pass through the critical period of stem emergence from the soil,


BIOGRAPHICAL SKETCH
Anthony Bennett Anderson was bom on July 13, 1950, in Oak Bluffs,
Massachusetts. In 1973 he received the degree of Bachelor of Arts in
natural science from Amherst College, Amherst, Massachusetts. After
spending 18 months as a Thomas J. Watson Fellow in Latin America, he was
employed as a researcher in the Botany Department of the Instituto
Nacional de Pesquisas da Amazonia (INPA) in Manaus, Brazil. In 1978, he
received the degree of Master of Science in tropical botany from INPA.
His thesis was on the ecology of white-sand vegetation in Amazonia.
After completing his doctorate, he will assume a position on the staff
of the Museu Paraense Emilio Goeldi in Belem, Brazil. Current research
interests include reproductive ecology of economic plants and land use
practices in the humid tropics.
194


Appendix B continued.
ABUNDANCE FREQUENCY DOMINANCE IMPORTANCE
3 P E C- I fi ..§stems/i-i3 1 2 or/.ha 2 2
30. Inga alba Willd. 1
31. Diospvros guianensis (Aubl.) Gurke* 1
TOTAL 366
0.27
1.09
0.008
0.03
0.46
0.27
1.09
0.007
0.03
0.46
99.95
100.03
25.803
99.99
100.00
Note: Asterisks indicate species that also occur on 1 ha of primary forest nearby.


63
In 21 female inflorescences observed, overlap of receptivity (0-48 h
after bract opening) with pollen release from a male inflorescence on
the same palm occurred only once.
The possibilities of inducing self-pollination in babassu are un
known. As inflorescences containing pistillate flowers are either func
tionally female or (more rarely) protogynousand opportunities for
pollen transfer between inflorescences on the same palm are rare self-
pollination is probably unusual under natural conditions.
Insect pollination. Presentation of complete data on abundance and
behavior of insect visitors to inflorescences of babassu awaits identi
fication of the specimens sent to specialists. A high diversity and
abundance of insects were observed visiting male infloresences, which
are odoriferous and produce copious pollen. Conversely, relatively low
insect diversity and abundance were found on female inflorescences, the
pistillate flowers of which give off no noticeable scent and do not
produce nectar. The only insect that occurred in abundance on both
inflorescence types was an undescribed species of Mystrops. This ca. 2-
mm long nitidulid beetle was found in sticky traps placed among the
leaves of palms that were not bearing inflorescences, which suggests
that it lives in the crowns of babassu. Prior to opening of either
inflorescence type, the beetles were observed to gather in the crown by
the hundreds or sometimes thousands. In a sample of flowers obtained
from an open male inflorescence, I consistently found more than one
beetle per flower; subsequent calculations produced an estimate of
48,300 staminate flowers on the inflorescence. The beetles were ob-


1 (o)
o
Figure 33. Palms per m (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in irrigation experiment.
Bars represent 1 standard error.


39
Methods
Phenology
I carried out monthly phenological observations of babassu from
January 1981 to March 1982. All adult palms at each site were labelled
and mapped. On the intensive study sites at Lago Verde, the sample
consisted of all adults in a 0.5-ha area; in each of the satellite sites
(Pindare Mirim and Caxias), 60 adults were randomly selected for obser
vation. The former site had been recently converted to pasture, whereas
the latter was in fallow following its use for shifting cultivation in
1979-80. These sites are representative of current management practices
of babassu stands in the seasonally wet savanna (Pindar Mirim) and dry
woodland (Caxias) ecological zones of Maranhao. On the site at Pindare
Mirim, leaves of babassu had been harvested for thatch ca. 2 yr prior to
initiation of the study.
To permit observations of leaf production and loss, the palms were
climbed and green leaves were labelled by spraying paint on the under
sides of the rachises. To enhance the visibility of newly emerged
leaves, ca. 10 basal leaflets per side (representing ca. 5% of the
total) were removed from marked leaves. I calculated leaf production
per month by counting the newly formed (unlabelled) leaves that had com
pletely expanded in the crown; loss per month was determined by counting
the number of labelled leaves that liad fallen to the ground. Fallen
leaves were cut in half after being counted so as to prevent being
recounted in subsequent months, and the leaves of adjacent palms were


Table 6. Germination of babassu fruits in four
experiments. Asterisks indicate significant
differences (* for p S05, ** for p.S.01).
EXPERIMENT
.QJO.M..I N A T I 0 N (ft)
Treatment Control
Removal of Mesocarp
54

34
Intensive Burn
10
3*
54
Moderate Burn
58
54
Shading
66
*
0


98
. Palms per (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in insecticide experiment.
Bars represent 1 standard error.
Figure 35


32
A first-hand familiarity with the taxa in the field provided a
sound basis for reviewing the extensive and often contradictory litera
ture on nomenclature. Results pertaining to the morphology and distri
bution of taxa in the babassu complex represent a synthesis of data
already published, data obtained in the field, and morphological
comparisons of specimens in the laboratory.
Results and Discussion
Examination of data on specimens obtained over a substantial area
indicates that the babassu complex is comprised of one highly variable,
widely distributed species and two associated species that appear to
hybridize with the former. Babassu is henceforth used exclusively in
reference to the principal species of the complex, which was the subject
of the ecological research described in the following chapters.
Nomenclature
The scientific name of babassu tentatively utilized in this study
is Orbignva martiana Barb. Rodr. (Table 2). Its valid publication by
Barbosa Rodrigues (1898) represents the first taxonomic description that
can be definitely attributed to babassu. Previously published names
associated with babassu fAttalea speciosa (Mart.) Barb. Rodr. and Orbig
nva lvdiae Drude] were based on incomplete collections, and their accom
panying descriptions were either insufficient or incorrect; neither can
be definitely attributed to babassu. Barbosa Rodrigues (1903) subse
quently reduced 0. martiana to synonymy under the new combination, 0.
speciosa (Mart.) Barb. Rodr. Although widely accepted in Brazil, the
latter name is invalid because the same combination (Qi_ speciosa Barb.


109
Table 16. Soil chemistry of four treatments at end of nutrient
experiment. Values are mean (standard deviations) of
five samples from 0-15 cm, each a composite of 10 cores.
FERTILIZED
NOT FERTILIZED
Weeded Non-weeded
Weeded Non-weeded
C (%)
0.59
(0.10)
0.58
(0.10)
0.53
(0.13)
0.67
(0.13)
Organic matter (%)
1.02
(0.17)
1.00
(0.17)
0.90
(0.22)
1.12
(0.20)
N (%)
0.07
(0.01)
0.07
(0.01)
0.06
(0.02)
0.07
(0.01)
C/N
8.4
8.3
8.8
9.6
P205 (ppm)
3.24
(0.60)
2.89
(0.83)
1.65
(1.30)
1.13
(0.35)
pH in H2O
5.58
5.92
5.62
5.92
pH in KC1
4.94
5.24
4.82
5.40
Ca + Mg (meq/lOOg)
2.97
(0.50)
3.22
(1.36)
2.50
(0.82)
3.80
(1.02)
Na (meq/lOOg)
0.08
(0.06)
0.07
(0.01)
0.08
(0.05)
0.13
(0.12)
K (meq/100g)
0.24
(0.06)
0.35
(0.14)
0.24
(0.08)
0.42
(0.06)
H (meq/100g)
2.04
(0.60)
3.01
(1.60)
1.84
(0.25)
1.75
(0.53)
A1 (meg/100g)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
0.00
(0.00)
Total Bases (meq/100g)
5.34
(0.94)
5.32
(0.92)
4.66
(1.01 )
6.09
(1.21)
Base Saturation (%)
62.0
(6.5)
66.4
(10.3)
59.6
(5.7)
70.4
(9.3)


65
than in the field or pasture (Fig. 25). In the secondary forest, fruits
were probably removed by pacas (Agouti paca) and agoutis (Dasyprocta
punctata). In the field and pasture, removal, was effected either by
domesticated animals or by runoff.
Predation. The only non-human seed predator of babassu observed at
my study sites was a species of bruchid beetle, Pachymerus nucleorum.
Attack of fruits and seeds by this predator (Fig. 26) occurred only
after fruit abscission (0 d). The rate of colonization increased sharp
ly at 10 d and leveled off at 40 d. Although ca. 70% of fruits were
colonized, only ca. 40% of seeds were destroyed; an average of 1.9 out
of a total of 3.1 seeds per fruit escaped predation.
Among fruits collected 40 d or longer after abscission, 20.1%
suffered predation of all seeds. Colonization of these fruits by P.
nucleorum was approximately the same regardless of the number of seeds
per fruit (Table 5). Because each larva was limited to one seed, how
ever, lower proportions of seeds were eaten as the number of seeds per
fruit increased (Table 5). These results suggest that selection by seed
predaceous beetles favors many-seeded fruits in babassu.
Seed fate. In my experiment on seed fate over time (Fig. 27), the
percentage of defective (aborted or rotted) seeds did not vary signifi
cantly among sites. After 1 yr, most of the seeds on the secondary
forest and pasture sites had succumbed to predation; removal of palm
trees that harbor seed-predaceous bruchid beetles probably resulted in
significantly lower levels of predation in the field (p .1.05, one-way
AN0VA). Largely due to differences in predation, percentage of quies-


118
rates were doomed and did not use them to calculate mean growth rates or
residence times. Independent correlation of age estimates with a his
torically known event (see Chapter 6) indicates that this assumption is
probably reasonable.
Growth rates in stages 4-12 were calculated as change in height per
year. Height of the palm was defined as the distance from the ground to
the sheath apex of the newest expanded leaf; this was initially measured
to nearest 5 cm either directly (stages 4-8) or with an inclinometer
(stages 9-12). After 1 yr, the height of palms in stages 4-8 was
remeasured directly. For taller palms, mean increment per leaf was
calculated as the length of stem covered with leaf bases divided by the
total number of leaf bases. Growth rates were then calculated as the
product of the mean increment per leaf multiplied by the number of
leaves produced per year.
Productivity
Lack of branching and above-ground stem thickening greatly facili
tated the measurement of productivity in babassu. Whereas most studies
estimate annual production of leaves and reproductive structures indi
rectly through sampling of litter fall, in babassu it was possible to
measure these components directly. My measurements consequently avoided
potential underestimates due to reabsorption, decomposition, and leach
ing prior to abscission. Losses due to herbivory were not measured but
were probably negligible.
Population productivity. I measured production of above-ground
components in all life stages of babassu present on the primary and


FRUITS REMOVED
66
Figure 25.
Removal of babassu fruits on three study sites
Verde. Shaded portions of bars show percentage
removed > 4 m.
at Lago
of fruits


Table 26. Annual changes in number of palms per life stage on the 1-ha secondary forest site
at Lago Verde. Life stages are defined in Table 18.
LIFE STAGE
INITIAL N
GAINS
LOSSES
NET CHANGE
FINAL N
Number and Description
Recruitment
from Last Stage
Recruitment
to Next Stage
Mortality
1 Seeds on ground
10,850
59,280
5,560
28,610
+25,110
35,960
2 Seedlings
10,360
5,560
140
1,820
+3,620
13,980
3 Pre-established juveniles
1,250
140
40
20
+80
1,330
4 Established juveniles
570
40
10
10
+20
590
5
115
10
5
5
0
115
6
72
5
10
2
-7
65
7 Mature palms
136
10
16
0
-6
130
8
86
16
0
0
+16
102
9
0
0
0
0
0
0
10
1
0
0
0
0
1
11
0
0
0
0
0
0
12
0
0
0
0
0
0
9VT


94
an amount dependent on the leaf's deviation from the horizontal (Ewel et
al. 1982).
Irrigation Experiment
This experiment followed a 2 x 2 completely randomized design
involving 20 plots. On alternate days beginning 15 January 1981, each
of 10 plots was irrigated with 40 L of water, which was equivalent to an
irrigation of 0.4 cm depth.
Fertilizer Experiment
This experiment followed a 2 x 2 completely randomized design
involving 20 plots. Following advice of local agronomists, fertilizer
dosages equivalent to 40 kg N (as NH^), 40 kg P (as P20^), and 20 kg K
(as KC1) per ha were applied to 10 plots on 28 December 1980. Identical
dosages of N were applied at 2-mo intervals, and of P and K at 4-mo
intervals. Composite soil samples from all 20 plots, comprised of 10
subsamples per plot, were obtained at the conclusion of the experiment
(15 May 1982).
Insecticide Experiment
This experiment followed a 2 x 2 randomized complete block design
involving 20 plots. Application of insecticide and nematicide was
blocked to minimize between-plot contamination. At weekly intervals
beginning on 30 January 1981, 10 plots were sprayed with a 4 cc per 10-L
solution of the non-systemic insecticide, synthetic pyrethrum (known
commercially as Decis). At monthly intervals beginning on 28 December
1980, 2 g/m2 of the systemic nematicide Furadan 5 G FMC were applied.


142
(lx) could thus be determined simply by counting the number of indi
viduals present at a given stage or age.
Initial ages in the life table were calculated from ovule fertili
zation rather than seed germination. The time between these events was
estimated to be 1 yr, assuming an average of 9 mo for fruit maturation
and 3 mo for germination (see Chapter 3). I used seeds produced per yr
instead of seeds on the ground (stage 1) in the first row of the life
table. The latter reflects seed numbers at one point in time, whereas
the former integrates seed dynamics over a 1-yr period; due to the
rapidity of seed dynamics, I considered the annual rate of seed produc
tion to be a better indicator of population structure. The values under
survivorship in the first and subsequent rows are 2-yr means. Total
seeds produced per yr (18,600) thus represent the mean of the 1980-81
(13,640) and 1981-81 (23,560) harvests. Likewise, totals in subsequent
rows represent means of the first (April 1981) and final (April 1982)
population censuses (see Tables 25-27), which were carried out at the
end of babassu's fruiting season. Mortality (dx) was calculated as the
difference between total survivorship in a given row and that in the
following row. Mortality divided by the time in yr spent in each row
gave mortality per yr (qx). The time spent in the final row (stage 12)
was estimated at 62 yr (see Table 19). To calculate values for the
final three columns in stage 12, the sole representative was assumed to
die at the conclusion of the study. Life expectancy (ex) was calculated
from the initial age in each row. Details on how life expectancy is
calculated are provided by Brewer and McCann (1982).


Table 21. Partitioning of above-ground dry matter produc
tion of babassu over an estimated life span of
184 yr.
j£_ _I_
LEAVES
Leaflets
1,141
12.8
Leaf Axes
2,386
26.7
Subtotal Leaves
3,527
39.4
STEMS
1,412
1 5.8
REPRODUCTIVE STRUCTURES
Male Inflorescences
439
4.9
Female Inflorescence Axes
437
4.9
Female Inflorescence Bracts
265
3.0
Subtotal Ancillary Structures
1a.14.1-.
12.8
Fruit Walls
2,661
29.8
Seeds
200
2.2
Subtotal Fruits
2.861
32.0
Subtotal Reproductive Structures
4,002
44.8
TOTAL
8.941
100.0


This study is dedicated to the late Harold E. Moore, Jr.
who inspired it.


21
Shifting cultivation. Throughout Maranhao, sharecropping is the
means by which the vast majority of cultivators gain access to land.
Crops are raised principally by shifting cultivation (Table 1). The
forest is felled by axes and burned toward the end of the dry season.
Planting usually takes place in December with the first wet season
rains. In the study region, the principal crops are upland rice, cassa
va, maize, and beans. As rice is sensitive to drought, and cassava is
intolerant of prolonged waterlogging, farmers often plant both, and
intercropping is the rule. Cultivated areas are frequently weeded until
the April-June rice harvest; cassava may be harvested up to 2 yr after
planting. Cultivated plots on upland sites are virtually never used for
two crops in a row, so new plots are required each year. Short cropping
periods make this form of land use extremely extensive. Shifting culti
vation of this sort can only remain viable in areas of low population
densities. Over large portions of Maranho, this is no longer the case.
From 1950 to 1975, land area in Maranhao utilized for temporary
cropping increased threefold, while the ratio of fallowed to cropped
land was cut in half (Table 1). The immediate implications of shortened
fallows for crop yieldsand long-term implications for site quality
are obvious. Increased population densities are the principal cause of
shortened fallows throughout Maranhao. But exceptionally short fallows
in the study region are a consequence of widespread conversion to pas
ture


136
bypassing the production of permanently accruing structures such as
stems is a highly successful strategy in many grass-dominated agricul
tural ecosystems, which are the most productive of terrestial sites
(Loomis and Gerakis 1975). As has been shown for babassu, palms may be
more similar to grasses than to other trees in terms of their allocation
of dry matter production. The productivity of wild stands or planta
tions of palms can thus be extremely high yet not as conspicuous as that
of plantations comprised of fast-growing trees such as Eucalyptus spp.
and Gmelina arbrea, in which dry matter production is allocated pri
marily to stems.
Reproduction
Studies of resource allocation in palms include estimates of dry
matter production over the potential life of Corypha el at a (Tomlinson
and Soderholm 1975), Jessenia ba taua (Balick 1980), and As trocar yum
mexicanum (Sarukhan 1980). Approximately 45% of the mass of above
ground components produced over a 184-yr lifetime in babassu was allo
cated to reproductive structures (Table 21). This is higher than simi
lar measures of allocation to reproductive structures in As trocaryum
mexicanum (37%), Jessenia bataua (34%), and Corypha elata (17%).
With the exception of Corypha elata, the figures cited above gen
erally exceed those reported for perennial herbs growing in temperate
areas (e.g., Struik 1965, Abrahamson 1979). The latter are subject to
winter dieback, which usually requires considerable allocation of re
sources to vegetative storage organs. Such allocation is not necessary
among annuals growing in temperate areas, and this has apparently led to


169
Beach, J. H. In press. The reproductive biology of the pejibaye palm
(Bactris gasipaes) and B¡_ porschiana in the Atlantic lowlands of
Costa Rica. Principes.
Beard, J. S. 1944 Climax vegetation in tropical America. Ecology 25:
127-158.
Black, C. A. 1965. Methods of soil analysis. Part 2. Chemical and
microbiological properties. American Society of Agronomy, Madison,
Wisconsin, USA.
Bondar, G. 1936. Notas biolgicas sobre Bruchdeos observados no Brasil.
Arquivos do Instituto Biolgico de Vegetapo (Rio de Janeiro
Brazil) 3: 1-44*
. 1954a. Nova especie de Orbignya, produtora do leo de
babacu. Arquivos do Jardim Botnico (Rio de Janeiro, Brazil) 13:
57-59.
. 1954b. A entomologa de babassu. Boletim do Instituto de
Biologia (Salvador, Brazil) 1(1): 9-17.
. 1957. Novo gnero e nova especie de palmeiras da tribo
Attaleine. Arquivos do Jardim Botnico (Rio de Janeiro, Brazil) 15:
49-55.
. 1964. Palmeiras do Brasil. Instituto de Botnica de Sao
Paulo, Sao Paulo, Brazil.
Bradford, D. F., and C. C. Smith. 1977. Seed predation and seed number
in Scheelea palm fruits. Ecology 58: 667-673.
Braga, H. C., and D. C. Dias. 1968. Aspectos socio-econmicos do babas
su. Instituto de leos, Ministrio de Agricultura, Rio de Janeiro,
Brazil.
Brewer, R., and M. T. McCann. 1982. Laboratory and field manual of
ecology. Saunders College Publishing, Philadelphia, Pennsylvania,
USA.
Bullock, S. H. 1980. Demography of an undergrowth palm in littoral
Cameroon. Biotropica 12: 247-255.
. 1981. Notes on the phenology of inflorescences and pol
lination in some rain forest palms in Costa Rica. Principes 25:
101-105.
Burret, M. 1929. Die Palmengattungen Orbignya, Attalea, Scheelea und
Maximiliana. Notizblatt 10: 493-543, o51-701.


176
Sarukhn, J. 1978. Studies on the demography of tropical trees. Pages
163-184 in P. B. Tomlinson and M. H. Zimmermann, editors. Tropi
cal trees as living systems. Cambridge University Press, Cam
bridge, England.
. 1980. Demographic problems in tropical systems. Pages
168-188 in 0. T. Solbrig, editor. Demography and evolution in
plant populations. University of California Press, Berkeley,
California, USA.
Shodt, L. L., and W. A. Mitchell. 1967. The pollination of Cocos
nucfera L. in Hawaii. Tropical Agriculture (Trinidad) 44s 133
141.
Siegel, S. 1956. Nonparametric statistics for the behavioral sciences.
McGraw-Hill, New York, New York, USA.
Silvertown, J. W. 1981. Seed size, life span, and germination date as
coadapted features of plant life history. American Naturalist 118:
86O-864.
. 1982. Introduction to plant population ecology.
Longman, New York, New York, USA.
Smith, N. 1974* Agouti and babassu. Oryx 12: 581-582.
Snedecor, G. W., and W. G. Cochran. 1978. Statistical methods. Sixth
edition. Iowa State University Press, Ames, Iowa, USA.
Stafleu, F. A. 1972. International code of botanical nomenclature.
Utrecht, The Netherlands.
Stebbins, G. L. 1971. Adaptive radiation of reproductive character
istics in angiosperms. II. Seeds and seedlings. Annual Review of
Ecology and Systematics 2: 237-260.
Steward, J. H. 1963. Handbook of South American Indians. United
States Government Printing Office, Washington, District of
Columbia, USA.
STI. 1979. Coco de babassu: matria-prima para produpao de alcool e
carvo. Secretaria de Tecnologia Industrial, Ministerio da Indus
tria e do Comercio, Brasilia, Brazil.
Stuik, G. J. 1965. Growth patterns of some native and perennial herbs
in southern Wisconsin. Ecology 46: 401-420.
Syed, R. A. 1979. Studies on oil palm pollination by insects. Bulle
tin of Entomological Research 69: 213-224.


78
by local inhabitants to feed on the mesocarp (but not the seeds) of
babassu fruits include porcupines (Coend prehen si lis and Coend sp.),
spiny rats (Proechimys longicaudatus and Mesomys hispidus), and pec
caries (Tayassu ta.1 acu and T. pcari, both locally extinct). None of
these species appears to have a significant role in dispersal of babassu
fruits.
My field observations indicate that water-mediated dispersal is a
significant factor on relatively open sites such as pastures. Here soil
compaction results in increased surface runoff, by which fruits were
observed to be carried distances of ca. 25 m. The potential role of
water-mediated dispersal is far greater. Dissemination by rivers large
ly accounts for the present distribution of babassu over nearly half a
continent (see Chapter 2).
Although it remains undocumented, the role of humans in the disper
sal of babassu is potentially significant. As discussed in Chapter 2,
the present occurrence of babassu in the Brazilian state of Cear may be
due to dissemination by indigenous peoples. Today settlers throughout
the Amazon region commonly transplant babassu to areas where it did not
formerly occur. On a local scale in Maranho, gathering of babassu
fruits is a widespread activity, particularly among women. Intensity of
fruit gathering is especially high on relatively open, accessible sites
such as pastures. Although the seeds from almost all gathered fruits
are ultimately consumed, some accidental dispersal occurs during gather
ing and transport. More significantly, perhaps, is the role of human
gatherers of babassu fruits as agents of selection. People prefer-


PALMS (no./ha)
148
10,000
1,000
I 00
I 0
I
10,000
1,000
I 00
I 0
I
I 0,000
1,000
! 00
I 0
m SEEDLINGS
H PRE-ESTABLISHED
^ JUVENILES
H ESTABLISHED JUVENILES
MATURE PALMS
Figure 41. Stage distributions of babassu on the primary forest,
secondary forest, and pasture sites at Lago Verde.


132
vertically. Thus, a stemless palm is able to place photosynthetic
tissue ca. 8 m above the ground; the leaf axes are also green and
consequently provide additional photosynthesis.
Exclusive allocation of above-ground productivity to leaves during
the stemless phase may provide an important competitive edge to babas
su, as well as to rainforest palms in general. These leaves act essen
tially as "throw-away" branches that permit stemless palms to occupy
large volumes at a comparatively low metabolic cost (Givnish 1978).
Channelling of above-ground productivity to photosynthetic tissues may
thus enable palms to persist for extensive periods in the understory,
until events such as gap openings provide opportunities for vertical
growth.
Because the apical meristem of pre-established juveniles is buried,
babassu's stem begins to grow long before it emerges from the soil.
Allocation of productivity to the stem actually begins at germination.
Vertical growth commences at ca. 38 yr in the primary forest site. The
stem does not emerge from the soil until the palm is ca. 50-yr old, by
which time the shoot is ca. 178-cm long. Until the onset of reproduc
tion, allocation of above-ground productivity to the stem rapidly in
creases (Figs. 37-38), and stem growth attains its highest rates (Table
19).
Reproduction begins when the palm is still below the forest canopy.
As the palm reaches the canopy at ca. 90 yr of age, partitioning of
productivity to reproductive structures increases rapidly. At the same
time, total productivity peaks and subsequently undergoes a steep de-


Table 2. Nomenclatura! history of babassu and its suspected hybrids
scientific NAME PUBII CATION
Martlus
1826
Drude
1861
B. Rodrigues
1898
B. Rodrigues
190)
Burret
1929
Burret
1932
Burret
1938
Bondar
195*
Bondar
1957
W. Boer
1965
Classman
1977
Attalea speciosa
Mart.
Described.,
..- 0. mart lana. .,
*
..- 0. barboslana.
Orblgnya lydlae
Orude
Described.
...- A. lydlae
Orblgnya mart lana
Barb. Rodr.
Described
...- 0. mart lana....
..- 0. olelfera.,
Orblgnya macrostachya
Drude ex Barb. Rodr.
- 0. speclosa1.
... 0. dammerlana..
..- 0. cohune
..Part ....
(Mart.) Barb. Rodr.
0. martlana
0. barboslana
0. olefera
Orblgnya damner lana Described Part | Part -
Bath. Rodr. £. spec losa 0. cohune
Attalea lydlae Described £. lydlae 0. barboslana
TOrudel Barb. Rodr.
Orblgnya macropetal a Described -A. spec losa 0. barboslana
Bur ret
Orblgnya huebnerI
Burret
Described,
Uncertain species
ObIgnya barbos lana
Burret
Proposed,
A. speclosa 0. barboslana
Orb I gnya olelfera
Burret
Orbignya te IxeI rana
Bondar
Mark Ieya dahlqrrnlana
Bondar
Attalea dahlgrenlana
(Bondar) W. Boer
Described
barboslana
Described.
Described....- A. dahlgrenlana..Uncertain species
Described M. dahlgrenlana
I.
0.
speclosa (Hart.) Barb,
currently reduced to
Rodr. Is
Synonymy
not to be confused with 0. spec losa Barb. Rodr.,
under 0. cohune (Mart.) Dahigrn ex Standi.
a name
2. Possible hybrid.
U>
O


34
outer (abaxial) surface of the sheath; large inflorescences that are
either exclusively staminate or hermaphroditic, each containing a
persistent, woody bract with a prominent apex; staminate inflorescences
with numerous, erect branches (rachillae), each branch bearing two to
four rows of staminate flowers on its lower (abaxial) surface only;
staminate flowers with a small calyx, two or rarely three dentate-tipped
petals, approximately 24 irregularly coiled and twisted stamens, and a
pistillode; hermaphroditic inflorescences with numerous branches, each
branch bearing one to rarely three pistillate flowers at the base and a
variable number of staminate flowers at the tip, the latter either
inviable or opening after the pistillate flowers; and a massive, oblong
drupe with a thick, woody indehiscent endocarp and usually two to five
oily seeds.
Due to its apparent propensity to hybridize, the distinction
between babassu and at least two closely related species (Orbignva
eichleri and Maximi liana maripa) is somewhat blurred. I suspect that
these latter species hybridize with babassu. Morphological comparisons
(Appendices D and E) illustrate the intermediate morphology of the
purported hybrids ("Orbignva teixeirana11 and "Markleva dahlgreniana").
Germination experiments have shown that both of the latter produce
viable seeds and thus are capable of regenerating; however, they are
only known to occur in association with their apparent progenitors. The
distribution of babassu and its purported hybrids is shown in Figure 15.
Anatomical, palynological, and electrophoretic studies of these
complexes have been initiated by colleagues at the New York Botanical


¡Table 24. Life table of the population of babassu
Life stages are defined in Table 18.
on the 1-ha
primary forest site at
Lago Verde.
LIFE STAGE
AGE (yr)
SURVIVORSHIP
MORTALITY
LIFE
Number and Description
Initial Mean
N
%
N
per yr
EXPECTANCY (yr)
- Seeds produced per yr
0
0.5
18,600
100
12,970
12,970
2.2
2 Seedlings
1
2.7
5,630
30
4,930
822
5.0
3 Pre-established juveniles
7
19
700
3.8
480
15.0
22
4 Established juveniles
39
45
220
1.2
155
15.5
13
5
49
52
65
.35
29
2.23
21
6
56
60
31
.17
0
0
28
7 Mature palms
63
68
36
.19
10
1.11
21
8
72
77
26
.14
19.5
1.50
18
9
85
102
4
.022
0
0
44
10
106
119
6.5
.035
5.5
0.204
23
11
133
-
0
0
0
0
-
12
168
185
1
.0054
1
0.0161
17
144


Table 20. Estimated growth rates and ages of babassu palms in the secondary
forest
number
site at Lago Verde,
of palms measured.
Stages
are defined
in Table 18.
N =
STAGE
N
LEAFLETS PER SIDE
HEIGHT
(cm)
TIME IN
AGE AT END
Range
in Stage
Der Year
Range
in Stage
Der Year
STAGE (vr)
OF STAGE (vr)
min.
max.
x (s.d.)
min.
max.
x (s.d.)
x (s.d.)
X
2
200
4.0
9.5
.98 ( .57)



5.6 ( 3.3)
5.6
3
64
10.0
100.0
3-89 (3.23)



23.1 (19.3)
28.7
4
57



0
99
8.6 ( 6.8)
11.5 ( 9.2)
40.2
5
9



100
199
8.2 ( 3.9)
12.1 ( 5.8)
52.3
6
22



200
499
44.7 (21.1)
6.7 ( 3.2)
59.0
7
62



500
999
41.9 (14.5)
11.9 ( 4.1)
70.9
8
41
__
__
1000
1499
40.5 (11.4)
12.3 ( 3.5)
83.2
123


92
(Fig. 32), 60 entire leaves of varying lengths were harvested from
babassu seedlings; the leaves were then traced and the cut-out tracings
passed through an area meter (Li-Cor Model 3100). The LAI of babassu
was determined by measuring all leaf lengths, converting to areas and
dividing the sum by the total area of the plot. To permit accurate
measures of individual growth without bias due to newly germinated
seedlings, I calculated leaf area per palm using only those seedlings
present since the first measurement (ca. 10 mo after initiation of the
experiments).
An additional measure of growth was obtained after the experiments
had run ca. 18 mo (May 1982) by harvesting above-ground biomass of
babassu (which consisted entirely of leaves as stems had not emerged)
within the central 4 x 1 m area of each plot. (A 50-cm border was left
to reduce edge effects.) The harvested material was oven-dried at 70C
and weighed to nearest gram.
A measure of competitiveness was obtained by comparing the biomass
of babassu with that of other species on the non-weeded plots. Biomass
of species other than babassu was separated into the following compo
nents: (1) flowers and fruits; (2) green leaves, green branches, and
green stems; (3) brown branches and brown stems; and (4) standing dead.
Competitiveness was further measured by periodically determining the LAI
of babassu relative to its competitors. The LAI was measured by
lowering a plumb bob onto 20 points distributed throughout each 5 x 2 m
plot, excluding margins; at each leaf contact, the species and 50-cm
height interval were recorded. This method underestimates true LAI by


Figure
1
LIST OF FIGURES
Approximate distribution of high-density stands of babassu
(Orbignya martiana) in Brazil 2
2 Aerial view of babassu stands in the county of Bacabal,
Maranhao 3
3 Specimen of babassu (0;_ martiana) in Pirapora, Minas
Gerais 3
4 Deforestation in the Tropical Moist Forest zone of
Maranhao 4
5 Regrowth of babassu on a deforested site in the Tropical
Moist Forest zone of Maranhao 4
6 Production of babassu kernels in Maranhao and Brazil, 1920-
1979 5
7 Flowchart of products derived from babassu fruits by indus
tries in Bacabal, Maranhao, during 1980 6
8 Flowchart of products derivable from babassu fruits via
current (1983) technology
9 Maranhao and neighboring states, with locations of research
sites 10
10 Major soil groups of Maranhao, according to the Brazilian
Classification System 11
11 Principal ecological zones of Maranhao 12
12 Coverage of babassu stands in Maranhao 13
13 Monthly rainfall at four research sites 18
14 Distribution of land and households according to size of
landholding in rural areas of Maranhao during 1950 and
1975 20
15 Distribution of specimens of Orbignya martiana and its
suspected hybrid complexes 31
ix


99
2
Figure 36. Palms per nr' (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in climate experiment. Bars
represent 1 standard error.


76
series shows that inflorescence types (male or female) can be clearly
distinguished up to 12 leaves prior to the newest emerged leaf. Given
an average production of ca. four leaves per year, this means that sex
determination occurs at least 3 yr prior to leaf expansion in the crown,
4 yr prior to flowering, and nearly 5 yr prior to fruiting. Detecting
possible relationships between climate and yield will therefore require
long-term studies of babassu's phenology.
Pollination
My results indicate that reproduction in babassu is predominantly
or perhaps exclusively via outcrossing, and that pollination is effected
by nitidulid beetles and wind. Such a combination of pollination sys
tems has been observed in other species of palms (Uhl and Moore 1977,
Moore and Uhl 1982). The coconut (Cocos nucfera) appears to be polli
nated by either honey bees or wind in Hawaii (Shodt and Mitchell 1967).
The oil palm was thought to be exclusively wind-pollina ted (Hartley
1977), but a recent study (Syed 1979) has implicated weevils as pollina
tors within the palm's natural range in West Africa and thrips on the
Malay peninsula. Likewise, Myers (1981) suggested that both wind and
insects effect pollination in Raphia taedigera in Costa Rica. Nitidulid
beetles are rarely cited as pollinators (Faegri and van der Pijl 1979),
but their role in the pollination of palms has either been documented
(Essig 1971) or implicated (Dransfield 1982). Wind pollination in palms
has been reported since antiquity, according to Faegri and van der Pijl
(1979).


CHAPTER 1
INTRODUCTION
Numerous species of palms dominate the landscape over extensive
areas of the lowland tropics (Beard 1944, Richards 1952, Moore 1973a).
Palm-dominated stands often occur on edaphically limited sites where
competition is apparently reduced (Myers 1981). Several species, how
ever, form nearly monospecific stands on sites originally covered by
forests of high biological diversity. Such secondary palm forests
originate from and are apparently maintained by human activities.
People, in turn, depend on the palms as a source of food, fuel, shelter
and income.
This study concerns one of the most economically important and
certainly the most ubiquitous of such secondary forest palms: Orbignya
martiana, known in Brazil as babassu. The tree-sized babassu palm forms
high-density stands over extensive areas of Brazil (Figs. 1-3). The
stands attain their greatest extent in the state of Maranho, where they
cover an area estimated at 102,970 km^ (Anonymous 1981), or roughly
equivalent to the U.S. state of Virginia. Babassu flourishes on sites
subject to intensive disturbances such as cutting and burning (Figs. 4-
5). The palm is a crucial resource for rural poor, who utilize the
fruits, leaves, and stem for a wide range of subsistence products.
Manual extraction of kernels from the extremely hard fruits represents
the state's largest cottage industry (involving over 400,000 families)
and its most important income source for women. The kernels are util-
1


INFRUCTESCENCES PER PALM
56
1981
F M A M J J A
1982-
SON D J F M
tu
3
i-
<
2
K
3
H
<
2
2
ui
£E
3
I-
<
2
ui
oc
3
I-
<
2
2
ui
£E
3
H
<
2
UI
cc
3
H
<
2
2
0.5
0.5
1.0
0.5
0.5
1.0
0.5
0.5
.0
1.5
Figure 22. Mean monthly number of immature and mature inflorescences
per palm at three study sites at Lago Verde. N = number
of palms observed per site.


46
wk intervals, the frames were lifted to check for germination. To test
the effects of fire, I exposed fruits to each of the following treat
ments: (1) a prolonged, high-intensity fire, simulating a forest fallow
burn; (2) a rapid, low-intensity fire, simulating a pasture burn; and
(3) no fire. To test the effect of consumption of the fruit mesocarp by
herbivores, I compared germination in fruits that had been dehusked
(epicarp and mesocarp removed, simulating the effect of herbivory) and
fruits that had not been dehusked. To test the effect of shading, I
planted 100 fruits on an exposed plot and 100 fruits on a plot shaded as
described as above.
Results
Phenology
Leaf turnover. Figure 16 shows data on leaf turnover of babassu at
the wet (Pindar Mirim), intermediate (Lago Verde), and dry (Caxias)
sites; similar data from the primary forest, secondary forest, and
pasture sites are shown in Figure 17. At all sites, leaf production
peaked during the two rainy seasons and leaf loss peaked during the
single dry season of observations. As in palms generally, leaf produc
tion in babassu is synonymous with growth; the rainy season is, there
fore, the growing season for the palm. Over a 1-yr period after initia
tion of observations, no difference was detected among leaf production
rates per palm at the wet (x = 3.9), intermediate (x = 41), and dry (x
= 3.7) sites. By contrast, Duncan's Multiple Range test revealed signi
ficant differences (p ^.05) in rates of leaf production per palm at the
three Lago Verde sites; the order was pasture (x = 5.1) > primary forest


Appendix A continued.
SPECIES
ABUNDANCE
stems/ha
FREQUENCX
%
DOMINANCE
m/ha
IMPORTANCE
1
15. Astrocarvum vulcare Mart.*
9
2.33
1.93
0.127
0.48
1.58
16. OmDhalea diandria L.
6
1.55
1.93
0.052
0.20
1.23
17. Eschweilera odora (PoeDD.) Miers.
5
1.30
1.93
0.108
0.41
1.21
18. Palicourea sd.
5
1.30
1.93
0.067
0.25
1.16
19. Unknown tree No. 1
4
1.04
1.45
0.217
0.83
1.10
20. Manilkara cf. surinaraensis (Mia.) Ducke
3
0.78
1.45
0.141
0.54
0.92
21. Jacaratia SDinosa A.DC.*
2
0.52
0.97
0.320
1.22
0.90
22. Dialium euianensis Aubl.1 Sandw.
3
0.78
1.45
0.107
0.41
0.88
28. Dueuetia sd.
3
0.78
0.97
0.042
0.16
0.63
24. Psvchotria sd.
3
0.78
0.97
0.028
0.11
0.62
25. Luehea Daniculata Mart.*
2
0.52
0.97
0.059
0.23
0.57
26. Sterculia sd.*
1
0.26
0.48
0.252
0.96
0.57
27. Richardella sd.
2
0.52
0.97
0.053
0.20
0.56
28. Cedrela odorata L.
2
0.52
0.97
0.040
0.15
0.55
29. Hirtella raceinosa Lam.*
2
0.52
0.97
0.037
0.14
0.54
180


124
between germination and initiation of vertical stem growth. Slower
growth rates in the primary forest are probably indicative of greater
suppression of seedlings and pre-established juveniles in this habitat.
Percentage of palms in these stages exhibiting zero or negative growth
rates was higher in the primary forest (65.2%) than in the secondary
forest (58.9%).
The high variances in early life stages (Tables 19 and 20) suggest
that mean growth rates are not a good measure of success. Palms that
ultimately became established were probably situated in light gaps,
where growth rates were enhanced. This may have been especially true in
the primary forest, where established palms (stages 4-8) almost invari
ably grew faster than in the secondary forest. In both habitats, verti
cal growth rates were initially slow until the palm surpassed ca. 200
cm, whereupon they increased rapidly, peaking at stage 7 in the primary
forest and stage 6 in the secondary forest. These stages immediately
preceeded or coincided with initiation of inflorescence production, when
resources were diverted from growth to reproduction. In both habitats,
growth rates steadily declined in subsequent stages. The mean growth
rates obtained for these later stages had far lower variances than in
earlier stages, which is probably indicative of greater chances of
survival following establishment.
Productivity
The data on growth indicate that babassu is both slow growing and
long-lived. The tallest palm (31.4m) on the primary forest site had an
estimated age of 184 yr. Over the entire lifetime of this palm, a large


110
Table 17. Soil chemistry of experimental sites at Lago Verde
and Belem. Samples were obtained after clearing and
before burning the fallow. Values are means
(standard deviations) of four samples from 0-15 cm,
each a composite of 10 cores.
Lago Verde
(Drv Climate)
Belem
(Wet Climate)
C (%)
0.6 (0.1)
2.2 (0.3)
Organic Matter (%)
1.0 (0.2)
1.6 (0.4)
N (%)
0.1 (0.0)
0.2 (0.0)
C/N
6
11
P205 (ppm)
3.8 (2.9)
1.2 (0.5)
pH
5.6
3-9
Ca + Mg (meq/100 g)
3.0 (0.8)
0.1 (0.1)
Na (meq/100 g)
0.1 (0.1)
0.0 (0.0)
K (meq/100 g)
0.4 (0.1)
0.1 (0.1)
H (meq/100 g)
1.6 (0.4)
3-7 (0.9)
A1 (meq/100 g)
0.0 (0.1)
1.7 (0.2)
Total Bases (meq/100 g)
5.1 (0.8)
5.6 (0.7)
Base Saturation ()
68.2 (6.9)
7.7 (3-3)


35
Garden in the U.S. and the Centro Nacional de Recursos Genticos
(CENARGEN) in Brazil.
Distribution and Origin
Babassu has a widespread distribution in South America (Fig. 15).
Dissemination of the large, heavy fruits has probably been exclusively
by mammals (either living or extinct) or water. Among living mammals,
rodents such as pacas (Agouti paca) and agoutis (Dasvprocta punctata)
are effective, short-range dispersal agents (see Chapter 3); humans and
possibly monkeys disseminate the fruits over greater distances.
Coevolution between hard-fruited palms such as babassu and large,
Pleistocene mammals seems likely (Janzen and Martin 1982), but a paucity
of paleontological evidence makes it impossible to assess the potential
contribution of these now extinct animals to the current distribution of
babassu.
Markley (1971) argued convincingly that the babassu palm originated
on the Brazilian Shield in the state of Gois, a major center of
diversity for the entire Attalea alliance. Since its origin during the
Precambrian, the Brazilian Shield has undergone considerable erosion by
numerous river systems. With few exceptions (discussed below), babassu
areas from the Paraguay River to the Amazon Basin are connected by
watersheds to the Brazilian Shield in Gois (Figs. 1 and 15).
Throughout its range, babassu is largely confined to floodplains, river
valleys, and former deltas, which lends support to the idea of
dissemination by rivers. Although their high specific gravity (ca. 1.4
g fresh weight/cm^) prevents the fruits from floating, they can be


Appendix B continued.
§ p
E C I E S
AmDAJCg,
stems/ha
FREOUENCY
i
DOMINANCE
nr/ha %
IMPORTANCE
%
15.
Enterolobium contortisiliauam (Veil.)
Morong. 1
0.27
1.09
0.128
0.50
0.62
16.
Gustavia auRusta L.*
2
0.55
1.09
0.016
0.06
0.56
17.
Licania aDetala (E. Mever) Fritsch*
1
0.27
1.09
0.081
0.31
0.56
18.
GeniDa americana L.
1
0.27
1.09
0.041
0.16
0.51
19.
Manilkara sd.*
1
0.27
1.09
0.038
0.15
0.50
20.
Cocoloba Daniculata Meissn.
1
0.27
1.09
0.035
0.13
0.50
21.
InRa edulis Mart.
1
0.27
1.09
0.022
0.08
0.48
22.
Luehea Daniculata Mart.*
1
0.27
1.09
0.012
0.05
0.47
23-
Mvrcia rufiDila McVauch
1
0.27
1.09
0.011
0.04
0.47
24.
Cordia sd.*
1
0.27
1.09
0.010
0.04
0.47
25.
Bauhinia Ruianensis (Aubl.) M.J.Baliq.
1
0.27
1.09
0.009
0.04
0.47
26.
Tabebuia cf. barbata (E. Mever) Sandw.
1
0.27
1.09
0.009
0.04
0.47
27.
Croton caiucara Benth.
1
0.27
1.09
0.009
0.03
0.46
28.
Hirtella racemosa Lam.*
1
0.27
1.09
0.009
0.03
0.46
29.
Unknown Vine No. 1
1
0.27
1.09
0.009
0.03
0.46
185


192
Orbignva martiana. "Markleva dahlgreniana". AND Maximlliana
maripa.
Orbignva martiana "Markleva dahlgreniana" Maximiliana maripa
CHARACTER (possible hybrid)
LEAF
Arrangement of leaf bases
in crown
spiralled
spiralled, in 5
vertical rows
spiralled, in 5-7
vertical rows
Sheath
length in adults (cm)
90-200
94
45-90
Petiole
length in adults (cm)
0-50
130
80-200
Rachis
length in adults (cm)
520-990
600-800
350-700
Pinnae
number per side
145-240
205-270
170-270
orientation
in same plane
in same plane to
slightly crispate
strongly crispate
disposition
regular
clustered in groups
of 2-5
clustered in groups
of 2-10
INFLORESCENCE
Peduncular bract
curvature
slight
moderate
pronounced
length of acumen (cm)
16-32
20-30
20-80
length excluding acumen
(cm) 140-300
80-170
70-100
width
12-41
30
40


CHAPTER 5
GROWTH AND PRODUCTIVITY
This chapter completes the life cycle of babassu with an analysis
of the palm's growth and productivity. Growth refers to the annual
increment of a morphological parameter (stem height or leaflets per
leaf) in individual palms. Productivity is here defined as the above
ground mass produced per unit time either by a population or by indi
viduals. Both growth and productivity are relatively easy to measure in
palms. As in other monocotyledons, secondary growth does not occur in
palms due to an absence of a lateral cambium, and primary growth above
the ground is generally limited to a single shoot apex. For practical
purposes, growth of the stem can thus be measured simply as changes in
height over time (Tomlinson 1963). Likewise, because leaves and repro
ductive structures tend to be few and large, direct measures of produc
tivity are relatively easy in palms. They are therefore ideal subjects
for studies of both growth and productivity.
In this chapter, I examine growth and productivity in individual
palms of different ages, as well as productivity in populations growing
under differing environmental conditions. The objective is to determine
how resources are allocated among the various components of the palm or
population; the goal is a deeper understanding of how growth and repro
duction are controlled in babassu.
115


Appendix A continued.
SPECIES
ABUNDANCE
stems/ha%
FREQUENCY
%
DOMINANCE
mVha%
IMPORTANCE
i
60. Brosimum ulearum Mildbr.
1
0.26
0.1*8
0.008
0.03
0.26
61. DiosDvros cf. euianensis (Aubl.) Gurke*
1
0.26
0.48
0.008
0.03
0.26
62. Vitex sd.
1
0.26
0.48
0.007
0.03
0.26
63. AmDhirrhox surinamensis Eich.
1
0.26
0.48
0.007
0.03
0.26
TOTAL
386
100.07
99.93
26.253
100.00
100.00
Note: Asterisks indicate species that also occur on 1 ha of secondary forest nearby,


INFLORESCENCE PRODUCTION PER PALM
51
0.5
9
0
d
0.5
1.0
1981 1982
J FMAM JJ ASOND J FM
PRIMARY FOREST
(N=24)


1
U U
0.5
9
0
d1
0.5
1.0
1.5
SECONDARY FOREST
(N=75)
itB-" cror
Figure 19. Mean monthly production of male and functionally female
inflorescences per palm at three study sites at Lago
Verde. N = number of palms observed per site.


PERCENT ATTACKED
67
Figure 26. Attack of babassu fruits and seeds by the bruchid beetle,
Pachymerus nucleorum, over time. Bars represent 1 standard
error.


113
growth of early successional vegetation was likewise not enhanced by
application of fertilizers (Harcombe 1977). I predict that growth of
babassu (on competition-free plots) and its competitors (on non-weeded
plots) would be enhanced by application of fertilizers. The results of
the climate experiment provide indirect support for this prediction.
Growth of babassu (on competition-free plots) and its competitors (on
non-weeded plots) was significantly enhanced under what were presumed to
be less favorable climatic conditions at the dry site (Lago Verde). The
relatively high soil fertility at this site compared to the wet site
(Belem) seems to be the most likely cause of this response (Table 17).
Seedling Recruitment
As mortality of babassu was extremely low in these experiments, I
shall consider density to be essentially synonymous with seedling
recruitment. The latter term is more useful in the discussion that
follows.
In the experiments described above, responses in recruitment were
much more pronounced than responses in growth. Recruitment was consist
ently enhanced on plots that were not weeded, irrigated, or located in
the wet climate. The response was due to increased germination, which
probably resulted from higher soil moisture levels on these plots.
Competitor presence had a consistently greater effect on recruitment
than either irrigation or the wet climate. Likewise, recruitment
contributed more substantially to babassu's overall performance (as
measured by LAI) than did growth. Thus competitor presence was the key
factor governing babassu's performance during the 18 mo experiments.


141
under varying degrees of disturbance (e.g., Marks 1974, Cook and Lyons
1983).
In this chapter I examine how stands or populations of babassu
respond to different disturbances. My objective is to compare the
structure and dynamics of three populations: one in primary forest, one
in secondary forest, and one in pasture. The structure of these popula
tions reflects past as well as present patterns of land use and thus
provides clues as to how high-density stands of babassu originate.
Likewise, the dynamics of these populations indicate their present
capacity to regenerate. A comparison of the structure and dynamics of
populations subjected to different land uses thus gives insight as to
the origin and possible future of the palm forests.
Methods
Population structure of babassu was determined by censusing samples
of each life stage (see Table 18) present on the primary forest, second
ary forest, and pasture sites. The areas allocated for censusing were
0.05 ha for stages 1-2, 0.1 ha for stage 3, 0.2 ha for stages 4-5, 0.5
ha for stage 6, and 1.0 ha for stages 7-12. The palms within these
areas were labelled and mapped. Following an initial census in April
1981, the populations were recensused twice at 6-mo intervals, during
which each stage's mortality and recruitment were recorded.
A static or time-specific life table was constructed for the pri
mary forest population (Table 24). To construct the life table, the
population was assumed to be numerically stable, which implies that the
age distribution and survivorship curves were identical. Survivorship


103
Table 10. Palms per m2 (a), leaf area per palm (b), and
leaf area index (c) of babassu at end of
fertilizer experiment. Asterisks indicate
significant differences (** for p .01)
between main effect means.
(a)
p
Palms der nr
No Fertilizers
Fertilizers
Mean
Not Weeded
27.8
25.4
26.6
*
Weeded
2.5
3.8
3.2
Mean
15.2
14.6
(b)
Leaf Area per Palm (cm
2)
No Fertilizers
Fertilizers
Mean
Not Weeded
248.2
302.6
275.4
*
Weeded
444.0
476.6
460.3
Mean
346.1
389.6
(c)
Leaf Area Index
No Fertilizers
Fertilizers
Mean
Not Weeded
.26
.25
.26
Weeded
. 10
.14
.12
Mean
.18
.20


predation. Germination required 3 mo and was enhanced in shade. Stem-
less palms persisted in the understory of a primary forest for about 50
yr, permitting build-up of high densities (6,000 per ha). Escape from
seed predators, shade tolerance, and an estimated lifespan of 184 yr
account for babassu's importance in such forests. Due to burial of
their apical meristems via cryptogeal germination, stemless palms sur
vive when sites are burned for shifting cultivation. During the subse
quent fallow, these palms are released and form virtually monospecific
stands. One such stand was still increasing in numbers after a fallow
of 33 yr. Annual above-ground dry matter production of this stand was
ca. 25 tha yr 70% of which was allocated to leaves. With fallows
of 4-5 yr, enough fuel accumulates in the form of leaves to permit the
hot fire necessary for weed control and nutrient release. Thus shifting
cultivation can be practiced with minimal thinning of the stands.
Babassu's dominance of the landscape results from its persistence
over a wide gradient of disturbances. Because of its capacity to form
monospecific stands, its low maintenance requirements, its integration
into major land uses, and its current importance in market and
subsistence economies, babassu seems most amenable for use as an
extensive, moderate-yielding crop.
xiii


CHAPTER 3
REPRODUCTIVE BIOLOGY
In autecological studies of plants, the various components of
reproduction (e.g., phenology, pollination, dispersal, predation, ger
mination) are often analyzed as separate entities. Yet these events are
obviously interdependent. The timing of flowering, for example, may be
related to competition for animal pollinators (e.g., Rathcke in press).
Tight interdependence between the often conflicting demands of seed
dispersal, escape from seed predators, germination, and ultimately
establishment has been repeatedly documented (e.g., Stebbins 1971, Har
per 1977, Silvertown 1981, Willson 1983).
This chapter commences the life history of babassu with an analysis
of the palm's reproduction under a range of ecological conditions.
Thus, comparative phenological observations were conducted in the major
ecological zones of Maranhao where babassu forests occur. The wide
spread changes in land uses currently underway in Maranhao are likely to
have a profound impact on the reproductive biology of local populations
of the palm (e.g., see Moore 1977, Janzen 1978a). I therefore carried
out studies of babassu's phenology, floral biology, dispersal, preda
tion, germination, and seed fate under a range of land uses associated
with babassu.
38


154
Annual cutting and burning were initiated after the pasture had been
established. Under this regimen, exposure of apical meristems at or
near the soil surface resulted in high mortality among stages 4-5, thus
accounting for their relatively low frequencies on the site.
Seed dynamics in the pasture differed from those in the forested
habitats (Table 27). In the pasture, over 98% of the annual seed pro
duction either died or was removed from the site by fruit-gathering
humans. Consequently, only a small percentage (ca. 1%) was available
for seedling recruitment. Among seedlings and juveniles, recruitment
during the 1-yr period was insufficient to cover mortality losses (Fig.
44)* Due to the relative homogeneity of the pasture, I believe that
this pattern is representative of long-term population dynamics of
babassu on the site. Annual cutting and burning of the understory are
standard management practices in pastures. Juvenile palms consequently
suffer recurring defoliations and eventual damage to their emerging
apical meristems, leading to high mortality rates. As a result, I
predict that continuation of present land use practices will lead to the
eventual demise of the babassu population on the pasture site.
Discussion
Most of the species that comprise forest communities depend on some
degree of disturbance for their long-term maintenance (Watt 1947, Whit
more 1982). In the absence of major disturbances, some species rely on
the formation of small gaps in the forest canopy (Hartshorn 1978). When
subjected to major disturbances such as a cyclone (Webb 1958) or fire
(Maissurow 1941) the site is typically occupied by a different set of


77
In babassu, insect pollination is probably predominant in forests,
while wind pollination is likely to be enhanced in open habitats such as
pastures. Neither system is exclusive of the other, and both may have
come into play during babassu's evolutionary history. Predominance of
male inflorescences; small (20 ym in diameter), monosulcate pollen
(personal observation); and lack of nectaries all point to pollination
by wind (Moore and Uhl 1982). Yet the strong scent produced by babas
su's staminate flowers and the bright inner surface of its inflorescence
bracts suggest adaptation to insect pollinators as well. The combina
tion of two pollination systems clearly enhances the palm's adaptability
to a wide range of ecological conditions.
Dispersal
The fruits of many palms are fleshy and colored, and obviously
adapted for dispersal by animals (Corner 1966, van der Pijl 1969, Kiltie
1981). Birds are probably the principal dispersal agents of palms
(McKey 1975), although mammals such as rodents and primates (including
humans) also feed on their fruits and disperse their seeds (Burtt 1929,
Janzen 1 971).
The dispersal of babassu fruits by animals other than humans occurs
almost exclusively within forested habitats, where pacas (Agouti paca)
and agoutis (Dasyprocta punctata) are abundant. These rodents carry
fruits to less exposed micro-sites (often > 4 distant, Fig. 25), where
the starchy mesocarp is partially or completely consumed and the fruits
abandoned. Scatterhoarding of babassu fruits by these rodents apparent
ly does not take place. Other mammals that I observed or were reported


Table 3. Fruit yields of babassu per ha and per palm during
1980-81 and 1981-82 on three sites. Values are
oven-dried mass in kg.
SITE
PER HECTARE
PER
PALM
1980-81
1981-82
1980-81
1981-82
Primary Forest
473
732
9.9
15.2
Secondary Forest
1328
1650
8.8
10.9
Pasture
1356
2866
15.4
32.6


64
served to feed on pollen and probably use both flower types as breeding
sites. Although they were far less abundant on female inflorescences,
more than one beetle was consistently observed entering each pistillate
flower. Entry was made either along the petal margins or, more common
ly, between the basally fused stigmas at the tip of the flower. Pollen
of babassu was found in isolated collections of My strops sp. obtained
from female inflorescences that contained no viable staminate flowers,
indicating pollen transport from another inflorescence. The beetles
were observed in both secondary forest and pasture, and similar speci
mens of My strops were collected from the crown of a babassu palm in
Pirapora, Minas Gerais, indicating that the association is widespread.
Wind pollination. Transport of babassu pollen by wind was re
peatedly confirmed by visual observations in both the pasture and sec
ondary forest sites. Preliminary analysis of pollen trapping data indi
cates a higher abundance of wind-borne pollen in the open pasture than
in the closed secondary forest. Variances in quantities of pollen-
trapped in both habitats were extremely high. Insect exclusion from one
female inflorescence in each habitat resulted in 57% fruit set in the
secondary forest (total =180 flowers) and 92% fruit set in the pasture
(total = 167 flowers). These results indicate that wind pollination
occurs in both habitats but may be more effective on the relatively open
pasture site.
Fruit Biology
Dispersal. The proportion of babassu fruits removed > 4 m was
significantly higher (p .01, one-way ANOVA) in the secondary forest


172
Gonsalves, A. D. 1955. 0 babapu: considerapes cientficas, tcnicas e
econmicas. Servico de Informapo Agrcola, Ministerio de Agri
cultura, Rio de Janeiro, Brazil.
Hamilton, W. D. 1966. The moulding of senescence by natural selection.
Journal of Theoretical Biology 12: 12-45.
Harcombe, P. A. 1977. The influence of fertilization on some aspects
of succession in a humid tropical forest. Ecology 58: 1375-1383.
Harper, J. L. 1977. Population biology of plants. Academic Press, New
York, New York, USA.
, and J. White. 1974. The demography of plants. Annual
Review of Ecology and Systematics 5: 419-463.
Hartley, C. W. S. 1977. The oil palm. Longman, London, England.
Hartshorn, G. S. 1972. The ecological life history and population
dynamics of Pentaclethra macroloba, a tropical wet forest dominant,
and Stryphnodendron excelsum, an occasional associate. Disser
tation. University of Washington, Seattle, Washington, USA.
. 1978. Tree falls and tropical forest dynamics. Pages
617-638 in P. B. Tomlinson and M. H. Zimmermann, editors. Tropical
trees as living systems. Cambridge University Press, Cambridge,
England.
Holdridge, L. R. 1967. Life zone ecology. Revised edition. Tropical
Science Center, San Jose, Costa Rica.
Holttum, R. E. 1955. Growth habits in monocotyledons: variations on a
theme. Phytomorphology 5: 399-413.
Hoopes, R. L. 1974. Flooding as a result of Hurrican Agnes, and its
effect on a macrobenthic community in an infertile headwater stream
in central Pennsylvania. Limnology and Oceanography 19: 853-857.
Hutchinson, G. E. 1957. Concluding remarks. Cold Springs Harbor
Symposium. Quantitative Biology 22: 415-427.
IBGE. 1956. Recenseamento geral do Brasil: Maranhao. Fundapo Insti
tuto Brasilierio de Geografia e Estatstica, Rio de Janeiro,
Brazil.
. 1979. Censo agropecuario: Maranhao. Fundapao Instituto Bra-
sileiro de Geografia e Estatstica, Rio de Janeiro, Brazil.
. 1981a. Anurio estatistico do Brasil. Fundapao Instituto Bra-
sileiro de Geografia e Estatstica, Rio de Janeiro, Brazil.


95
Climate Experiment
This experiment followed a 2 x 2 completely randomized design in
two blocks (dry and wet sites), thus involving a total of 40 plots. To
test for possible ecotypic differences between locally adapted
populations, fruit origin was superimposed as a treatment. In addition
to Lago Verde, fruits were obtained from a population of babassu near
the village of Tracuateua, ca. 200 km E of Belem (Fig. 9).
The variances of the data on babassu's performance (Tables 9 to 15)
were not homogeneous according to Bartlett's test (Snedecor and Cochran
1978) but were generally proportional to the square of the means. The
values were therefore logarithmically transformed. Statistical analyses
were performed on the transformed data; means are reported for
untransformed data. Statistical analyses involved application of the
general linear model (Freund and Littell 1981) followed by Duncan's
Multiple Range test when significant differences were detected.
Results
A consistent outcome of the irrigation (Fig. 33), fertilizer (Fig.
34), insecticide (Fig. 35), and climate (Fig. 36) experiments was higher
seedling recruitment (as measured by density) on the non-weeded plots.
Conversely, seedling growth (as measured by leaf area per palm) was
consistently higher on weeded plots. However, the second effect
consistently failed to compensate for the first effect, and babassu's
overall LAI was consequently higher on the non-weeded plots in all
experiments.


Figure 2. Aerial view of babassu
stands near Bacabal, Haranhao
Figure 3. Specimen of babassu
(CM martiana) in Pirapora,
Minas Gerais.
u>


59
Fruit yields per ha and per palm (Table 3) were highest in the
pasture, probably due to the moderate density and relative openness of
the babassu stand on this site. Per-ha yields were lowest in the pri
mary forest and intermediate in the secondary forest. This is probably
due to differences in palm densities, which were low in the primary
forest (48 mature palms/ha) and high in the secondary forest (150 mature
palms/ha). Conversely, per-palm yields were intermediate in the primary
forest and low in the secondary forest, probably due to increased shad
ing of mature palms on the latter site.
All three sites showed substantial increases in yields between the
1980-81 and 1981-82 fruiting seasons. On all sites the principal source
of these higher yeilds was an increase in the number of palms that bore
fruit; increased numbers of infructescences per fruiting palm consti
tuted the second most important source (Table 4).
Floral Biology
Fruit set. In the secondary forest, 78.4% (standard deviation =
24.3%) of flowers became fruits, 19-9% (24.3%) either aborted or were
not fertilized, and 1.7% (1.7%) were apparently attacked by the larvae
of an unidentified weevil. In the pasture, the figures were 82.1%
(20.7%), 17.1% (18.8%), and 0.8% (2.1%) respectively. These numbers
indicate high breeding success of babassu in both habitats.
Parthenogenesis. Fruit set consistently failed to occur in flowers
that were not exposed to pollen sources (n = 31 on three inflores
cences), as well as in flowers in which the stigmas had been removed


SEEDS PER PALM
135
Figure 40. Annual number of seeds produced per palm as a function
of mean age per life stage. Bars represent +1 standard
error. Arrow indicates age of youngest mature palm on
primary forest site, where data were obtained.


23
Grazing. In contrast to the rest of Maranho, grazing in the study
region is carried out predominantly on improved pastures. This has been
a recent development; in 1950, improved pastures comprised only 0.5% of
Maranhao's total occupied area (Table 1). The formation of improved
pastures has enabled grazing to encroach into areas once used exclusive
ly for shifting cultivation.
Conversion to pasture is generally carried out as part of shifting
cultivation. On upland plantations of crops such as rice and maize (but
rarely cassava, due to its relatively long cropping period), seeds of
the African grass, Hyparrhenia rufa (known locally as la.j eado), are
introduced just prior to harvest (March-May). Use of fertilizers is
rare, and legume-grass associations are apparently unknown. Pastures
are usually burned during the dry season, and the burn often gets out of
control; wildfires are a conspicuous feature of the landscape throughout
Maranho.
Babassu. The study region is characterized by exceptionally high
coverage of babassu. Spontaneous occurrence of the palm appears to be
limited only on steep hilltops (where fruit dispersal may limit colon
ization) and waterlogged soils (where root respiration may be inhib
ited). Elsewhere the palm forest virtually blankets the landscape,
except where it is thinned or clearcut by shifting cultivators and
ranchers. Although clearcutting of babassu is illegal in Maranho, the
practice appears to be increasingly common, especially in areas where
pasture conversion is widespread.


159
Shifting cultivators traditionally leave the palm forests intact or
carry out only moderate thinning (Fig. 45). A high proportion (ca. 70%)
of above-ground dry matter production in babassu stands is allocated to
leaves, which probably accounts for the preservation of the palm forests
under shifting cultivation. According to shifting cultivators on babas
su-dominated sites in Maranhao, a minimum fallow period of 4-5 yr re
sults in sufficient accumulation of fuel in the form of leaves to pro
duce the hot fire necessary for nutrient recycling and weed contol (Nye

and Greenland 1960). The minimum fallow period coincides with the
average leaf lifespan in babassu (4.4 yr); because leaf production
appears to remain constant after defoliation (see Chapter 3), palms
require this period to regain their full complement of leaves.
In short, babassu is extremely well adapted to traditional systems
of shifting cultivation. The palm forests form spontaneously, require
little or no maintenance, exhibit relatively few pest problems (see
Bondar 1954b) compared to plantation monocultures such as oil palm
(Hartley 1977) and coconut (Child 1974) promote site recovery during
the fallow (see Furley 1 975), and are already tightly integrated into
market and subsistence economies.
Limiting Factors
In recent years, conversion of formerly cultivated lands to cattle
pastures has occurred at an accelerating rate in frontier areas of
Brazil. With increased land pressures on shifting cultivators, average
fallow periods in the state of Maranhao declined from 8.7 yr in 1950 to
5.0 yr in 1975; in the study region (the central Mearim Valley), the


FRUIT PRODUCTION (kg/ha)
Figure 23. Monthly fruit production per ha on three study sites
at Lago Verde. Values represent air-dried mass.


2
3
4
5
6
7
8
9
10
11
12
Table 19. Estimated growth rates and ages of babassu palms in the primary
forest site at Lago Verde. Stages are defined in Table 18. N =
number of palms measured.
N LEAFLETS PER SIDE HEIGHT (cm) TIME IN AGE AT END
Range in Stage per Year Range in Stage per Year STAGE (vr) OF STAGE (vr)
min. max.x (s.d.) min. max.x (s.d.)x (s.d.)x
6.0 ( 3.1) 6.0
92
4.0
9.5
.92
( .48)
24
10.0
100.0
2.85
(1.73)
25
a
8
14
14
10
2
2
2
1
0
99
9.4
( 6.5)
100
199
14.0
( 9.5)
200
499
46.6
(28.7)
500
999
53.4
(13.1)
1000
1499
37.3
(13.4)
1500
1999
23.5
( 3.5)
2000
2499
18.7
( 6.1)
2500
2999
14.4
( 2.8)
3000
3499
8.0
( )
31.6
(
19.3)
37.6
10.5
(
7.4)
48.1
7.1
(
4.8)
55.2
6.4
(
4.0)
61.6
9.3
(
2.3)
70.9
13.4
(
4.8)
84.3
21.2
(
3.2)
105.5
26.7
(
8.7)
132.2
34.7
(
6.7)
166.9
62.4
(
)
229.3
122


133
cline (Fig. 37), probably due to the increased metabolic costs of main
taining a long stem (Zimmermann 1973). This decline is absorbed princi
pally by the palm's vegetative structures, especially the leaves; pro
ductivity of reproductive structures decreases moderately but continues
to increase on a relative basis.
At the outset of reproductive activity, most inflorescences pro
duced are male (Fig. 39). As the palm approaches the canopy, produc
tivity of female inflorescences and seeds per palm increases rapidly,
peaking at 104 yr. Productivity of female inflorescences (Fig. 39) and
seeds per palm (Fig. 40) subsequently undergoes moderate decline, while
productivity of male inflorescences increases. As wind pollination
occurs in babassu (see Chapter 3), pollen dispersal from emergent palms
is thus enhanced, which probably results in increased outcrossing.
At high densities, populations of babassu may attain high produc
tivity. The population on the secondary forest site, for example,
produced ca. 25 t of dry matter per yr (Table 23). Although the esti
mated productivity of the babassu stand is less than that of the entire
secondary forest, it is nonetheless one of the highest figures recorded
for a forested site, whether native or planted (Kira 1975, Cannell
1982). Among the few forested sites that I have found with higher
above-ground dry matter production (31.3 t*ha *yr ) is a plantation of
African oil palm in Malaysia (Ng et al. 1968). A common feature of both
babassu stands and oil palm plantations is that most of their produc
tivity is allocated to leaves, which are not permanently accruing and
thus may be relatively inexpensive to produce (Givnish 1978). In fact,


16
17
18
19
20
21
22
23
24
25
26
27
Soil chemistry of four treatments at end of nutrient
experiment 109
Soil chemistry of experimental sites at Lago Verde and
Belem 110
Definition of 12 life stages of babassu used in study . .117
Estimated growth rates and ages of babassu palms on the
primary forest site at Lago Verde 122
Estimated growth rates and ages of babassu palms on the
secondary forest site at Lago Verde 123
Partitioning of above-ground dry matter production of
babassu over an estimated life span of 184 yr 126
Annual above-ground dry matter production of mature babassu
palms on three study sites at Lago Verde 127
Annual above-ground dry matter production of all babassu
palms on the primary and secondary forest sites at Lago
Verde 128
Life table of the babassu population on the 1-ha primary
forest site at Lago Verde 144
Annual changes in number of palms per life stage on the 1-
ha primary forest site at Lago Verde 145
Annual changes in number of palms per life stage cn the 1-
ha secondary forest site at Lago Verde 146
Annual changes in number of palms per life stage cn the 1-
ha pasture site at Lago Verde 147
viii


4
Figure 5. Regrowth of babassu on a deforested site in the Tropical
Moist Forest zone of Maranhao.


FEMALE INFLORESCENCES
PER PALM (%)
Figure 20. Distributions of female inflorescences per palm in six
populations of babassu. Forest and pasture sites = Lago
Verde; wet site = Pindar Mirim; dry site = Caxias;
northerly site = Tocantinpolis, Gois; and southerly
site = Pirapora, Minas Gerais. N = number of palms ob
served per site.


72
PALM
NUMBER
I I I I I I 1 I 1
0 20 40 60 80 100
GERMINATION (%)
Figure 29. Germination in fruits obtained from 10 babassu
palms. Each hatched bar represents one palm;
intervals represent il standard error.


19
the most agriculturally productive areas in Maranhao. Yet agricultural
productivity has not brought prosperity to most of the people living
there.
Land tenure. The main front of agricultural expansion in Maranhao
swept westward through Bacabal during the 1950s. But the pattern by
which the study region was settled can still be seen today in other
areas of frontier expansion throughout Brazil (Velho 1972). As settle
ment pushes forward, once isolated areas become increasingly accessible,
and the process of land concentration soon commences. The aftermath of
this process can be seen from the data on land distribution in Maranhao
(Fig. 14). In 1950, 61.8% of the state's occupied lands were owned by
2.0% of its households. Despite a rapid expansion of area under occupa
tion, there had been no discernible improvement by 1975, when 41*4% of
the land was controlled by 0.5% of the households. In absolute terms,
the amount of land in smallholdings (<100 ha) increased by a factor of
4.0; however, the number of smallholdings increased by a factor of 5.6.
Smallholdings became, on the average, smaller, and the number of land
less increased.
As throughout most of Maranhao, land tenure in the study region is
now well defined. The vast majority of rural inhabitants live in
thatched huts strung like beads along the roads, footpaths, and water
courses that form a network throughout the region. Behind these hamlets
and villages, fences are invariably present, and beyond these fences lie
the resources upon which people depend.


Appendix A continued.
SPECIES
ABUNDANCE
stemg/ha%
FREQVENCX

DOMINANCE
nr/ha %
IMPORTANCE
i
45. Amaioua euianensis Aubl.
1
0.26
0.48
0.033
0.12
0.29
46. Unknown Vine No. 1
1
0.26
0.48
0.026
0.10
0.28
47. Himatanthus sucuuba (SDruce) Wood.
1
0.26
0.48
0.024
0.09
0.28
48. Mvrcia cf. falax (Rich.) DC.
1
0.26
0.48
0.022
0.09
0.28
49. Casearia 3D.
1
0.26
0.48
0.019
0.07
0.27
50. Erisma cf. uncinatum Warm.
1
0.26
0.48
0.018
0.07
0.27
51. Manilkara sd.*
1
0.26
0.48
0.018
0.07
0.27
52. Unknown Vine No. 2
1
0.26
0.48
0.016
0.06
0.27
58. Protium Dallidum Cuatr.
1
0.26
0.48
0.015
0.06
0.27
54. Unknown Vine No. 3
1
0.26
0.48
0.014
0.05
0.27
55. Trichilia micrantha Benth.
1
0.26
0.48
0.013
0.05
0.26
56. Pouteria sd.
1
0.26
0.48
0.012
0.05
0.26
57. Simaba cedrn Planch.
1
0.26
0.48
0.011
0.04
0.26
58. Unknown Vine No. 4
1
0.26
0.48
0.010
0.04
0.26
59. Anacardium cf. microcamum Ducke
1
0.26
0.48
0.008
0.03
0.26
182


Table 15. Oven dried mass of babassu and competitors in four
experiments. Vegetation was ca. 18-mo old at time of
harvest (May 1982). Values are g per m^; mean percent
of total dry mass in parentheses. Significant
differences are shown as (p £.05).
Irrigation Fertilizers Insecticides Climate
No
Yes
No
Yes
No
Yes
Drv
Wet
COMPETITOR
Fruits
9.8
5.8
37.8
37.3
7.6
24.0
16.5
21.3
Leaves
247.5
285.0
391.0
516.0
476.5
740.0
457.0
427.2
Stems
744.5
664.5
724.8
930.8
927.0
3,013.0
1,690.0
421.2
Standing Dead
749.6
167.4
105.8
32.4
88.3
106.7
72.3
31.2
Total Competitors
1,751.4
1,122.7
1,259.4
1,516.4
1,499.4
3,883.7
2,235.8
900.9
BABASSU
33.0
32.5
34.6
32.6
40.8
33.2
28.3
36.1
(2.5)
(2.8)
(3.3)
(2.7)
(4.3)
(1.4)
(2.3)
* (4.3)
TOTAL
1,784.4
1,155.2
1,294.0
1,549.0
1,540.2
3,916.9
2,264.1
937.0
108


LAND (%)
1950
(95,382 Km2)
1975
(124,091 km2)
Figure 14. Distribution of land and households according to size of
landholding in rural areas of Maranhao during 1950 and
1975. Size of holding defined by effective occupation of
the land. Amount of land increased from 1950 to 1975 due
to frontier expansion. Sources: IBGE 1956, 1979.


37
lower Tertiary) may have forced the Tocantins to change to its present
course.
A final anomaly in babassu's distribution is its disjunct
occurrence in the Brazilian state of Cear (Fig. 1). Here the palm is
confined to isolated and geologically very old uplifts, and there are no
river systems that run between these structures and the highlands of
Gois. The uplifted areas to which babassu is confined in Cear have
long served as favorable sites for human settlement and agriculture, due
to their relatively cool temperatures, high rainfall, and fertile soils.
Markley (1971) argued that babassu's occurrence in Cear is due to
dissemination by humans. Early aboriginal groups may have introduced
fruits obtained from the Parnaiba valley ca. 200-300 km distant.


INFLORESCENCES PER PALM
134
4.0 h
3.0-
2.0-
.0-
TOTAL INFLORESCENCES
O FEMALE INFLORESCENCES

60
ai:iiiiiiiIr
80 100 120 140
MEAN AGE (yr)
60
I I
180
Figure 39. Annual number of inflorescences produced per palm
as a function of mean age per life stage. Arrow
indicates age of youngest mature palm on the pri
mary forest site, where data were obtained.


82
beetle larvae apparently ceased (Fig. 26). Approximately 15-18 d fol
lowing oviposition, tiny (0.75 x 0.10 mm) larvae emerge and proceed to
the base of the fruits. Penetration takes place through the germination
pores, which offer minimal resistance. Prior to abscission, fruits are
attached at the bases and their germination pores are thus protected,
which explains why fruits are only attacked after they fall. Approxi
mately 2-3 d after penetration begins, the larvae reach the seeds.
Observations of fruit colonization revealed a maximum of one larva per
seed chamber (locule), suggesting that late arrivals are probably canni
balized (see Janzen 1971); movement between seed chambers is prohibited
by the thickness of the inner endocarp walls. One seed provides suffic
ient food for a larva's complete development. Adults emerge ca. 2.5 mo
after oviposition by burrowing a ca. 8-mm-wide exit hole through the
endocarp.
In a closely related palm (Scheelea rostrata) in Costa Rica, Brad
ford and Smith (1977) found that ca. 60% of the fruits lost all their
seeds to a predaceous bruchid beetle (Caryobruchus buscki) or to ro
dents; in babassu, the total figure was ca. 20% (excluding potential
losses to humans). A key component of this high predator escape is the
predominance of many-seeded fruits in babassu (Table 5); in contrast,
most (ca. 95%) fruits in the Costa Rican population of Sh rostrata were
one-seeded (Bradford and Smith 1977). In both species, the proportion
of seeds eaten by predators declined as the number of seeds per fruit
increased. Thus, in S;_ rostrata mean seed losses to predators in one-,
two-, and three-seeded fruits were 61 %, 46%, and 40%, respectively


163
Possible Solutions
In 1979, the government of Maranho established a research insti
tute, the Instituto Estadual do Babapu (INEB) dedicated exclusively to
developing viable management of the palm forests and increasing their
integration into the market economy. Since its founding, this institute
has set up a germplasm bank for babassu located near the town of Baca-
bal. Comprised of seeds collected over the geographic range of babassu
and its suspected hybrids, the bank will eventually provide genetic
material for selection and breeding. The goal is to develop precocious,
high-yielding strains for planting. If other crops such as oil palm
serve as a model (Pyke 1972), such strains will ultimately be utilized
in intensively cultivated plantations, which will then form the basis
for an efficient production system capable of providing centralized
industries with a steady flow of raw materials.
Babassu's high genetic variability indicates that the potential for
developing improved strains of the palm are excellent. Present annual
fruit yields of babassu stands in Maranho average 1-2 t/ha (Anonymous
1981), which is similar to those of wild stands of oil palm in Nigeria
(Zeven 1967); selection in the latteras well as management subsidies
have increased yields by a factor of 10 (Purseglove 1972). Like oil
palm, sex ratios in babassu appear to be under genetic control and thus
may be amenable to selection. Likewise, growth rates of babassu are
extremely slow in closed forests; although considerably higher on more
open sites, palms may still require more than a decade to begin fruit
production, according to local informants. This time period can probab-


190
APPENDIX D
SELECTIVE COMPARISON OF GROSS MORPHOLOGICAL CHARACTERS IN
Orbignva martiana. "0. teixeirana", AND CL. eichleri.
CHARACTER
0. martiana
"0. teixeirana"
(possible hybrid)
STEM
Length in adults (cm)
500-3,000
0-800
DBH (cm)
19-50
19-42
LEAF
Number per palm
10-25
8-13
Sheath
length in adults (cm)
90-220
30-100
striations on abaxial surface
generally present
present or absent
Petiole
length in adults (cm)
0-50
17-40
Rachis
length in adults (cm)
520-990
152-745
cross section
4-sided
4-sided
abaxial surface
weakly to strongly
lepidote
weakly lepidote
Pinnae
number per side
145-240
96-190
orientation
in same plane
in same plane to
crispate
disposition
regular
clustered to regular
adaxial surface
dull to lustrous
dull
abaxial surface
weakly glacous
weakly glacous to
smooth
0. eichleri
0
0
3-8
20-33
absent
50-80
120-253
triangular
smooth
63-128
crispate
clustered
dull
mostly smooth


166
desirable strains of babassu may soon be feasible. The potential exists
for inducing self-pollination in palms with promising characteristics
such as a high proportion of female inflorescences or precocity. Whether
or not selfing can be induced in babassu is not presently known, but
incompatibility does not occur in other cultigens such as coconut and
oil palm, and the techniques for artificial selfing in these palms are
simple (Child 1974 Hartley 1977).
Desirable strains of babassu could be introduced either under
stands that presently exist or on deforested sites. With proper fire
control, underplanting would be especially appropriate in pastures fol
lowing elimination of seedlings and juveniles. Such controlled regen
eration would enable ranchers to improve their stands genetically while
maintaining constant forest cover; considerable precedent exists for
combining plantations of palms and pastures (e.g., Plucknett 1979).
Alternatively, desirable strains of babassu could be used for reforesta
tion of degraded sites in Amazonia and other tropical regions. With
minimal effort, high-density stands could be established over extensive
areas, providing economic return while promoting site recovery.
Long-term maintenance of the babassu economy ultimately depends on
the solidity of its base. If current technologies for processing fruits
were scaled down to the community level, the bottleneck of manual fruit
breaking could be bypassed with fewer negative effects on the people who
depend on babassu for their livelihood. Communities could more effi
ciently produce their own subsistence products such as oil, fuel, and


42
of a random sample of 10 infructescences obtained in each of the two
habitats.
Studying pollination in babassu required recognition of female
inflorescences before they opened. As no morphological differences
could be discerned between the two inflorescence types prior to anthe-
sis, I distinguished them by opening a small slit in the bract and
examining the flowers within. In female inflorescences, the slit was
subsequently sealed with tape to prevent contamination.
To determine whether self-pollination occurs in babassu, sleeves
were designed to exclude external sources of pollen from female inflor
escences. Each sleeve was made of light-weight canvas in the form of a
tapered cylinder with a length of ca. 2.5 m and a maximum diameter of
ca. 0.5 m. Circular wires were sewn on the inside of the sleeve to
provide support, and a small polyethylene window was placed in the
middle to allow observation of the inflorescence. Entry by insects was
prevented by painting bands of insect trapping adhesive over the base
and tip of the inflorescence bract; one end of the sleeve was placed
over each band and securely shut with wire.
Small bags were designed to isolate individual pistillate flowers
from pollen sources. To reduce chances of contamination, the entire
inflorescence was covered with a sleeve prior to opening. Each flower
was individually bagged after the inflorescence had opened, and the
sleeves were subsequently removed. Time of pollination was estimated by
exposing pistillate flowers to external sources of pollen at different
periods following opening of the bract. To test whether parthenogenesis


31
80 70 60 50 40
Figure 15. Distribution of specimens of Orbignya martiana and its
suspected hybrid complexes. Circles represent (). mar
tiana; squares represent the Orbignya-Maximiliana com
plex; triangles represent the JO. martiana-O. eichleri
complex. Sources: Wessels Boer (1965), Glassman (1977),
and M. J. Balick and A. B. Anderson (unpublished).


156
during which the risk of death by cutting and burning is greatest. The
fallow also provides shade for enhanced germination and seedling re
cruitment. Periodic alternations of forest clearing and regrowth thus
lead to the formation of exceptionally dense stands of babassu.
Following the initial clearing of primary forest, a population of
babassu will continue to flourish even if subsequent clearing is delayed
considerably, as indicated by continued regeneration of the secondary
forest population after a fallow of 33 yr (Fig. 44)* The time required
to return to the original forest composition prior to disturbance may be
extremely prolonged. As fallows >10 yr are exceptional, in zones cur
rently dominated by babassu, one can safely predict that once a popula
tion captures a site, it can maintain itself indefinitely. The stand
will only be replaced by intentional eradication, or by annual cutting
and burning on a long-term basis.
When subjected to sporadic disturbances, populations of babassu
consequently thrive. However, when a stand is subjected to more fre
quent disturbances such as annual pasture fires, regeneration is no
longer sufficient for long-term maintance of the population. This
results from a combination of low seedling recruitment due to fruit-
gathering humans and high mortality of juveniles due to repeated expo
sure of their apical meristerns to fire. Because of babassu's longevity,
the ultimate demise of pasture stands will not occur in the near future.
Nevertheless, widespread conversion of formerly cultivated lands to
pastures throughout Maranho has important implications for the future
of the palm forests.


105
Table 12. Palms per m2 (a), leaf area per palm (b), and
leaf area index (c) of babassu at end of
climate experiment, with climate and weeding
as main effects. Asterisks indicate
significant differences (** for p £.01)
between main effect means. Where there is
interaction between main effects, superscripts
are used to compare treatment means. Values
(a)
that do not share the same superscript differ
significantly (p £.05).
p
Palms Der nr
Drv Climate
Wet Climate
Mean
Not Weeded
25.0a
26.5a
25.8
Weeded
3-9
13-9b
8.9
Mean
14.4
20.2
(b) Leaf Area per Palm (cm2)
Drv Climate
Wet Climate
Mean
Not Weeded
231.2b
254.2b
242.7
Weeded
433.0a
278.7b
355.8
Mean
332.1
266.4
(c)
Leaf Area Index
Drv Climate
Wet Climate
Mean
Not Weeded
.25
.39
.32
*
Weeded
.13
.19
.16
Mean
.19
** .29


100
Irrigation experiment
In addition to the presence of other plants (henceforth referred to
as competitors), irrigation had a pronounced effect on germination in
babassu, as indicated by the high density of palms on irrigated plots
(Table 9a). Weeded plots that were not irrigated contained a signifi
cantly lower (p _^.01) palm density than any other treatment combination.
But irrigation was a less effective means of promoting germination than
competitor presence, as indicated by the significantly lower (p.05)
density of palms on weeded, irrigated plots than either treatment combi
nation on non-weeded plots. Irrigation did not influence growth (Table
9b), and babassu's substantially higher LAI (Table 9c) on irrigated
plots is consequently due to increased density of the palm under irriga
tion.
Effects of irrigation on babassu's competitors appear to be
minimal. Competitor LAI was substantially higher on non-irrigated plots
(Table 14). However, dry mass of competitor leaves was marginally
greater on irrigated plots (Table 15). The higher overall mass of
competitors on non-irrigated plots is almost entirely attributable to
standing dead.
Fertilizer experiment
Application of substantial levels of N, P, and K had no effect on
the performance of either babassu or its competitors (Tables 10, 14, and
15). No differences in the relatively mobile N and K were detected
between treatments, ca. 2 mo after the final application of fertilizers


91
The Belem site consisted of ca. 8-yr-old secondary forest in which
babassu was absent. The soil was a Dystrophic Yellow Latosol (Brazilian
classification system), equivalent to a Xanthic Ferrasol (FAO Legend) or
Haplorthox (U.S.D.A. Soil Taxonomy). Monthly rainfall for the site is
p
shown in Figure 13. An area of ca. 1,600 m was cut on 1 October and
burned on 10 November 1980. Subsequent site preparation was identical
to that used at Lago Verde. Twenty plots were set up and babassu fruits
planted on 19-21 November 1980.
At each site, four composite soil samples (each comprised of 10
cores) were obtained after clearing and before burning the fallow.
After air drying, all soil samples were analyzed at the soil laboratory
of the Centro de Pesquisa Agropecuria do Trpico Umido (CPATU) in
Belem. Soil pH was determined in a 1:1 aqueous solution. Organic C was
determined by the Walkley-Black (193*0 wet digestion method. Total
soil N was determined by the regular macro-Kjeldahl method (Black 1965).
Cations and P were extracted by the double acid extraction method (0.5 N
HC1 and 0.025 N H2S0^). Cations were analyzed by atomic absorption
spectrophotometry; P was analyzed colorimetrically.
On half the plots at Lago Verde and Belem, competitors of babassu
were periodically weeded. Time intervals between weedings ranged from 4
wk during the wet season to 10 wk during the dry season.
Babassu's density, leaf area per palm, and leaf area index (LAI)
were measured ca. 10, 14, and 18 mo after initiation of the experiments.
Density of babassu was calculated as the number of germinated fruits per
m^. To determine the relationship between leaf length and leaf area


86
7i<>ure 31. Cryptogeal germination in babassu. from left to right,
seedlings were 1-, 4-, 8-, and 14-wk old. Scale 15 cm.


PRODUCTIVITY (kg/yr)
130
LENGTH OF STEM (m)
O I 25 10 15 20 30
I 20
I OOH
80H
REPRODUCTIVE
STRUCTURES
60 H
40H
20^
20 40 60 80 100 120 140 160 180 200
AGE (yr)
Figure 37. Absolute allocation of annual above-ground dry matter
production per babassu palm as a function of age.


75
Leaf and inflorescence production in babassu are strongly cor
related due to the close physical association of the two structures.
Nevertheless, most phenological studies of palms in their natural habi
tats (e.g., Mora Urpi and Solis 1980, Bullock 1981, Myers 1981, Beach in
press) have dealt exclusively with reproductive components. To my
knowledge, studies of both leaf and inflorescence production in palms
have been limited to cultigens under plantation conditions (e.g., Child
1974, Hartley 1977). In the African oil palm, for example, leaf produc
tion occurs throughout the year, which results in nearly continuous
flowering (Corley et al. 1976); in babassu, leaf production and flower
ing tend to be seasonal. Likewise, the minimum annual rate of leaf
production in oil palm recorded in the literature is ca. 18 leaves
(Corley et al. 1976), compared to a mean of ca. four leaves in babassu.
Probably as a result of its high rate of leaf production, minimum leaf
lifespan in wild stands of oil palm in Nigeria is only 2.1 yr (Zeven
1967); in babassu, mean leaf lifespan is 4*4 yr. I believe that this
latter figure has important implications for management of babassu under
shifting cultivation, a topic explored further in Chapter 7.
As has been documented in many other species (e.g., Harper 1977,
Janzen 1978b, Pinero and Sarukhan 1982), fruit yields in babassu vary
considerably from year to year. As is the case in oil palm (Hartley
1977), I suspect that climatic factors such as rainfall may have a
significant effect on annual variability in babassu's fruit yields. The
influence is probably most pronounced at the time of inflorescence
differentiation rather than at fruit set. Examination of developmental


116
Methods
Growth
To measure growth in babassu, I defined 12 life stages based on
distinctive morphological characters (Table 18). On the primary and
secondary forest sites, a sample of palms from each stage was labelled
and mapped. Total sampling area allocated for each stage was 0.05 ha
for stage 2, 0.1 ha for stage 3, 0.2 ha for stages 4-5, and 0.5 ha for
stages 6-12.
In stages 2 and 3, vertical stem growth had not begun and leaf
measurements provided the only indicators of growth. Leaf size in
babassu continued to increase through stage 5 and thus appeared to be
positively correlated with growth in early life stages. On the newest
fully expanded leaf of each palm in stages 2 and 3, I measured the
length of the rachis and the number of leaflets on each side; both were
remeasured after 1 yr. Annual changes in rachis lengths produced a far
greater coefficient of variation than did annual changes in leaflets per
leaf. I therefore selected the latter as a more reliable measure of
growth. This measure was used to calculate a mean growth rate of palms
in stages 2 and 3. In both the primary and secondary forests, a high
percentage (ca. 50%) of these palms exhibited negative or negligible
growth rates. If all individuals measured were used to derive mean
growth rates, I calculate that the average time for a palm to pass
through stages 2 and 3 would be 207 yr in the secondary forest and 738
yr in the primary forest. As these numbers are clearly untenable, I
arbitrarily assumed that individuals exhibiting negative or zero growth


177
Taher, M. M., and R. C. Cooke. 1975. Shade-induced damping-off in
conifer seedlings. I. Effects of reduced light intensity on infec
tion by necrotrophic fungi. New Phytologist 75: 567-572.
Tomlinson, P. B. 1963. Measuring growth rates in palms. Principes 7:
40-44.
. 1979. Systematics and ecology of the Palmae. Annual
Review of Ecology and Systematics 10: 85-107.
, and P. K. Soderholm. 1975. The flowering and fruiting
of Corypha elata in South Florida. Principes 19: 83-99.
, and M. H. Zimmermann. 1966. Anatomy of the palm
Rhapis excelsa. III. Juvenile phase. Journal of the Arnold Arbor
etum 47: 301-312.
Turner, T. 1983. Facilitation as a successional mechanism in a rocky
intertidal community. American Naturalist 121: 729-738.
Uhl, N. W., and H. E. Moore, Jr. 1973. The protection of pollen and
ovules in palms. Principes 17: 111-149.
, and H. E. Moore, Jr. 1977. Correlations of inflorescence,
flower structure, and floral anatomy with pollination in some
palms. Biotropica 9: 170-190.
Valverde, 0. 1957. Geografia econmica e social do babapu no meio
norte. Revista Brasileira de Geografia 19: 381-420.
van der Pijl, L. 1969. Principles of dispersal in higher plants.
Second edition. Springer-Verlag, New York, New York, USA.
Van Valen, L. 1975. Life, death, and energy of a tree. Biotropica 7:
260-269.
Velho, 0. G. 1972. Frentes de expanso e estrutura agricola. Zahar
Editores, Rio de Janeiro, Brazil.
Viveiros, F. F. 1943. 0 babacu nos estados do Maranhao e Piaui.
Boletim do Ministerio de Agricultura (Rio de Janeiro, Brazil) 32:
1-43.
Walkley, A., and I. A. Black. 1934* An examination of the Degtjareff
method for determining soil organic matter, and a proposed modifi
cation of the chromic acid titration method. Soil Science 37: 29-
38.


Appendix D continued
CHARACTER
0. martiana
INFLORESCENCE
Bracts
length of prophyll (cm)
length of acumen of peduncular
bract (cm)
abaxial surface of
peduncular bract
Peduncle
length (cm)
Rachis
length (cm)
curvature
50-150
16-32
tan to rust-
brown
56-185
48-130
not recurved
Rachillae
attachment to rachis
throughout
length of subtending
bracteole
Staminate flowers
petals per flower
fusion of petals
ca. 2-4 mm
2 (-3)
separate
stamens per flower
(22-)24-26(-30)
FRUIT
Staminoidal ring
weakly to strongly
defined
0. eichleri
"0. teixeirana"
(possible hybrid)
0. eichleri
30-33
11-15
20
6-12
blood red
blood red
37-95
37
34-110
weakly to strongly
recurved
20-37
strongly recurved
on abaxial side
only
ca. 14-18 mm
on abaxial side
only
ca. 15-50 cm
2
separate
(2-)3(-4)
separate to partially
connate
20-23
13-16
weakly to moderately
defined
weakly defined


16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Mean monthly gain and loss of leaves per palm on three
sites in Maranho 47
Mean monthly gain and loss of leaves per palm at three
study sites at Lago Verde 48
Mean monthly production of male and functionally female
inflorescences per palm at three sites in Maranho 50
Mean monthly production of male and functionally female
inflorescences per palm at three sites at Lago Verde 51
Distributions of female inflorescences per palm in six
populations of babassu 54
Mean monthly number of immature and mature infructescences
per palm at three sites in Maranho 55
Mean monthly number of immature and mature infructescences
per palm at three sites at Lago Verde 56
Monthly fruit production per ha on three study sites in
Lago Verde 57
Monthly fruit production per palm on three study sites at
Lago Verde 58
Removal of babassu fruits on three study sites at Lago
Verde 66
Attack of babassu fruits and seeds by the Bruchid beetle,
Pachymerus nucleorum, over time 67
Seed fate over time in three study sites at Lago Verde. ... 69
Cumulative germination of babassu fruits over time 71
Germination in fruits obtained from 10 babassu palms 72
Force required to break palm fruits 84
Cryptogeal germination in babassu 86
Leaf area as a function of leaf length in seedlings of
babassu 93
Palms per m^ (a), leaf area per palm (b), and leaf area
index (c) of babassu over time in irrigation experiment ... 96
x


112
(p £.01 and p £ .05, respectively). These responses are largely
attributable to decreased vigor of competitors. Under the dry climate,
stem and total biomass of competitors were substantially higher (Table
15)and competitor LAI significantly higher (p £ .01, Table 14) than
under the wet climate. The latter site, however, was characterized by a
significantly higher LAI of grasses (p £ .05, Table 14). Lower soil
fertility (Table 17) at the wet climate site may account for the lower
overall competitive vigor and a shift to greater dominance by grasses,
with concomitant increases in babassu's relative performance.
Discussion
Seedling Growth
In the experiments described above, babassu's performance was
separated into two components: growth (leaf area per palm) and density
(palms per m2). As in other palms (e.g., Rees 1963), growth in babassu
was consistently enhanced on weeded plots. Growth likewise showed
significant responses to ecotype. Enhanced growth by locally obtained
palms at each site indicates differential adaptation between populations
of babassu, rather than mere acclimitization to local conditions.
Experimental application of subsidies in the form of irrigation,
fertilization, or insecticides did not enhance growth in babassu. The
lack of response in either babassu or its competitors to fertilizers was
particularly striking. Under the relatively high soil fertility
characteristic of the Lago Verde site, nutrients may not be limiting to
plant growth. Alternatively, a nutrient other than N, P, or K may have
been limiting on the site. On fertile volcanic soils in Costa Rica,


Table 7. cont'd.
OCCURRENCE
TAXON
Stems
Leaves Inflorescences
Fruits
Stored
Seeds
Coleptera (cont'd)
Cucujidae
14. OrvzaeDhilus surinamensis*
+
Curculionidae
Baridinae
15. Tonesia babassu
+
Cholinae
16. Horaalinotus coriaceus*
+
17. B, validus
+ +
+
Petalochilinae
18. Celestes babassu
+
19. Cf mosesi
+
20. C. orbienvae
+
21. C. rosentali
+
22. C, teixeira-leitei
+
Pyropinae
21. Belopoeus sp.
+
24. B. orbienvae
+
25. Rhinostomus barbirostris
+
26. RhvnchoDhorus Dalmarum*
+
27. SDhenoDhorus sdd.
+
Nitidulidae
28. CarDODhilus sdd.*
+


Appendix A continued.
S...E JLJL-L-E-J5
30. Casearia svlvestris Sw.*
31. Seguieria macrophvlla Benth.
32. Ocotea sp.
33. Cordia scabrifolia A.DC.
34. Protium insigne (Tr. ex PI.) Engl.
35. Sclerolobium paniculatum Vog.
36. Licania cf. aptala (E. Meyer) Fritsch*
37. Cocoloba latifolia Lam.*
38. Calvcophvllum spruceanum Benth.
39. Schizolobium amazonicum Hub. ex Ducke
40. Couratari cf. macrosperma A.C. Smith
41. Astronium cf. fraxinifolium Schott.
42. Christiana africana DC.
43. Citharexvlum cf. macrophvllum Poeb.
44. Mouriri cf. guianensis Aubl.
ABUNDANCE FREQUENCY DOMINANCE IMPORTANCE
stems/ha% mfVhi %
2
0.52
0.97
0.025
0.09
0.53
2
0.52
0.97
0.022
0.09
0.52
3
0.78
0.48
0.077
0.29
0.52
2
0.52
0.48
0.057
0.22
0.41
2
0.52
0.48
0.042
0.16
0.39
1
0.26
0.48
0.102
0.39
0.38
1
0.26
0.48
0.093
0.35
0.37
2
0.52
0.48
0.022
0.08
0.36
1
0.26
0.48
0.084
0.32
0.35
1
0.26
0.48
0.080
0.30
0.35
1
0.26
0.48
0.067
0.26
0.33
1
0.26
0.48
0.059
0.22
0.32
1
0.26
0.48
0.052
0.20
0.31
1
0.26
0.48
0.040
0.15
0.30
1
0.26
0.48
0.040
0.15
0.30
oo


175
, and N. W. Uhl. 1982. Major trends of evolution in palms.
Botanical Review 48: 1-69.
Mora Urpi, J., and E. M. Solis. 1980. Polinizacin en Bactris gasipaes
H. B. K. (Palmae). Revista Biologia Tropical 28: 153-174-
Myers, R. L. 1981. The ecology of low diversity palm swamps near
Tortuguero, Costa Rica. Dissertation. University of Florida,
Gainesville, Florida, USA.
Ng, S. K., S. Thamboo, and P. de Souza. 1968. Nutrient contents of oil
palms in Malaya. II. Nutrients in vegetative tissues. Malaysian
Agricultural Journal 46: 332-390.
Nye, P. H., and D. J. Greenland. 1960. The soil under shifting culti
vation. Commonwealth Agricultural Bureaux, Farnham Royal, Bucks,
England.
Pinero, D., and J. Sarukhn. 1982. Reproductive behaviour and its
individual variability in a tropical palm, Astrocaryum mexicanum.
Journal of Ecology 70: 461-472.
, J. Sarukhan, and P. Alberdi. 1982. The costs of repro
duction in a tropical palm, Astrocaryum mexicanum. Journal of
Ecology 70: 473-481.
Plucknett, D. L. 1979- Managing pastures and cattle under coconuts.
Westview Press, Boulder, Colorado, USA.
Purseglove, J. ¥. 1972. Tropical crops: Monocotyledons. Longman,
London, England.
Pyke, M. 1972. Technological eating, or where does the fish-finger
point? John Murray, London, England.
Rathcke, B. J. In press. Patterns of flowering phenologies: test
ability and causal inference using a random model. In D. R.
Strong, D. Simberloff, and L. G. Abele, editors. Ecological com
munities: conceptual issues and the evidence. Princeton University
Press, Princeton, New Jersey, USA.
Rees, A. R. 1963- An analysis of growth of oil palm seedlings in full
day-light and in shade. Annals of Botany 27: 325-337.
Richards, P. W. 1952. The tropical rain forest. Cambridge University
Press, Cambridge, England.
Rizzini, C. T. 1963. Sobre a distinpo e distribuipao de duas especies
de babapu (Orbignya). Revista Brasileira de Geografia 25: 313-326.


APPENDIX C. MORPHOLOGICAL DESCRIPTION OF Orbignva
martiana Barb. Rodr.
Stem large, solitary, columnar, to ca. 30 m high and 19-50 cm in
diameter; leaf scars inconspicuous, ca. 8-30 cm apart; leaf bases
persistent just below the crown.
Leaves ca. 10-25, erect-arching; sheath ca. 90-220 cm long, often
with yellow, longitudinal striations on abaxial surface that may extend
to base of rachis; petiole 0-50 cm long; rachis 520-990 cm long, base
trough-shaped in cross section, middle more or less 4-sided in cross
section and generally channelled longitudinally above insertion of
pinnae, apex triangular in cross section, abaxial and lateral surfaces
at middle of apex weakly to densely white- or brown-lepidote, abaxial
surface smooth; pinnae 145-240 per side, inserted at regular intervals
and in same plane, rigid, dull to lustrous above, glaucous beneath,
midvein prominent; basal pinnae 80-185 cm long, 0.5-2.0 cm wide; middle
pinnae 95-168 cm long, 1.5-7.0 cm wide; apical pinnae 22-70 cm long,
0.5-3-0 cm wide.
Inflorescences androdioecious, interfoliar; prophyll ca. 50-150 cm
long; peduncular bract woody, persistent, ca. 140 to over 300 cm long
and 12.5-41 cm wide, bearing an acumen 16-32 cm long, adaxial (interior)
surface white- to yellow-lepidote at anthesis, becoming tan- to rust-
brown over time; all axes of inflorescences white- to yellow-lepidote at
anthesis; peduncle 56-185 cm long; rachis 48-130 cm long; staminate
inflorescences bearing up to 648 rachillae, the latter ca. 10-25 cm
long, erect, subtended by a ca. 4 mm long bracteole, bearing 17-102
187


28
Thus it is no surprise that babassu is often confused with other species
and even genera, notably Attalea. Oil palms commercially referred to as
babassu include Orbignva agrestis in Para, Attalea olefera in Gois and
Minas Gerais, A. geraensis in Minas Gerais, and A. pindobassu in Bahia
(Markley 1971). Botanists are not immune to generic confusion.
Aboriginal dispersal was evoked to account for the disjunct occurrence
of babassu in the state of Sao Paulo (Ferri 1974, 1980), although the
population has been identified as pertaining to Attalea (J. T. de M.
Costa, pers. comm.).
To resolve intergeneric confusion in the Attalea alliance, Wessels
Boer (1965) recommended reducing Orbignva and related genera to the
single genus Attalea. but his suggestion has not been followed by other
contemporary taxonomists (Moore 1973b, Glassman 1977). Countervailing
arguments include the weight of taxonomic tradition, and the observation
that Wessels Boer's scheme merely reduces the confusion to the specific
level (H. E. Moore, pers. comm.). Due to the uncertain limits between
genera, the possible occurrence of intergeneric hybrids, and the
practical difficulty of distinguishing genera in the field, I believe
that the current splitting of the Attalea alliance into five genera
(Moore 1973b) is not justified. Final resolution of the confused
generic limits within the Attalea alliance will require exhaustive study
of the entire group. As such a study is not on the horizon at present,
I shall follow taxonomic tradition and consider babassu within the genus
Orbignva.


62
(n = 8 on one inflorescence). These results indicate that parthenogen
esis does not occur in babassu.
Self-pollination. Inflorescences in babassu are generally either
male (containing exclusively staminate flowers) or female (containing at
least one pistillate flower and a variable number of staminate flowers
that are usually inviable). The former are far more numerous than the
latter. During a 3-wk period in March 1981, 21 (12%) of 173 inflores
cences examined contained pistillate flowers. Of these 21 inflores
cences, only three (14^) contained staminate flowers that shed pollen
and thus appeared to be viable. Consequently, most inflorescences that
contain pistillate flowers are exclusively female in a functional sense.
In a pistillate flower bagging experiment, fruit set occurred only
in flowers exposed 0-48 h after opening of the inflorescence; subse
quently exposed flowers did not set fruit. Thus pistillate flowers
appear to be receptive for a maximum of 48 h after inflorescences open.
In the three female inflorescences with functional staminate flowers
that I observed, pollen release invariably began more than 48 h after
opening of the inflorescence. In one of these inflorescences, fruit set
did not occur when external sources of pollen were excluded by bagging.
Thus pollen transfer from viable staminate flowers to receptive pistil
late flowers on the same inflorescence appears to be prevented by pro-
togyny.
Pollen transfer from male inflorescences to receptive female in
florescences on the same palm likewise appears to be rare. On a given
palm, simultaneous opening of the two inflorescence types is unlikely.


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
THE BIOLOGY OF Orbignya martiana (PALMAE),
A TROPICAL DRY FOREST DOMINANT IN BRAZIL
By
Anthony B. Anderson
December, 1983
Chairman: John J. Ewel
Major Department: Botany
Monospecific stands of the babassu palm (Orbignya martiana) occupy
widespread areas of Brazil, attaining their greatest coverage (102,970
km^) in the state of Maranho. Near Lago Verde, Maranhao, I conducted
comparative studies of babassu's life history to determine how the palm
is maintained in mature forests as well as how it dominates disturbed
sites.
Flowering in babassu peaked during the rainy season (January-March)
and coincided with a flush of new leaves (x = 4.1 per yr); fruiting
peaked during the following dry season (October-December). Outcrossing
was predominant and was effected by wind and nitidulid beetles (Mystrops
sp.). Dispersal of the massive ( >100 g) fruits was mediated primarily
by gravity and by rodents. Due to the hard fruit endocarp, seed preda
tion was virtually limited to bruchid larvae (Pachymerus ncleo rum),
which entered through the germination pores; ca. 40% of seeds suffered
xii


24
Study Sites
Lago Verde. The ecological research was centered on an 800-ha farm
in the county of Lago Verde, ca. 25 km NW of Bacabal (Fig. 9). The farm
is flat and dissected by several intermittent streams. Soils are gener
ally fertile and well-drained. The predominant soil type is Eutrophic
Red-Yellow Podzolic. Approximately 75% of the farm has been converted
to pasture, and the rest is covered by various ages of forest fallow.
I selected three 1-ha sites for ecological studies, representative
of undisturbed primary forest, old secondary forest, and annually burned
pasture. The forest sites were located in an 11-ha reserve maintained
for hunting and for harvest of fallen trees as a source of lumber. Each
study site measured 80 x 125 m. For orientation, wooden stakes were
used to define twenty 20 x 25 m blocks; on the forested sites, the
blocks were delimited by 2-m wide paths. Community characteristics of
each site were obtained by measuring abundance, frequency per block, and
diameter at breast height (dbh) of all trees with dbh 10 cm. Unknown
species were collected and identified in the herbarium of the Museu
Paraense Emilio Goeldi in Belem. Community characteristics of the
forested sites are summarized in Appendices A and B.
The primary forest site is probably representative of the original
vegetation that once covered the region. Radiocarbon dating of charcoal
fragments found at ca. 15 cm depth indicate that the site was burned 650
+ 250 yr before present (letter dated 4 August 1983 from Richard R.
Pardi, Radiocarbon Laboratory, Queens College, Flushing, New York, USA).
I believe that the site has suffered no major disturbance since. Local


179
APPENDIX A. SPECIES ABUNDANCE, FREQUENCY, DOMINANCE, AND IMPORTANCE
IN 1 HA OF PRIMARY FOREST, LAGO VERDE, MARANHAO.
$ p
E C I E S
ABUNDANQB
stems/ha%
FREQUENCY
... %
DOMINANCE
nr/ha 1
IMPORTANCE
%
1.
Orbienva martiana Barb. Rodr.*
78
20.21
9.18
7.201
27.43
18.94
2.
Eucenia sd.
35
9.07
7.25
2.984
11.37
9.23
3-
SDondias mombira Urb.*
13
3.37
4.35
4.542
17.30
8.34
4.
Gustavia aucusta L.*
45
11.66
6.28
1.185
4.52
7.47
5.
Swartzla DOlvDhvlla A.DC.
15
3.89
4.35
2.867
10.92
6.38
6.
Hieronima alchornioides (Friere) Allem.*
29
7.51
6.76
0.433
1.65
5.31
7.
Protium heDtaDhvllum (Aubl.) March.
21
5.44
5.31
0.377
1.43
4.06
8.
Cvnometra marcinata Benth.
13
3.37
4.83
0.524
2.00
3.40
9.
Croton of. diasii Pires
12
3.11
3-38
0.639
2.44
2.98
10.
SaDium curuDita Hub.
7
1.81
2.90
0.599
2.28
2.33
11.
Cordia sd.*
9
2.33
2.90
0.403
1.53
2.25
12.
Richeria crandis Vahl.
2
0.52
0.48
1.414
5.39
2.13
13.
Bombax sd.
6
1.55
2.90
0.179
0.68
1.71
14.
Trichilia auadri iuea H.B.K.
8
2.07
2.42
0.101
0.39
1.62 1


CHAPTER 4
ESTABLISHMENT
Establishment from seed consists of seedling recruitment and
subsequent growth. This constitutes a particularly vulnerable phase of
a plant's life cycle (Harper 1977). Following successful dispersal to a
microsite suitable for germination, most seedlings face death due to
such causes as shading, damage by falling litter, herbivory, disease, or
drought. Combinations of these causes may occur; for example, shaded
seedlings are more susceptible to disease than non-shaded seedlings
(Taher and Cooke 1975).
Due to the success of babassu in attaining community dominance, the
palm's performance during the vulnerable establishment phase is of
special interest. In this chapter, I describe experiments designed to
test babassu's performance during establishment. Performance includes
both adaptiveness and competitiveness. The former is delimited by
physiological limits of tolerance; the latter is controlled by the
intensity of competition (Hutchinson 1957). Adaptiveness and competi
tiveness of babassu were respectively monitored on weeded and non-weeded
plots. The experimental treatments consisted of the presence or absence
of subsidies in the form of (1) irrigation, (2) fertilization, (3)
insecticide applications, and (4) climate. The hypothesized outcomes
(Table 8) predicted that adaptiveness would be higher under subsidized
conditions, whereas competitiveness would be higher under non-subsidized
conditions. Precedent exists for this latter prediction. Stresses such
88


162
The key bottleneck in the babassu economy is and always has been
supply (Gonsalves 1955, Valverde 1957, Braga and Dias 1968, STI 1979).
Incorporating the palm's full energetic potential into the market econo
my requires that centralized industries be supplied with a steady flow
of raw materials in the form of fruits. The required continuity of
supply is illusory due to the seasonal nature of the harvest. Because
of relatively low yields per unit areaas well as the low value per
unit productcosts of transporting entire fruits to a centralized
facility quickly become prohibitive. Thus the required volume of supply
is likewise illusory. Furthermore, people who harvest babassu fruits
generally extract the kernels, most of which are ultimately sold to
local vegetable oil industries; the fruit husks are then used to make
charcoal. These people are understandably reluctant to part with what
is usually their exclusive source of fuel. Thus there is considerable
social resistance to incorporating babassu's full energetic potential
into the market economy (Anderson and Anderson 1983). Even traditional
industries based on vegetable oil are chronically plagued by insuffi
cient supply. In addition, the high costs of raw materiallargely due
to labor-intensive, manual extraction of the kernelsreduce babassu's
competitiveness in relation to other vegetable oils. As a result, most
of the industries dependent on babassu are currently in crisis, and the
continued viability of the palm as a source of market products is in
doubt


CHAPTER 7
IMPLICATIONS FOR MANAGEMENT
People and the Palm Forest
The adaptability of babassu to a wide range of ecological condi
tions has been a common theme throughout this study. High genetic
variability is manifest in babassu's propensity to hybridize and its
capacity to form locally adapted ecotypes. Broadly synchronized flower
ing over wide geographic areascombined with breeding via outcrossing
provide ample opportunities for genetic recombination. Babassu's pol
lination by wind as well as beetles may have facilitated its successful
transition from closed forests to more open sites. Although massive,
the fruits of the palm have been dispersed over nearly half a continent
by flooding rivers and runoff, and on a more local scale by humans and
rodents. Large, now-extinct mammals such as giant ground sloths and
gomphotheres may also have had a role as agents of dispersal, as well as
agents of selection in the evolution of babassu's extraordinarily thick
endocarp. Other than axe-wielding humans, there are no mammals current
ly on the scene capable of penetrating this formidable structure. Ba
bassu's other important seed predator, a bruchid beetle, skirts the
endocarp by entering as tiny larvae through the germination pores.
However, because there are usually several seeds per fruitsand one is
sufficient to satisfy each larvasome almost always escape.
In closed forests, germination in babassu typically takes place 3
mo after fruit fall. Due to high shade tolerance, seedlings and stem-
157


6
Figure 7. Flowchart of products derived from babassu fruits by industries
in Bacabal, Maranhao, during 1980. Diamonds represent "either-
or" options. Numbers are percentages of fruit weight. Adapted
from Mendes and Carioca (1981) and Empresa Industrial de Baca
bal (personal communication).


PALMS (no./ha)
Figure 42. Age distribution of babassu on the primary forest site at Lago Verde.
149


137
the general observation that annuals allocate proportionately more pro
ductivity to reproductive structures than do perennials (Harper and
White 1974). The generalization obviously does not extend to the humid
tropics, where evergreen perennials are not subject to die back each
year.
It is no surprise that the semelparous (i.e., monocarpic) Corypha
elata allocates a much lower proportion of lifetime productivity to
reproductive structures than iteroparous (i.e., polycarpic) palms such
as babassu, in which reproduction occurs annually over ca. 120 yr. The
sheer size of the single reproductive event in Ch elata is nonetheless
impressive. In absolute terms, it is equivalent to ca. 18 yr of repro
duction in an average mature palm of babassu.
Compared to other iteroparous palms such as Jessenia bataua and
Astrocaryum mexicanum, babassu's allocation to reproductive structures
is high. This is probably due to the exceptionally long period of
senescence in babassu, during which vegetative growth decreases sharply
while allocation to reproductive structures undergoes only a moderate
decline (Figs. 37-38). Babassu's pattern of increasing senescence with
age conforms to that predicted for trees by Harper and White (1974)
However, Sarukhan (1980) found no evidence for senescence in a number of
tropical trees, including Astrocaryum mexicanum (Pinero et al. 1982).
Although it has been argued that senescence in plants is inevitable
(Hamilton 1966), many species probably fail to manifest senescence under
natural conditions. I suspect that the presence or absence of senes
cence in species of palms may be primarily a function of their maximum


15
of babassu dominate the landscape. During the past 25 yr, rapid influx
of settlers from the arid Brazilian northeastcombined with an ambi
tious road-building program throughout the state have pushed human
settlement deep into Tropical Moist Forest, and extensive areas are
presently occupied by young stands of babassu. The principal land uses
throughout the moist or dry forest zone are shifting cultivation and
cattle ranching. In contrast to other ecological zones in Maranhao,
ranching here is carried out primarily on improved pastures.
The palm forest. In the major ecological zones of Maranhao des
cribed above, babassu stands are generally found on the sites most
intensively utilized by people. On such sites, crops are established
and pastures maintained by periodic cutting and burning. These prac
tices account for the origin of the babassu stands in Maranhao.
Prior to the arrival of Europeans, Maranhao had long been occupied
by numerous indigenous groups (described in Steward 1963). Among the
dominant sedentary tribes (e.g., Tupinamba, Timbira), agriculture was
generally confined to the moist bottomlands, and villages were estab
lished nearby. During the dry months, bottomland stands of babassu and
buriti (Mauritia flexuosa) palms provided an important dietary supple
ment. As with the more nomadic tribes, territorial rights were identi
fied with access to these stands, which served as a source of fiber and
fuel, as well as a hedge against famine. High densities of babassu were
probably limited to these bottomland sites. The original nature of the
far more extensive upland forests can be discerned in the remnants that


85
Aotus trikingatus. according to local informants) successfully prey on
babassu by pulling off immature fruits and consuming the liquid endo
sperm within (see Izawa and Mizuno 1977). Thus seed predators either
are completely deterred or successfully bypass babassu's thick endocarp.
I suspect that there is no predator currently exerting sufficient selec
tive pressure to account for the evolution of this extraordinary struc
ture. Coevolution with large, Pleistocene mammals (Janzen and Martin
1982) thus provides the most satisfactory explanation for the origin of
the thick, massive fruits characteristic of babassu.
Germination
The results of my experiments indicate that germination in babassu
is least successful on disturbed sites such as pastures and shifting
cultivation plots, where shading is minimal, fires either frequent or
intensive, and mesocarp-removing dispersal agents rare. While germina
tion success of babassu is thus reduced, its peculiar mode of germina
tion confers upon the palm an extraordinary capacity to recover from the
stresses inherent to these sites.
Germination in babassu occurs when one or more embryonic axes
emerge from the germination pores at the base of the fruit. Each axis
enters the soil and penetrates to a variable depth, where it subse
quently differentiates into roots and leaves (Fig. 31). This so-called
cryptcgeal mode of germination is found throughout the Attalea alliance
and in distantly related palm genera such as Acanthococos and Sabal
(Corner 1966), as well as in many other plant families (Jackson 1974).
In plants that exhibit cryptogeal germination, the shoot arises from


143
A stage projection matrix (Lefkovitch 1965) was not constructed for
this population. Such a matrix requires transfer coefficients that are
both constant (Silvertown 1982) and positive between all stages of the
population. During a 1-yr period, six of the 12 life stages received no
recruitment (Table 25). A longer period of observations would thus be
required to construct a projection matrix for this population.
Results
Primary Forest
In the primary forest population of babassu, annual mortality rates
as calculated in the life table (Table 24) closely paralleled those
obtained during the 1-yr period of field observations (Table 25). This
lends support to the initial assumption of stability in the population.
Babassus survivorship as a function of age in the primary forest (Table
24) showed a reverse J-shaped curve characteristic of many mature forest
tree populations and indicative of a stable age population (Silvertown
1982). Palm in older life stages almost consistently declined in fre
quency, yet virtually all stages were represented (Fig. 41) one would
expect such a pattern in a stable population. Age distribution in the
primary forest likewise followed a reverse J-shaped curve (Fig. 42).
Seed dynamics showed a similar pattern in both the primary and
secondary forests (Tables 25-26). Approximately half the annual seed
production succumbed to mortality, a tenth germinated, and the rest went
to soil storage. On both sites, seedling recruitment was more than
sufficient to maintain seedling populations.


KERNEL PRODUCTION (IOOOI)
Figure 6. Production of babassu kernels in Maranho and Brazil,
1920-79. Sources: IBGE (1981a and previous volumes to
1916).


104
Table 11. Palms per m2 (a), leaf area per palm (b), and
leaf area index (c) of babassu at end of
insecticide experiment. Asterisks indicate
(a)
significant differences (** for p
between main effect means.
Palms Der m2
.01)
No
Insecticides
Insecticides
Mean
Not Weeded
30.8
27.6
29.2
**
Weeded
5.8
5.4
5.6
Mean
18.3
16.5
(b)
Leaf Area
Der Palm (cm2)
No Insecticides
Insecticides Mean
Not Weeded
350.0
226.4 288.2

Weeded
552.8
402.8 477.8
Mean
451.4 **
314.6
(c)
Leaf Area Index
No Insecticides
Insecticides
Mean
Not Weeded
=r
on

.29
.31
*
Weeded
.17
.11
.14
Mean
.25
.20


83
(Bradford and Smith 1977); the corresponding figures for babassu62%,
53%, and 42% (Table 5)are remarkably similar. The trend suggests that
in both species, selection by seed predators favors many-seeded fruits.
However, such selection probably operates within strict constraints.
Bradford and Smith (1977) found that in rostrata, individual seeds in
one-seeded fruits contained 46% and 123% higher energy content than
those in two- and three-seeded fruits, respectively. Seedlings from
one-seeded fruits should therefore have a competitive advantage, In
addition, seedlings emerging from the same many-seeded fruit compete
intensively with each other, further increasing the competitive advan
tage of seedlings from one-seeded fruits. The number of seeds per fruit
thus represents a compromise between opposing selective forces.
The endocarp of babassu is exceptionally thick (x = 49.8 mm, s.e. =
1.1) and resistant to crushing (Fig. 30) compared to other species of
palms. It would thus appear to provide effective deterrence against
seed predators. Yet the figures cited above indicate that levels of
predation in one- to three-seeded fruits are almost identical in babassu
and 5k rostrata, despite the latter's comparatively thin endocarp (4-5
mm, according to Bradford and Smith 1977). As described above, seed-
predaceous beetles penetrate the fruits of babassu via the germination
pores, where the endocarp is least resistant; entry is effected in 2-3 d
(Bondar 1936). The thick endocarp of babassu thus appears to provide
little deterrence against the palm's most effective seed predator.
Conversely, the endocarp completely deters other potential predators
such as rodents and squirrels; only monkeys (e.g., Cebus apella and


LEAF PRODUCTION PER PALM
48
1981 1 1982
J FMAMJJ A SONDJ FM
+ 1.5
+ 1.0
+ 0.5
0
-0.5
-1.0
-1.5
PRIMARY FOREST
j (n = ;
: nd n
21]

llllliili luu
1
1
1
C
C
TJi=r
Figure 17. Mean monthly gain and loss of leaves per palm on three
study sites at Lago Verde. N = number of palms observed
per site. ND = no data.


Table 7. Preliminary checklist of insects that feed on babassu. Asterisks
indicate economically important pests. Source: Costa Lima (1967-68).
LCULLLLm
Stored
TAXON Stems Leaves Inflorescences Fruits Seeds
Homoptera
Asterolecaniidae
1. Asterolecanium sp. +
Diaspididae
2. Aspidiotus destructor +
Lepidoptera
Phycitidae
3. Cadra cautella +
4. Plodia interpunctella +
Coleptera
Bruchidae
5. Carvobruchus lipismatus*
6. C_t pergandei
7. Pachvmerus nucleorum*
8. £j_ olearius*
Bostrichidae
9. Dinoderus minutus
Cerambycidae
10. Macrodontia cervicornis +
Chrysomelidae
11. Coraliomela brunnea
Cleridae
12. Necrobia rufipes*
Dermestidae
13 Dermestes ater*
+ +
+
+ +
+
+
+
+