INFORMATION TO USERS

INFORMATION TO USERS
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis arvj
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bteedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these will be noted.
Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, t)eginning at the upper left-hand comer and continuir)g
from left to right in equal sections with small overiaps.
Photographs included in the original manuscript have t)een reproduced
xerographically in this copy.
Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations appearing
in this copy for an additional charge. Contact UMI directly to order.
Bell & Howell lnformatk>n arxj Learning
300 North Zeeb Road, Ann Arbor, Mi 48106-1346 USA
800-521-0600
<g>
UMI
MORE USERS AND MORE USES;
CHOOSING BETWEEN LAND AND FOREST EN MALAWI'S PROTECTED AREAS
by
Barron Joseph Orr
Copyright © Barron Joseph Orr 2000
A Dissertation Submitted to the Faculty of the
GRADUATE INTERDISCIPLINARY PROGRAM IN
ARID LANDS RESOURCE SCIENCES
In Partial Fulfillment of the Requirements
For the Degree of
DOCTOR OF PHILOSOPHY
In the Graduate College
THE UNIVERSITY OF ARIZONA
2000
UMI Number: 9992123
Copyright 2000 by
Orr, Barron Joseph
All rights reserved.
®
UMI
UMI Microform 9992123
Copyright 2001 by Bell & Howell Information and Leaming Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
Bell & Howell Information and Leaming Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor. Ml 48106-1346
2
THE UNIVERSITY OF ARIZONA t)
GRADUATE COLLEGE
As members of the Final Examination Committee, we certify that we have
read the dissertation prepared by
entitled
More Users and More Uses:
0"
Choosing between Land and Forest
in Malawi's Protected Areas
and recommend that it be accepted as fulfilling the dissertation
requirement for the Degree of
Doctor of Philosophy
Dace
Charles F. Hutchinson
Si S c p i Dace
z- ZDace
uy R, McPherson
-T
q-
n..
Date
Thomas R. McGulre
Date
Final approval and acceptance of this dissertation is contingent upon
the candidate's submission of the final copy of the dissertation to the
Graduate College.
I hereby certify chat I have read this dissercation prepared under my
direction and recoanend that it be accepted as fulfilling the dissertation
requirement.
Dissertation Director
Charles F. Hutchinson
Date
3
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of requirements for an
advanced degree at The University of Arizona and is deposited in the University Library
to be made available to borrowers under rules of the Library.
Brief quotations from this dissertation are allowable without special permission,
provided that accurate acknowledgment of source is made. Requests for permission for
extended quotation from or reproduction of this manuscript in whole or in part may be
granted by the copyright holder.
SIGNED:
4
ACKNOWLEDGMENTS
There were a number of guiding and driving forces that led me to and through graduate
school. My mother, Jo, taught me how to communicate when words fail, and gave me the
gifts of instinct, vision and desire. My father. Bill, gave me both the will to start and
finish, and the analytical tools to get the job done. My brothers, Goehrig and Wade, my
sister Elise, and my step-mother Donnie helped shape me and remind me of where I
came. And my partner in crime, my best pal and best guide is my wife, Patricia, who not
only endured the whole process, but even took the wheel when and where I could not.
I have had other mentors along the way. My cousin Terry set both the example and the
standard. My grandmother Liney Rubins and great uncle and aunt, Capt. Barron Gray and
Teeny Lowrey taught me the importance of family and history. Emil and Virginia
Hansen, Ron and Marilyn Hartman, and the Rev. W.H. and Jody Fogg showed me how to
find and encourage the strengths of others. Lou Moir advised me during and long after
my undergraduate work. Abdulah and Zayda Bousta taught me dignity and the value of
the written word. Peter Vechey demonstrated how success comes from empowering one's
colleagues rather than competing with them. Joyce Mtawale Andersen opened my eyes to
community and the power of women. Kasweswe Mwafongo, Barbara Eiswerth, Susan
Skirvin, Alejandro Leon, and Dr. Kevin Carmichael taught me how to make a difference.
I had a number of informal mentors who helped me stick to the task; Dr. Herb Carter,
Maura Mack, Cynthia Lindquist, Sadoon Younis, Kebe Brahim, Karim Batchily, Wim
van Leeuwen, Hsain Ilahiane, Len Milich, Eric Weiss, John Maingi. Committee members
before and during my dissertation were invaluable. Bill Bull and Guy McPherson taught
me how to pose a question and how to go about answering it. Stuart Marsh honed my
analytic and management skills, and taught me humility. Tim Finan and Tom McGuire
saw the anthropologist in me and brought it out. Tony Seymour taught me how to
negotiate and facilitate where the stakes were high and the rules less than clear. My
committee chair and dissertation advisor. Chuck Hutchinson, showed me how to bring all
the disparate, often complex pieces together to tell the seemingly simple story. And Sonia
Economou, Olivia Mendoza, and Chiyo Yamashita-Gill kept me on task.
I wish to express my deep gratitude to the inhabitants of villages adjacent to protected
areas in Malawi who gave of their time and knowledge, and to the Malawi Government
and USAED, who sponsored the Malawi Public Lands Utilization Study. A very special
note of thanks is due to the researchers on that study who persevered under extremely
difficult conditions and to whom credit for a successftil exercise belongs; Luke Malembo,
Elizabeth Malimero, Gertrude Songo, Gerald Matola, Joseph Magalasi, Ireen Mithi,
Edward Sambo, Lista Sichali, Neverson Masanja, Stephanie Jayne, Humphrey Chapama,
Tawana Chauluka, Harry Banda, Dennis Kambuzi, Etness Kayoyo, Kent Burger, Patricia
Rodriguez, John Rogan, Diego Valdez-Zamudio, Sam Drake, Doug Rautenkranz, Jessica
Walker, David Halmo, Michele Companion, and Mamadou Baro. Cindy Wallace gave of
her time and skill in creating the map layouts in both PLUS and this dissertation.
DEDICATION
I dedicate this work to my wife and best friend,
Padraigin Ni h-AIlumhziran Orr
and to the memory of those who passed before,
well before their time...
Jo Demarest Orr
Lin Choo-shiu
Mark O'Brien
Peter Vechey
Harry Banda
Tawana Chauluka
Soloman Chipompha
Harriet Wall in
Simon Phiri
Wongani Phiri
John the Driver
John Mphande
Flyson Bonda
Welbon Kasweswe Mwafongo
6
TABLE OF CONTENTS
LIST OF FIGURES
8
LIST OF TABLES
10
ABSTRACT
12
CHAPTER 1: INTRODUCTION
14
EXPLANATION OF THE PROBLEM AND ITS CONTEXT
14
STUDY SITE DESCRIPTION
17
MALAWI
17
MULANJE FOREST RESERVE
19
LIWONDE NATIONAL PARK
20
DZALANYAMA FOREST RESERVE AND RANCH
22
VWAZA MARSH WILDLIFE RESERVE
23
EXPLANATION OF DISSERTATION FORMAT
27
RELATIONSHIP OF THE APPENDED MANUSCRIPTS
27
CONTRIBUTION OF THE AUTHOR
29
CHAPTER 2: PRESENT STUDY
31
SUMMARY
31
METHODS
31
RESULTS
36
CONCLUSIONS
38
REFERENCES
43
A. APPENDIX A. INTEGRATING METHODOLOGIES FOR LIVELIHOOD
SECURITY AND ETHNOECOLOGICAL ASSESSMENTS IN RURAL
MALAWI
49
A.1 ABSTRACT
49
A.2 INTRODUCTION
50
A.3 STUDY AREAS
52
A.3.1 MULANJE FOREST RESERVE
52
A.3.2 LIWONDE NATIONAL PARK
53
A.3.3 DZALANYAMA FOREST RESERVE AND RANCH
54
A.3.4 VWAZA MARSH WILDLIFE RESERVE
55
A.4 METHODS
55
A.4.1 APPROACH
55
A.4.2 KEY VARIABLES: INCOME AND NATURAL RESOURCE
UTILIZATION
63
A.4.3 VALUATION IN THE MALAWIAN CONTEXT
64
A.4.4 SPECIES IDENTIFICATION
66
A.4.5 QUANTITY AND WEIGHT ISSUES
68
A.4.6 PRICE
76
A.5 DATA SUMMARY
79
A.5.1 DOMESTIC PRODUCTION
79
A.5.2 PROTECTED AREA NATURAL RESOURCE UTILIZATION
80
7
A.6 RESULTS AND DISCUSSION
81
A.6.1 RELIANCE
82
A.6.2 RELIANCE AND TOTAL INCOME
83
A.6.3 DISTRIBUTION OF INCOME AND POVERTY
85
A.7 CONCLUSIONS
93
A.8 LITERATURE CITED
94
A.9 FIGURE LEGENDS
105
A.IO FIGURES
106
B. APPENDIX B. QUANTIFYING PROTECTED AREA RESOURCE
DEMAND AT THE SPECIES LEVEL: ANOTHER CONSIDERATION FOR
INTEGRATED CONSERVATION AND DEVELOPMENT PROJECTS
112
B.l ABSTRACT
112
B.2 INTRODUCTION
113
B.3 STUDY AREAS
116
B.3.1 MULANJE FOREST RESERVE
117
B.3.2 LIWONDE NATIONAL PARK
118
B.3 3 DZALANYAMA FOREST RESERVE AND RANCH
119
B.3.4 VWAZA MARSH WILDLIFE RESERVE
120
B.4 METHODS
120
B.4.1 APPROACH
120
B.4.2 SPECIES IDENTIFICATION
124
B.5 RESULTS: SPECIES USE
126
B.5.1 DOMESTIC PRODUCTION
126
B 5.2 PROTECTED AREA NATURAL RESOURCE UTILIZATION
129
B.6 RESULTS: SUSTAINABILITY ANALYSIS
144
B.7 CONCLUSIONS
145
B.8 LITERATURE CITED
148
B.9 FIGURES LEGENDS
155
B.10 FIGURES
162
C. APPENDIX C: CHOOSING BETWEEN FOREST AND AGRICULTURE:
THE EVOLVING ECONOMIC ROLE OF PROTECTED AREAS
180
C.1 ABSTRACT
181
C.2 INTRODUCTION
182
C.3 STUDY AREA
185
C.4 METHODS
188
C.5 RESULTS
193
C.6 DISCUSSION
201
C.7 REFERENCES AND NOTES
203
C.8 FIGURES
210
8
LIST OF FIGURES
FIGURE 1.1. MALAWI IN AFRICA (UNITED NATIONS MAP 3858, NOVEMBER
1994)
25
FIGURE 1.2. LOCATION OF STUDY AREAS IN MALAWI
26
FIGURE A.I. THE RELATIVE IMPORTANCE OF ALL INCOME ELEMENTS. .106
FIGURE A.2. PROTECTED AREA AND NON-PROTECTED AREA COMPONENTS
OF PER CAPITA INCOME FOR ALL SAMPLED HOUSEHOLDS
107
FIGURE A.3. SIMPLE LINEAR REGRESSION FOR PER CAPITA INCOME AS A
PREDICTOR FOR RELIANCE ON PROTECTED AREA RESOURCES
108
FIGURE A.4. LORENZ CURVES FOR INCOME INCLUSIVE AND EXCLUSIVE OF
PROTECTED AREA PROCEEDS
109
FIGURE A.5 DIFFERENCES IN RELATIVE INCOME BETWEEN INCOME
GROUPS
110
FIGURE A.6. CHANGE IN RELATIVE INCOME SHARE RESULTING FROM THE
ADDITION OF PROTECTED AREA-BASED PROCEEDS TO HOUSEHOLD
INCOME
111
FIGURE B. 1 PER CAPITA PRODUCTION FOR MAJOR CROPS
162
FIGURE B.2. PER CAPITA PRODUCTION FOR IMPORTANT MINOR CROPS. .163
FIGURE B.3. PER CAPITA LIVESTOCK PRODUCTION
164
FIGURE B.4. PER CAPITA WOOD PRODUCTION ON AGRICULTURAL LAND. 165
FIGURE B.5 OVERALL PROTECTED AREA UTILIZATION BY CATEGORY OF
USE (PER CAPITA QUANTES AND RELATIVE PERCENTAGES)
166
FIGURE B.6. OVERALL PROTECTED AREA UTILIZATION BY CATEGORY OF
USE (PER CAPITA VALUE AND RELATIVE PERCENTAGES)
167
FIGURE B.7 PER CAPITA UTILIZATION OF MAJOR PLANT AND MUSHROOM
SPECIES FOR FOOD
168
FIGURE B.8 PER CAPITA UTILIZATION OF A SELECTION OF THE MOST
IMPORTANT MINOR PLANT FOOD SPECIES
169
FIGURE B.9 PER CAPITA UTILIZATION OF MAJOR FISH SPECIES
170
FIGURE B. 10 PER CAPITA FOOD UTILIZATION OF WILD ANIMALS,
INCLUDING INSECTS AND HONEY
171
FIGURE B. 11 PER CAPITA PROTECTED AREA FUELWOOD UTILIZATION 172
FIGURE B 12. PER CAPITA PROTECTED AREA CONSTRUCTION UTILIZATION.
174
FIGUHE B. 13 PER CAPITA PROTECTED AREA FIBER UTILIZATION
175
FIGURE B. 14. PER CAPITA PROTECTED AREA TOOL AND HANDICRAFT
SPECIES UTILIZATION
176
FIGURE B. 15. PER CAPITA PROTECTED AREA THATCH UTILIZATION
178
FIGURE B 16. PROJECTED DECLINE IN THE SUSTAINABLE SUPPLY OF WOOD
(MAI ROUNDWOOD EQUIVALENT) AS POPULATION GROWS
179
9
FIGURE C. I. AGRICULTURE SUITABILITY MODEL (40,27)
211
FIGURE C.2. POPULATION PRESSURE SPATL\L MODEL (ADAPTED: 27)
212
FIGURE C.3. VEGETATION CLASSIFICATION AND CHANGE DETECTION
MODEL (27)
213
FIGURE C.4. REGRESSION FOR PER CAPITA INCOME AS A PREDICTOR FOR
RELL\NCE ON PROTECTED AREA RESOURCES
214
FIGURE C.5. CHANGE IN RELATIVE INCOME SHARE RESULTING FROM THE
ADDITION OF PROTECTED AREA-BASED PROCEEDS TO HOUSEHOLD
INCOME
215
FIGURE C.6. WOOD DEMAND VERSUS SUSTAINABLE SUPPLY IN MALAWI.
216
FIGURE C.7. LAND DEMAND VERSUS AVAILABILITY IN MALAWI
217
FIGURE C 8. PROJECTED DECLINE IN THE SUSTAINABLE SUPPLY OF WOOD
(MAI ROUNDWOOD EQUIVALENT) AS POPULATION GROWS
218
FIGURE C 9. HYPOTHETICAL DECLINE IN AGRICULTURALLY SUITABLE
LAND INSIDE THE PROTECTED AREAS AS POPULATION GROWS
219
FIGURE CIO. LANDSAT THEMATIC MAPPER FALSE COLOR IMAGES OF
LIWONDE NATIONAL PARK IN 1984 AND 1994 (27)
220
FIGURE C 11. EASTERN-MOST PORTION OF LIWONDE NATIONAL PARK IN
1984 AND 1994
221
FIGURE C. 12. AGRICULTURAL SUITABILITY OF LAND IN LIWONDE
NATIONAL PARK (27)
222
FIGURE C. 13. POPULATION PRESSURE AROUND LIWONDE NATIONAL PARK
(27)
223
FIGURE C. 14. LAND COVER CLASSIFICATION AND CHANGE DETECTION IN
LIWONDE FROM 1984 TO 1994
224
10
LIST OF TABLES
TABLE A. 1. SUMMARY OF DATA COLLECTION METHODS
57
TABLE A.2. NUMBER OF SPECIES AND HOUSEHOLD OBSERVATIONS OF USE
67
TABLE A.3 VOLLHVIE CONVERSION FACTORS AMONG LOCAL UNITS OF
MEASURE
69
TABLE A 4. WEIGHT CONVERSIONS AND PRICES FOR AGRICULTURAL
CROPS
70
TABLE A.5 WEIGHT CONVERSIONS AND PRICES FOR PROTECTED AREA
FOODS AND MEDICINES, GROUPED INTO CLASSES BY AGRICULTURE
PRODUCT SIZE-EQUrVALENTS
72
TABLE A.6. LIVE WEIGHTS AND PRICES OF ANIMAL SPECIES
73
TABLE A. 7 CONVERSION UNITS FOR WOOD, THATCH AND FIBER
75
TABLE A.8. SINGLE TREE VOLUME EQUATIONS
76
TABLE A.9 PRICES FOR BIOMASS USED TO CREATE MAJOR WOOD AND
HANDICRAFT PRODUCTS
78
TABLE A. 10. POPULATION, LAND, AND AGRICULTURAL PRODUCTION
79
TABLE A ll BARRIERS TO PROTECTED AREA ACCESS AND OVERALL
UTILIZATION
80
TABLE A. 12. PERCENTAGE OF TOTAL FOOD, WOOD, AND MEDICINES THAT
WAS DERIVED FROM PROTECTED AREA NATURAL RESOURCS
83
TABLE A. 13. SUMMARY OF REGRESSION RESULTS
84
TABLE A. 14 POVERTY MEASURES AND ONE POLARIZATION MEASUHE FOR
THE GENERAL QUADRATIC PARAMETERIZATION OF THE LORENZ
CURVE.*
86
TABLE A. 15. DESCRIPTIVE STATISTICS AND INEQUALITY INDICES FOR
INCOME WITHOUT AND WITH PROTECTED AREA PROCEEDS
88
TABLE A. 16. POVERTY INDICES.*
91
TABLE A. 17. POVERTY ANALYSIS
92
TABLE B. 1 SUMMARY OF DATA COLLECTION METHODS
121
TABLE B.2. NUMBER OF SPECIES AND HOUSEHOLD OBSERVATIONS OF USE.
125
TABLE B.3 POPUXATION, LAND, AND AGRICULTURAL PRODUCTION
126
TABLE B 4. LIVESTOCK AND WOOD PRODUCTION ON AGRICULTURAL
LAND
128
TABLE B.5. BARRIERS TO PROTECTED AREA ACCESS AND OVERALL
UTILIZATION
130
TABLE B.6. PER CAPITA PROTECTED AREA FOOD UTILIZATION
132
TABLE B.7. PER CAPITA PROTECTED AREA FUELWOOD AND
CONSTRUCTION UTILIZATION
136
11
TABLE B.8 PER CAPITA PROTECTED AREA FIBER, TOOL AND HANDICRAFT
UTILIZATION
139
TABLE B 9 PER CAPITA PROTECTED AREA THATCH AND MEDICINAL
PLANT UTILIZATION
141
TABLE B. 10. THE FIFTEEN MOST FREQUENTLY REPORTED SPECIES USED
FOR MEDICINAL PURPOSES IN EACH PROTECTED AREA
143
TABLE B. 11 SUSTAINABILITY OF WOODY PROTECTED AREA RESOURCES.
145
TABLE C 1 SUMMARY OF DATA COLLECTION METHODS
191
TABLE C.2. RELIANCE (PERCENTAGE OF PROTECTED AREA INCOME
RELATIVE TO TOTAL INCOME), LAND, AND POPULATION
194
TABLE C 3 MEAN INCOME, INCOME EQUALITY, AND POVERTY WITHOUT
AND WITH PROTECTED AREA-BASED INCOME
195
TABLE C.4 KEY INPUTS FOR THE SUSTAINABILITY ANALYSIS
198
TABLE C.5 SPATL\L RELATIONSHIP BETWEEN AREAS THAT EXPERIENCED
A DECLINE IN BIOMASS (1984 - 1994) TO BOTH POPULATION OUTSIDE
AND AGRICULTURALLY SUITABLE LAND INSIDE EACH PROTECTED
AREA
200
TABLE C.6 THE CONSUMPTIVE USE VALUE OF AGRICULTURE ON
CUSTOMARY LAND AND NATURAL RESOURCE UTILIZATION ON
PROTECTED AREA LAND.*
201
12
ABSTRACT
Local inhabitants risk the loss of ecological resources when land is cleared for
cultivation as population densities and the demand for land resources increase. This
dilemma is investigated through an interdisciplinary socioeconomic and ethnoecological
assessment of427 households in communities adjacent to four protected areas in Malawi.
This study introduces a multidimensional approach that captures baseline socioeconomic
information and resource utilization in a quantitative, integrated manner. Household
income was derived from a "sum of the parts" aggregation of income elements including
species-level agricultural production and resource utilization data. Regression analysis (R^
= 0.84) demonstrated that poorer households are more reliant on protected area-based
income than are wealthy households. Lorenz curve analysis demonstrated that income
distribution equality improves when proceeds from protected areas are included in
household income. Poverty threshold analysis indicates that exploitation of protected area
resources is a livelihood strategy that halved the number of households that otherwise
would have remained beneath a basic needs poverty threshold. Ecological resources are
shown to meet demand for more people and for a longer time frame than converting the
same lands to agriculture. However, conversion is more likely because per hectare values
are 2 to 3.5 times greater for agriculture than for consumptive ecological resource use.
Spatial analysis suggests points of negative land cover change (1984-94) were not
associated with the proximity of population but with the agricultural suitability of the land.
The results suggest the kinds of decisions people will make under extreme stress, when
13
consideration of potential impacts is overwhelmed by the need to survive. This study
demonstrates that protected area resources play a pivotal role in poverty alleviation, and
by extension, efforts to make sustainable use and sustainable development compatible.
14
CHAPTER 1: INTRODUCTION
EXPLANATION OF THE PROBLEM AND ITS CONTEXT
Developing nations with predominantly agrarian economies face a difficult natural
resource dilemma. Food and Agriculture Organization (FAQ) estimates suggest these
countries lost 14.9 million hectares of forest area annually between 1980 and 1995 (FAQ
1997). Particularly in Africa, the driving force was generally the expansion of subsistence
agriculture. As population densities and the demand for land resources increase, local
inhabitants risk the loss of ecological resources when land is cleared for cultivation. This
dilemma has direct bearing on efforts to alleviate poverty, encourage sustainable
development, and promote community-based natural resource management, particularly
where currently protected land is being considered for conversion to agriculture.
As the amount of total forested land declines, natural resources in protected areas
become pivotal for both conservation and sustainable development. Poor rural households
rely on natural resources extracted from protected parks and reserves, despite the
potential for detrimental ecological impact. This deliberate livelihood strategy is intended
to balance income against expenditure in normal times (Arnold and Falconer 1989) while
providing a final safety net in times of crisis (Garine and Koppert 1988). Despite the
importance of wild resources in household economies, baseline assessments of livelihood
security often fail to capture their role in anything other than broad or aggregate terms.
Conversely, ethnoecological research that might meet this need is often conducted at a
spatial and temporal scale unsuitable for sustainable development planning.
15
Evolving societal concerns have increased pressure to meld the goals of
conservation and sustainable development. The desire to improve upon conventional
"fences and fines" approach to conservation (Barrett and Arcese 1995) has coincided with
a more general trend towards including local communities in research, planning, and
management of development initiatives (Brandon and Wells 1992). This has fostered
interdisciplinary research that integrates socio-economic and ecological perspectives
where concerns for protection of natural areas and economic development of local
communities converge (Munasinghe 1992).
Although the potential of community-based management of natural resources has
come under increased scrutiny (Wainwright and Wehrmeyer 1998), the long-term financial
benefits of conservation to local people have the potential to outweigh those of agriculture
or logging (e.g. Kremen et al. 2000). In spite of potential benefits, local populations in
developing countries tend to receive a small share of the use and non-use benefits assessed
by outsiders (Godoy ef al. 2000).
To rectify this inequality, community-based natural resource management and
integrated conservation and development projects (IDCPs) have become widespread.
Their goal is to exploit the role that human consumption of protected natural resources
can play in both conservation and development by directing the returns derived from
conservation back into the community (Brandon and Wells 1992). This would suggest the
potential for harmony between ecological and livelihood sustainability, and indeed IDCPs
are viewed as having greater appeal than the exclusionary, "fences and fines" conservation
strategies that preceded them (Barrett and Arcese 1995). Nevertheless, they require far
16
more ecological, economic, social, and institutional monitoring, an investment difficult for
developing nations and their inhabitants to make. Moreover, the ecological and
socioeconomic aspects of sustainability are not always compatible.
This research presents a methodology to capture the role of protected area
resources in rural household economies that integrates ethnoecological resource valuation
techniques (Godoy, Lubowski, and Markandya 1993) with a combination of rapid and
participatory, livelihood security assessment tools (Chambers 1990; 1994, Campbell,
Luckert and Scoones 1997; Finan 1996; Woodson 1997). The methodology was tested
over 12 months between 1996 and 1997 as part of the Malawi Public Lands Utilization
Study. The data gathered were then used answer three fundamental questions;
(1) What are the per capita quantities and consumptive use-values of each species used
for various uses, whether agriculturally produced or collected from the protected area?
(2) Are poorer households more reliant on protected area-based income?
(3) Do protected area proceeds influence the overall equality in the distribution of
household income?
(4) Does the current supply of protected area resources meet current demand?
(5) Based on current patterns of use and population growth, when will demand outstrip
sustainable supply?
(6) If protected area land suitable for agriculture was opened to cultivation, when would
it be consumed, based on current patterns of land use and population growth?
(7) What is the per hectare value of protected area land used for a) natural resource
product utilization, and b) agriculture?
17
(8) Does spatial analysis of land cover change suggest negative impacts are associated
with proximity to higher population densities (as would be the case if flielwood
extraction dominated the process) or with the distribution of agriculturally suitable
land?
STUDY SITE DESCRIPTION
The study sites used for this research are all in Malawi, in southeastern Africa
(Figure 1.1 and 1.2). Each Appendix is a manuscript that includes a description of Malawi
and the four protected study areas in a manner appropriate to the manuscript's theme.
Sites are more thoroughly described in this introductory chapter to provide a more
comprehensive geographic context for the entire study.
MALAWI
In 1997, 19% of Malawi's 94,000 km^ of land was under protection in the form of
four wildlife reserves, five national parks, and seventy-seven forest reserves (Orr et al.
1998). This percentage is not exceptional in Afiica or elsewhere. However, in the same
year, 86% of the country's 9.65 million people lived in rural areas (Malawi Government
1998). Malawi's average population density of 103 persons/km^ of land was three times
that of its neighbors (United Nations Population Division 2000), resulting in twice the per
capita pressure on both protected and non-protected wooded areas. With a mean land
holding size at or below 1.0 ha for a family of five in rural Malawi (BDPA/AHT 1998;
House and Zimalirana 1992) and the importance of traditional agriculture in the dominant
livelihood systems, the demand for agricultural land is substantial. Furthermore, 98% of
18
mral and 94% of urban energy demand is satisfied through flielwood and charcoal
(Arpaillange 1996), a level of dependence exceeded internationally only by Nepal (Pearce
and Turner 1990).
As a result of this population pressure, the forested area in Malawi was reduced by
half between 1946 and 1996 (FAO 1981; FAO 1999; Millington e/<ar/. 1989; Openshaw
1996, Willan 1947). This has had the effect of increasing the importance of remaining
protected area resources as both common resource base and potential agricultural land
(Mwafongo 1994; Mwafongo and Kapila 1999).
The population of Malawi is predominantly rural, with smallholder farmers
(averaging 1.0 ha of land under cultivation) constituting 90% of the nation's poor (World
Bank 1998).The agroecosystems in Malawi are dominated by a sub-humid tropical, unimodal rainfall system (ranging fi'om 700 to 1400 mm annually), with loamy sand soils
characterized as having "low" to "suflBcient" nutrient levels (Snapp 1998). High
population growth rates have led to small land holdings (House and G. Zimalirana 1992),
and continuous cultivation (limited or no fallow), whereas in 1938, plots were farmed only
3 to 5 years before extended rest (Berry and Petty 1992).
In response to the acknowledged population pressure to convert protected areas to
agricultural lands, the Government of Malawi commissioned a Public Lands Utilization
Study (PLUS) in 1996 to study both the environmental risks of conversion and the
importance of the reserves and their resources to adjacent communities (Orr et al. 1998).
The study was guided by a national Steering Committee on Land, which was made up of
60 government and non-govemment stakeholder agencies. The Committee selected four
19
protected areas for intensive study, including Mulanje Forest Reserve, Liwonde National
Park, Dzalanyama Forest Reserve and Ranch, and Vwaza Wildlife Reserve.
MULANJE FOREST RESERVE
Mulanje Mountain was first gazetted as a forest reserve in 1927, with the nearby
Michesi Mountain added in 1929. Centered on 15°57'S, 35°39'E, the reserve now covers
56,314 ha of mostly mountainous terrain. Geologically, the massif consists of a large
syenitic intrusion, rising from the surrounding plain from 600m to 3,000m above sea level.
Steep slopes and shallow dystric-fersialic soils limit the agricultural suitability of the forest
to some lowlands on the southern and eastern edge of the reserve (Paris 1991a; Pike and
Rimmington 1965).
The miombo woodland (mesic-dystrophic savanna, here dominated by
Julbernardia and Brachystegia species) on the lower slopes has a mean annual rainfall of
800 mm and a mean annual temperature of 21°C. By contrast, the montane forests and
grasslands of the plateau average 2500mm rainfall and 16°C, dropping to an average
minimum of 5°C in the montane forests and grasslands on the plateau and southern slopes.
The reserve protects a number of catchment sources from erosion. It also shelters
considerable biological diversity that includes a greater variety of wildlife than any other
forest reserve in Malawi and over thirty endemic flora species (Chapman 1962; Edwards
1985; Wild 1964), including the threatened Khaya anthotheca (Welw.) C. DC., Philippia
nyasana Aim and T.C.E. Fries (lUCN Species Survival Commission 2000).
20
The reserve also serves as a tourist attraction and a source of high quality timber,
including the Widdringtonia whytei Rendle, which is now considered endemic to Mulanje
(Pauw and Linder 1997) "Mulanje Cedar is Malawi's national tree and is considered
"vulnerable" on the 2000 lUCN Red List (lUCN Species Survival Commission 2000).
Pimts patula Schiede ex Schltdl. & Cham., an exotic originally planted for timber, is part
of the threat as it was become invasive on the mountain.
A large Eucalyptus plantation in the southeastern portion of Mulanje was intended
to supply flielwood and charcoal locally and to urban centers, though the high cost of
transport has limited progress towards reaching objectives. The average population
density around the reserve in 1996 was 211 persons km"^, with the greatest concentration
on the southern side of the reserve near Malawi's most productive tea estates. This
population was 46% male in 1996, and 21% had attended school beyond junior primary.
Household size averaged 5.3 persons holding 0 .8 ha of land, and 29% of these households
were female-headed. The dominant ethnic groups were Lomwe (58%) and Manganga
(29%).
LIWONDE NATIONAL PARK
In the Upper Shire River valley, the riverine-lacustrine Liwonde plain, originally
defined as a Controlled Area for managed game hunting, was constituted as a National
Park in 1972. In 1976 a corridor was added in the northeast to facilitate elephant
movements to and from Mangochi Forest Reserve In addition, a 1 km wide by 35 km long
strip was added to the western bank of the Shire to act as a buffer for wildlife. Centered
21
on I4°50'S, 35°2rE, today the Park encompasses 54,633 ha of predominantly flat
topography, broken by the Chiunguni Hills, a foyaitic ring-complex rising from 450m to
900m above sea level in the southern end of the Park. Along the Shire, Liwonde has
extensive marsh and floodplain areas that include both palm {Borasus aethiopum Mart.)
and reed {Phragmites mauritianus Kunth.) communities. In the main body of the Park,
mopanic and gleyic soils ill suited for agriculture cover alluvium and gravelly colluvium
underlain by Pre-Cambrian gneiss (Pike and Rimmington 1965; Venema 1991). The mean
annual rainfall in the Park is 1000mm, while the mean annual temperature is 24°C, with a
minimum average of 14°C. Liwonde protects wildlife in the Upper Shire and is a Malawian
example of Mopane Woodland, broad-leafed, lowland, drought deciduous woodland and
savanna dominated by mopane (Colophospermum mopane (Kirk ex Benth.) Kirk ex
Leonard). Three years after being declared extinct in Malawi, the first black rhinoceros
{DicerOS bicontis L. minor) were introduced to a sanctuary within the Park in 1993
(Bhima and Dudley 1997), and in 2000 the population is five. Once in decline, the Park is
home to the only elephant {Loxodonta africana Blumenbach) population in the country
that has grown significantly in the past 20-25 years (Bhima and Bothma 1997). Tourism is
the official primary use of the park, with a primary objective to attract high revenue
tourists to Malawi. A wire fence on the more densely populated western edge of the Park
and security measures combine to enforce stricter protection in Liwonde than in all other
protected areas in Malawi. The Park also serves a critical catchment protection role for the
Shire River, the primary source of the country's electricity. The average population
density around the Park in 1996 was 166 persons km'^, with heavier concentrations along
22
the northeastern corridor. The boundary population was 48% male in 1996, and 14% had
attended school beyond junior primary. Household size averaged 4.7 persons holding 0.9
ha of land, and 28% of these households were female-headed. The Yao ethnic group
accounted for 78% of the boundary population in 1996.
DZALANYAMA FOREST RESERVE AND RANCH
The history of protection in Dzalanyama, which means "fiill of wild animals" in
Chichewa (the primary indigenous language), actually began with the creation of a game
reserve called Central Angoniland in 1911. The decline in big game and concern over
water shifted protection emphasis to water and soil, and the Dzalanyama Forest Reserve
was constituted in 1922. In 1966 the Malingunde dam was built, the first of two dams fed
from Dzalanyama tributaries that account for 30% of all water needs for Lilongwe. Four
years later Dzalanyama became the largest protected area in Malawi to be managed by
two different government agencies when 66,574 ha of its lowlands were opened to a
livestock scheme called Dzalanyama Ranch. Centered on 14°20'S, 33°22'E, the reserve
today encompasses 98,827 ha of terrain ranging from low lying, open Miombo and dambo
areas (grasslands in seasonally inundated drainage lines) at 1,100m, to the more closedcanopy Miombo of the Dzalanyama Hill Range that peak at nearly 1,700m in elevation
(Ngalande 1995). Eutric-fersialic soils cover much of the alluvium and colluvium that is
underlain by a Precambrian Basement Complex of granulites, gneiss and schists (Brown
and Young 1965; Lorkeers and Venema 1991). The lowland soils outside the dambos are
largely suitable for agriculture (Orr et al. 1998). The lowlands of Dzalanyama average
23
lOSOmm in annual rainfall, while the highlands average 1350 mm. The mean annual
temperature for the reserve is 20°C, with a minimum average of 10°C. The natural
woodland and almost 5,000 ha of mainly plantations in Pinus kesiya Royle ex Gordon,
Eucalyptus camaldulensis Dehnh., and E. tereticomis Sm. supply fiielwood locally and to
the Lilongwe metropolitan area. The western edge of the reserve runs along a forested
area in Mozambique that has seen limited use over the past 25 years. The rest of the
reserve is bounded by a population averaging 119 persons km"^, with heavier
concentrations near tertiary roads that lead to the main Mchinji-Lilongwe-Dedza highway.
The boundary population was 47% male in 1996, and 19% had attended school beyond
junior primary. Household size averaged 5.1 persons holding 1.6 ha of land, and 22% of
these households were female-headed. The Chewa ethnic group accounted for 99% of the
boundary population in 1996.
VWAZA MARSH WILDLIFE RESERVE
Originally a controlled hunting area, what is today Vwaza Marsh Wildlife Reserve
was officially constituted in 1977. Centered on 11°00'S, 33°28'E, four major landscape
classes have been identified within Vwaza's 98,214 ha. From east to west, these include;
(a) a hills with shallow, rocky, paralithic soils, underlain by a biotite gneiss basement
complex; (b) gently sloping pediments of Karroo sediments, biottite gneiss and dolerite
dykes, covered by sandy, eutric-ferralic soils; (c) alluvial plains that are inundated most of
the year, comprised of mostly fine textured, gleyic and mopanic soils; and (d) plateau areas
underlain by deeply weathered gneiss and quartzite, forming high infiltration, eutric-
ferralic soils (McShane 1985; Paris 1991b; Young and Brown 1962). Elevations in Vwaza
range 1050m in elevation at Lake Kazuni to over 1600 m at Mohohe Hill in the northeast.
Rainfall in the south and west of the reserve averages at 800mm, and reaches 1,100 in the
northeast. . The mean annual temperature for the reserve is 20°C, with a minimum average
of I0°C.
The majority of Vwaza Marsh is made up of Brachystegia-Julbemardia Miombo
woodland, though some montane woodland, dambo grassland, mopane woodland, and
thicket associated with perched water tables are also present. The reserve is rich in both
flora and fauna, with 1,200 plant, 427 bird, 85 mammal, 34 reptile, and 22 amphibian
species. Vwaza has a long history of trade related to wildlife dating back to the 18"*
century. The elephant population, estimated at 250 in 1985 has declined to the point that
sighting assessments are no longer considered valid. Vwaza also contains the tsetse fly
species Glossina morsitam and G. pallidipes, though Trypanosomiasis of cattle is
endemic and an increasing number of sleeping sickness cases among humans have been
reported since 1980 (McShane 1985). The eastern, Zambian side of the reserve is sparsely
populated, while the land surrounding the Malawian boundary averaged 95 persons km"^
in 1996, a density due in part to the creation of numerous tobacco estates over the past 20
years, limiting customary land expansion. The boundary population was 48% male in
1996, and 39% had attended school beyond junior primary. Household size averaged 5.5
persons holding 1.8 ha of land, and 25% of these households were female-headed. The
Tumbuka ethnic group accounted for 95% of the boundary population in 1996.
25
MALAWI
L^VJ;u
BL.MnU
UNITED REP
OF
TANZANIA
MPRTHERN'
ZAMBIA
l\itLK
MOZAMBIQUE
XcemtITAI^U^^ 1 \
'^y
_
ULUlJ.
•^OI^TH
FIGURE I.I. MALAWI IN AFRICA (UNITED NATIONS MAP 3858, NOVEMBER
1994).
LOCATION OF STUDY AREAS IN MALAWI
UWONDE NATIONAL PARK
VWAZA WILDUFE RESERVE
Ukulunqwa
Kapalala
^Balakasl
Kapemra
0)atama
O
Thomas Mkandawire
Kamwendo
DZAUNYAMA FOREST RESERVE
MULANJE FOREST RESERVE
Mthangombe
Kamphambanya
Kanjinga
I
Kambenje
O
-Kadewere
MIelemba
Tsumbi
FIGURE 1.2. LOCATION OF STUDY AREAS WITHIN MALAWI.
K)
OS
27
EXPLANATION OF DISSERTATION FORMAT
The main body of this dissertation is contained in three appendices (A, B, and C).
These are individual research manuscripts that are logically connected and integrated into
the dissertation as a whole. All three manuscripts are based on data collected from 17
communities adjacent to four protected areas in Malawi (Figure 1.1) between 1996 and
1997, as well as spatial analysis of digital data layers that were created for the Public
Lands Utilization Study over the same time frame.
RELATIONSHIP OF THE APPENDED MANUSCRIPTS
The first manuscript details the multidimensional field methodology that provided
both baseline socio-economic information concerning household production and detailed,
species level protected area resource utilization. It documents how the annual harvests of
agricultural and protected area natural resource products were a) identified, b) measured
(by volume in local units of measure), converted to weights in scientific units (kg), verified
against on-site physical measurements and ancillary, key respondent interviews, and then
valued.
The methodology represents a unique blend of tools used to gather livelihood
security information and ethnoecological data. This combination permitted the assessment
of the role protected area resources play in the livelihoods of those living in adjacent
communities. In particular, the first manuscript assesses how the proceeds derived from
28
protected area resource utilization impact equity in the distribution of income, and provide
a means to overcome poverty for the poorest households.
The second manuscript focuses on how protected area resources represent a
bridge between the goals of sustainable development and efforts to promote communitybase natural resource management. It expands on the documentation of methods in the
first manuscript by summarizing the quantity (kg) and value (Malawian Kwacha) of both
agricultural production and protected area resource utilization. These results are presented
on per capita basis, by species for each category of use for the four protected areas under
study. They are then aggregated and analyzed in conjunction with projected population
totals and then compared with an assessment of sustainable protected resource supply.
The results demonstrate how the methodology can be useful in providing critical resource
demand information at the species level, while also contributing to defining the context
within which a resource management system that permits controlled extraction can be
implemented.
The third manuscript brings the results of the first two manuscripts together into a
Malawian case study application of the Boserup (1965) hypothesis on how higher
population densities may spur spontaneous agricultural intensification, and the TifTen,
Mortimore, and Gichuki (1994) extension of this hypothesis that suggests communityinitiated environmental amelioration may also result. It demonstrates how, despite high
population densities, conditions in the Malawian context are not ripe for spontaneous
agricultural intensification. Instead, poor smallholder households are basing both farming
and resource utilization decisions on the absolute, short-term need to survive.
29
The strategy for poverty alleviation identified in the first manuscript proves,
cetenis peribus, to have the potential for sustainability, as demonstrated in the second
article. The third article applies the valuation results of the first two articles on a per
hectare basis, comparing proceeds fi^om agricultural against those derived fi'om protected
area use. Though protected area proceeds are clearly a strategy being used by poor
households to overcome poverty, converting the land beneath those ecological resources
into agriculture would provide much higher returns. The choice between land and forest
proves to be obvious in short term economic terms, but far more costly in the long run,
because agriculturally suitable land would be exhausted by the growing population in
relatively short order, and the option of alleviating poverty through the exploitation of
protected area resources would not longer be available.
CONTRIBUTION OF THE AUTHOR
All three papers were the result of research conducted in the framework of the
Malawi Public Lands Utilization Study (PLUS), which the author of the dissertation
coordinated from 1996 to 1998. PLUS provided a rich and comprehensive database of
biophysical and socioeconomic data used throughout the dissertation. However all
analysis, results and conclusions in the dissertation are original and unique, extending
beyond those reported in the PLUS final report (Orr et al. 1998) submitted to the
Malawian government and the study sponsor, the U.S. Agency for International
Development.
30
With guidance from the dissertation committee, the research design, methodology,
field logistics, data analysis, and interpretations for all three manuscripts were original
contributions provided entirely by the author. The author collaborated with the co-authors
of the first two manuscripts in data collection, supported by a team of eight enumerators
and two vegetation specialists from the Forestry Research Institute of Malawi. The author
collaborated with the co-author on conceptual design of the third manuscript. The staff of
the Arizona Remote Sensing Center conducted the primary satellite image processing and
spatial database construction that was used as a foundation for the analysis in the third
manuscript. Finally, the author collaborated with the co-authors of all three manuscripts in
the verification of results.
31
CHAPTER 2: PRESENT STUDY
SUMMARY
The literature review, data, methods, results, discussion and conclusions of this
study are presented in the appended papers. The following is a summary of the most
important findings.
METHODS
To capture the quantity and value of protected area resource demand, it was
necessary to develop a multidimensional approach that documented baseline
socioeconomic information and resource utilization in a quantitative, integrated manner.
This approach involved the integration of qualitative and quantitative ethnobiological and
socioeconomic information concerning household agricultural production and the use of
plant and animal species fi^om protected areas, as well as spatial analysis of biophysical
data (see Appendix A for detailed summary). Central to this effort was a quantitative
survey conducted in 1996-97 of427 households comprised of 2,205 individuals in across
the four protected areas. The survey was based on respondent recall of production and
resource utilization activities, particularly at the species level. A coincident market survey
was done to permit conversion of local volumetric measures to weights, and to establish
retail prices for each species. Overlapping key respondent interviews captured specialized
resource use (i.e., small enterprises involving fiielwood, charcoal, wild foods, hunting,
handicrafts, tool making, healing, etc.). Data were collected through interviews conducted
in the villages and in the protected areas during resource extraction. These surveys were
32
essential in linking the size and price of final products back to the physical quantity of
species used. This was particularly important for uses where the relationship between the
weight of raw materials were greater than the weight of the product actually sold, or the
unit of sale included the services of a specialist (i.e., a healer converting roots into a cure
for an ailment and then selling both that cure and advice).
With the exception of the single formal survey, data were gathered through
participatory methods; local inhabitants carried out the investigation, presentation, and
preliminary analysis under the guidance and training of the research team (Campbell,
Luckert, and Scoones 1997; Chambers 1990, 1994). The field research team was made up
of four Malawian male and female enumerators conversant in the key local languages
(primarily Chichewa, Chiyao, and Chitumbuka). The resulting data were coded and
entered in the SPSS (Statistical Package for the Social Sciences 1998), where all
aggregation and statistical analysis was conducted.
Species reported by households were physically verified (where possible) in 136
resource assessment plots (eight per village). A field botanist was present at all the
assessment plots, working with local inhabitants to confirm all local names and match
them to scientific equivalents for each species identified. Because scientific species names
did not always correspond directly with local (and sometimes polysemic) names, a
particular effort was made to incorporate local perceptions and classifications into all
instrument responses and into how survey questions were posed (Martin 1995). The
Forestry Research Institute of Malawi (FRIM) and the National Herbarium provided
indispensable assistance with local species names. Other national experts, and Malawi's
33
rich tradition of gathering ethnographic biological information, proved invaluable where
local confirmation was not possible. The extensive plant dictionary of Binns (1972) was
used for confirming Latin names. The works of Williamson (1975), Morris and Msonthi
(1996), and Morris (1990) were essential for addressing gaps in local plant, fiaiit, and
mushroom descriptions, and provided the foundation for evaluating species use.
The socio-economic analysis was intended to capture two key variables. The first
addressed household well being, expressed in terms of both direct and non-direct
household income. The second addressed household protected area natural resource
utilization, broken into seven major use categories; food, fijelwood, fiber, tools, medicinal
plants, and both wood and thatch for construction. Income estimates fi-om ail sources
(direct and indirect) were compiled for each individual household and converted into per
capita values (15 Malawi Kwacha = 1 USD).
For the first manuscript, the aggregate results provided an estimate of household
income including that associated with protected area resources fori996. From this it was
possible to calculate "reliance," or the proportion of total income derived fi-om resources
taken fi-om the protected area. Variability in reliance among households was then analyzed
to determine the role of protected area resources as an income generating activity, and its
impact on poverty and the distribution of income. Reliance was obtained by assessing the
distribution of income without protected area proceeds first, and then comparing that with
an assessment with those proceeds. The specific tools used were standard quantitative
methods for measuring inequality and poverty. These included a Lorenz curve, a Gini
coefficient and polarization index derived fi"om the Lorenz curve, and three poverty
34
indices. The mathematical methods selected to parameterize the Lorenz curve and
determine the underlying indices are defined in Appendix A. The actual calculations were
conducted in POVCAL software, developed specifically for this purpose by the World
Bank (Chen, Datt, and Ravallion 1992).
In the second manuscript, the results were evaluated at the species level by use.
They were also aggregated to create a per capita estimate of protected area resource
demand. A geographic information system (ESRI 1997; 2000) was employed to assess the
sustainability of use over time. Rural population estimates based on population totals and
growth trends derived fi-om the 1998 census were mapped in a 5 km zone of influence
adjacent to each protected area. Total protected area resource demand equaled the total
population in the zone of influence multiplied by the mean per capita resource utilization
estimates (kg) for 1996.
The threshold for sustainable protected area resource extraction was defined as the
volumetric (m^ ha'' yr ') mean annual increment (MAI) of all woody biomjiss in the
protected area, based on a roundwood equivalent (wood in its natural state as felled or
otherwise harvested). Vegetation classifications created fi^om 1994 Landsat Thematic
Mapper satellite imagery during PLUS (Orr et al. 1998) were converted into biomass
maps using MAI estimates for each vegetation type developed by FRIM (Masamba and
Ngalande 1997). The threshold for sustainable extraction was the total of the MAI
estimates for each vegetation class multiplied by the total hectares for each class across the
protected area.
35
Estimates for total protected area resource demand per capita in 1996 were held
constant and applied to population estimates based on 1987 through 1998 growth rates
(Malawi Government 1998). The total demand for protected area resources was compared
to the estimated sustainable supply through time to determine when that sustainable supply
would be exhausted.
The third manuscript involved the spatial comparison of the value per hectare that
could be derived from using protected area land for ecological resources versus the value
that could be obtained from cultivating agriculturally suitable land. This involved the
spatial integration of the results obtained in the first two manuscripts and an additional
spatial analysis of agricultural suitability. The latter involved analysis of PLUS Maps of the
agricultural suitability of land in all four protected areas that were generated from a Land
Resources Evaluation Project (LREP) model and data produced by the Government of
Malawi and FAO (Eschweilere/a/. 1991; Orre/a/. 1998). This manuscript also took
advantage of vegetation classifications for 1984 and 1994 and a change detection analysis,
both products of PLUS (Orr et al. 1998).
Analysis for the third manuscript involved three key comparisons; (1) resource
utilization vs. agricultural land use values; (2) sustainability of ecological vs. land
resources in the face of population pressure and competing uses; (3) the spatial correlation
of negative change with points of high population pressure vs. sites of agriculturally
suitable land within each protected area.
36
RESULTS
Each manuscript is appended (Appendices A, B, and C) Their findings are
summarized here in direct reference to the seven fundamental research questions posed.
(1) What are the per capita quatUitiesa n d consumptive use-values o f each species used
for various uses, whether agricidturally produced or collected from the protected
area?
In all, 694 unique "used" species were identified in the formal survey. The quantities
and values associated with that use are reported on a per capita basis for each category
of use for the most important species in Appendix A. The result are presented
graphically so that each of the four protected areas can be compared.
(2) Are poorer households more reliant on protected area-based income?
Regression analysis demonstrated that poorer households are more reliant on
protected area-based income than wealthy households. The relationship between
income and reliance was exponential, with a reasonable fit (R^ = 0.84). On the basis of
the high probability of significance of the t-statistic, the null hypothesis that per capita
income (independent variable) has no impact on reliance was rejected. The shape and
direction of the exponential curve fit suggests an inverse income-reliance relationship.
On average, for every 100 MK increase in per capita income, the portion of income
that is protected area-based can be expected to decline by 0.1%.
(3) Do protected area proceeds influence overall ec[uality in the distribution of income?
Equality in the distribution of income, with and without protected area proceeds were
37
evaluated with the Gini index (derived from a Lorenz curve), and then compared.
When proceeds from protected areas are excluded from the rest of household income,
the pattern of income distribution among households results in a Gini index of 56.3%.
The ratio declines to 50.9% (more equity) when the protected area proceeds are
included in the income totals.
(4) Does the current supply of protected area resources meet current demand?
The current demand of protected area ecological resources more than meets demand
in all four protected areas.
(5) Based on current patterns of use and population growth, when will demand outstrip
sustainable supply?
Using 1996 as a base year, and hold current utilization rates constant, the sustainable
supply of protected area ecological resources can be maintained for 50 years in
Mulanje, Liwonde, and Vwaza, and 100 years in Dzalanyama.
(6) If protected area land suitable for agriculture was opened to cultivation, when would
it be consumed, based on current patterns of land use and population growth?
Using 1996 as a base year, the results show rapid conversion of agriculturally suitable
land in Mulanje (5 years), despite low population growth rates, because only a small
percentage of the mountainous reserve is suitable for agriculture. Liwonde's
agriculturally suitable land would last only 8 years, while much larger Vwaza and
Dzalanyama (which also have much larger per capita land holdings) would last 30, and
58 years, respectively.
38
(7) What is the per hectare value of protected area land used for a) natural resource
product litilization, and b) agriculture?
The per hectare value of protected area natural resource product utilization ranges
from 44.78 USD in Dzalanyama to 67.64 USD in Mulanje. The agricultural values are
significantly higher, ranging from 188.84 USD in Liwonde to 316.23 USD in
Dzalanyama.
(8) Does spatial analysis of land cover change suggest negative impacts are associated
with proximity to higher population densities (as would be the case if fuehvood
extraction dominated the process) or with the distribution of agriculturally suitable
land?
Spatial analysis of land cover change between 1984 and 1994 shows limited spatial
association between concentrations of high population adjacent to sites of negative
land cover change. By contrast, there is a high association between agriculturally
suitable lands and negative land cover change.
CONCLUSIONS
This research introduces a multidimensional field methodology that can provide
both baseline socio-economic information concerning household production and detailed,
species level protected area resource utilization. Using the example of four protected areas
in Malawi, I have demonstrated how the annual harvests of agricultural and protected area
natural resource products were; (a) identified; (b) measured (by volume in local units of
39
measure); (c) converted to weight units, (d) verified against on-site physical measurements
and key respondent interviews; and (e) valued in monetary units.
When collected data are aggregated to the household level, the importance of
protected area-based income is evident. Poor households are generally more reliant on
protected area resources than other households. In turn, that reliance has a major influence
on the distribution of income between poor and rich households. Exploitation of protected
area resources is a livelihood strategy that halved the number of households that would
have remained beneath the basic needs poverty threshold. These findings have major
policy ramification, not only for environmental management and conservation, but also for
poverty alleviation.
In a society that is almost entirely dependent on agriculture households, and in
particular poor households, low income households have created income-generating
alternatives to maintain livelihood security. That these alternatives are based on protected
area proceeds represents a challenge and an opportunity to those attempting to balance
sustainable use and sustainable development simultaneously.
The methodology described here represents a unique blend of tools used to gather
livelihood security information and ethnoecological data that have not been used
traditionally to capture the role of protected area resources in household economies. This
provides the opportunity to address and monitor the influence of those resources on local
communities and gauge the ecological impact of resource demand.
The quantity and consumptive use value of protected area species can be used to
inform natural resource management. Knowledge of the level of demand for individual
40
species can be compared with field survey analysis of its ecological status to estimate the
impact of extraction. Where rates of demand are deemed excessive, use of alternative
species in the protected area can be promoted. Moreover, the role individual species plays
in overall household income can be assessed. Substitutes among agricultural crops or
livestock can be identified and, where desired, alternative income generating activities can
be promoted to reduce pressure on specific species.
Variation was high among four Malawian protected areas in products extracted
from them and the species that were used. In some cases this was due to biophysical
conditions that favored specific species, or protection policies that target selected species,
such as large mammals. In other cases, local preferences or the availability of alternatives
explained the variation.
Despite this variation, for each use category, a few species dominated, none more
than the multipurpose fiiiit tree, masuku {Uapaca kirkiana Mull. Arg.). This species was
the most important protected area source of food, was very important for fuelwood, and
was used for construction, fiber, and medicinal purposes. Our findings support the
conclusion of both Ngulube (1995) and Malembo, Chilanga, and Maliwichi. (1998) who
both found the pervasive use of masuku in the wild as in important factor supporting its
candidacy for domestication as an agroforestry species.
Aggregation of woody species utilization data can also provide an estimate of
overall protected area resource demand. In three of the protected areas studied, demand
will exceed the sustainable supply of woody biomass within 50 years, and within 100 years
in the fourth. This suggests that, barring any additional demand pressure, there is time to
41
enhance community-based natural resource initiatives that are promoted for all four
protected areas. In addition to their inherent institutional and management complexities,
community conservation efforts must take these time constraints into account.
An analysis of aggregate sustainable use provides a context for natural resource
decision making. The analysis, particularly when combined with species-level demand
assessments and physical resource inventories, provides a set of monitoring tools for
community management and conservation. Without regularly updated estimates of
protected area resource demand, threats to individual species can only be assessed by
impact on the resource, a metric applied only after damage is done. Providing resource
demand information enhances management options and offers an understanding of the
problem from the perspective of the user. Working without this information where
managed use is encouraged limits the tools available to resource managers to find
alternatives for threatened resources.
Poor households who actually use protected area proceeds to overcome poverty
would still choose to covert those protected area land base to agriculture if given the
option. There is strong indication in both the qualitative and quantitative data that those
surveyed were aware of the longer term risks of converting protected areas to agriculture,
and that they are equally aware of the immediate differences in consumptive use values of
both the land and the ecological resources. This would seem to refute Carrasco's (1993)
position that knowledge of the value of ecological resources might lead to conservation is
unlikely when the difference between the value of land cultivation and consumptive use of
natural resources is so large. The fact that protected area resources provide a critical
42
income flow tliat compensates for insufficient land in many cases does not compensate for
short-term risk. Our findings are more in line with Reardon (1998)- that poor households
in Malawi are thinking about survival first, that they recognize conditions are not ripe for
agricultural intensification as a solution, that they are aware that protected area resources
may be only a stop-gap. The short-term risk in investing in either farming innovation or
conservation overwhelms longer-term considerations.
Unfortunately, using protected area-based income for poverty alleviation has
longer-term limitations as a livelihood strategy. The protected area results assume a
system limited to those living within a 5 km zone of influence around each protected area.
Yet supply and demand pressures on both land and ecological resources at the nationallevel suggest unsustainable demand, undermining the notion of a closed system. .
Protected areas supply products not only to local populations, but society as a whole.
Moreover, the total value of protected area resources to the whole of Malawi goes beyond
the economics of simple consumptive use, to include genetic diversity, cultural, religious,
aesthetic, and intrinsic natural value (Rolston 1985).
Demand for land and the demand for ecological resources will result in difficult
choices at the local level. Reliance on protected area resources can provide much-needed
alternatives to agricultural income in the short run where land is scarce. But in the longer
term, this reliance may expose poorer households to the strong possibility that land
demand will eventually overtake the resource base providing the alternative income
stream. Government and donor environmental policies and their prescriptive interventions
must consider replacing the poverty-alleviation function of protected area resources with
43
income alternatives that are neither land nor forest-based. Furthermore, community
conservation initiated by donor projects that do not address the large financial differences
between the value of land and the value of ecological resources may actually accelerate the
risk faced by the very communities to gain from such measures. As suggested by
Shyamsundar and Kramer (1997), proceeds from protection (ecotourism, etc.) need to be
reinvested in local communities to compensate for the value of foregone agricultural lands
and secondary protected area products. In a country like Malawi, where conditions are not
ripe for autonomous agriculture intensification, and income diversification is often through
the exploitation of natural resources, investment must be directed towards the
development of alternative income streams that give poorer households options. It is
essential that these options are not entirely dependent on natural resources, and that they
do not exacerbate already critical land shortages.
REFERENCES
Arnold, J.E.M., and J.E. Falconer. 1989. Household food security and forestry; an analysis
of socioeconomic issues. Community Forestry Note 1. FAO, Rome.
Arpaillange, J. 1996. Urban household energy; demand side strategy. SEED/Ministry of
Energy and Mining, Lilongwe, Malawi.
Barrett, C.B., and P. Arcese. 1995. Are integrated conservation-development projects
(ICDPs) sustainable? World Development 23(7)1073-1084.
Bawa, K.S. and S. Dayanandan. 1997. Socioeconomic factors and tropical deforestation.
Nature 386;562-563.
BDPA/AHT. 1998. Customary land utilization study. BDPA/AHT, Lilongwe, Malawi.
Berry, V. and C. Petty (eds ). 1992. The Nyasaland Survey Papers, 1938-1943,
Agriculture, Food and Health. Academy Books. London, UK.
44
Bhima R., and CO. Dudley. 1997. The influence of the Shire River on Liwonde National
Park, Malawi, with special reference to elephant movements. Koedoe 40:9-18.
Bhima R., and J. du P. Bothma. 1997. Age structure of elephants in Liwonde National
Park. Koedoe 40:1-8.
Binns, B. 1972. Dictionary of plant names in Malawi. Government Printer, Zomba,
Malawi.
Boserup, E. 1965. The Conditions of Agricultural Growth: The Economics of Agrarian
Change Under Population Pressure. Aldine Publishing Company, New York, USA.
Brandon, K.E., and M. Wells. 1992. Planning for people and parks: design dilemmas.
World Development 20(4)557-570.
Brown, P. and A. Young. 1965. The physical environment of Central Malawi: with special
reference to soils and agriculture. Government Printer, Zomba, Malawi.
Carrasco, D.A. 1993. Constraints to sustainable soil and water conservation: a Dominican
Republic Example. Ambio 22, 347-350.
Chapman, J.D. 1962. The vegetation of the Mlanje Mountains, Nyasaland: a preliminary
account with particular reference to the Widdringtonia forests. Nyasaland
Government Printer, Zomba, Nyasaland (Malawi).
Campbell, B.M., M. Luckert, and I. Scoones. 1997. Local-level valuation of savanna
resources; a case study from Zimbabwe. Economic Botany 51 ;59-77.
Chambers, R. 1990. Rapid and participatory rural appraisal. Appropriate Technology
16(4); 14-16.
Chambers, R. 1994. The origins and practice of participatory rural appraisal. World
Development 22:953-969.
Chen, S , G Datt, and M. Ravallion. 1992. POVCAL: a computer program for calculating
poverty measures from grouped data. Policy Research Department, World Bank,
Washington DC.
Edwards, L 1985. Conservation of plants on Mulanje Mountain, Malawi. Oryx 19:86-90.
Eschweiler, J.A., Paris, S., Venema, J.H., Lorkeers, A.J.M. and Green, R.L 1991.
Methodology for Land Resources Survey and Land Suitability Appraisal. LREP,
Field Doc. No. 30, MOA/UNDP/FAO, Lilongwe, Malawi.
45
ESRI. 1997. Arclnfo version 7.1.2. Environmental Systems Research Institute, Inc.,
Redlands, Washington, USA.
ESRI. 2000. ArcView version 3.2. Environmental Systems Research Institute, Inc,
Redlands, Washington, USA.
FAO. 1981. Forest resources of tropical Africa; tropical forest resources assessment
project. FAO, Rome.
FAO. 1997. State of the world's forests. FAO, Rome.
FAO. 1999. State of the world's forests. FAO, Rome.
Finan, T.J. 1996. Anthropological research methods in a changing world. Pages 101-124
in Moran, E.F., ed.. Transforming societies, transforming anthropology. University
of Michigan Press, Ann Arbor.
Garine, I. de, and G. Koppert. 1988. Coping with seasonal fluctuations in food supply
among savanna populations: the Massa and Mussey of Chad and Cameroon. Pages
210-259 in I. De Garine and G.A. Harrison, eds.. Coping with uncertainty in food
supply. Clarendon Press, Oxford.
Godoy, R., R. Lubowski, and A. Markandya. 1993. A method for the economic valuation
of non-timber tropical forest products. Economic Botany 47:220-233.
Godoy, R., D. Wilkie, H. Overman, A. Cubas, G. Cubas, J. Demmer, K. McSweeney, and
N. Brokaw. 2000. Valuation of consumption and sale of forest goods from a
Central American rain forest. Nature. 406: 62-63.
House, W.J. and G. Zimalirana. 1992. Rapid population grouth and poverty generation in
Malawi. The Journal of Modem African Studies 30:141-161.
lUCN Species Survival Commission. 2000. The 2000 lUCN Red List of Threatened
Species. Online. Available HTTP: http://redIist.cymbiont.ca/splash.html. November
2000.
Kremen, C., J.O. Niles, M.G. Dalton, G.C. Daily, P R. Ehrlich, J.P. Fay, D. Grewal, and
R.P. Guillery. 2000. Economic incentives for rain forest conservation across
scales. Science 288(5472): 1828-1832.
Lorkeers, A.J.M., and J.H. Venema. 1991. Land Resources Evaluation Project: Land
Resources Appraisal of Lilongwe Agricultural Development Division.
MLW/85/011 Field Document No. 24. Malawi Govemment/UNDP/FAO,
Lilongwe, Malawi.
46
Malawi Government. 1998. Report of Preliminary Results; 1998 Population and Housing
Census. National Statistics Office, Zomba, Malawi.
Martin, G.J. 1995. Ethnobotany; a methods manual. People and Plants Conservation
Manuals, Volume 1. Chapman and Hall, London.
Masamba, C., and J. Ngalande 1997. Inventory data of biomass growing stock and supply.
FRIM Technical Report. Forestry Research Institute of Malawi, Zomba.
McShane, T.O. 1985. Vwaza Marsh Game Reserve; a baseline ecological survey.
Department of National Parks and Wildlife, Rumphi, Malawi.
Millington, A. Townsend, J., Kennedy, P., Saull, R., Prince, S., and Madams, R. 1989.
Biomass assessment; woody biomass in the SADCC region. Earthscan, London.
Morris, B. 1990. An annotated check-list of the macrofungi of Malawi. Kirkia 13; 323364.
Morris, B. and J.D. Msonthi. 1996. Medicinal plants of Malawi. Pages 191-557 in B.
Morris ed., Chewa medical botany; a study of herbalism in Southern Malawi.
International African Institute, London.
Munasinghe, M. 1992. Biodiversity protection policy; Environmental valuation and
distribution issues. Ambio 21;227-236.
Mwafongo, W.M.K. 1994. Enviommental degradation and resource management
strategies in post-independence Malawi. Malawi Journal of Social Science 16:4155.
Mwafongo, W.M.K. and M.L.M. Kapila. 1999. Environmental management in Malawi lessons from failure. In; Environmental Planning, Policies and Politics in Eastern
and Southern Africa.Macmillion Press, Houndsmill, Basingstoke, Hampshire, UK.
Ngalande, J.D. 1995. An integrated management plan for Dzalanyama Forest Reserve,
Malawi. M.S. thesis. University of Aberdeen, Scotland.
Openshaw, K. 1997. Urban biomass fuels; production, transportation, and trading study.
Alternative Energy Development, Inc/Malawi Ministry of Energy and Mining,
Lilongwe, Malawi.
Orr, B., B. Eiswerth, T. Finan, and L. Malembo. 1998. Malawi public lands utilization
study. Malawi Ministry of Lands, Housing, Physical Planning, and
Surveys/University of Arizona/Forestry Research Institute of Malawi, Lilongwe,
Malawi.
47
Paris, S. 1991a. Land Resources Evaluation Project: Land Resources Apprsiisal of
Blantyre Agricultural Development Division. MLW/85/011 Field Document No.
24. Malawi Govemment/UNDP/FAO, Lilongwe, Malawi.
Paris, S. 1991b. Land Resources Evaluation Project: Land Resources Appraisal of Mzuzu
Agricultural Development Division. MLW/85/011 Field Document No. 27. Malawi
Govemment/UNDP/FAO, Lilongwe, Malawi.
Pauw, C.A., and H.P. Linder. 1997. Tropical African cedars {Widdringtonia,
Cupressaceae); systematics, ecology and conservation status. Botanical Journal of
the Linnean Society 123:297-319.
Pearce, D.W., and R.K. Turner 1990. Economics of natural resources and the
environment. Harvester Wheatsheaf, London.
Pike, J.G., and G.T. Rimmington. 1965. Malawi: a geographical study. Oxford University
Press, London.
Reardon, T. 1998. African Agriculture: Productivity and Sustainability Issues. In:
International Agricultural Development, Third Edition. Eicher, C.K. and Staatz,
J.M. (eds.). Johns Hopkins University Press, Baltimore, USA.
Rolston III, H. 1985. Valuing Wildlands. Environ. Ethics 7:23-48.
Shyamsundar, P. and Kramer, R. 1997. Biodiversity Conservation—At What Cost? A
Study of Households in the Vicinity of Madagascar's Mantadia National Park.
Ambio 26:180-184.
Snapp, S.S. 1998. Soil Nutrient Status of Smallholder Farms in Malawi. Communications
in Soil Science Plant Analysis. 29:2571-2588.
Statistical Package for the Social Sciences. 1998. SPSS base 8.0 for Windows: user's
guide. SPSS Inc., Chicago.
Tiffen, M., Mortimore M. and Gichuki, F. 1994. More People, Less Erosion:
Environmental Recovery in Kenya. Wiley, Chichester, UK.
United Nations Population Division. 2000. Infonation: statistical data for member states of
the United Nations (United Nations CyberSchoolBus). Online. Available HTTP:
http://www.un.org/Pubs/CyberSchoolBus/infonation/e_infonation.htm . November
2000.
48
Venema, J.H. 1991. Land Resources Evaluation Project; Land Resources Appraisal of
Liwonde Agricultural Development Division. MLW/85/011 Field Document No.
22. Malawi Govemment/UNDP/FAO, Lilongwe, Malawi.
Wainwright C., and W. Wehrmeyer. 1998. Success in integrating conservation and
development? a study from Zambia. World Development 26:933-944.
Wild, H. 1964. The endemic species of the Chimanimani Mountains and their significance.
Kirkia. 4:125-159.
Willan, R.G.M. 1947. Annual report of the Forestry Department for the year ended 31"
December, 1946. Nyasland Protectorate, Zomba, Malawi.
Williamson, J. 1974. Useful plants of Malawi: revised and extended edition. Montfort
Press, Limbe, Malawi.
Woodson, D.G. 1997. Lamanjay, food security, securite alimentaire\ a lesson in
communication for BARA's mixed-methods approach to baseline research in Haiti,
1994-1996. Culture and Agriculture 19:108-122.
World Bank 1998. Malawi: Human Resources and Poverty. Profile and Priorities for
Action. No. 15437-MAI. World Bank, Washington D.C., USA.
Young, A. and P. Brown. 1962. The physical environment of northern Nyasaland: with
special reference to soils and agriculture. Government Printer, Zomba, Malawi.
49
A. APPENDIX A. INTEGRATING METHODOLOGIES FOR LIVELfflOOD
SECURITY AND ETHNOECOLOGICAL ASSESSMENTS IN RURAL
MALAWI
A.l
ABSTRACT
Orr, Barron J. (Office of Arid Land Studies, University of Arizona, 1955 E.
Street,
Tucson, AZ 85719, USA, barron!^g.arizona.edu), L.N. Malembo, and Humphrey T.
Chapama (Forestry Research Institute of Malawi, P.O. Box 270, Zomba, Malawi,
frim^^alawi.net). INTEGRATING METHODOLCXJIES TO MEASURE THE ROLE OF PROTECTED
AREA RESOURCES IN LIVELIHOOD SECURITY. This study introduces a multidimensional
approach that captures baseline socioeconomic information and resource utilization in a
quantitative, integrated manner. It was based on a quantitative survey of427 households
across four protected areas in Malawi. Household income was derivedfrom a "sum of
the parts" aggregation of income elements including species-level agriculture production
and resource utilization data. Regression analysis (R^ = 0.84) demonstrated that poorer
households are more reliant on income derived from protected resources than wealthy
households. Lorenz curve analysis demonstrated that income distribution equity improves
when proceeds from protected areas are included in household income. Poverty threshold
analysis indicates that exploitation of protected area resources is a livelihood strategy
that halved the number of households that otherwise would have remained beneath the
basic needs poverty threshold. This study demonstrates that protected area resources
play a pivotal role in poverty alleviation, and by extension, efforts to make sustainable
use and sustainable development compatible.
50
Key Words: protected area resources; valuation; livelihood security; inequality; income
distribution; reliance; poverty threshold, participatory appraisal; Malawi.
A.2 INTRODUCTION
Poor rural households rely on natural resources extracted from protected
parks and reserves, despite the potential for detrimental ecological impact. This deliberate
livelihood strategy is intended to balance income against expenditure in normal times
(Arnold and Falconer 1989; Guijt, Hinchcliffe, and Melnyk 1995). When basic livelihood
security is threatened by unanticipated economic events, wild resource utilization also
serves as a final safety net survival strategy (Garine and Koppert 1988; Luoga, Witkowski,
and Balkwill 2000; Nurse 1975; Luckert et al. 2000; Zinyama, Matiza, and Campbell
1990). Despite the importance of wild resources in household economies, baseline
assessments of livelihood security often fail to capture their role in anything other than
broad or aggregate terms (Giuliano, Ilahiane, and Orr 1998), or ignore them altogether
(Fleuret 1979).
Livelihood security assessments would benefit from a basic understanding of
ethnobiological and economic aspects of ecological systems. Equally, biological
conservation assessments would be enhanced by an understanding of the immediate causes
of environmental degradation through analysis of regularly available, locally relevant
socio-economic data (Bawa and Dayanandan 1997). Baseline livelihood security
assessments and follow-up monitoring studies are conducted routinely in many developing
nations for decision making or as a diagnostic tool for projecting and monitoring the
51
impact of development projects. These assessments tend to target large areas, with
relatively brief visits to study sites and thus could not replace long-term site and resourcespecific ethnobiological research. However, methodologically integrated, regularly
updated livelihood assessments could provide a unique and essential monitoring
component to resource utilization and valuation studies
The data gathered in baseline livelihood security assessments are commonly used
to measure poverty. By integrating methods to value plant and animal use, it is possible to
quantify the relative impact of proceeds derived from protected areas on the distribution of
income among a population. It also permits the assessment of the importance of protected
area resources in the poverty alleviation strategies pursued by the poorest households.
Within this framework we developed a multi-dimensional approach to maximize
the rapid collection of baseline socio-economic data drawn from livelihood security
assessment methodology (Woodson 1997) and participatory rapid appraisal (Campbell,
Luckert, and Scoones 1997; Chambers 1990, 1994) and combine them with
ethnobiological resource valuation methods (Godoy, Lubowski and Markandya 1993).
The approach was tested in communities around four protected areas in Malawi. The data
collected permitted us to address four essential questions;
(1) What are the per capita consumptive use-values of agricultural production and
protected area resource utilization among rural households?
(2) What is the relationship between protected area proceeds and all other major income
components within rural households?
(3) Are poorer households more reliant on protected area-based income?
52
(4) Do protected area proceeds influence overall equality in the distribution of income?
A.3 STUDY AREAS
In 1997, 86% of Malawi's 9.65 million people lived in rural areas, translating into
an average population density of 103 persons/km^ of land, three times that of its neighbors
(United Nations Population Division 2000). At the same time, 19% of Malawi's 94,000
km^ of land is under protection in the form of four wildlife reserves, five national parks,
and seventy-seven forest reserves (Orr et al. 1998), resulting in twice the per capita
pressure on both protected and non-protected, wooded areas
To better understand the ramifications of this pressure, the Government of Malawi
commissioned a Public Lands Utilization Study (PLUS) in 1996 to study both the
environmental risks of converting protected land to agriculture and the importance of the
reserves and their resources to adjacent communities (Orr et al. 1998). Four of the areas
selected for intensive study are the subject of this research; Mulanje Forest Reserve,
Liwonde National Park, Dzalanyama Forest Reserve and Ranch, and Vwaza Wildlife
Reserve.
A.3.1 MULANJE FOREST RESERVE
Mulanje Forest Reserve, located in southeastern Malawi (centered on 15°57'S,
35°39'E, covering 56,314 ha of mostly mountainous terrain) has vegetation ranging fi-om
montane woodland and grassland on the plateau to miombo woodland (mesic-dystrophic
savanna, dominated by Julbemardia and Brachystegia species) on the lower slopes. Steep
slopes and shallow dystric-fersialic soils limit the agricultural suitability of the forest to
53
some lowlands on the southern and eastern edge of the reserve (Pike and Rimmington
1965; Paris 1991a). The reserve protects a number of catchments from erosion and
supplies both hard and softwood timber. It also shelters considerable biological diversity
that includes a greater variety of wildlife than any other forest reserve in Malawi and over
thirty endemic flora species (Edwards 1985). Mulanje lies in the most densely populated
area of the four reserves (211 km"~ in 1996), with the greatest concentration on the
southern side of the reserve near Malawi's most productive tea estates. This population
was 46% male in 1996, and 21% had attended school beyond junior primary. Household
size averaged 5.3 persons holding 0.8 ha of land, and 29% of these households were
female-headed. The dominant ethnic groups were Lomwe (58%) and Manganga (29%).
A.3 2 LIWONDE NATIONAL PARK
Liwonde National Park is located in south central Malawi (centered on 14°50'S,
35°21 'E, encompassing 54,633 ha) in the Upper Shire River valley, on a predominantly
flat, riverine-lacustrine plain covered in mostly mopanic and gleyic soils ill suited for
agriculture (Venema 1991). Liwonde protects biodiversity and wildlife in the Upper Shire
and one of the few Malawian examples of mopane woodland (a broad-leafed, drought
deciduous woodland and savanna dominated by Colophospermum mopane (Kirk ex
Benth.) Kirk ex Leonard) (Bhima and Dudley 1997). It is the most important location for
ecotourism in Malawi. Enforcement of protection is stricter in Liwonde than anywhere
else in Malawi, due in part to a wildlife fence on the more densely populated western edge
of the Park. It is also the base of operation for the nation's wildlife scout training program.
54
The average population density around the Park in 1996 was 166 persons km"^, with
heavier concentrations along the northeastern corridor. The boundary population was 48%
male in 1996, and 14% had attended school beyond junior primary. Household size
averaged 4.7 persons holding 0.9 ha of land, and 28% of these households were femaleheaded. The Yao ethnic group accounted for 78% of the boundary population in 1996.
A.3 3 DZALANYAMA FOREST RESERVE AND RANCH
Dzalanyama Forest Reserve is located in the central, western part of Malawi,
(centered on 14°20'S, 33°22'E). It is Malawi's largest forest reserve, encompassing
98,827 ha of terrain. The western third of this area is close-canopy, upland miombo that
borders a forested area of Mozambique that has seen limited human use over the past 25
years (Ngalande 1995). The rest is shared with a government agricultural scheme called
Dzalanyama Ranch, located in a mostly low lying area of open miombo and dambo
(grasslands in seasonally inundated drainage lines). The lowlands are dominated by eutricfersialic soils that are suitable for agriculture (Lorkeers and Venema 1991). The portion of
the reserve boundary facing into Malawi is bounded by a population averaging 119
persons km"^, with heavier concentrations near tertiary roads that lead to the main
Mchinji-Lilongwe-Dedza highway. The boundary population was 47% male in 1996, and
19% had attended school beyond junior primary. Household size averaged 5.1 persons
holding 1.6 ha of land, and 22% of these households were female-headed. The Chewa
ethnic group accounted for 99% of the boundary population in 1996.
55
A.3.4 VWAZA MARSH WILDLIFE RESERVE
Vwaza Marsh Wildlife Reserve (centered on 11°00'S, 33°28'E) primarily serves to
protect biodiversity and wildlife, and to promote ecotourism (though current levels are
well below that of Liwonde). Vwaza also serves to contain the tsetse fly; Trypanosomiasis
of cattle is endemic and an increasing number of sleeping sickness cases among humans
have been reported since 1980 (McShane 1985). The majority of Vwaza Marsh is situated
on eutric-ferralic soils that are suitable for agriculture (Paris 1991b. Vegetation consists of
mostly miombo woodland, though some montane woodland, dambo grassland, mopane
woodland, and thicket are also present. The eastern, Zambian side of the reserve is
sparsely populated, while the land surrounding the Malawian boundary averaged 95
persons km"^ in 1996. This density is greater than much of northern Malawi, due in part to
the creation of numerous tobacco estates over the past 20 years, limiting customary land
expansion. The boundary population was 48% male in 1996, and 39% had attended school
beyond junior primary. Household size averaged 5.5 persons holding 1.8 ha of land, zmd
25% of these households were female-headed. The Tumbuka ethnic group accounted for
95% of the boundary population in 1996.
A.4 METHODS
A.4 1 APPROACH
This study was conducted with a multidimensional approach using on baseline
livelihood security data collection methods developed in the Sahel and Haiti (Finan 1996;
Woodson 1997), and then adapted to evaluate the flow of protected area resource product
56
utilization as part of the Malawi Public Lands Utilization Study (Orr et al. 1998). A
livelihood security approach uses a "rapid" survey of individuals and groups within an area
that is conducted over a period of weeks. It is intended to capture both a quantitative and
qualitative picture of household food stocks, flows, and patterns of consumption. While
sufficient and useful for studying relatively large areas quickly, it is subject to some
limitations associated with respondent recall over the period about which questions are
asked. (Bernard et al. 1984; Nachman 1984).
In this study, rather than strictly food, we focused on natural resource utilization.
This allowed us to develop a "muhidimensional" approach that was based on other
independent observations of resources and behaviors to corroborate survey respondent
responses (Table A.1). We used a mix of overlapping qualitative and quantitative
techniques both inside the home (where both male and female members participated), and
inside the protected area, where activities could be observed and measured. Ultimately,
results based on respondent recall were calibrated with an inventory analysis of resource
utilization zones. Other than the formal survey, data collection was participatory in nature;
local inhabitants carried out the investigation, presentation, and preliminary analysis under
the guidance and training of the research team (Campbell, Luckert, and Scoones 1997;
Chambers 1990, 1994).
57
TABLE A 1. SUMMARY OF DATA COLLECTION METHODS
Data Gathering Activity
Primary Objectives
Rapid Appraisal (138 villages)
Interviews with traditional officials
Village list, locations, resource patterns
Community meetings
Access to infrastructure, services, resources
Agricultural calendar
Off-farm income generating activities
Patterns of protected area use
Crop, livestock, wild species list
Focus group interviews
(men and women separately)
Qualitative specialized use
Land and resource tenure
Attitudes towards protection
Changes in resource availability/access
Intensive study (17 villages)
Participatory mapping
(men and women separately)
Village spatial extent and infrastructure
Protected area vegetation
Present and past resource utilization
Key respondent interviews
Quantitative specialized use (223 respondents)
Land and resource tenure
Village history/crisis/environmental change
Alternatives to protected area resources
Local unit volume and weight conversions
Local market retail prices
Resource assessment
(136 plots)
Total canopy cover
Ground cover (point frequency)
Species identification and use
Local name / Latin name verification
Woody species:
Diameter at breast height (DBH)
Height, species abundance
Herbaceous species
Species cover
Intra-household variability
Demographics
Income (agriculture production, livestock.
remittances, oflF-farm activities, etc.)
Protected area resource utilization
Formal survey
(427 households)
•
58
Data collection began in April 1996 with a three month pilot study at Zomba and
Malosa Forest Reserves, and was completed in March 1997. The socio-economic research
team included four men and four women to reduce gender bias. All members of the
research team spoke Chichewa, as did all the respondents in Dzalanyama and Mulanje, and
the majority in Liwonde. The team also included one Chiyao speaker for elderly
respondents around Liwonde, and four native Chitumbuka speakers, for Vwaza Marsh
respondents.
A resource assessment team comprised of a botanist and two forestry mensuration
specialists worked inside each protected area adjacent to the villages selected for intensive
study. During the pilot study these researchers were trained in the theoretical background
of the livelihood systems approach and all participated in the resource assessment to gain
familiarity with the ethnoecological aspects of the study. Both qualitative and quantitative
survey instruments were created with feedback from the entire team. Each instrument was
tested and modified during the pilot study. Because scientific species names did not always
correspond directly with local (and sometimes polysemic) names, a particular effort was
made to incorporate local perceptions and classifications into all instrument responses and
into how questions were posed (Martin 1995).
A.4. LI Rapid appraisal
Research in each protected area began with a rapid appraisal to capture variation
among localities and agroclimatic zones (Beebe 1995; Bruce 1989; Freudenberger 1995).
Similar to the findings of Brouwer et al. (1997) in central Malawi, a "distance of use"
59
survey revealed that villages further than 5km from the rarely used protected area
resources. Due to uncertainties regarding topographic survey sheets and census maps, a
list of villages within that distance was confirmed through meetings with the District
Commissioner and with each Traditional Authority Chiefs (TA) responsible for land
adjacent to the protected areas.
A random sample of these villages was selected, and the TA (or a designate)
traveled with the authors and introduced them to village chiefs. The village chiefs helped
arrange a community meeting and focus group meetings designed to elicit generally shared
information on village and household livelihoods, including access to infrastructure,
resources, and services, and the role of public lands and resources in local livelihood
systems.
Community interviews focused on the history of the village and the use of
protected area resources for food, medicine, fuel, and construction materials, and how
availability and use of such resources has changed through time. Focus group interviews
collected data on village infrastructure, access to resources, and livelihood activities,
including the nature of agricultural practices (crops, seasonality, land, labor, conservation
strategies), livestock, other income-generating activities (such as wage labor), periods of
food insecurity and coping strategies for dealing with them, and patterns of public land
resource use. With these qualitative data gathering techniques, 138 villages were sampled
during the study. The data collected, including local species names and uses, was
subsequently incorporated into participatory and formal survey instruments specific to
each protected area.
60
A.4.1.2 Intensive Study
Patterns identified in the rapid appraisal were compiled and evaluated in
conjunction with secondary agroclimatic and ecological data about each protected area.
This was followed by a purposive selection of villages for more intensive study, designed
to capture biophysical variability and its impact on livelihood systems and protected area
resource use. Seventeen villages were selected across the four protected areas under
intensive study.
Fieldwork in each village began with a multipurpose meeting with local residents.
The three goals of these meetings were; (a) to generate a list frame of household heads for
the formal survey, (b) to identify specialized users and other key respondents for
interviews, and (c) to map resource utilization zones within the protected area. It was
necessary to generate a timely list frame when it became apparent that census and
drought-related food aid household lists were out-dated.
A.4.1.3 Key respondent interviews
The rapid appraisal revealed that some essential resource utilization and land
tenure information held by a relatively small number of people in each village risked not
being captured in a quantitative survey. Therefore, key respondents were targeted for
interviews about specialized resource use (i.e. small enterprises involving fiielwood,
charcoal, wild foods, hunting, handicrafts, tool making, healing, etc.), land and resource
tenure, alternatives to protected area resources, village history and environmental change,
prices, and local units of measure. These interviews were semi-structured, often taking
61
place while the respondent was conducting an activity of interest. Where possible, the
production process for converting raw resources into products for sale was observed and
input/output quantities were measured.
A.4.1.4 Participatory mapping
The participatory mapping exercise was conducted with groups of men and
women; each group was composed of individuals of varying ages and experience in
resource use (Gupta et al. 1989; Momberg, Atok, and Sirait 1996). After delineating
infrastructure, soils, and land cover (as part of context and scale training), each group
mapped present land use and resource utilization on clear acetate over 1994 aerial photos
(1;25,000). The resulting maps were used to evaluate the spatial extent and location of
different resource utilization activities, and to determine where physical measurements
should be taken.
A.4.1.5 Resource assessment
In order to verify species names, resource utilization patterns, and to assess
impact, a resource assessment was conducted. The participatory mapping participants
identified one male and one female inhabitant deemed most experienced in the utilization
of resources from protected areas. These local experts accompanied the resource
assessment team to each protected area vegetation type/resource utilization zone identified
in the participatory mapping process. Together, they sampled a total of eight 100 m^ plots
per village and took a variety of ecological measurements in each (Table A.1). In support
of the socio-economic analysis, the botanist and local representatives identified woody and
62
herbaceous species in each plot, documenting the local and scientific name for each as well
as observed and potential use. This information was then added to the quantitative survey
questionnaire.
A.4.1.6 Quantitative survey
The sample for the formal survey was made up of 25 to 30 households drawn
randomly from the list frame, with the gender of selected household heads in the same
proportion as in the overall population. Recognizing the variable structure and strategies
of households (Guyer and Peters 1987; Vaughan 1985; Wilk and Netting 1984), the social
and physical unit in this study consisted of members who performed activities that
sustained the unit and/or regularly took meals together, irrespective of permanent
residence. It is necessary to further note that much of the survey analysis distinguishes
between jointly-managed and female-headed households, since this latter group tends to
constitute a relatively disadvantaged segment of the rural population. Approximately 28
percent of the sampled households were managed by single women, most of whom were
either divorced or widowed. Households where the woman was considered the de jure
legal and customary head (widowed, divorced, separated, or unmarried women) or the de
facto head (where the male was absent for more than half the time) were considered
female-headed (cf Kennedy and Peters 1992).
The final data set for the formal survey included 427 households comprised of
2,205 individuals. Designed to capture intralocality variation, this survey instrument
collected information in three critical areas; the household asset base (family labor, land.
63
and animals), access to non-agricultural income, and the detailed use of resources from the
protected areas. These data were coded and entered in the SPSS (Statistical Package for
the Social Sciences 1998) software system in Malawi, then analyzed at the University of
Arizona using this same system.
A.4 2 KEY VARIABLES; INCOME AND NATURAL RESOURCE UTILIZATION
Much of the analysis focused on two variables that define the primary research
questions presented above. The first variable was that of household income, expressed in
terms of both direct and non-direct household income. The second variable was that of
household utilization of protected area natural resources, broken into seven major
categories; food, fijelwood, fiber, tools, medicinal plants, and both wood and thatch
construction.
Income estimates from all sources (direct and indirect) were compiled for each
individual household and converted into per capita values. Direct income was defined as
the monetary compensation received by household members for wage labor, sale of
agricultural products, remittances, and a variety of oflf-farm income generating activities,
including sales of protected area natural resources. Indirect income was the composite of
activities that have utility to the household, each being assigned an estimated value derived
from local retail market prices. The income from these activities included the value of
goods intended for consumption within the household, such as subsistence maize
production and protected area resource utilization.
64
For assessment of both income and utilization, the methodology required annual
household totals for categorized items. These were quantities (measured by weight) of
agricultural outputs and protected area resources in some instances, and monetary values
in others. Requesting annual totals was often contrary to local accounting methods. To
calculate reliable annual totals, the modular design of the questionnaire encouraged
interviewers to collect data in local units over time periods expressed by the respondent.
These locally expressed quantities and time frames were then converted to standard
values. All monetary values are reported in Malawian Kwacha (MK), where 1 USD =15
MK during 1996 and 1997.
A.4.3 VALUATION IN THE MALAWIAN CONTEXT
The objectives of the study warranted assigning economic values based on a "total
value of production" approach. This decision was based on the limited alternatives to
smallholder agriculture and the essential role the collection of protected area resources
plays in the study areas and in Malawi as a whole. This results from a lack of alternative
labor markets, lack of arable land, limited technology, and limited access to agricultural
inputs that have led many to pursue supplemental income through entrepreneurial
activities.
In Malawi, the agriculture sector provides self-employment to 92% of the
population. In Africa, only Burundi and Rwanda report a lower proportion of total
population in urban areas than Malawi's 14% (World Bank 2000a). There are essentially
no local labor alternatives to subsistence agriculture.
65
The vast majority of agricultural production and natural resource collection is
conducted with manual technology on limited land holdings. It is estimated that 40-55% of
smallholders in Malawi have land holdings of less than 0.7 ha. The Malawian government
has determined that this amount of land is inadequate to produce enough food for
subsistence needs (Chilowa 1998). Becker (1990) demonstrated that restricted land
availability in Malawi forces rural families to diversify their labor supply to meet off-farm
opportunities that might finance subsistence requirements rather than investing in new
technologies or replacing farming income all together. Furthermore, subsistence activities
that involve protected area resources (i.e. household fiielwood collection) in support of
home production are not market replaceable under current land and labor conditions in
Malawi (Engberg, Sabry, and Beckerson 1987).
The impact of agricultural inputs on returns to labor in Malawi is also limited. The
use of agricultural inputs, free or purchased, is on the decline in Malawi. Following the
removal of farm input subsidies in the early 1990's, the smallholder sector decreased its
use of fertilizer by almost 75%, purchased only enough hybrid seed to plant 7% of the
maize area, and reimbursed creditors for only 20% of the inputs purchased on loan (Carr
1997).
There are few or no opportunities for substitution within existing Malawian land
and labor conditions, nor are there alternative technologies or agricultural production
methods that might improve agricultural returns. Furthermore, protected area resource
utilization is an irreplaceable element of household production. Therefore, the valuation
66
employed in this research was based on total (rather than net) production and protected
area resource utilization.
A 4 4 SPECIES IDENTIFICATION
The Forestry Research Institute of Malawi (FRIM) and the National Herbarium
provided indispensable assistance with local species names. A field botanist was present at
all 136 resource assessment plots, working with local inhabitants to confirm all local
names for each species identified. These plots were not spatially nor temporally sufficient
to capture all species (and their products) identified on village and household surveys.
National experts, and Malawi's rich tradition of gathering ethnographic biological
information, proved invaluable where local confirmation was not possible. The extensive
plant dictionary of Binns (1972) was used for confirming Latin names. The works of
Williamson (1975), Morris and Msonthi (1996), and Morris (1990) were essential for
addressing gaps in local plant, fiaiit, and mushroom descriptions, and provided the
foundation for evaluating species use.
Nomenclature for mammals was taken fi^om Ansell and Dowsett (1988), birds fi-om
Benson and Benson (1977) and McShane (1985). Nomenclature for insects was taken
from Sweeney (1970), and fish from Ribbink et al. (1983) and Tweddle and Willoughby
(1979). These were all was supported by detailed ethnobiological descriptions of use
(CCAM, 1992; Hayes 1978; Kelly 1993; Morris 1998).
A summary of the species by type and lifeform is contained in Table A.2. Of the
694 species encountered during the study, 101 local identifications, accounting for 4.6%
67
of the total protected area species used by households, could not be verified. The majority
of these were annual forbs and insects physically unavailable at the time of data collection.
TABLE A.2. NUMBER OF SPECIES AND
HOUSEHOLD OBSERVATIONS OF USE.
Wild Use
No. of
Domestic Use No. of
Lifeform
Species
Lifeform
Species
Mammal
Bird
Fish
Insect
Honey
*Mushroom
Tree
Shrub
Climber
*grass
Forb
34
9
18
28
1
9
235
54
39
46
141
Animal
Field crop
Fruit tree
Wood tree
10
44
15
ii
Total species
614
Total species
80
Household
Household
observations
12,604
observations
3,720
of use
of use
^several of households could not provide individual
species names for mushrooms and grasses.
Due to cultural and linguistic variation in protected areas and nearby villages, a
number of local names were obtained for each species identified, the vast majority of
which were available in species dictionary (Binns 1972). We therefore limited reporting
here to the most common name cited during the data collection process.
68
A.4.5 QUANTITY AND WEIGHT ISSUES
Throughout the entire data collection period, local market surveys were carried
out for the purpose of converting local units of measure into kilogram weights for a wide
range of both domestic commodities and public land resources. Farming families rarely
recalled agricultural harvest or forest utilization in standard units of measure. It was
therefore necessary to identify common local units of measure, and then determine an
average size for each of those units. Some local units of measure had a standard (e.g.,
"No. 10 Plate"), whereas others (e.g., "basket") were much more variable. During the
pilot study measurements of each unit were made in a number of villages and markets to
determine a typical unit. Examples of each local unit were then purchased for reference
during data collection. Weighing scales were used opportunistically during household
interviews and during the market surveys. However, it was not possible to capture all
commodities in all possible local units of measure. To compensate, conversion factors
among local units were determined by physical comparison tests, using a variety of
commodities (Table A.3). "Between-unit" conversion factors are valid only where the
volume of an individual item does not overwhelm the volume of the local unit (i.e. "plates
of papayas").
These factors were used to calibrate physical weights obtained during the market
survey in order to create a standard conversion table of local measures to kilograms. The
conversions for agricultural crops are listed in Table A.4. Protected area species were
more problematic because of limited samples, or no samples for rare or out-of-season
items.
69
TABLE A. 3 VOLUME CONVERSION FACTORS AMONG LOCAL UNITS OF
MEASURE.
Local Unit Med. No. 10 Basin Lichelo Basket 50 kg 90 kg DenguOxCart
of Measure Plate Plate
Bag Bag
Medium Plate
mm
No. 10 Plate
1.77
4.33
0.57
•
6.20 15.15 26.62 47.91 75.42 633.57
2.45
3.50
8.56 15.04 27.07 42.62 358.01
Basin
0.23
0.41
M
1.43
3.50
6.14 11.05 11.05
Lichelo
0.16
0.29
0.70
mm
2.44
4.29
Basket
0.07
0.12
0.29
0.41
50 kg Bag
0.04
0.07
0.16
0.23
•1 1.76
0.57
90 kg Bag
0.02
0.04
0.09
0.13
0.32
Dengu
0.01
0.02
0.06
0.08
0.20
92.83
7.73 12.17 102.22
3.16
4.98
41.81
1.80
2.83
23.80
0.56
»
1.57
13.22
0.35
0.64
wm.
Ox Cart
0.002 0.003 0.01
0.01 0.02 0.04 0.08
*"lichelo" is a basket lid, and "dengu" is an extra-large basket.
8.40
0.12
Wm.
To compensate, resources extracted for food and medicinal purposes were
grouped in size-equivalent classes based on other products identified by respondents as
sharing similar size, shape and consistency. Physical descriptions of each protected area
product were obtained in the field and then crosschecked with the ethnoecological
literature. Average weights of plant parts extracted for medicinal purposes were derived
from data collected in 28 key respondent interviews with healers. Table A. 5 provides
kilogram conversions for such size-equivalent class categories for food and medicinal
items, with an example of protected area species that corresponds to each. Table A.6
provides live weights of livestock and wildlife species obtained fi"om the literature
(Cockbum 1982; Hayes 1978; Mason and Maule 1960; Owen 1975; Wollnye/a/. 1998).
TABLE A.4 WEIGHT CONVERSIONS AND PRICES FOR AGRICULTURAL CROPS
Latin Nuinc
Abelmoxchus e.iculctUus (I-.) Mociich
Allium cepa 1,.
Ammas comosus (1,.) Mcrr.
Arachls hypof>aea 1,.
Artocarpus
Mcrr.
Brassica chinensis I,.
lirassica Hopii.t 1.. vnr olcil'crn DC
Brassica napiis 1,. vnr esculcnta DC
Brassica oleracea L. var bullata DC
Cajanus cajan Millsp.
Camellia sinensi (I..) Kunt/c
Capsicum annuum 1,.
Carica papaya 1,.
Casimiroa eduiis l,a Lluvc & l,cx.
Ciitvs aurantifolia (Christin.) Swingle
Cilnis limonium (L.) Bumi. 1".
Cilivs sinensis (L.) Osbcck
Cocos nucifera L.
Colocasia esculenta (L.) SchotI
Cucumis salivus 1,.
Cucurhita maxima Ducli. cx l,am
Cucurhilapepo I,, (leaves)
Daucus carola I,.
Dioscorea spl.
lileusine coracana (L.) (lucrtn.
Glycine max (L.) Men
Gossypium harhadcnse /G. hirsutum I-.
Iklianthus annuus L.
Ipomea batatas (1,.) l.am.
I.ahtah purpureus (1,.) Sweet
i-oeiii/linglisli Name
A
C.
C
D
li
1
B
!•
11
-- weights (kg) per local unit of measure --
J
Price
MK kg'
llicrere/okra
0.06
I.I
2.0
4.8
6.9 16.9 29.8
54
84
708
21.32
nnye/.i/onion
0,17
1.2
2.2
5.3
7.5
58
92
771
6.50
nnniisi/pincap|)lc
0.81
1.8
3.2
7.8 11.1 27.1 47.6
intedza/groundiuils
0.05
0.7
1.3
3.1
4.5
11.0 19.3
35
55
460
4.1
3.6
5.8 14.3 25.0
45
71
596
4.83
5.1
40
62
524
4.30
20 31
265
83 131 1,099
4.25
4.32
jaklhiit/jacklruil
18,4 32.4
86 135 1,134
25.00
7.90
14.53
0.40
lanapozi/Chinese cabbage
0.06
0.9
1.7
swidi/Swedish luniip
Oil
0.8
1,5
mpini wotuwa/rape
kabiclii/cabbagc
0.05
1.08
0.7
3.1
nandolo/pigeon pea
0.4
1.7
1.5
2.6
1.8
2.6 6.3 1 1 1
7.5 10.8 26.3 46.2
6.3
9.0 22 1 38.8
chayi/tca
0.3
0.6
1.5
2.1
5.1
9.0
70 110
923
16
26
215
9.78
2().(K)
44
69
584
14,75
tst»bola/l'eppcrs
0.09
0.9
1,6
4.0
mpapaya/papaya
0.96
1.7
3.1
7.6
10.8 26.4 46.4
83 131 1,104
1.92
masuku a chi/ungii/Mexican apple
0.13
2.1
3.8
9.2
13.2 32.3 56.7 102 161 1,350
2,35
ndimwe/llinc
0.07
1.8
3.8
7.8
11.2 27.4 48.0
86 136 1,144
3,32
ndimu/lenion
0.20
1.9
3.4
8.3
11.9 29.1 51.1
92 145 1,217
5.20
malaanji/orangc
0.18
2.2
3.8
9.4
13.4 32,7 57.5 103 163 1,368
6.95
nkoko/coconut
0.27
0.8
3.5
5.0
0.2
1.4
0.3
0.7
1.0
2.0
3.5
8.6
koko/coco yam leaves
mankhaka/cucunibcr
0.20
thenged/a/pumpkin
0.31
mussa/pie pumpkin (leaves)
5.7
12.5 22.0
14.0 24.5
12.2 21.4
38
4.3
2.4
12.3 30.0 52.6
8
7.48
102
12
95 149 1,253
10.11
58
91
767
2.71
4.3
8
12
102
3.50
70 110
920
6.89
2.1
5.2
7.5
0.3
0.7
1.0
6.3
9.0 22.0 38.6
2.4
0.08
1.5
2.6
mpama/yam
0.29
1.9
3.4
8.4
12.0 29.4 51.7
1.4
2.5
6.1
8.7 21.2 37.2
12.0 29.3 51.5
93 146 1,230
67 106
11.56
93 146 1,227
35
19.07
20.(K)
2.68
1.9
3.4
8.4
thonje/cotton
0.1
0.1
0.2
sanifulawa/sunllower
1.9
3.3
8.1
11.6 28.3 49.6
89 141 1,181
2.0
3.5
8.7
12,4 30,3 53.2
96 151 1,266
1.5
2.6
6.3
9.0 22,1 38.8
mbalata/swcct potato
guza/liyacinth bean
0.30
0.8
1.5
1.55
886
s»ya beati/st)ya bemi
0.3
1.20
18.3 32.2
1.2
0.2
canol/carrol
lipoko/ringcr millet
508
61
3
70
4
no
923
5,42
11.58
-J
o
TABLE A.4 - Conlintied
l.atin Name
l.acliica saliva I.,
Lagenaha siceraria (Molina) Standley
Lycopersicon esuleiUum Mill.
Macadamia integrifoHa Maiden and Betche
Mains domestica Borkh.
Slangifera indica 1„
Manihot escuhnta Grant/,
Manihot esculanta Grant/
Xhntx alba I,.
*\lu.sa paradisiaca ' M. sapientum L.
^NicoUana tabacum l„
Oiyza saliva l„ (Giram)
PemUsc'tum americanum (L) K. Schuni,
Perxea americana Mill.
Phaxeolus vulf'aris 1,.
Physalis peniviana 1,.
Pisum sativum L. var. an'ense (iams.
Prunus persica (1,.) Stokes
Psidium guajava L.
XSacchamm officinanim 1.,
Solanum melongena 1,.
Solanum tuberosum L.
Sorghum bicolor (1. ) Moench
Stizolobium alerrinum I'ipcr and Tracey
Vigna radiata (I„) Wilc/.ek
Vigna unguiculata (1„) Walp,
Voandzeia subtenanea Thou,
y.ea mays L,
I)
(i
I.ocal/lin(!lish Name
saladi/Iettucc
A
0.33
B
0.8
G
1.5
3.6
5.2 12.7 22.3
11
40
1
63
J
MK kfi'
532
5,10
mphonda/gourd
0,31
1,2
2.1
5.2
7.5
18.3 32.2
58
91
766
2,71
mapwctckere/tomato
0,13
2.0
3.5
8.7
12.4
30.3 53.2
96 151 1,267
12,65
1.9
3,4
8.3
11.9 29.1 51.1
92 145 1,217
13,01
macadamia/macadamia
!•;
!••
apulo/apple
0.14
1.7
3.0
7.5
10.7 26.1 45.8
82 130 1,090
4,07
mango/mango
0,27
1.8
3.2
7.7
11.1 27.0 47.5
chinanga/cassava rcK)!
0.01
2.1
3.7
9.0 12.9 31.5 55.3
85 135 1,131
99 157 1,316
3.26
chinanga/cassava leaves
0.25
0.2
0.3
0.7
1.0
1,3
2.3
5.7
8.2 20,0 35,1
2,3
0,1
4.8
0.2
10,1
0,4
mapulesi/mulberrj'
ntlux:hi/l)anana
fodya/tobacco
4,3
8
12
102
3.55
63 100
836
7.58
8.44
21.75
mpungn/ricc
2,7
uchewere/bulrush millet
1,5
14.5 35,4 62,1 112 176 1,479
0,6 7.6 21.1 38 60
502
4.8 11.6 16,6 40,7 71.4 129 202 1,701
2.6 6.4
9.1 22,3 39.2 7! 111
933
ovocado/av(x:ado
0.12
0.10
2,4
0.85
2.09
14.76
9.26
1,9
3.4
8.3
11.9 29,2 51,2
khwanya/haricot or kidney bean
1.4
2.5
6.0
8.7 21.1 37.1
67 105
92 145 1,219
884
6.67
jamu/g(X)seberr>'
1.4
2.5
6.2
8,9 21.7 38.1
69 108
906
6.99
tuware/pea
1.3
2.3
5.8
8.2 20.1 35.3
64 100
841
13.00
15.09
pichesi/peach
0.15
1.8
3,1
7.7
11.0 26.8 47.1
85 133 1,121
4.52
guwava/guava
0.16
1.6
2.9
7.1
10.1 24.7 43.5
78 123 1,034
3.89
nzimbe/sugar cane
0.50
magringana/egg plant
0.10
1.2
2.1
5.1
mbatata kachewere/potato
0.11
1.7
3.0
7,5
82 130 1,091
6,42
mapilo/sorgiium
chitedze/velvet bean
2.1
1,6
3.8
11,83
mphod/a/gram bean
1,4
2.5
9,2 13,2 32.3 56.7 102 161 1,350
6.8 9,7 23.7 41,6 75 118
989
6.0 8,7 21.1 37,1 67 105
884
khobwe/cowpea
1,5
2.6
6.5
9,3 22.7 39,8
72 113
948
13.11
nzama/liambarra groundnut
0,6
1.1
2.8
4,0
31
404
15.63
chimanga/mai/.e
1.9
3.3
8.1
90 142 1,191
2.55
54
2.8
7.3
17.9 31.4
10,7 26.1 45.8
9.7
17.0
11,7 28.5 50,0
57
89
48
454
1.50
748
16.93
6.55
9,80
lAwal units of measure: A = individual item; B = medium plate, C = No. 10 Plate; 1) = basin; li = lichelo; I-' = basket; (i = 50kg bag; 11 = 90kg bag; I =
dengu; J = ox cart * M. puradisiaca and M. sapientum are also locally tabulated in "nikoko" or "bunches" (12.25 kg / mkoko). t N- tabacum is also locally
tabulated in "bundles" (100.0 kg / bundle), and "boards" (12.0 kg / board). J S. otVicinarum is also kx;ally tabulated in "headloads" (9.0 kg / headload).
TABLE A.5. WEIGHT CONVERSIONS AND PRICES FOR PROTECTED AREA FOODS AND MEDICINES, GROUPED
INTO CLASSES BY AGRICULTURE PRODUCT SIZE-EQUIVALENTS.
Size
Specie.s lisample for Hach Size licjuivalenl
1'Equivalent
Latin Species Name
l.ocal/l-nglish Name
A
n
C
1)
I-
!•
(i
11
1
.1
-- weights (kg) per l(x;al unit of measure -cl«si>si>A)lack-jack
0.2
0,3
mpungaziwe/red rice
2.7
4.8 11.6
mpalapala/snowberry
1.3
2.3
5.7
8,2 20,0 35,1
63 100
836
1..33
1 40
hard bean
fleshy bean
Fif;na spp.
chitambeAvild beans
sonkwe/Deccan hemp
mateme/elephant orange
grain
mulberry
goo.sebcrry
masuku
Hibiscus cannahinus I,.
guava
Siiychnos spinosa l,am.
mango
Myrianthus holstii l:ngl.
C(xx)nut
Adamonia digitata 1,.
sweet potato llabenaria waller! Reichb. f
papaya
Telfairiapedata (Sims) 1l(X)k.
breadfruit
Treculio africana Decne.
peppers
0.7
1.0
102
1 55
16,6 40.7 71,4 129 202 1,701
0,37
2,4
4.3
8
MK kg'
Otyza lonffistaminala Chev. and Roehr.
Sccurinega viima (Roxb. ex Willd.) Huill.
S}'Z}'gium amialum 1 kx:hst. ex Krauss
IJapaca kirkiana MClll. Arg.
Tamarimhts imUca 1,.
•fixxl leaves Hidem pilosa L.
O.Ol
Price
12
1.4
2.5
6.2
8,9 21,7 38,1
69 108
906
masukuAvild kK|uat
0.03
1.5
2.7
6,6
9,5 23.2 40.7
73 115
969
1 31
bwemba/tamarind
0.05
0.7
1,3
3,1
4,5
35
460
1.38
0.05
1.5
2.6
6.3
0.9
1,6
4,0
0.16
1.6
2.9
7.1
10.1 24.7 43.5
70 110
923
584
44 69
78 123 1,034
3 43
0.09
makwakwa
0.27
I.K
3.2
7.7
II.1 27.0 47,5
85 135 1,131
1,12
malaml>eAiaobab
0.27
0.8
1.4
3.5
5.0 12,2 21,4
508
0.93
chikandeAvild yam
0.30
2.0
3.5
8.7
12.4 30,3 53,2
96 151 1,266
0.25
1.7
3.1
7.6
10,8 26,4 46,4
mchisu/water berry
11.0 19.3
9,0 22.1 38.8
14.0 24.5
5,7
38
55
61
1.09
1,22
matundu/oy.ster nut
0.96
83 131 1,104
0,29
nJale/African breadfruit
11.0 11,0 11,0 22.0 22.0 33.0 58,0 104 164 1,380
0,09
0.04
1,47
mushrixims
general mushrtxmis
bowa/nkhowani/liwawa
insects
TemiUdae spp.
ngumbiAvinged termite
thoney
honey
uchi/ng'oma/lioney
0.3
0.6
1,5
2.1
5.1
9,0
16
26
215
0.2
0.4
1,1
1,5
3,8
6,6
12
19
158
8.04
13.90
small fish
malemba/smull cypriaid
0.09
chambo/chambo
0.90
2.7
3.6
7.2
12,6
16,7 40,9 71.9 129 204 1,711
18,0 44,0 77,3 139 219 1,840
1.85
medium fish
mlaniba/sharpttx>th catfish
3.58
1.8
3.6 10,7
17.9 43,8 76,9 138 218 1,830
1,40
mpoloni/carrot tree
0.01
0.2
0.3
0,7
1.0
2,4
4,3
8
12
102
changaluche
0.11
0.5
0.9
2,1
3.0
7,3
12,9
23
37
307
20,30
mchalamira/lieartwcxxi
0.01
0.2
0.3
0.7
1.0
2,4
4,3
8
12
102
58,44
chibowachamuchitsa
0.04
0.3
0.6
1.5
2,1
5.1
9,0
16
26
215
58,44
liarhus spp.
Oreochromis spp.
large fish
CInrias gnriepinus Burchcll
*med. leaves Steganolaenia amiiacea 1 lochst.
•med. rixMs Zanha africana (Radlk.) lixell
•med. bark
Cassia ahbreviala Oliv.
med. fungi
Coprinus africanus Pelger
4.8 n,7
1.11
58,44
1-ocal units of measure; A = individual item; B = medium plate; C = No. 10 Plate; 1) = basin; li = lichelo; 1* = basket; C) = 5()kg bag; 11 = 90kg bag; 1 = dengu.
•leaves, r(X)ts, bark used for medieinal puqioses and leaves u.sed for HHXI are also collected and tabulated in small "bunches" or "bundles" (0.25 kg / bundle),
t the local units for honey were litre bottles (1.44 kg / litre)
-J
N)
73
TABLE A.6 LIVE WEIGHTS AND PRICES OF ANIMAL SPECIES.
Latin Species Name
Live
Local/English Species Name Weight Price
kg M K k g '
Domesticated Animals
Af:as spp.
Bos indicus Linnaeus
Capra spp.
Cavia porcellus Linnaeus
Gallus domesticus (Linnaeus)
Lepiis spp.
Ovis spp.
Siiis spp.
Treron spp.
bakha/duck
ng'ombe/cattle
mbuzi/goat
mbila/guinea pig
nkhuku/chicken
kalulu/rabbit
nkhosa/sheep
nkhumba/pig
nkunda/pigeon
2.7
195.5
19.4
0.9
3.6
1.8
40.0
125.0
0.5
14.72
11.46
12.45
17.76
10.70
16.23
10.00
3.01
14.75
59.0
0.5
10.0
8.2
0.9
0.5
0.9
3.6
2268
272.2
226.8
20.4
272.2
0.5
6000
0.5
40.8
0.2
0.2
63.5
90.7
4.1
0.5
11.3
77.1
136.1
18.1
680.4
680.4
0.9
68.0
3.80
5.49
4.75
4.75
5.49
5.49
5.49
5.49
1.05
2.90
2.90
4.75
2.90
5.49
0.10
5.49
3.80
5.49
5.49
3.80
3.80
5.49
5.49
4.75
3.80
2.90
4.75
1.67
1.67
5.49
3.80
Wild Mammals
Aepyceros melampus Lichtenstein
Atelerix albiventris Wagner
Cercopithecus albogularis Sykes
Cercopithecus pygerythrus F. Cuvier
Cricetomys gambianus Waterhouse
Crocidura hirta Peters
Heliophobius argentocinereus Peters
Herpesies ichneumon Linnaeus
Hippopotamus amphibius Linnaeus
Hippotragus equinus Desraarest
Hippotragtis niger Harris
Hystrix africae-australis Peters
Kobus ellipsiprymnus Ogilby
Lepus saxatilis F. Cuvier
Loxodonta africana Blumenbach
Muridae family
Papio cynocephalus Linnaeus
Paraxerus cepapi A. Smith
Petrodromus tetradactvlus Peters
Phacochoerus aethiopicus Pallus
Potamochoerus porcus Linnaeus
Procavia capensis Pallus
Prondolagus rupestris A. Smith
Raphicerus sharpei Thomas
Redunga arundinum Boddaert
Sigmoceros lichtensteinii Peters
Sylvicapra grimmia Linnaeus
Syncerus caffer Sparrman
Taurotragus oryx Pallas
Thryonomys swinderianus Temminck
Tragelaphus scriptus Pallus
nswala/^pala
kanungu/four-toed hedgehog
nchima/blue monke\'
pusi/vervet monkey
ngwime/giant rat
fvvilif\vili/red musk shrew
nifuko/silver\' mole rat
nyenja/large gre>' mongoose
mvTiu/hippopatamus
chilemb we/roan antelope
mphalapala/sable antelope
nungu/porcupine
chuzu/waterbuck
kalulu/scrub hare
njobvu/African elephant
mbewa/category of mice
nyani/yellow baboon
benga/bush squirrel
sakhwi/4-toed elephant shrew
kaphulika/wart hog
nguluwe/bush pig
mbira/rock hvTa.\
kafiimbwe/red rock hare
mtungwa/Sharpe's grysbok
mphoyo/southem reedbuck
nkhoziA^ichtenstein's hartebeest
gvvape/grey duiker
njati/African buffalo
tsefu/eland
nchezi/cane rat
mbawala/bushbuck
Birds
Anthreptes colaris (Vieillot)
Centopus spp.
Francolinus afer (MUller)
songue/collared sunbird
nkuta/coucal
nkhwali/red necked francolin
0.2
0.5
0.9
5.49
5.49
5.49
74
TABLE A.6. - Continued.
Latin Species Name
Live
Local/English Species Name Weight Price
kg M K k g '
Xeciariniia senegalensis (Linnaeus)
songwe/scarlel-chested sunbird
\'umida meleagris (Linnaeus)
nkhanga/helmeted guinea fowl
Streptopelia deciphens (Hartlaub and Finsch) njiwa/mouming dove
Treron australis (Linnaeus)
nkunda/African green pigion
Turtur chalcospilos (Wagler)
katundula/emerald-spotted wood dove
Fish fand a freshwater crab"*
Bagrus meridionalis GOnther
Barbus spp.
Barbus brevicauda Keilhack
Clarias gariepinus Burchell
Engraulicypris sardella Gttnlher
Hippopotamyrus discorhynchus Peters
Labeo mesops GOnther
Marcusenius macrolepidotus Peters
Marcusenius nyasensis Worthington
Opsaridium microlepis Ganther
Oreochromis spp.
Rhamphochromis spp.
Serranochromis robustus GOnther
Synodontis njassae Keilhack
Potomonautes monttvagus Chace F.
kampango/kampoyo cat fish
matemba/small c\priaids
kadyakoio
mlamba/sharptooth cat fish
usipa/Iake sardine
ngundamwala/Zambezi parrotfish
nchila/African carp
mphuta/buildog
samwamowa
mpasa/lake salmon
chambo/tilipia
ncheni/tigerfish
nkakafodyaA'ellow-bellied bream
chikolokolo/Malawi squeaker
nkhanu/freshwater crab
0.2
1.4
0.5
0.5
0.5
5.49
5.49
5.49
5.49
5.49
3.6
0.1
1.7
5.2
0.05
0.5
1.4
0.5
0.5
2.3
0.8
0.8
2.0
0.5
0.05
•0.72
•0.72
*0.72
*0.72
*0.72
*0.72
*0.72
*0.72
*0.72
*0.72
*0.72
*0.72
*0.72
*0.72
1.85
*The 0.72 MK kg"' fish price was used for all but Liwonde National Park, where market
surveys showed a significantly higher price of 4.47 MK kg"' due high demand from and
easy access to nearby urban customers.
Assessing the quantity of biomass extracted for wood (fiiel, construction, and
tools) fiber, thatch, and handicrafts required a number of conversion factors. Wood
volume to mass conversions were obtained from an urban biomass fijel study (Openshaw
1997), while all other local unit measurements were derived fi-om the average of recorded
measurements across all four protected areas in this study (Table A.7). Direct physical
measurements were adequate to estimate quantity where a one-to-one relationship existed
between biomass extracted and biomass sold or consumed (food, thatch, and fiielwood).
Estimating the quantity of biomass extracted for medicinal purposes, fiber, tools, and
handicrafts required additional information about the quantities extracted (often by plant
75
part) used to create a final product. Though the random sample of the formal survey
captured some of these specialized uses, it was necessary to conduct key respondent
interviews to obtain more detailed estimates. This included interviews with 18 healers, 10
timber cutters, and 27 artisans making utilitarian handicrafts (baskets, mats, hand tools,
etc.).
TABLE A.7. CONVERSION UNITS FOR WOOD, THATCH AND FfflER.
Local Unit of Measure
kg Local Units of Measure Conversion Factors
*Solid cubic meter
667
9.7 headloads per ox cart
*Stacked cubic meter
367
25 wood poles per ox cart
Headload of fiielwood
3.6 wood poles per headload
32.6
Headload of thatch
23.6
12 bamboo poles per headload
Headload of bamboo
5 thatch bundles per headload
21.3
Wood pole
9.1
5 fiber bundles per wood headload
Bamboo pole
48.5 bundles per ox cart
1.4
Palm frond
1.3
tHerbaceous fiber
4.7
bundle
tPalm fiber bundle
4.0
t Agave fiber bundle
3.9
•{"Woody fiber bundle
6.5
* based on air dry wood (15% moisture content) from Openshaw (1997)
t this is the weight of the biomass extracted in order to make a bundle of fiber.
Though wood utilization was generally reported in "headloads" or "poles," a
number of observations were reported as "whole mature trees" either on farmland or
within a protected area. To estimate the biomass of a standardized individual tree, the
average height (H) and diameter at breast height (DBH) for these species was calculated
from resource assessment data, and then used in single tree volume equations obtained
from the literature (Table A S). Recent empirical research by Abbot, Lowore and Werren
(1997) in Malawi provided single tree volume equations for Miombo woodland settings.
76
addressing the majority of species identified in the study that were extracted whole.
Eucalyptus and Pinus spp. and Colophospermum mopane were also reported as being
extracted whole.
TABLE A.8. SINGLE TREE VOLUME EQUATIONS
Type
Single Tree Volume Estimation Volume
(and data source)
Miombo
(Canopy Species)
DBH
cm
H
m
logio V = -3.98 + 2.6 logio D
(Abbot, Lowore, and Werren 1997)
19.1 11.6
Miombo
logio V = -3.87 + 2.43 log,o D
(Understory Species) (Abbot, Lowore, and Werren 1997)
18.7 10.8
Eucalyptus spp.
V = 0.000032141 D*
(Shiver and Brister 1992)
» H'
Pinus spp.
logio V = -4.674 + 1.8644 logio D + 1.3246 logio H
(Malimbwi and Philip 1989)
14.8 21.2
21.2 20.1
Colophospermum
mopane
Dried biomass (kg) = 0.85 * (-326 + 31.3 (D))
21.5 12.2
(Mushove et al. 1995)
V = total overbark volume (m3); D = diameter at breast height (cm), H = height (m)
A.4.6 PRICE
Items sold by households were valued directly by the retail price obtained fi'om
their sale in the market. Households valued subsistence production and protected area
resource utilization on the price they would have paid if the items had been purchased
(Chibnik 1978; Mellor 1966). These items were therefore assigned value based on retail
prices reported by households or obtained during the local market surveys. Prices were
captured when products were at market and averaged over the year.
The average retail prices for food and medicinal plants are reported in the final
column of Tables 4 and 5. The average retail prices for meat (Table A.6) were based on
77
live animal weights of species which were sold whole, or for the saleable meat weights for
larger species butchered prior to sale (Fa and Purvis 1997; Martin 1984). Prices for all
major wood, fiber, and handicraft categories are listed in Table A.9.
Local market retail prices for protected area food, fuel, thatch, and fiber products
tended to be based more on the measurement unit of sale, regardless of the species sold
(i.e. 5 MK per plate, 4 MK per headload, etc.). It was therefore possible to use the local
market retail price of the closest substitute for those protected area resources for which a
local price was unavailable (Godoy, Lubowski and Markandya 1993).
Key respondent interviews with healers revealed some variability in price among
species used for medicinal purposes. However these data were not sufficient to assign
differential prices by species. A price estimate generalized across species was possible by
relating the revenue generated from the practice of healing to the weight of the plant part
used to generate that revenue. A similar strategy was used for major utilitarian handicrafts
(brooms, mats, baskets, wooden and bamboo hand tools), curios made for tourists, and
timber. The pricing differential among species was captured through a cost structure
assigned by the Malawian Forestry Department based on a royalty or volumetric
assessment (Table A.9).
TABLE A ,9. PRICES FOR BIOMASS USED TO CREATE WOOD AND HANDICRAFT PRODUCTS.
Vegetation Type
JRoyalty
Species lixample for l-ach Vegetation Type
--MK--
Latin name
!,(x;al/l'nglish Name
Whole Tools/ Pole Heud- Tiber 1 lundi
Tree Timber
load
-craft
-- price (MK kg') --
Miombo canopy
15.00 Pcricopsis anffolensis (Dak.) van Meeuwen
mwanga/afromiosia wo(xl
0.10
0.63
0.21 0.13
0.31
0,63
Miombt) canopy
50.00 Vilex (Joniana Sweet
ntonogoli/black plum
0.33
2.09
0,70 0.13
0.31
2,09
85.(X) Termituilia sericea Hurch. cx DC.
1(K).(X) I'terocarpus an^olenslx DC.
nuphini/silvcr tcrminalu
0.57
3.55
1.18 0.13
0.31
3.55
Miombo canopy
inlombwa/African teak
0,67
4,17
1.39 0.13
0.31
4.17
Miombo canopy
170.00 Khaya nyasica Stapf
mbawu/rcd mahogany
Miombo understory
15.00 PiUosUgma ihonningii (Schumach.) Milne-Redh. chilimbc/camelfoot
l(K).00 Apo(fyles (iimiJiala H. Mcy. ex. Am.
mu/a/u/white jwar
Miombo canopy
Miombo understory
i.Ol
6.10
2.02 0.13
0.31
6.30
0.13
0.84
0.28 0,13
0.31
0,84
O.W
5.62
1.87 0,13
0.31
5,62
0,55
Miombo shrubs
15.00 Hyrsocarpiis orientalis (Baill.) Hak.
kamenenambu/i
0.06
0.06
0.06 0.13
0,31
Mopane
50.00 C. mopane
lsanya/l)utterlly-tree
0.17
1.06
0.35 0.13
0.31
1.06
mkungud/a/Mulanje cedar
bulugamu/tlcKxled gum
0.71
0.10
4.45
0.66
1.48 0.13
0.22 0.13
0.31
0.31
4,45
0.66
paini/kesiya pine
0.22
1.35
0.45 0.13
0.31
1,35
0.45 0.11
0.50
1,25
•Mulanje ccdar
*Eucalyptus spp.
*Pinus spp.
Palm
tBamb(X)
155.72 fV. notUJJora
10.59 hiicalyplus grandis W. 1 till ex Maiden
48.35 P. kesiya
40.00 Borassus aethiopiim Mart.
1.00 Oxytenanlhera ahyssinica (A. Rich.) Munro
mla/a/Deleb fan palm
msungwi/common bamb(X)
0.70
0.70
0.70 0.70
Agave
n/a Agave sisalana (lingl.) I'erriiie
khonje/sisal
Reeds
n/a Phragmites mauritianus Kunth.
mbango/matete/reed grass
0.11
Thatch
n/a Hyparrhenia nyassae (Rcndle) Stapf
kamphe/bush thatch grass
0.11
I Icrbacious crafts
n/a Si</a acuta l^umi. f
masache/brtxmi plant
0.11
1,25
0,51
1,25
1,25
1,25
0.43
0.55
•The Forestry Department prices species commonly used to make timbei by the cubic meter, here converted to a "standard tree" as defined in Table 7.
tl^amkx) royalties arc charged by individual poles rather than whole plants.
J The royalty for individual mnxl and bambw> species is assigned annually be the Malawian l-'orestry Department for sales to timber and carpentry amipanies.
-J
00
79
A.5 DATA SUMMARY
By integrating data collection methods for the assessment of livelihood security and
protected area resource utilization, aggregated species-level results can be presented from
the perspective of the household. In order to normalize for variance in household size,
most results are reported "per capita," and averaged to reflect each overall protected area.
A.5 1 DOMESTIC PRODUCTION
Domestic production included field crop and fruit trees cultivation, wood lots for
flielwood and timber, and livestock production. The mean per capita value of total
domestic production in Dzalanyama (3,975 MK) and Vwaza (4,587 MK) was between
100 and 150% greater than that of Mulanje (1735 MK) and Liwonde (1705 MK). The
major difference between the communities surrounding these reserves is access to land,
primarily a function of population density. High population density around Mulanje and
Liwonde limits farmers to half the land area for cultivation available to farmers around
Dzalanyama and Vwaza (Table A.10).
TABLE A. 10. POPULATION, LAND, AND AGRICULTURAL PRODUCTION
Protected
Mean
Population
Mean
Mean
Area
Crop
Density
Land Holding
Crop
Production
Production
kg capita ' yr"' MK capita"' yr"'
persons km^
ha capita"'
Mulanje
308.0
.146
1658.36
211
365.1
Liwonde
.194
1553.23
166
819.8
3238.98
Dzalanyama
119
.316
581.8
Vwaza
95
.329
2424.46
80
A.5 2 PROTECTED AREA NATURAL RESOURCE UTILIZATION
Though many factors influence the level of natural resource utilization, inhabitants
around all four protected areas raised the issue of barriers to access as most important.
These included government agency efforts at protection (i.e. policing with forest guards,
wildlife reserve and national park scouts, and fencing), and physical barriers (primarily
distance from household, elevation and slope, and flooding). Table A. 11 provides a
relative assessment of barriers to access along with mean per capita figures for overall
protected area resource utilization for each protected area.
TABLE All. BARRIERS TO PROTECTED AREA ACCESS AND OVERALL
UTILIZATION.
Protected
Relative
Relative Ease
Mean
Mean
Area
Level of
of Physical
Total
Total
Utilization
Protection
Access
Utilization
kg capita * yr ' MK capita ' yr"'
Mulanje
873
388.51
Low
Very DiflScult
Liwonde
379.35
Very High
484
Difficult
Dzalanyama
520.94
Low
1135
Easy
Vwaza
681.90
Medium
Easy
1449
Residents around Liwonde National Park reported the lowest natural resource
utilization, due to protection, and at least in part, to fears of government detection of their
activities. The fear may have inhibited some respondents, resulting in an underestimate of
Liwonde utilization. However it should be noted that the most common request during the
Liwonde rapid appraisal was for a tour of the Park in order to see, for the first time, the
large wildlife species inside. While scouts are not constant fixtures on the boundaries,
Liwonde is the base of training for wildlife scouts throughout Malawi, and is fenced on its
more populated western boundary. One kilometer east of that boundary is the Shire River
81
and its flood plain, encouraging a high fishing trade, but serving as an elective barrier to
most other uses.
The exceptionally steep slopes of Mulanje Mountain and local beliefs about the
plateau itself limited access to the lower slopes. Dzalanyama had the fewest physical
barriers, and like Mulanje, is a forest reserve with fewer use restrictions and limited agency
personnel available for protection. Inhabitants around Vwaza Marsh reported high levels
of policing near a small government base camp at Lake Kuzuni, but much lower levels
away from that camp. Flooding limited, but did not prohibit, access to parts of the reserve
during the rainy season.
A.6 RESULTS AND DISCUSSION
Species-level data on the quantity and consumptive use value of protected area
resources can be aggregated to assess the importance of protected area resource
utilization for household livelihood strategies. In each category of use are essential
elements of household production (i.e. fuelwood consumption), strategies to vary the diet
or extend food stocks during vulnerable times (i.e. food), and the actions taken to
diversify income risk (i.e. handicrafts). Taken in aggregate and compared with all other
income, these data provide a sense of local community re//awce on protected area
resources. At the protected area-level, reliance on specific categories of resources can help
the natural resource manager qualify the importance of demand in terms of alternatives.
Comparing reliance across households can provide insight on how protected area
resources influence both the struggle against poverty and the distribution of income.
82
A.6.1 RELIANCE
The most direct measure of the importance of protected area resources to the
livelihoods of those living in adjacent communities is income. By capturing all elements of
household production, including subsistence production and natural resource utilization, it
is possible to calculate the proportion of total income that is protected area based. The
relationship between these key variables - total income and income derived from protected
areas - can be termed reliance. It describes the relationship of all income to that portion
of income derived from protected areas, including protected area employment, direct petty
commerce (i.e. sale of wild foods), indirect petty commerce (i.e. those proceeds from
healing or brewing related to protected area resources), and home use of protected area
natural resource products. Figure A. 1 portrays mean per capita elements of income,
including that, which is protected area based. It also displays the relative proportions
provided by each element of income in percentage terms. Households adjacent to Mulanje
and Liwonde generate less than half the income of those surrounding Dzalanyama and
Vwaza. Yet the proportion of total income that was protected area derived is much higher
(20.3 and 15.8% for Mulanje and Liwonde compared to 10.4 and 13.3% for Dzalanyama
and Vwaza, respectively).
Some insight is gained by calculating reliance for major resource categories (Table
A. 12). Though protected area food prices tended to be lower than those for domestic
agricultural food products, the percentage of plant and animal food that was derived from
the protected area was high. Enforcement of government restrictions on resource use
limited the plant food utilization around Liwonde, whereas adjacent to Vwaza, much
83
higher per capita income resulted in few households requiring food supplements during the
rainy season. Low levels of domestic livestock production were compensated by high
levels of hunting in protected areas in all but Dzalanyama, where domestic livestock
production was much higher than the other reserves. The importance of protected area
wood and medicinal resources was high in all four protected areas.
TABLE A. 12. PERCENTAGE OF TOTAL FOOD, WOOD, AND MEDICINES THAT
WAS DERIVED FROM PROTECTED AREA NATURAL RESOURCS.
Wood
Wood Medicinal
Plant
Animal Animal
Plant
Value
Food
Biomass Value
Food
Food
Food
Biomass Value Biomass Value
-%-
Mulanje
Liwonde
Dzalanyama
Vwaza
31
8
23
10
10
I
9
4
82
93
27
93
63
85
17
72
83
74
69
88
91
82
77
94
68
91
76
87
A.6.2 RELIANCE AND TOTAL INCOME
More insight into overall reliance by local populations on the protected areas is
gained by investigating variation between households. McGregor (1995), using a wealth
ranking exercise (poor vs. rich) and a diet study, found that poorer households were more
dependent on miombo woodland resources than wealthy households in a communal area in
Zimbabwe. Her research suggests that further differentiation of wealth might reveal
patterns beyond the scope of a binary assessment. On that basis, we ranked and plotted
estimates of per capita income for all 427 households, stacking the portion of income
associated with protected areas on top of all non-protected area income (Figure A.2).
84
Household protected area resource utilization strategies vary considerably across
the data set, however the pattern displayed suggests a greater portion of total income is
protected area based for poorer households. We used the statistical technique of simple
regression model with an exponential curve fit to confirm our visual interpretation. The
regression model, Y =
, where Y = dependent variable, reliance (the percentage of
income that is protected area based), X = independent variable, per capita income, b and c
are constants, and e = equals the base of the natural logarithm. This was run on 30 equal
groups of households (Table A.13). The values for each group were based on the mean
per capita income and the mean portion of that income that was protected area-based to
calculate the reliance percentage. The regression was run at the 95% confidence level.
TABLE A. 13 SUMMARY OF REGRESSION RESULTS
Model Summary
Exponential
Adjusted
Curve Fit
R
R?
R?
0.919(»)
0.844
0.834
* predictors; (constant), per capita income
Standard Error
of the estimate
0.182
ANOVA(t)
Sum of
Squares
df
Regression
5.036
1
Residual
0.928
28
Total
5.964
29
* predictors: (constant), per capita income
dependent variable: reliance
Mean
Square
F
5.036 151.899
.0332
Coefficients*
(Constant)
per capita income
Coefficients Std. Error t-stat
0.295
0.013
22.302
-8.79E-05 7.13E-06 -12.325
* predictors: (constant), per capita income
Sig.
0.000
0.000
Sig.
.000(*)
85
The overall model, tested with the F-statistic, was significant at the 0.001 level.
The significance of the independent variable (per capita income) was tested with the tstatistic, and proved to be significant at a probability level of 0.001. While these two
statistics provide evidence in favor of the model, the role of the
statistic is limited to
providing some indication of goodness of fit of the sample in a body of data (Gujarati
1995). In this analysis, the
statistic of 0.84 suggests a reasonable fit of the exponential
curve fit model (Figure A.3).
On the basis of the high probability of significance of the t-statistic, the null
hypothesis that per capita income has no impact on reliance can be rejected. The shape and
direction of the exponential curve fit suggests an inverse income-reliance relationship. On
average, for every 100 MK increase in per capita income, the portion of income that is
protected area-based can be expected to decline by 0.1%.
A 6 3 DISTRIBUTION OF INCOME AND POVERTY
The results of the reliance model suggest inhabitants of nearby communities rely on
protected areas for a substantial portion of their income, and that reliance is greater for
poorer households. These results, however, do not fully explain the influence of reliance
on protected areas as a strategy against poverty. Does the fact the poorer households
show greater reliance actually impact the distribution of income, and does it play a role in
poverty alleviation?
One method (proposed by Loomis 1993) of addressing these questions is to
conduct a "without and with" analysis. In this case, we assessed the distribution of income
86
without protected area proceeds first, and then compared the resuhs to an assessment with
those proceeds. The specific tools used were standard quantitative methods for measuring
inequality and poverty. These included a Lorenz curve, a Gini coefficient and polarization
index derived from the Lorenz curve, and three poverty indices. The mathematical
methods selected to parameterize the Lorenz curve and determine the underlying indices
are listed in Table A. 14. The actual calculations were conducted in POVCAL software,
developed specifically fiar this purpose by the World Bank (Chen, Datt, and Ravallion
1992).
TABLE A. 14. POVERTY MEASURES AND ONE POLARIZATION MEASURE FOR
THE GENERAL QUADRATIC PARAMETERIZATION OF THE LORENZ CURVE.*
Equation of the general quadratic
Lorenz curve (L(p))
Ul - L ) = a ( p ' - L ) + b L ( p - 1) +c(p - L)
or.
Villaseftor and Arnold
(1989)
= - [bp = e + imp' + np + ^")''"]/2
Headcounl index { H )
H = [n +
Poven\' gap index iPG)
PG
Foster-Grcer-Thorbecke
povert\- index {P:)
P: = IPG-H-iu- /r)[aH + bUH)
-(r/16)ln{(l -///ji) / ( l - H / s z ) / } ]
Wolfson polarization index ( I f J
IF =
+ 2z'fi) {(6 + Iz'fiy - m
Dun and Ravallion
(1992)
Dutt and Ravallion
(1992)
Foster, Greer, and
Thorbecke (1984)
Wolfson (1994)
*Adapted from Dutt and Ravallion (1992).
z = the povertv- line; // = mean income; Z, = the Lx)renz curve; L(p) = parameterized Lorenz curv'e;
a, b , c, are constants in the general quadratic form; e = - { a + b + c + \ ) \ m = n ' - 4a; n = 2be - 4c;
r=in'- Ame'f '-, si=(r- n)/(2m)\i;= - (r + n)li2m)\ G = Gini coefficient;
= distribution-corrected
mean income =//(!- G);
= the mean income of the poorest half of the population; med = median income.
A.6.3.1 Measures of inequality
The Lorenz curve was first proposed by statistician Max Lorenz in 1905 as a
method to compare wealth through a cumulative income curve. Coined the "gold
standard" for the concept of inequality (Wolfson 1994), it is the most commonly used
distributional tool for assessing cumulative income in a population. The curve plots the
87
proportion of total income (vertical axis) received by the bottom 1%, 5%, 50%, etc.
income earners against the corresponding population proportion (horizontal axis) so that
mean income equals unity (Figure A-4). The curve must pass through the two comers and
be convex, and the state of perfect equality would lie exactly across the diagonal at 45°
(Amiel and Cowell 1999). For complete inequality, in which only one person has all the
income, the Lorenz curve would coincide with a right angle made by the lower and right
boundaries of the curve. In Figure A-4, we compare the distribution of income that is a)
inclusive of protected area proceeds and b) exclusive. Regardless of the difference in total
value, if these two measures of income are distributed in the same proportions across all
households, the two curves would coincide. Instead, we find that income inclusive of
protected area proceeds is more equitably distributed (that is, closer to the 45° egalitarian
line).
The most common measure of dispersion within group values represented by the
Lorenz curve is the Gini index (G), developed by Corrado Gini in 1912. It is defined as the
arithmetic average of the absolute values of differences between all pairs of incomes (Sen
1973, p. 31),
G = (l/2/iV)r Z"
/=/ j=/
Where n = number of persons, /= 1,...«, _y, = the income of person /, yj = the income of
person j, and // = the average level of income.
It is therefore the area of the lens-shaped piece formed by the diagonal line of
absolute equality and the Lorenz curve itself, divided by the area of the entire triangular
88
region underneath the diagonal. The larger the coefficient, the greater the degree of
dispersion, implying greater inequality. It can be generally stated that more highly
developed countries tend to have lower differentiation of income (expressed as
percentages, a Gini index of 25 to 40%) while developing nations tend to have higher
differentiation (45 to 60%).
The income distribution in this study results in a Gini index of 56.3% when
proceeds from protected areas are excluded (Table A. 15). The ratio declines to 50.9%
(more equity) when the protected area proceeds are included in the income totals.
TABLE A. 15 DESCRIPTIVE STATISTICS AND INEQUALITY INDICES FOR
INCOME WITHOUT AND WITH PROTECTED AREA PROCEEDS.
Inequality Indices
Descriptive Statistics
Dispersion Polarization
Per Capita Annual Income (MK)
Gini (G) Wolfson (fF)
Mean
Standard
Median
%
%
Cw)
Deviation
Without protected
56.3
57.1
3646.65
4783.31
2074.02
area proceeds
With protected
50.9
50.2
4320.25
4937.69
2706.02
area proceeds
The Gini index may not capture changes in income by the middle strata, a
phenomenon Wolfson (1994) calls the "disappearing middle class." To assess this
possibility, a polarization index was also calculated. WTien there is complete equality there
is no polarization. Polarization increases as the index increases, until a hypothetical
maximum is reached where half the population has zero income and the other half has
twice the mean. Here, the magnitude and direction of change in the polarization index
were very similar to the Gini index resuhs (57.1 and 50.2%, respectively). This suggests a
89
more equitable distribution of income associated with the addition of protected area
livelihood strategies was accompanied by a decline in polarization.
Although popular, the Gini index gives little insight into the location and
concentration of income inequality among high- versus low-income groups (Foster 1992,
p. 146). This problem can be addressed in part by calculating relative income shares
(Figure A.5), as the World Bank does in its World Development Reports (e.g. World
Bank 2000b). The inequitable distribution of income common in developing countries is
evident, and corresponds with the results of the Lorenz and Gini analyses. Note that the
addition of protected area-based income results in a more equitable overall distribution,
with a decline in the share held by the richest group and increases in the other four groups.
The percentage changes in relative income share are more evident illustrated in Figure
A .6.
A.6.3.2 Measures of poverty
Change in relative income share suggests that protected area livelihood strategies
play a central role in poverty alleviation for communities adjacent to the four protected
areas. However, to understand poverty, the assessment of the distribution of income must
be augmented by its relationship to a pre-determined poverty threshold. Two poverty
thresholds were considered, one based on basic human needs and the other based on an
arbitrary reference based on the national survey data that indicate 32% of Malawi
households are below the poverty threshold.
90
The "basic needs'' poverty threshold relates minimum nutritional requirements to
the energy provided by a primary diet staple, which in turn can be converted into a value
based on the market price of that staple. In the case of Malawi, the Government of
Malawi, the Food and Agriculture Organization, the World Health Organization combined
forces to determine the dietary and energy requirements needed to meet the nutritional
needs of healthy persons (Johnson 1996). The values assessed for all classes of people (i.e.
men, women, children, pregnant women, etc.) were multiplied against population
estimates for each category for all of Malawi to create a single per capita energy
requirement figure. The annual maize requirement necessary to supply that energy is 170.3
kg per capita of Malawian maize. Multiplied by the average price of maize over the study
period (2.55 MK), and increased to reflect that food represents about 65% of total
expenditures of poor households (World Bank 1996), we calculated the "basic needs
poverty threshold" in Malawi to be 668.10 MK in annual income in 1996/97.
Theoretically, the poorest third of households in Malawi should fall below this
threshold if the income data in our study were based on the same factors considered in the
national studies. However, our research included all aspects of subsistence production and
protected area resource utilization, the majority of which are not captured in the national
studies. Thus, we cailculated a second "poverty reference threshold" (1700.00 MK) to
include the poorest 32% of all households to reference the national prevalence of absolute
poverty reported by the World Bank (1996).
Next we selected three poverty measures from among the many available in the
literature, basing our decision on the practices of major development organizations such as
91
the World Bank. Assessing poverty by a single measure, such as the percentage of the
population below the poverty threshold relative to the population as a whole may ignore
one or more facets of poverty. For example, though a poverty head count quantifies the
number of poor, but ignores the depth of poverty (i.e. all poor are not equally poor). It
also does not relate poverty to economic inequality. For these reasons we added two other
measures that consider poverty as being proportional to the poverty gap and relate it to
the inequality captured in the Lorenz curve (Table A.16).
TABLE A. 16. POVERTY INDICES.*
Headcount index (Jf)
The proportion of population who are poor.
Poverty gap index {PG)
The aggregate income shortfall of the poor as a
proportion of the poverty threshold and normalized by
population size.
Foster-Greer-Thorbecke
Like PG, but based on the sum of squared poverty deficits
so that any increase in subgroup poverty must increase
total poverty.
* Adapted from Dutt and Ravallion (1992); equations are in Table A-14.
Using the maize equivalent "basic needs" poverty threshold (668.10 MK), a head
count provides a direct indication of the number of people who are able to meet minimum
needs in part due to protected area-based income. Across all four protected areas, an
estimated 353,438 people were living within 5 km of the reserve boundaries (Malawi
Government 1998; Orr et al. 1998). Applying the results of our sample to this population,
17.5%, or 61,820 of those people would fall below the basic needs poverty threshold
without access to protected area-based income (Table A. 17). When protected area
proceeds are included in the calculation of total income, only 8.8% of the total population,
or 31,064 people remain below the basic needs poverty threshold. In other words, fiilly
92
half the population who would fall below the basic needs poverty threshold without
protected area-based income is rise above the poverty threshold due to income generated
from protected areas.
TABLE A. 17. POVERTY ANALYSIS.
Poverty Threshold
Poverty Indices (%)
(MK)
(^)
Head Count Poverty Gap
(//)
Basic Needs
iPG)
Without protected
17.5
6.9
668.10
area proceeds
With protected
8.8
668.10
2.3
area proceeds
Reference
Without protected
43.6
1700.00
21.8
area proceeds
With protected
1700.00
31.9
13.5
area proceeds
*FGT = Foster-Greer-Thorbecke poverty gap index.
FGT
(P.)
3.6
0.8
13.9
7.5
The proportion of the total population that falls below the "poverty reference
threshold" (1700.00 MK) is obviously much higher (43.6 and 31.9%, respectively). This
translates into 154,146 people who would fall below the poverty reference threshold
without benefit of protected area proceeds. What is striking is that a much smaller
percentage (27%, as opposed to the 50% noted with the "basic needs" threshold) of those
people rise out of poverty with the benefit of protected area-based income. This suggests
that protected area resources and other associated protected area income impacts far more
of the poorest of the poor than those above them on the income ladder.
Though the initial percentages are lower, the results for the poverty gap index and
the Foster-Greer-Thorbecke poverty gap index show the same direction of change as the
93
head count. The magnitude of the difference "without" and "with" the benefit of protected
area proceeds (on the basic need poverty threshold, 66% for PG, and 77% for Pi) is even
larger than the headcount (50%). Like the head count index, the proportion of people who
rise out of poverty when measured on the basic needs poverty threshold is almost double
that of the higher, reference poverty threshold. These findings strongly suggest that
a) protected area-based income provides a livelihood strategy that can lift a great
number of the poor above the poverty threshold, and that
b) that benefit is more pronounced for the poorest of the poor.
A.7 CONCLUSIONS
This research introduces a multidimensional field methodology that can provide
both baseline socio-economic information concerning household production and detailed,
species level protected area resource utilization. Using the example of four protected areas
in Malawi, we have demonstrated how the annual harvests of agricultural and protected
area natural resource products were; (a) identified; (b) measured (by volume in local units
of measure); (c) converted to weight units, (d) verified against on-site physical
measurements and key respondent interviews; and (e) valued in monetary units.
When collected data collected are aggregated to the household level, the
importance of protected area-based income becomes quite evident. Poor households are
generally more reliant on protected area resources than other households. In turn, that
reliance has a major influence on the distribution of income between poor and rich
households. Exploitation of protected area resources is a livelihood strategy that halved
94
the number of households that otherwise would have remained beneath the basic needs
poverty threshold. These findings have major policy ramification, not only for
environmental management and conservation, but also for poverty alleviation.
In a society that is almost entirely dependent on agriculture households, and in
particular poor households, have created income-generating alternatives to maintain
livelihood security. That these alternatives are based on protected area proceeds
represents a challenge and an opportunity to those attempting to balance sustainable use
and sustainable development simultaneously. The methodology represents a unique blend
of tools used to gather livelihood security information and ethnoecological data in
conjunction with larger initiatives that traditionally have not captured the role of protected
area resources. This provides the opportunity to address and monitor the influence of
those resources on local communities and gauge the ecological impact of resource
demand.
A.8
LITERATURE CITED
Abbot, P., J. Lowore, and M. Werren. 1997. Models for the estimation of single tree
volume in four Miombo woodland types. Forest Ecology and Management 97:2537.
Amiel, Y., and F.A. Cowell. 1999. Thinking about inequality; personal judgement and
income distributions. Cambridge University Press, Cambridge.
AnselU W.F.H., and R.J. Dowsett. 1988. Mammals of Malawi; an annotated check list
and atlas. Trendrine Press, Zennor, United Kingdom.
95
Arnold, J.E.M., and J.E. Falconer. 1989. Household food security and forestry: an
analysis of socioeconomic issues. Community Forestry Note 1. FAO, Rome.
Becker, H. 1990. Labour input decisions of subsistence farm households in southern
Malawi. Journal of Agricultural Economics 41; 162-171.
Beebe, J. 1995. Basic concepts and techniques of rapid appraisal. Human Organization
54:42-51.
Benson, C.W. and Benson C.M. 1977. The birds of Malawi. Montfort Press, Blantyre,
Malawi.
Bernard, H.R., P. Killworth, D. Kronenfeid, and L. Sailor. 1984. The problem of
informant accuracy: the validity of retrospective data. Annual Review of
Anthropology 27:382-396.
Bhima R., and C.O. Dudley. 1997. The influence of the Shire River on Liwonde
National Park, Malawi, with special reference to elephant movements. Koedoe
40:9-18.
Binns, B. 1972. Dictionary of plant names in Malawi. Government Printer, Zomba,
Malawi.
Brouwer, I.D., J.C. Hoorweg, and M.J. van Liere. 1997. When households run out of
fuel: responses of rural households to decreasing fiielwood availability, Ntcheu
District, Malawi. World Development 25:255-266.
Bruce, J.W. 1989. Community forestry: rapid appraisal of tree and land tenure.
Community Forestry Note 5. FAO, Rome.
96
Campbell, B.M., M. Luckert, and I. Scoones. 1997. Local-level valuation of savanna
resources; a case study from Zimbabwe. Economic Botany 51:59-77.
Carr, S.J. 1997. A green revolution frustrated; lessons learned from the Malawi
experience. African Crop Science Journal 5; 93-98.
CCAM. 1992. Malawi's traditional and modem cooking. Chitukuko Cha Amayi
m'Malawi, Lilongwe, Malawi.
Chambers, R. 1990. Rapid and participatory rural appraisal. Appropriate Technology
16(4); 14-16.
Chambers, R- 1994. The origins and practice of participatory rural appraisal. World
Development 22;953-969.
Chen, S., G Datt, and M. Ravallion. 1992. POVCAL; a computer program for
calculating poverty measures from grouped data. Policy Research Department,
World Bank, Washington DC.
Chibnik M. 1978. The value of subsistence production. Journal of Anthropological
Research 34; 561-576.
Chilowa, W. 1998. The impact of agricultural liberalisation on food security in Malawi.
Food Policy 23;553-569.
Cockburn, H. 1982. Poultry keeping in East Africa. Kenya Literature Bureau, Nairobi.
Dutt, G., and M. Ravallion. 1992. Growth and redistribution components of changes in
poverty measures. Journal of Development Economics 38;275-295.
Edwards, L 1985. Conservation of plants on Mulanje Mountain, Malawi. Oryx 19;86-90.
97
Engberg, L.E., J.H. Sabry, and S.A. Beckerson. 1987. Production activities, food
supply and nutritional status in Malawi. Journal of Modem African Studies
25:139-147.
Fa, JE, and A. Purvis. 1997. Body size, diet and population density of afrotropical forest
mammals: a comparison with neotropical species. Journal of Animal Ecology
66:98-112.
Finan, T.J. 1996. Anthropological research methods in a changing world. Pages 101-124
in Moran, E.F., ed.. Transforming societies, transforming anthropology. University
of Michigan Press, Ann Arbor.
Fleuret, A. 1979. Methods for evaluation of the role of fruits and wild greens in Shambaa
diet; A Case Study. Medical Anthropology 3:249-269.
Foster, J., J. Greer, and E. Thorbecke. 1984. A class of decomposable poverty
measures. Econometrica 52:761-766.
Foster, P. 1992. The world food problem. Lynne Rienner Publishers, Boulder, Colorado.
Freudenberger, K.S. 1995. Tree and land tenure: using rapid appraisal to study natural
resource management. Community Forestry Case Study Series 10. FAO, Rome.
Garine, I. de, and G. Koppert. 1988. Coping with seasonal fluctuations in food supply
among savanna populations; the Massa and Mussey of Chad and Cameroon. Pages
210-259 in I. De Garine and O A. Harrison, eds.. Coping with uncertainty in food
supply. Clarendon Press, Oxford.
Giuliano, A.R., H. Dahiane, and B. Orr. 1998. Evaluation of USAID Title II Food Aid
Programs; Review of Ten CARE Country Programs. CARE, Atlanta.
98
Godoy, R., R. Lubowski, and A. Markandya. 1993. A method for the economic
valuation of non-timber tropical forest products. Economic Botany 47:220-233.
Guijt, I., F. Hinchclifle, and M. Melnyk. 1995. The hidden harvest; the value of wild
resources in agricultural systems. IIED, London.
Gujarati, D.N. 1995. Basic econonmetrics. Third Edition. McGaw-Hill, New York.
Gupta, A.K., and IDS Workshop. 1989. Maps drawn by farmers and extensionists.
Pages 86-92 in Farmer first; farmer innovation and agricultural research.
Intermediate Technology Publications, London.
Guyer J.L, and P.E. Peters. 1987. Conceptualizing the household - issues of theory and
policy in Africa - introduction. Development and Change 18; 197-214.
Hayes, G.W. 1978. A guide to Malawi's national parks and game reserves. Montfort
Press, Limbe, Malawi.
Johnson, J.S. 1995. An analysis of the extent, causes and effects of food insecurity in
Malawi, with an approach to improving food security. Smallholder Agricultural
Productivity Programme Occasional Paper No. 1. FAO MLW/92/006, Lilongwe.
Kelly, D. 1993. Aah . but those elephants; David Kelly paints Liwonde. Malawi Wildlife
Society, Blantyre, Malawi.
Kennedy, E. and P. Peters. 1992. Household food security and child nutrition; the
interaction of income and gender of household head. World Development
20;1077-1085.
99
Loomis, J.B. 1993. Integrated public lands management; principles and applications to
national forests, parks, wildlife refuges, and BLM lands. Columbia University
Press, New York.
Lorkeers,
and J.H. Venema. 1991. Land Resources Evaluation Project; Land
Resources Appraisal of Lilongwe Agricultural Development Division.
MLW/85/011 Field Document No. 24. Malawi Govemment/UNDP/FAO,
Lilongwe, Malawi.
Luckert, M.K., J. Wilson, V. Adamowicz, and A.B. Cunningham. 2000. Household
resource allocations in response to risks and returns in a communal area of western
Zimbabwe. Ecological Economics 33;383-394.
Luoga, E.J., E.T.F. Witkowski, and K. Balkwill. 2000. Differential utilization and
ethnobotany of trees in Kitulanghalo Forest Reserve and surrounding communal
lands, eastern Tanzania. Economic Botany 54;328-343.
Malawi Government. 1998. Report of Preliminary Results; 1998 Population and
Housing Census. National Statistics Office, Zomba, Malawi.
Malimbwi, R.E., and M.S. Philip. 1989. A compatible taper/volume estimation system
for Pinus patula at Sao Hill Forest Project, Southern Tanzania. Forest Ecology
and Management 27; 109-115.
Martin, G.H.G. 1984. Carcass composition and palatability of some wild animals
commonly used as food. World Animal Review 53;40-44.
Martin, G.J. 1995. Ethnobotany; a methods manual. People and Plants Conservation
Manuals, Volume 1. Chapman and Hall, London.
Mason, LL., and Maule J.P. 1960. The indigenous livestock of eastern and southern
Africa. Commonwealth Agricultural Bureaux, Famham Royal, England.
McGregor, J. 1995. Gathered produce in Zimbabwe's communal areas changing resource
availability and use. Ecology of Food and Nutrition 33:163-193.
McShane, T.O. 1985. Vwaza Marsh Game Reserve: a baseline ecological survey.
Department of National Parks and Wildlife, Rumphi, Malawi.
Mellor, J.W. 1966. The economics of agricultural development. Cornell University Press,
Ithaca, New York.
Momberg, F., K. Atok, and M. Sirait. 1996. Drawing on local knowledge: a community
mapping training manual. Ford FoundationA^ayasan Karya Sosial Pancur
KasihAVWF Indonesia Programme, Jakarta.
Morris, B. 1990. An annotated check-list of the macrofiingi of Malawi. Kirkia 13: 323364.
Morris, B. 1998. The power of animals: an ethnography. Berg, Oxford.
Morris, B. and J.D. Msonthi. 1996. Medicinal plants of Malawi. Pages 191-557 in B.
Morris ed., Chewa medical botany: a study of herbalism in Southern Malawi.
International African Institute, London.
Mushove, P.T., J.A.B. Prior, C. Gumbie, and D.F. Cutler. 1995. The effects of
different environments on diameter growth increments of Colophospermum
mopane and Combretum apiculatum. Forest Ecology and Management 72:287292.
101
Nachman, S.R. 1984. Lies my informants told me. Journal of Anthropological Research
40:536-555.
Ngalande, J.D. 1995. An integrated management plan for Dzalanyama Forest Reserve,
Malawi. M.S. thesis. University of Aberdeen, Scotland.
Nurse, G.T. 1975. Seasonal hunger among the Ngoni and Ntumba of Central Malawi.
Africa 45:1-11.
Openshaw, K. 1997. Urban biomass fuels: production, transportation, and trading study.
Alternative Energy Development, Inc/Malawi Ministry of Energy and Mining,
Lilongwe, Malawi.
Orr, B., B. Eiswerth, T. Finan, and L. Malembo. 1998. Malawi public lands utilization
study. Malawi Ministry of Lands, Housing, Physical Planning, and
Surveys/University of Arizona/Forestry Research Institute of Malawi, Lilongwe,
Malawi.
Owen, J.E. 1975. The meat-producint characteristics of the indigenous Malawi goat.
Tropical Science 17:123-138.
Paris, S. 1991a. Land Resources Evaluation Project: Land Resources Appraisal of
Blantyre Agricultural Development Division. MLW/85/011 Field Document No.
24. Malawi Govemment/UNDP/FAO, Lilongwe, Malawi.
Paris, S. 1991b. Land Resources Evaluation Project; Land Resources Appraisal of Mzuzu
Agricultural Development Division. MLW/85/011 Field Document No. 27. Malawi
Govemment/UNDP/FAO, Lilongwe, Malawi.
Pike, J.G., and G.T. Rimmington. 1965. Malawi: a geographical study. Oxford
University Press, London.
Ribbink, A.J., B.A. Marsh, A.C. Marsh, A.C. Ribbink, and B.J. Sharp. 1983. A
preliminary survey of the cichlid fishes of rocky habitats in Lake Malawi. South
Afncan Journal of Zoology 18:149-310.
Sen, A. 1973. On economic inequality. Norton & Company, New York.
Shiver, B.D., and G.H. Brister. 1992. Tree and stand volume functions for Eucalyptus
saligna. Forest Ecology and Management 47:211-223.
Statistical Package for the Social Sciences. 1998. SPSS base 8.0 for Windows: user's
guide. SPSS Inc., Chicago.
Sweeney, R.C.H. 1965. Animal life in Malawi: volume I, invertebrates. Institute for the
Publication of Textbooks, Belgrade.
Tweddle D., and N.G. Willoughby. 1979. An annotated checklist of the fish fauna of the
River Shire south of Kapachira Falls. Nyala 5:75-91.
United Nations Population Division. 2000. Infonation: statistical data for member states
of the United Nations (United Nations CyberSchooIBus). Online. Available HTTP:
http://www.un.org/Pubs/CyberSchoolBus/infonation/e_infonation.htm . November
2000.
Vaughan, M. 1985. Household units in Southern Malawi. Review of Afiican Political
Economy 34:35-45.
103
Venema, J.H. 1991. Land Resources Evaluation Project; Land Resources Appraisal of
Liwonde Agricultural Development Division. MLW/85/011 Field Document No.
22. Malawi Govemment/UNDP/FAO, Lilongwe, Malawi.
Wilk, R.R., and R. McC. Netting. 1984. Households; changing forms and functions
Pages 1-28 in R. McC. Netting, R.R. Wilk, E.J. Amould eds.. Households;
comparative and historical studies of the domestic group. University of California
Press, Berkeley.
Williamson, J. 1974. Useful plants of Malawi; revised and extended edition. Montfort
Press, Limbe, Malawi.
Wolfson, M.C. 1994. When inequalities diverge. The American Economic Review
84;353-358.
Wollny, C.B.A., A.G. Namwaza, T.S.W. Makamba, and K. Kuttner. 1998. Zum
stand der rinderzucht in Malawi. Archiv fiir Tierzucht. 41 ;33-44.
Woodson, D.G. 1997. Lamanjay, food security, securite alimentaire. a lesson in
communication for BARA's mixed-methods approach to baseline research in Haiti,
1994-1996. Culture and Agriculture 19; 108-122.
World Bank. 1996. Malawi; growth through poverty reduction. Update of the 1990
Country Economic Memorandum, 8140-MAI. World Bank, Lilongwe.
World Bank. 2000a. Afncan development indicators 2000. World Bank, Washington
DC.
World Bank. 2000b. World development report 2000/2001. Attacking poverty. Oxford
University Press, Oxford.
104
Zinyama, L.M., T. Matiza, and D.J. Campbell. 1990. The use of wild foods during
periods of food shortage in rural Zimbabwe. Ecology of Food and Nutrition
24:251-265.
105
A.9 FIGURE LEGENDS
FIGURE A.1 THE RELATIVE IMPORTANCE OF ALL INCOME ELEMENTS.
FIGURE A.2. PROTECTED AREA AND NON-PROTECTED AREA COMPONENTS
OF PER CAPITA INCOME FOR ALL SAMPLED HOUSEHOLDS.
FIGURE A.3. SIMPLE LINEAR REGRESSION FOR PER CAPITA INCOME AS A
PREDICTOR FOR RELIANCE ON PROTECTED AREA RESOURCES.
FIGURE A 4 LORENZ CURVES FOR INCOME INCLUSIVE AND EXCLUSIVE OF
PROTECTED AREA PROCEEDS.
FIGURE A. 5. DIFFERENCES IN RELATIVE INCOME BETWEEN INCOME
GROUPS.
FIGURE A.6. CHANGE IN RELATIVE INCOME SHARE RESULTING FROM THE
ADDITION OF PROTECTED AREA-BASED PROCEEDS TO HOUSEHOLD
INCOME.
106
A.10 FIGURES
MuJaive
Lhvonde
-7'
0
1000 2000 3000
4000
5000 6000
Per Capita Income (MK)
Mularye
Liwonde
/ V . v;
Dzalaityama
Vwaza
0%
20%
40%
60%
80%
100%
Percent of Total Income
O food crops
S emigration
non-food crops • wood
• estate
• other
I meat
I protected area
FIGURE A.L THE RELATIVE IMPORTANCE OF ALL INCOME ELEMENTS.
107
45.000
40,000
35.000
30,000
1 25,000
s
20.000
y
15,000
^ 10,000
5.000
CM
SO
f*^
sO
<*i
—
^
Households (in order of increasing income)
• non-protected area-based • protected area-based
FIGURE A.2. PROTECTED AREA AND NON-PROTECTED AREA COMPONENTS
OF PER CAPITA INCOME FOR ALL SAMPLED HOUSEHOLDS.
108
40%
35%
-9E^5x
•
y = 0.2949e
»
Wi
S
A
30% [••
£
•s 25%
R-= 0.8444
F=151.899
(Sig. .000)
I ^
£
.S
20%
"oJ —
OS IS
15%
e
t= 22.302
(Sip. .000)
10%
S5
5%
0%
5,000
10,000
15,000
20,000
25,000
Per Capita Income (MK)
reliance
predicted refiance
FIGURE A.3. REGRESSION FOR PER CAPITA INCOME AS A PREDICTOR FOR
RELIANCE ON PROTECTED AREA RESOURCES.
109
100%
80%
60%
40%
20%
0%
0%
20%
40%
60%
80%
100%
Cumulative Population (%)
a) income with protected area
^~b) income without protected area
hypothetical equally distributed income
FIGURE A.4. LORENZ CURVES FOR INCOME INCLUSIVE AND EXCLUSIVE OF
PROTECTED AREA PROCEEDS.
110
80%
e
2
s
60%
e
i-
40%
as
</3
Of)
ce
w«>
r« ri
20%
0%
1st Quintile
(Poorest)
2nd Quintfle
3rd Quintile
4th Quintile
5th Quintile
(Richest) .
Income Groups (fifths of the population)
• income without protected area contribution
• income with protected area contribution
FIGURE A. 5 DIFFERENCES IN RELATIVE INCOME BETWEEN INCOME
GROUPS.
Ill
100%
V
u
SB
£
CW
w
E
e
w
80%
w
>
r
20%
S
60%
40.1%
40%
21.9%
12.5%
5.0%
-7.9%
0%
-20%
c?
/
Income Gn>ii|)s (fifths of the population)
I
1
•change in share of income I
cT
FIGURE A.6. CHANGE EN RELATIVE INCOME SHARE RESULTING FROM THE
ADDITION OF PROTECTED AREA-BASED PROCEEDS TO HOUSEHOLD
INCOME.
112
B. APPENDIX B. QUANTIFYING PROTECTED AREA RESOURCE DEMAND
AT THE SPECIES LEVEL: ANOTHER CONSIDERATION FOR
INTEGRATED CONSERVATION AND DEVELOPMENT PROJECTS
B.l
ABSTRACT
GIT, Barron J. (Office of Arid Land Studies, University of Arizona, 1955 E.
Street,
Tucson, AZ 85719, USA, barron'%ag.arizona.edu), H.T. Chapama, and Luke N.
Malembo (Forestry Research Institute of Malawi, P.O. Box 270, Zomba, Sdalawi,
frim<%malawi.net). QUANTIFYING PROTECTED AREA RESOURCE DEMAND AT THE SPECIES
LEVEL: AN ESSENTL«LL ELEMENT OF INTEGRATED CONSERVATION AND DEVELOPMENT
PROJECTS. This study quantified resource utilization at the species level and overall
resource demand in order to evaluated sustainable use in four protected areas in Malawi.
The annual harvests of agricultural and protected area natural resource products from
427 households were identified, measured (by volume in local units of measure),
converted to weight units, verified against on-site physical measurements and key
respondent interviews, and valued in monetary units. Results for the major species used
for each category of use are documented. Overall woody resource demand is compared
to sustainable supply through spatial analysis of mean annual increment and population
pressure over time. If resource use is confined to the poptdation with a 5 km zone of
influence, the sustainable supply of total protected area ecological resources has the
potential to meet demand for 50 years in three of the reserves, and 100 years in the
fourth. The analysis of both species-level and aggregate resource dematidprovides a set
of monitoring tools for community management and conservation.
113
Key Words: protected area resources; valuation; resource demand; sustainable use;
sustainable development; participatory rural appraisal; ethnoecology; Malawi.
B.2 INTRODUCTION
In the past, the goals of conservation biology and economic development were
rarely integrated in projects despite their common human dimension. While researchers
from both groups studied natural resources, one group tended to focus on human impact
while the other assessed human need. With the emergence of the ecological and economic
reality that human natural resource demand may outstrip environmental capacity (Vitousek
et al. 1986), the call for sustainable development (where environmental, social and
economic conditions can be maintained or improved for future generations) became a
more common theme (World Commission on Environment and Development 1987).
Evolving societal concerns have increased pressure to meid the goals of
conservation and sustainable development. The desire to improve upon conventional
"fences and fines" approach to conservation (Barrett and Arcese 1995) coincided with a
more general trend towards including local communities in research, planning, and
management of development initiatives (Brandon and Wells 1992). Protected area natural
resources are increasingly viewed as the fiindamental link between conservation and
sustainable development (Wild and Mutebi 1997). This has fostered interdisciplinary
research that integrates socio-economic and ecological perspectives where concerns for
protection of natural areas and economic development of local communities converge
(Munasinghe 1992).
114
Although the potential of community-based conservation of natural resources has
come under increased scrutiny (Wainwright and Wehrmeyer 1998), the long-term financial
benefits of conservation to local people have the potential to outweigh those of agriculture
or logging (e.g. Kremen et al. 2000). In spite of potential benefits, local populations in
developing countries tend to receive a small share of the use and non-use values assessed
by outsiders (Godoy ei al. 2000). To rectify this inequality, community-based natural
resource management and integrated conservation and development projects (IDCPs)
have become widespread. Their goal is to exploit the role that human consumption of
protected natural resources can play in both conservation and development by directing
the returns derived from conservation back into the community (Brandon and Wells
1992).
Though IDCPs are viewed as having greater appeal than the exclusionary, "fences
and fines" conservation strategies that preceded them (Barrett and Arcese 1995), they
require far more ecological, economic, social, and institutional monitoring. Much of the
criticism of early IDCPs focused on; (1) project implementation with poor ecological data
relative to socioeconomic data describing changes over time; (2) limited ecological
understanding of the potential impacts of various harvest rates; and (3) poor assessment of
the economic, political, and social benefits of existing utilization (Barrett and Arcese 1995;
Gibson and Marks 1995).
The consumptive use-value of protected area natural resources is the portion of
forest resource value that has direct local benefit. The quantity and value of resource
demand tend to be roughly estimated or overlooked entirely when IDCPs are designed
115
(Barrett and Arcese 1995). As the number of uses and resource users increases, the
complexity and costs of managing sustainable use conspicuously increases (Cunningham
1994). This suggests a need for data on the quantity and value of resources consumed
from the protected area targeted by the ICDP as well as a consideration of the alternatives
derived from household production and expenditure. If such information were regularly
collected at the species level, resource demand could be monitored, making sustainable
use a more practical goal.
This research demonstrates how an integration of regularly collected social and
economic data can be augmented with ethnobiological information in order to capture
protected area resource demand. We used a multi-dimensional approach in communities
around four protected areas in Malawi to address three ftindamental questions at the
species level:
(9) Which species are used for household livelihood?
(10) What are the per capita quantities of each species used for various uses, whether
agriculturally produced or collected from the protected area?
(11) What are the per capita consumptive use-values of agricultural production and
protected area resource utilization?
In aggregate form, these data were then integrated with information on population
trends and compared to assessments of sustainable resource supply, to address two
questions:
(1) Does the current supply of protected area resources meet current demand?
116
(2) Based on current patterns of use and population growth, how long before
sustainable supply is outstripped by demand?
B.3 STUDY AREAS
In 1997, 19% of Malawi's 94,000 km" of land was under protection in the form of
four wildlife reserves, five national parks, and seventy-seven forest reserves (Orr et al.
1998). This percentage is not exceptional in Afiica or elsewhere. However, in the same
year, 86% of the country's 9.65 million people lived in rural areas (Malawi Government
1998). Malawi's average population density of 103 persons/km^ of land was three times
that of its neighbors (United Nations Population Division 2000). This translates into a
density that twice as high on both protected and non-protected wooded areas. The
demand for agricultural land is substantial because mean farm size is at or below 1.0 ha for
a family of five in rural Malawi (BDPA/AHT 1998; House and Zimalirana 1992) and
traditional agriculture is the dominant livelihood systems. Furthermore, 98% of rural and
94% of urban energy demand is satisfied through fiielwood and charcoal (Arpaillange
1996), a level of dependence exceeded internationally only by Nepal (Pearce and Turner
1990).
As a result of demand for wood and agricultural land, the forested area in Malawi
was reduced by half between 1946 and 1996 (FAO 1981; FAO 1999; Millington et al.
1989; Openshaw 1996, Willan 1947). This has increased the importance of remaining
protected area resources as both common resource base and potential agricultural land.
117
In response to pressure to convert protected areas to agriculture, the Government
of Malawi commissioned a Public Lands Utilization Study (PLUS) in 1996 to study the
environmental risks of conversion and the importance of the reserves to adjacent
communities (Orr et al. 1998). The study was guided by a national Steering Committee on
Land, which was made up of 60 government and non-government stakeholder agencies.
The Committee selected four protected areas for intensive study, including Mulanje Forest
Reserve, Liwonde National Park, Dzalanyama Forest Reserve and Ranch, and Vwaza
Wildlife Reserve.
B.3 1 MULANJE FOREST RESERVE
Mulanje Forest Reserve is located in southeastern Malawi, centered on 15°57'S,
35°39'E, covering 56,314 ha of mostly mountainous terrain. It has vegetation ranging from
montane woodland and grassland on the plateau to miombo woodland (mesic-dystrophic
savanna, dominated by Julbernardia and Brachystegia species) on the lower slopes.
Geologically, the massif consists of a large syenitic intrusion, rising from the surrounding
plain from 600m to 3,000m above sea level. Steep slopes and shallow dystric-fersialic soils
limit the agricultural suitability of the forest to some lowlands on the southern and eastern
edge of the reserve (Paris 1991a; Pike and Rimmington 1965).
The reserve protects a number of watersheds from erosion. It also shelters
considerable biological diversity that includes a variety of wildlife greater than any other
forest reserve in Malawi and more than thirty endemic flora species (Chapman 1962;
Edwards 1985). The reserve also serves as a tourist attraction and a source of high quality
118
timber. A large Eucalyptus plantation in the southeastern portion of Mulanje was intended
to supply fuelwood and charcoal locally, though the cost of transport has limited its
success. The average population density around the reserve in 1996 was 211 persons km'^,
with the greatest concentration on the southern side of the reserve near Malawi's most
productive tea estates.
B 3.2 LIWONDE NATIONAL PARK
Liwonde National Park is located in south central Malawi centered on 14°50'S,
35°2rE, encompassing 54,633 ha in the Upper Shire River valley. It occupies a flat,
riverine-lacustrine plain covered in mostly mopanic and gleyic soils ill suited for
agriculture (Venema 1991). Liwonde was intended to protect wildlife in the Upper Shire
and is a Malawian example of mopane woodland, broad-leafed, lowland, drought
deciduous woodland and savanna dominated by mopane iColophospermum mopane (Kirk
ex Benth.) Kirk ex Leonard). Along the Shire, Liwonde has extensive marsh and
floodplain areas that include both palm {Borasus aethiopum Mart.) and reed {Phragmites
mauritianus Kunth.) communities.
Three years after being declared extinct in Malawi, the first black rhinoceros
{Diceros bicornis L. minor) were introduced to a sanctuary within the Park in 1993
(Bhima and Dudley 1997). The population in 2000 is five. Once in decline, the Park is
home to the only elephant (Loxodonta qfricana Blumenbach) population in the country
that has grown significantly in the past 25 years (Bhima and Bothma 1997). Tourism is the
primary use of the park, with the objective of attracting high revenue tourists to Malawi.
119
A wire fence on the densely populated western edge of the Park and security measures
combine to enforce stricter protection in Liwonde than in all other protected areas in
Malawi. The Park also serves a critical watershed catchment protection role for the Shire
River, the primary source of the country's electricity. The average population density
around the Park in 1996 was 166 persons km'^ with heavier concentrations along the
northeastern corridor.
B 3 3 DZALA>IYAMA FOREST RESERVE AND RANCH
Dzalanyama Forest Reserve is located in the central, western part of Malawi,
centered on 14°20'S, 33°22'E. It is Malawi's largest forest reserve, encompassing 98,827
ha. The western third of this area is closed-canopy, upland miombo (Ngalande 1995). The
rest is shared with a government agricultural scheme, Dzalanyama Ranch, located in a
low-lying area of open miombo and dambo (grasslands in seasonally inundated drainage
lines). The lowlands are dominated by eutric-fersialic soils that are suitable for agriculture
(Lorkeers and Venema 1991).
The natural woodland and almost 5,000 ha of plantations in Phms kesiya Royle ex
Gordon, Eucalyptus camaldulensis Dehnh., and E. tereticomis Sm. supply fiielwood
locally and to the Lilongwe metropolitan area. Two dams fed from Dzalanyama streams
account for 30% of all water needs for Lilongwe. The western edge of the reserve runs
along a forested area in Mozambique that has seen limited use over the past 25 years. The
rest of the reserve is bounded by a population averaging 119 persons km'^, with heavier
concentrations near tertiary roads that lead to the main national highway.
120
B 3 4 VWAZA MARSH WILDLIFE RESERVE
Vwaza Marsh Wildlife Reserve (I TOO'S, 33°28'E) is composed of a BrachystegiaJulbemardia Miombo woodland, interspersed with montane woodland, dambo grassland,
mopane woodland, and thicket associated with perched water tables. The majority of
Vwaza Marsh is situated on eutric-ferralic soils that are suitable for agriculture (Paris
1991b).
Vwaza is rich in both flora and fauna, with 1,200 plant, 427 bird, 85 mammal, 34
reptile, and 22 amphibian species. It serves to protect biodiversity and wildlife, and to
promote ecotourism (McShane 1985). Dating from the 18th century, Vwaza was created
as a game reserve but also to contain the tsetse fly. The elephant population, estimated at
250 in 1985, has declined to the point that sighting assessments are no longer considered
valid. The eastern Zambian side of the reserve is sparsely populated. However, the land
surrounding the Malawian boundary averaged 95 persons km"^ in 1996, a density due in
part to the creation of numerous tobacco estates over the past 20 years.
B.4 METHODS
B.4.1 APPROACH
To capture the quantity and value of protected area resource demand, it was
necessary to develop a multidimensional approach that captured baseline socioeconomic
information and resource utilization in a quantitative, integrated manner. This approach
involved the integration of qualitative and quantitative ethnobiological and socioeconomic
information concerning household agricultural production and the use of plant and animal
121
species from protected areas, as well as spatial analysis of biophysical data (see Table B. I,
and Appendix A for detailed summary). Central to this effort was a quantitative survey of
427 households comprised of2,205 individuals across the four protected areas. The
survey was based on respondent recall of production and resource utilization activities,
particularly at the species level. A coincident market survey was done to permit
conversion of local volumetric measures to kilograms, and to establish retail prices for
each species.
TABLE B. 1. SUMMARY OF DATA COLLECTION METHODS.
Data Gathering Activity
Primary Objectives
Rapid Appraisal (138 villages)
Interviews with traditional officials
•
Community meetings
• Patterns of livelihood strategies
• Patterns of protected area use
• Crop, livestock, wild resource species list
Focus group interviews
(men and women separately)
• Qualitative specialized use
• Land and resource tenure
• Attitudes towards protection
• Changes in resource availability/access
Village list, locations, resource patterns
Intensive study (17 villages)
Participatory mapping
•
Present and past resource utilization
Key respondent interviews
•
•
•
Quantitative specialized use (223 respondents)
Local unit volume and weight conversions
Local market retail prices
Resource assessment
(136 plots)
Formal survey
(427 households)
•
•
•
Vegetation measurements
Local name / Latin name verification
Income (agriculture production, livestock.
remittances, off-farm activities, etc.)
Protected area resource utilization
•
122
With the exception of the single formal survey, data were gathered through
participatory methods: local inhabitants carried out the investigation, presentation, and
preliminary analysis under the guidance and training of the research team (Campbell,
Luckert, and Scoones 1997; Chambers 1990, 1994). The field research team was made up
of four Malawian male and female enumerators conversant in the key local languages
(primarily Chichewa, Chiyao, and Chitumbuka).
Overlapping key respondent interviews captured specialized resource use (i.e.,
small enterprises involving fiielwood, charcoal, wild foods, hunting, handicrafts, tool
making, healing, etc.). Data were collected through interviews conducted in the villages
and in the protected areas during resource extraction. These surveys were essential in
linking the size and price of final products back to the physical quantity of species used.
This was particularly important for uses where the relationship was not one to one (i.e.,
handicrafts, healing services). All monetary values are reported in Malawian Kwacha
(MK), where 1 USD = 15 MK during 1996 and 1997.
The aggregate results provided an estimate of protected area resource demand for
1996. A geographic information system was employed to assess the sustainability of use
over time. Rural population estimates based on population totals and growth trends
derived from the 1998 census were mapped according to Environmental Planning Areas
(EPA), the smallest administrative unit available for population data. It was necessary to
assume that the rural population was evenly distributed across each EPA and then limit the
analysis to the portion of each EPA adjacent to the protected areas of interest with a
resource utilization "zone of influence." The extent of influence was on average 5
123
kilometers, determined by surveys during the rapid appraisal and verified in independent
research (Brouwer et al. 1997). Finally, the population for urban centers located within
the zone of influence as added to the rural total. Total protected area resource demand
equaled the total population in the zone of influence times the mean per capita resource
utilization estimates (kg) for 1996.
The threshold for sustainable protected area resource extraction was defined as the
volumetric (m"* ha"' yr"') mean annual increment (MAI) of all woody biomass in the
protected area, based on a roundwood equivalent (wood in its natural state as felled, or
otherwise harvested). The threshold was tied to woody resource extraction because the
combination of forest clearing for agriculture and flielwood consumption overwhelm all
other biomass use, and generally represent the destruction of habitat. Vegetation
classifications created fi-om 1994 Landsat Thematic Mapper satellite imagery during PLUS
(Orr et al. 1998) were converted into biomass maps using MAI estimates for each
vegetation type developed by FRIM (Masamba and Ngalande 1997). The threshold for
sustainable extraction was the total of the MAI estimates for each vegetation class
multiplied by the total hectares for each class across the protected area.
Per capita estimates for total protected area resource demand in 1996 were held
constant and applied to population estimates based on 1987 through 1998 growth rates
(Malawi Government 1998). The total demand for protected area resources was compared
to the estimated sustainable supply through time to determine when that sustainable supply
would be exhausted.
124
B.4 2 SPECIES IDENTIFICATION
The Forestry Research Institute of Malawi (FRIM) and the National Herbarium
provided indispensable assistance with local species names. A field botanist was present at
all 136 resource assessment plots, working with local inhabitants to confirm all local
names for each species identified. Because scientific species names did not always
correspond directly with local (and sometimes polysemic) names, a particular eflfort was
made to incorporate local perceptions and classifications into all instrument responses and
into how survey questions were posed (Martin 1995).
These plots were not spatially nor temporally sufficient to capture all species (and
their products) identified on village and household surveys. National experts, and
Malawi's rich tradition of gathering ethnographic biological information, proved
invaluable where local confirmation was not possible. The extensive plant dictionary of
Binns (1972) was used for confirming Latin names. The works of Williamson (1975),
Morris and Msonthi (1996), and Morris (1990) were essential for addressing gaps in local
plant, fhiit, and mushroom descriptions, and provided the foundation for evaluating
species use.
Nomenclature for mammals was taken from Ansell and Dowsett (1988), birds from
Benson and Benson (1977) and McShane (1985). Nomenclature for invertebrates was
taken from (Sweeney 1970), and fish from and fish from Ribbink et al. (1983), and
Tweddle and Willoughby (1979). These were all was supported by detailed
ethnobiological descriptions of use (CCAM, 1992; Hayes 1978; Kelly 1993; Morris 1998).
125
A summary of the species by type and lifeform is contained in Table B.2. Of the
694 species encountered during the study, 101 local identifications, accounting for 4.6%
of the total protected area species used by households, could not be verified. The majority
of these were annual forbs and insects physically unavailable at the time of data collection.
TABLE B.2. NUMBER OF SPECIES AND
HOUSEHOLD OBSERVATIONS OF USE.
Wild Use
No. of
Domestic Use No. of
Lifeform
Species
Species
Lifeform
Mammal
Bird
Fish
Insect
Honey
*Mushroom
Tree
Shrub
Climber
*grass
Forb
34
9
18
28
1
9
235
54
39
46
141
Animal
Field crop
Fruit tree
Wood tree
10
44
15
ii
Total species
614
Total species
80
Household
Household
observations
12,604 observations
3,720
of use
of use
*several of households could not provide individual
species names for mushrooms and grasses.
Due to cultural and linguistic variation in protected areas and nearby villages, a
number of local names were obtained for each species identified, the vast majority of
which were available in species dictionary (Binns 1972). We therefore limited reporting
here to the most common name cited during the data collection process.
126
B.5 RESULTS: SPECIES USE
By integrating data collection methods for the assessment household production
and protected area resource utilization, results can be presented from the perspective of
the species (by use) as well as the household. In order to normalize for variance in
household size, most results are reported "per capita," and averaged to reflect each overall
protected area.
8 5 1 DOMESTIC PRODUCTION
Domestic production included field crop and fruit trees cultivation, wood lots for
fuelwood and timber, and livestock production. The mean per capita value of total
domestic production in Dzalanyama (3,975 MK) and Vwaza (4,587 MK) was between
100 and 150% greater than that of Mulanje (1735 MK) and Liwonde (1705 MK). The
major difference between the communities surrounding these reserves is access to land,
primarily a function of population density. High population density around Mulanje and
Liwonde limits farmers to half the land area for cultivation available to farmers around
Dzalanyama and Vwaza (Table B.3).
TABLE B 3 POPULATION, LAND, AND AGRICULTURAL PRODUCTION.
Protected
Mean
Population
Mean
Mean
Crop
Area
Density
Land Holding
Crop
Production
Production
kg capita ' yr ' MK capita ' yr '
persons km^
ha capita '
1658.36
Mulanje
308.0
211
.146
Liwonde
1553.23
166
365.1
.194
3238.98
Dzalanyama
119
819.8
.316
Vwaza
95
581.8
2424.46
.329
127
Crops represented over 70% of the mass and 90% of the value of domestic
production in all four protected areas. The variation in agroclimatic conditions around the
four reserves permit cultivation of cash, non-food crops (primarily tobacco) in some areas
and not in others. The proceeds from non-food crops made a major difference in Vwaza
(primarily tobacco, Nicotiana tabacum L.) representing 45% of the total value of
domestic production, compared to 14, 6, and 2% in Dzalanyama, Liwonde, and Mulanje.
Though farmers interviewed in this study tend to cultivate a wide variety of crops
(6.36 different species per household), maize {Zea mays L.) was the dominant food crop
in all but Mulanje, where fhiit species were equally important (Figure B.l). Tobacco was
the dominant cash crop. Agroclimatic conditions in Vwaza are particularly suited for the
Burley variety, which explains the much higher value of agricultural production around
that reserve. Households around all four protected areas cultivated smaller amounts of a
variety of other crops to supplement the maize production (Figure B.2). In Mulanje, the
combination of banana (Musa paradisiaca L. and M sapientum L.) and Ananas comosus
(L.) Merr.) accounts for 35% of agricultural crop income. Groundnuts (Arachis hypogaea
L .) provide 16% of Dzalanyama's and 20% of Vwaza's crop income, while rice (Oryza
saliva L. (Gram)) provides 14% in Liwonde. Tomatoes (Lycopersicon esulentum Mill.)
grown in communities Dzalanyama and Liwonde are regularly sold, providing 14% and
18% of agricultural proceeds, respectively.
Only 9% of Malawi's food protein is met by livestock production, compared to
20% for other developing countries (Banda and Kamwanja 1993). In this research,
livestock contributed only 2% of household income in Mulanje, Liwonde, and Vwaza, and
128
4% on Dzalanyama (Table B.4). Household livestock production is higher near
Dzalanyama in part because the protected area also serves as the largest cattle ranch in the
country. The same individuals hired to tend small herds of the Ranch's cattle also own
their own animals. In the other three reserves, chickens, goats, and pigs are the most
important livestock species (Figure B.3).
TABLE B.4 LIVESTOCK AND WOOD PRODUCTION ON AGRICULTURAL
LAND.
Protected
Mean
Mean
Mean
Mean
Livestock
Wood
Area
Wood
Livestock
Production
Production
Production
Production
kg capita"' yr"' MK capita"' yr"' kg capita"' yr"' MK capita"' yr"'
Mulanje
2.9
13.36
129.1
31.23
Liwonde
3.5
38.63
114.1
12.16
Dzalanyama
19.1
155.71
326.5
32.67
Vwaza
14.2
80.95
124.5
12.57
The well-documented importance of trees on farms in Malawi (e.g. Dewees 1995)
is evident in our findings. Adjacent to Mulanje, trees provide the majority of agricultural
income, while around Dzalanyama households plant and maintain over four times the
number of trees per capita for wood production than the other three protected areas. In
villages adjacent to Dzalanyama, 46 of the 47 trees capita ' found on agricultural land are
not for food. They contribute 327 kg capita"' yr"' of wood, most of which is sold (Table
B.4). By contrast, only 10 trees capita"' are found on agricultural land in the other three
protected areas, and between 20% (Liwonde and Vwaza) and 50% (Mulanje) of these are
maintained primarily for their fhiit.
Eucalyptus spp. dominate wood produced on agricultural land, most of which is
consumed as fuel wood or poles in house construction in Mulanje, Liwonde, and Vwaza
129
(Figure B.4). The exception is Dzalanyama where relatively large agricultural tracts are
dedicated to as many as 1000 fast-growing, fiielwood trees. A portion of these are
regularly harvested, cut into short sticks in preparation for the 60 kilometer bicycle
journey to Lilongwe where they are sold to meet urban demand.
B.5 2 PROTECTED AREA NATURAL RESOURCE UTE^IZATION
Of the 610 protected area species encountered during the study, 580 were
described as collected and used by at least one household. The quantities collected of the
most important species for each major category of use are reported below. Due to space
limitations, most use category values are reported in aggregate only.
B.5.2.1 Overall utilization
Though many factors influence the level of natural resource utilization, inhabitants
around all four protected areas raised the issue of barriers to access as most important.
These included government agency efforts at protection (i.e. policing with forest guards,
wildlife reserve and national park scouts, and fencing), and physical barriers (primarily
distance from household, elevation and slope, and flooding). Table B.5 provides a relative
assessment of barriers to access along with mean per capita figures for overall protected
area resource utilization for each protected area.
130
TABLE B 5 BARRIERS TO PROTECTED AREA ACCESS AND OVERALL
UTILIZATION.
Protected
Relative Ease
Relative
Mean
Mean
Area
Level of
of Physical
Total
Total
Protection
Access
Utilization
Utilization
kg capita"' yr"' MK capita ' yr''
Mulanje
388.51
Low
Very Difficult
873
Liwonde
379.35
484
Very High
Difficult
Dzalanyama
520.94
Low
Easy
1135
Vwaza
681.90
Easy
1449
Medium
Residents around Liwonde National Park reported the lowest natural resource
utilization, due to protection, and at least in part, to fears of government detection of their
activities. The fear may have inhibited some respondents, resulting in an underestimate of
Liwonde utilization. However it should be noted that the most common request during the
Liwonde rapid appraisal was for a tour of the Park in order to see, for the first time, the
large wildlife species inside. While scouts are not constant fixtures on the boundaries,
Liwonde is the base of training for wildlife scouts throughout Malawi, and is fenced on its
more populated western boundary. One kilometer east of that boundary is the Shire River
and its flood plain, encouraging a high fishing trade, but serving as an effective barrier to
most other uses.
The exceptionally steep slopes of Mulanje Mountain and local beliefs about the
plateau itself limited access to the lower slopes. Dzalanyama had the fewest physical
barriers, and like Mulanje, is a forest reserve with fewer use restrictions and limited agency
personnel available for protection. Inhabitants around Vwaza Marsh reported high levels
of policing near a small government base camp at Lake Kuzuni, but much lower levels
131
away from that camp. Flooding limited, but did not prohibit, access to parts of the reserve
during the rainy season.
Comparing total utilization across uses is problematic due to variation in life form
(plants vs. animals), or even parts within similar life forms (leaves for medicine vs. wood
for fuel). Therefore, the species-level results are presented by type of use. However, one
species stands out from all others in the data set.
In terms of the quantity of biomass extracted, fuelwood was the most important
natural resource product extracted from all four reserves (Figure B.5). Despite the
differences in total utilization between the reserves (Table B.5), the relative importance of
fuelwood to other uses was consistent. Fuelwood accounted for 48% of all biomass
utilization in Vwaza, 60% in Liwonde and Dzalanyama, and 64% in Mulanje.
Food resources provided the largest share of protected area resource value among
all use categories (Figure B.6). The combination of plants (including mushrooms) and
animals/animal products (including honey) represented 43% of utilization value in Vwaza,
57% in Mulanje, 63% in Liwonde, and 66% in Dzalanyama.
When totals are aggregated across all uses, the most important individual species
collected was the semi-deciduous fhiit tree known locally as masuku. At 82.6 kg capita"'
yr"', the amount of Uapaca kirkiana biomass extracted was nearly double that of its
nearest competitor, Pericopsis angolensis. On value, Uapaca kirkiana was 2.7 times more
important than the next greatest protected area species, mushrooms. By providing 83.11
MK per capita, masuku would rank as the tenth most important agricultural crop in the
data set, ahead of all domesticated fhiits but mango and banana. These exceptional
132
utilization levels and the multipurpose nature of this tree (food, fiiel, tools, medicine, etc.)
are why Ngulube (1995) has suggested this is a "Cinderella" tree, known well to those
who use it extensively, but generally overlooked by horticulturists and agroforestry
researchers.
B.5.2.2 Food utilization
Over 81% households surveyed reported collecting food products from the
protected areas. Of these households, 86% used at least part of that food to supplement
inadequate food stocks and 72% sold protected area food products for cash. Protected
area plant food utilization was highest in Dzalanyama and Mulanje, while meat utilization
was highest in Vwaza and Liwonde, home to much larger wildlife and fish populations
(Table B.6).
TABLE B.6. PER CAPITA PROTECTED AREA FOOD UTILIZATION.
Protected
Mean
Mean
Mean
Mean
Area
Meat*
Meat*
Plant Food
Plant Food
Utilization
Utilization
Utilization
Utilization
kg capita"' yr"' MK capita'* yr"' kg capita ' yr ' MK capita"' yr '
Mulanje
52.90
12.8
135.6
175.33
Liwonde
49.4
221.59
31.8
17.10
7.8
Dzalanyama
41.25
241.6
306.25
Vwaza
203.3
t
229.42
t
65.5
88.08
*wild animal meat utilization, including honey and insects used for food.
tVwaza's meat utilization is high in part due to one household that reported hunting 10
elephants, selling most of the meat for much less than other reported wildlife meat sales.
Though very rare in Liwonde National Park, and relatively uncommon in Vwaza,
the fruit of Uapaca kirkiana was still the most important food, collected by half of all
households surveyed, and representing over half the plant food biomass extracted from
both Mulanje and Dzalanyama (Figure B.7). It, and the second most commonly collected
133
fruit, Parinari ciiratellifolia, were ranked as the top two priority fruits identified by
Malawian farmers to be included in a domestication program proposed by researchers
from the International Centre for Research in Agroforestry (Malembo, Chilanga, and
Maliwichi 1998).
Over 65% of the households surveyed reported collecting mushrooms, and 35% of
those sold at least some for cash. Though individual species were rarely reported, the wide
variety included Amanita hemibapha (Berk, and Br.) Sacc., Coprinus qfricanus Pelger,
Russtila schizoderma Pat., Strobilomyces costatipora (Beeli) Gilb, and Termitomyces
clypeatus Heim.. In Liwonde, flood plain areas produce a wild rice {Oryza
longistaminata) close to neighboring villages, and is the important plant food collected
from within Park boundaries. Tubers of ground orchids were important in all four
reserves, particularly Habenaria walleri in Dzalanyama.
A large number of other tuber, fruit, and green leafy vegetable species were
collected from each protected area, the most important of which are reported in Figure
B.8. Four different wild yams {Dioscorea spp.) were reported in Mulanje and another in
Liwonde as necessary for meeting food needs when stores of maize run out during the
rainy season. Species of the genus Dioscorea have been found to have adequate nutritional
value to permit populations living in the African tropical rainforest to live independently of
agriculture (Hladik and Dounias 1993).
The total quantities of green leafy vegetables collected are lower than fhiits and
tubers on the basis of weight. This underestimates their importance to Malawian diet as
they are regularly served with the maize-based staple food called nsima. Across all four
134
protected areas, 72 different vegetable species were observed as used by at least one
household, and of these 40 were described as species used specifically to help bridge the
"hungry season" from the onset of rains to dry season harvest. The most important of
these included Bidenspilosa L., Amaranthus thunbergii Moq., and Solanum nigrum L..
The elevation in Mulanje provides habitat for Momordicafoetida Schumach. and Thonn.,
and Thimbergia lancifolia T. Anders., which explains their importance for communities
near the mountain and their absence in the other three protected areas.
Proximity to Lake Malombe and the Shire River contributed to high overall meat
utilization in communities around Liwonde National Park (Table B.6; Figure B.9). In both
Liwonde and Vwaza the most important fish species utilized was sharptooth catfish
(Clarias gariepinus Burchell). Despite similarly high quantities collected in Vwaza, the
value of fish was much higher adjacent to Liwonde due to the proximity of major urban
markets in Blantyre and Zomba. Utilization of large mammals was higher in Vwaza and
Liwonde because these species are simply not prevalent in either forest reserve (Figure
B.IO).
The largest single contributor to overall wild animal utilization was actually 109.5
kg capita ' of Loxodonta qfricana, due to a single household in Vwaza that reported
hunting 10 elephants in 1996. The vast majority of the 27% of households that fished
inside the protected areas sold some portion of their catch. Hunters were present in 48%
of households, and those hunting large mammals also sold some of their quarry. However,
almost 40% of the households that collected meat from the protected areas concentrated
135
on smaller mammals like the edible rodents (classified locally as mbewa, mostly of the
Muridae family).
The most common insects collected from the protected areas for food were
caterpillars of the Lepidoptera order, winged termites (Macroiermes spp.) and
grasshoppers {Acanthacris ruficomis (Audinet-Serville)) and Cyrtacanthacris aeruginosa
(Stcll). Other species included the sand cricket {Brachytrypes membranaceus), the black
flying ant {Carebara vidua F. Smith), the red locust {Nomadacris septemfasciata
(Serville)), cicadas {Cicada spp.) and a shield bug {Sphaerocoris sp ). Though 36% of all
households surveyed reported collecting insects, only 11% of these reported any sales. In
addition, 11% of households (mostly in Vwaza and Dzalanyama) collected honey from the
protected area, half of which some sold for cash. The miombo woodland bee species
responsible for this honey are Meliponula bocandei Spin, Trigona spp., and the
domesticated Apis mellifica adansonii (Latr.) (Parent, Malaisse, and Verstraeten 1978).
Though the reported sales of insects and honey were lower than other secondary food
products, when queried, local inhabitants did not dismiss their economic potential. This
supports the suggestion by Munthali and Mughogho (1992) that caterpillar utilization and
bee keeping in Malawi could provide almost 60% more income per hectare than traditional
agriculture in ecologically appropriate locations, if the economic incentives to develop
such enterprises were made available.
136
B.5.2.3 Fuehvood and comtniction utilization
During the rapid appraisal, inhabitants of villages adjacent to Mulanje,
Dzalanyama, and Vwaza reported obtaining the majority of their fiielwood far from the
protected area, whereas Liwonde residents suggested that less than half their needs were
met by the National Park. The quantitative assessment of total fiielwood collected in the
villages studied intensively corroborated with the rapid appraisal (Table B.7.), with
Liwonde residents collecting only half that of the other three reserves, primarily because
prohibition on protected resource use is well enforced. The Mulanje fiielwood quantity of
560.8 kg capita'' yr ' is similar to results obtained by Abbot and Homewood (1999) for
Lake Malawi National Park (c. 10.1 kg capita"' yr"' or 525.2 kg capita"' yr"'). The amounts
for Dzalanyama (677.2 kg capita"' yr"') and Vwaza (696.5 kg capita"' yr"') correspond well
with 18 estimates for southern Africa reviewed in 1993 by Shackleton (687 kg capita"' yr"'
± 48.8 kg capita"' yr"').
TABLE B.7. PER CAPITA PROTECTED AREA FUELWOOD AND
CONSTRUCTION UTILIZATION.
Protected
Mean
Mean
Mean
Mean
Area
Fuelwood
Construction
Construction
Fuelwood
Utilization
Utilization
Utilization
Utilization
kg capita"' yr"' MK capita"' yr"' kg capita"' yr"' MK capita"' yr"'
Mulanje
560.8
3.24
7.2
72.91
Liwonde
10.00
288.3
26.9
37.48
Dzalanyama
677.2
35.7
11.50
88.03
Vwaza
696.5
66.99
208.8
90.54
The species extracted from the protected areas for fiielwood varied from village to
village due to biophysical conditions or the location of government plantations. In all, 181
different species were observed as being used for fiiel. While no individual species
137
overwhelmingly dominated, the top 10 species collected accounted for 40% of all
flielwood used, and the top 30 species accounted for 73% (Figure B. 11).
Fuelwood species preferences corresponded to those found by Abbot et al. (1997),
working in central Malawi's Chimaliro Forest Reserve. In a community ranking procedure
they found Julbemardia paniculata and Pericopsis angolemis, to be the most preferred
of sixteen miombo woodland species studied. Using a fuelwood index, they found
Pericopsis angolensis to be the highest quality fuelwood. In the current study, Pericopsis
angolensis was the most popular fuelwood species, accounting for 38.1 kg capita ' yr ', or
6.8 % of all fuelwood collected from the four protected areas, vjhAe Julbemardia
paniculata ranked third at 24.4 kg capita"' yr ', or 4.3%.
Though all species of the genus Brachystegia evaluated by Abbot et al. (1997)
ranked among the less fuel-efficient and of lower community preference, their widespread
availability in Mulanje, Dzalanyama and Vwaza made these species important for meeting
fuel needs. The combined utilization of the ten Brachystegia species encountered in this
study (including Liwonde National Park, where the genus is uncommon) was 86.6 kg
capita ' yr ', or 15.4% of all fuelwood collected. In Liwonde, Colophospermum mopane
was the most important protected area fuelwood species, followed by Combretum spp.
(which were also selected frequently for fuel in Vwaza).
Exotic plantations of Eucalyptus spp. and Pittus spp. within protected area
boundaries met 24% of the reported protected area fuelwood demand in Mulanje. The
only other reserve-based plantations encountered in the sample were of E. camaldulensis
138
and E. tereticomis in Dzalanyama, accounting for 5% of the reported fiielwood utilization
there.
Demand for protected area wood for construction purposes (poles, timber,
carpentry, etc.) ranged from only 7.2 kg capita"' yr ' around Mulanje to 208.8 kg capita"'
yr"' around Vwaza (Table B.7). Inhabitants adjacent to Mulanje indicated that mountain's
steep slopes discouraged all but construction specialists from sourcing poles or timber
from the reserve. In Vwaza, limited physical barriers and exceptional demand led to very
high utilization rates. The demand resulted from the liberalization of tobacco legislation in
the early 1990's, permitting smallholders to grow a Burley variety that requires drying
sheds for air curing, commonly constructed as a latticework of poles. Smallholders
surrounding Vwaza have shifted rapidly into Burley cultivation, precipitating a heavy
demand for poles to construct sheds.
The greatest construction demand placed on an individual species in one protected
area occurred in Liwonde (Figure B. 12). Colophospermum mopane, used at a rate of 20.4
kg capita"' yr"', accounted for 77% of the poles and timber extracted from the Park by
neighboring villages. In the other three protected areas, no individual species dominated.
Limited alternatives and the light weight of bamboo species Arundinaria alpina
placed it in higher demand than any wood species extracted for construction in Mulanje.
The most common protected area bamboo species used for construction in Liwonde and
Dzalanyama was Oxytenanthera abyssinica.
139
B.5.2.4 Fiber, tool, and handicraft utilization
During the rapid appraisal it became clear that fiber, primarily for maldng rope,
was an important natural resource product that was often secured from protected areas.
The quality varied from "high" when used to bind poles during house construction to
"low" when used to secure headloads of fuelwood. Fiber was most commonly obtained by
stripping the bark of a sapling (1.5 - 10 cm diameter) of a preferred woody species and,
separating the fiber layer from the cortex. Key respondent interviews with house
construction specialists indicated that 30 to 50 saplings yielded the fiber necessary for
construction of a typical earthen, thatched roof home (9 m^). This is comparable to the 35
saplings found to be necessary by Peham (1996) in research on Liwonde National Forest,
just south of Liwonde National Park.
Protected area fiber utilization ranged from 4.8 kg capita"' yr"' in Liwonde to 21.1
kg capita ' yr"' in Vwaza (Table B.8). The lower figures reported in Liwonde are in part
due to a long tradition of extracting the nylon linings from discarded tires. The strips
(called linja) are used for construction in general and by fishermen to make seine nets.
TABLE B.8. PER CAPITA PROTECTED AREA FIBER, TOOL AND HANDICRAFT
UTILIZATION.
Protected
Mean
Mean
Mean
Mean
Area
Tools/Crafts
Fiber
Fiber
Tools/Crafts
Utilization
Utilization
Utilization
Utilization
MK
capita"' yr"'
kg capita"' yr"' MK capita"' yr '
kg capita"' yr"'
Mulanje
6.9
58.02
2.19
61.9
Liwonde
34.18
4.8
1.50
29.5
Dzjilanyama
9.9
3.17
12.21
12.9
Vwaza
43.19
6.62
21.1
43.0
140
In Dzalanyama, higher fiber usage was attributed to the construction of animal pens for
livestock, while in Vwaza, fiber is used to bind the pole-based drying sheds for air curing
of Burley tobacco.
Species selection depended on availability and use, with preference being given to
woody species (i.e. Brachystegia spp.) of known strength and durability (Figure B. 13).
Fiber derived fi"om sisal leaves {Agave sisalana) was used to make a high-quality string,
and an herbaceous species such as Cissampelos mucronata was used for binding in basket
making.
The manufacture of some utilitarian products also required fiber (e.g. hand
brooms), however, the portion of biomass utilized was accounted for under the category
"handicrafts" listed in the right-most columns of Table B.8. Over 61% of Mulanje's 61.9
kg capita"' yr"' utilization in this category came from a carpentry enterprise, and another
36% was spread across several highly specialized hand-broom enterprises on the southern
/
side of the mountain that supply markets throughout Malawi. The majority of Liwonde's
utilization was associated with basket-making enterprises servicing nearby Mangochi. In
contrast to this, Vwaza's totals were spread across the majority of households surveyed;
66% made mats, baskets and/or utilitarian wicker tools, and 81% made wooden hand
tools. Dzalanyama's utilization in this category was the lowest (12.9 kg capita * yr"'), with
a limited handicraft industry. Almost three-quarters of the utilization reported there was
associated with the manufacture of wooden hand tools.
The quantity of biomass extracted fi-om major species used for making crafts and
tools is documented in Figure B.14. The key carpentry species extracted in Mulanje
141
included Brachestegia spp., Burkea qfricana, Pericopsis angolensis, and Uapaca
kirkiana. The hand brooms involved a combination of Xerophyta splendens and Sida
acuta, generally supplemented with Psychotria zombamontana, Hypoxis nyasica Bak, or
Triumfetta rhomboidea. The baskets in Liwonde and Vwaza were predominantly made
from the fronds of Borassus aethiopum. Mats in all protected areas were made with P.
mauritiatms if available. Where not available, they were made with Pennisetum pttrpurium
or Hyparrhenia rufa. Tool handles were most commonly made from the durable, nonsplitting suffrutices of Acacia polyacantha, Annona senegalensis, Brachestegia
spiciformis, Julbemardia paniculata, and Xeroderris stuhlmannii.
B.5.2.5 Thatch and medicinal plant utilization
Utilization of protected area grasses for thatch roofs ranged from Liwonde's 52.0
kg capita' yr' to almost four times that amount in Vwaza (Table B.9). Most of the
residents who did not use protected area thatch could obtain the necessary grasses from
customary land. Others had lower overall demand because they did not have thatch roofs:
over 18% of the households surveyed in Mulanje had installed corrugated iron roofs
versus only 1% in Vwaza.
TABLE B.9. PER CAPITA PROTECTED AREA THATCH AND MEDICINAL
PLANT UTILIZATION.
Protected
Mean
Mean
Mean
Mean
Area
Medicinal
Thatch
Thatch
Medicinal
Utilization
Utilization
Utilization
Utilization
kg capita"' yr"' MK capita"' yr"' kg capita"' yr"' MK capita"' yr"'
Mulanje
9.55
14.38
86.8
0.4
Liwonde
51.77
5.72
52.0
1.4
Dzalanyama
42.20
148.4
16.32
1.7
Vwaza
205.0
134.52
22.55
5.5
142
The protected area utilization quantities of the most important major and minor
thatch species are displayed in Figure B.15. The genus Hyparrhenia accounted for 57% of
the thatch grass extracted from the protected areas. Only in Mulanje did a nonHyparrhenia species {Heteropogon contortus) predominate. A number of households
were unable to report thatch usage on the basis of individual species. In those cases, thatch
was reported in aggregate as "udzu," or general grass.
A special effort was made to elicit information about the use of protected area
species for their medicinal properties. During the pilot study and throughout the rapid
appraisal a view was expressed by local inhabitants that researchers (in general) might
exploit local knowledge about medicinal plants. The research team therefore agreed to
focus on aggregated quantities and value rather than the specifics of medicinal use,
particularly since exceptional detail about these products and their uses in Malawi has
already been documented by Morris and Msonthi (1996) and Williamson (1974). Despite
the agreement to limit disclosure, the research team felt that particularly in Mulanje and
Liwonde, protected area medicinal plant usage was largely under-reported. Only in Vwaza
did the majority of households feel comfortable in disclosing this information.
The aggregated quantities and value of reported protected area medicinal natural
resource product utilization is reported in the right-most columns of Table B. 10. Across
the four protected areas, 272 different species were reported as used for medicinal
purposes. For most species, far more biomass was extracted in order to create root or
stem-based medicines than for leaf, bark, or fruit based medicines. To account for this.
143
species importance was determined by the number of households using the species, rather
than the total biomass extracted to create those medicines. The top fifteen protected area
species utilized for medicinal purposes fcv each protected area is reported in Table B. 10.
In order to respect our agreement with local inhabitants, species are ranked only.
TABLE B. 10. THE FIFTEEN MOST FREQUENTLY REPORTED SPECIES
USED FOR MEDICINAL PURPOSES IN EACH PROTECTED AREA.
Medicinal Plant
M L
D
Species
--rankAlbizia zimmermcmii Harms
Allophyliis africanus Beauv
is
•
Annona senegalensis Pers.
Antidesma venosum E. Mey. ex Tul.
8
*
•
Asparagus africanus Lam.
—
Borassus aethiopum Mart.
9
Bridelia carthatica Bertol. f
5
*
Burkea qfricana Hook.
Byrsocarpus orientalis (Baill.) Bak.
Cassia abbreviata Oliv.
J
*
Cassia sp.
2
Catunarecam spinosa (Thunb)
11 •
*•
Cyphostemma junceum (Webb) Descoings
12
Cyphostemma nierieme
14
2
*
Dalbergia nitudula Welw. ex Bak.
*
Desmodium velutitmm (Willd.) DC.
8
Dicoma anomala Sond.
5
1
Dicoma sessiliflora Harv. Sub sp. sessiliflora
6
Euphorbia hirta L.
9 •
Fagara macrophylla (Oliv.) Engl.
14
*
Fiais capensis Thunb.
Flacourtia indica (Burm. f ) Merr.
11
Heteromorpha trifoliata (Wendl.) Eckl. and Zeyh.
4
Iboza riparia (Hochst.) N. E. Br.
13
Leptactina benguelensis (Welw. ex Benth. & Hook.f) Good
Melia azedarach L.
13
*
Mondia whytei (Hook. F.) Skeels
5
Olax obtu^oHa De Wild.
10
Olinia usambarensis Gilg
1
Ozoroa reticulata (Bak. f ) R. and A. Fernandez
3
*
»
Pterocarpus angolensis DC
4
—
—
—
13
—
—
—
10
~
~
—
—
—
V
—
*
15
*
—
8
9
—
J
—
—
—
—
—
*
—
—
~
—
—
—
—
5
*
6
~
—
—
—
4
—
~
12
—
—
—
*
—
—
—
—
~
—
~
—
—
—
—
—
—
7
~
*
—
144
Medicinal Plant
M L
D V
Species
—rank—
3
Piliostigma thonningii (Schumach.) Milne-Redh.
2 13 1
Pseudolachnostylis maproimeifolia Pax
15 *
*
*
Psorospermum febrifugum Spach.
1
*
Rhus longipes Engl.
6 •
7 12 7 11
Stegcmotaenia araliacea Hochst.
*
7
Schrebera alata (Hochst.) Welw.
*
Sclerocarya cqffra Sond.
10 *
*
9
Seciiridaca longepedunculata Fresen.
14
*
Strombosia scheffleri Engl.
15
*
14
Terminalia sericea Burch. ex DC.
Terminalia stenostachya Engl, and Diels
11
Venwnia amygdalina Delile
8
Vigna fischeri Harms
10
6
Zanha africana (Radlk.) Exell
4
2
*
*
Zizyphiis mucronata Willd.
12
M = Mulanje; L = Liwonde; D = Dzalanyama; V = Vwaza
* = reported as used, but not among top 15; ~ = not reported as used for medicine
~
—
—
~
—
~
—
—
if
—
~
—
—
—
—
~
—
~
—
B.6
RESULTS: SUSTAINABILITY ANALYSIS
Protected area resource utilization data collected at the species level can be
aggregated in order to evaluate sustainable use for the entire protected area over time. In
the case of Malawian protected areas, the two predominant pressures are felling trees in
preparation for farming, and extracting wood for fiiel, construction, etc. While poaching
and the over-utilization of fruits and medicinal plants are equally important, the loss of
habitat as a consequence of woody species extraction threatens both wildlife and rare
plants. Our research therefore focused on sustainable supply of wood as a measure of
overall sustainable use.
Sustainability analysis at the protected area level shows that in 50 years, the local
demand for protected area wood will exceed sustainable supply in all but Dzalanyama,
145
where supplies will last another 50 years (Table B.11 and Figure B.16). Resource demand
was based on the average wood utilization per capita multiplied by the population in the
zone of influence for each protected area. Each protected area has very different projected
rates of population growth and wood consumption, resulting in varied rates of resource
decline. For example, the sustainable supply of woody biomass in highly protected
Liwonde is projected to last a few years longer than Vwaza, despite the much larger
resource base available inside Vwaza Wildlife Reserve.
TABLE B 11. SUSTAINABILITY OF WOODY PROTECTED AREA RESOURCES.
—Forecast After 1996—
— 1996 EstimatesWhen
Estimated How Many
Would
Per
Sustainable Total Capita
Zone of Population More People Sustainable
Wood
Wood Wood
Influence
Growth
Could Be Supply Be
Supply Demand Demand Population
Rate
Supported? Exhausted?
people
people
year
-m %
37,765
M 163,233 127,686 628
135,648
0.5
2046
155,547
2049
L 125,122 48,844 327
99,604
1.8
D 395,413 81,995
730
74,868
286,177
2096
1.6
V 233,937 61,763 951
43,316
2.7
120,751
2046
M = Mulanje, L = Liwonde, D = Dzalanyama, V = Vwaza
B.7 CONCLUSIONS
The quantity and consumptive use value of protected area species can be used to
inform natural resource management in each protected area. Knowledge of the level of
demand for an individual species can be compared with field survey analysis of its
ecological status. Where rates of demand are deemed excessive, alternative species in the
protected area can be promoted. Moreover, the role each individual species plays in
overall household income can be assessed. Substitutes among agricultural crops or
146
livestock can be identified, and where necessary, alternative income generating activities
can be promoted to reduce the pressure on specific species.
Variation was high among four Malawian protected areas in products extracted
from them and the species that were used. In some cases this was due to biophysical
conditions that favored specific species, or protection policies that target selected species,
such as large mammals. In other cases, local preferences or the availability of alternatives
explained the variation. For example, freshwater crabs are relatively common throughout
Malawi, but are generally favored for consumption in two areas Mulanje National Forest,
and Zomba National Forest, where the crab species Potomonaittes montivagus Chace F. is
found. A similar example is the bush pig (Potamochoerusporcus L.), which is also
relatively widespread, but not reported as consumed in villages adjacent to Liwonde. That
region is predominantly of Yao ethnicity, mostly practitioners of Islam, which does not
permit the consumption of this species.
Despite this variation, for each use category, a few species dominated, none more
than the multipurpose fhiit tree, masuku (Uapaca kirkiana Mull. Arg.). This species was
the most important protected area source of food, was very important for fuelwood, and
was used for construction, fiber, and medicinal purposes. Our findings support the
conclusion of both Ngulube (1995) and Malembo, Chilanga, and Maliwichi. (1998) who
both found the pervasive use of masuku in the wild as in important factor supporting its
candidacy for domestication as an agroforestry tree.
Aggregation of woody species utilization data provides an estimate of overall
protected area resource demand. In three of the protected areas studied, demand will
147
exceed the sustainable supply of woody biomass in 50 years, and in 100 years in the
fourth. This suggests that, barring any additional demand pressure, there is time to
enhance community-based natural resource initiatives that are promoted for all four
protected areas. In addition to their institutional and management complexities,
community conservation efforts must take these time constraints into account.
This analysis of protected area resource demand and sustainable supply has
caveats. The analysis does not account for the spatial impact of encroachment. Though
theoretically barred, encroachment of agriculture into protected areas would effectively
reduce the overall biomass base, thereby accelerating the decline in sustainable supply.
Equally, this analysis does not account for spatial variation in use that might result from
differential access, or the distribution of favored resources. Pressure points that might
receive higher than sustainable utilization can result in habitat fragmentation, a significant
threat to sustainability.
An analysis of aggregate sustainable use provides a context for natural resource
decision making. The analysis, particularly when combined with species-level demand
assessments and physical resource inventories, provides a set of monitoring tools for
community management and conservation. Without regularly updated estimates of
protected area resource demand, threats to individual species can only be assessed by
impact on the resource, a metric applied only after the damage is done. Providing resource
demand information enhances the management options and offers an understanding of the
problem from the perspective of the user. Working without this information where
148
managed use is encouraged limits the tools available to resource managers to find
alternatives for threatened resources.
B.8
LITERATURE CITED
Abbot, J.I.O., and K. Homewood. 1999. A history of change; causes of miombo
woodland decline in a protected area in Malawi. Journal of Applied Ecology
36:422-433.
Abbot, P., J. Lowore, C. Khofi, and M. Werren. 1997. Defining firewood quality; a
comparison of quantitative and rapid appraisal techniques to evaluate firewood
species fi-om a southern Afnca savanna. Biomass and Energy 12;429-437.
Ansell, W.F.H., and R.J. Dowsett. 1988. Mammals of Malawi; an annotated check list
and atlas. Trendrine Press, Zennor, United Kingdom.
Banda, J.W., and L.A. Kamwanja. 1993. Dairy/beef production systems research
programme in Malawi. Pages 27-32 in J.A. Kategile and S. Mubi, eds. Future of
livestock industries in east and southern Afiica. Proceedings of the workshop held
at Kadoma Ranch Hotel, Zimbabwe, 20-23 July 1992. International Livestock
Centre for Afnca, Addis Ababa.
Barrett, C.B., and P. Arcese. 1995. Are integrated conservation-development projects
(ICDPs) sustainable? World Development 23(7)1073-1084.
BDPA/AHT. 1998. Customary land utilization study. BDPA/AHT, Lilongwe, Malawi.
Benson, C.W. and Benson C.M. 1977. The birds of Malawi. Montfort Press, Blantyre,
Malawi.
149
Bhima R., and C.O. Dudley. 1997. The influence of the Shire River on Liwonde
National Park, Malawi, with special reference to elephant movements. Koedoe
40:9-18.
Bhima R, and J. du P. Bothma. 1997. Age structure of elephants in Liwonde National
Park. Koedoe 40:1-8.
Binns, B. 1972. Dictionary of plant names in Malawi. Government Printer, Zomba,
Malawi.
Brandon, K.E., and M. Wells. 1992. Planning for people and parks: design dilemmas.
World Development 20(4)557-570.
Brouwer, LD., J.C. Hoorweg, and M.J. van Liere. 1997. When households run out of
fuel: responses of rural households to decreasing fuel wood availability, Ntcheu
District, Malawi. World Development 25:255-266.
Campbell, B.M., M. Luckert, and I. Scoones. 1997. Local-level valuation of savanna
resources: a case study from Zimbabwe. Economic Botany 51:59-77.
CCAM. 1992. Malawi's traditional and modem cooking. Chitukuko Cha Amayi
m'Malawi, Lilongwe, Malawi.
Chambers, R. 1990. Rapid and participatory rural appraisal. Appropriate Technology
16(4):14-16.
Chambers, R. 1994. The origins and practice of participatory rural appraisal. World
Development 22:953-969.
150
Chapman, J.D. 1962. The vegetation of the Mlanje Mountains, Nyasaland: a preliminary
account with particular reference to the Widdringtonia forests. Nyasaland
Government Printer, Zomba, Nyasaland (Malawi).
Cunningham, A.B. 1994. Integrating local plant resources and habitat management.
Biodiversity and Conservation 3:104-115.
Dewees, P.A. 1995. Trees on farms in Malawi; private investment, public policy, and
farmer choice. World Development 23:1085-1102.
Edwards, I. 1985. Conservation of plants on Mulanje Mountain, Malawi. Oryx 19:86-90.
FAO. 1981. Forest resources of tropical Africa: tropical forest resources assessment
project. FAO, Rome.
FAO. 1999. State of the world's forests. FAO, Rome.
Finan, T.J. 1996. Anthropological research methods in a changing world. Pages 104-120
in Moran, E.F., ed.. Transforming societies, transforming anthropology. University
of Michigan Press, Ann Arbor.
Gibson, C.C., and S.A. Marks. 1995. Transforming rural hunters into conservationists:
an assessement of community-based wildlife management programs in Africa.
World Development 23(6):941-957.
Giuliano, A.R., H. Dahiane, and B. Orr. 1998. Evaluation of US AID Title II Food Aid
Programs: Review of Ten CARE Country Programs. CARE, Atlanta.
Godoy, R., D. Wilkie, H. Overman, A. Cubas, G. Cubas, J. Demmer, K.
McSweeney, and N. Brokaw. 2000. Valuation of consumption and sale of forest
goods from a Central American rain forest. Nature. 406: 62-63.
151
Hayes, G.W. 1978. A guide to Malawi's national parks and game reserves. Montfort
Press, Limbe, Malawi.
HIadik, A., and E. Dounias. 1993. Wild yams of the African forest as potential food
resources. Pages 163-176 in C M. Hladik, A. Hladik, O F. Linares, H. Pagezy, A.
Semple, and M. Hadley, eds. Tropical forests, people and food: biocultural
interactions and applications to development. Man and the Biosphere Series,
Volume 13. UNESCO and The Parthenon Publishing Group, Paris.
House, W.J. and G. Zimalirana. 1992. Rapid population grouth and poverty generation
in Malawi. The Journal of Modem African Studies 30:141-161.
Kelly, D. 1993. Aah...but those elephants: David Kelly paints Liwonde. Malawi Wildlife
Society, Blantyre, Malawi.
Kremen, C., J.O. Niles, M.G. Dalton, G.C. Daily, P.R. Ehrlich, J.P. Fay, D. Grewal,
and R.P. Guillery. 2000. Economic incentives for rain forest conservation across
scales. Science 288(5472): 1828-1832.
Lorkeers, A.J.M., and J.H. Venema. 1991. Land Resources Evaluation Project: Land
Resources Appraisal of Lilongwe Agricultural Development Division.
MLW/85/011 Field Document No. 24. Malawi Govemment/UNDP/FAO,
Lilongwe, Malawi.
Malawi Government. 1998. Report of Preliminary Results: 1998 Population and
Housing Census. National Statistics OfiBce, Zomba, Malawi.
Malembo, L.N., T.G. Chilanga, and C.P. Maliwichi. 1998. Indigenous miombo fruits
selected for domestication by farmers in Malawi. Pages 16-39 in J.A. Maghembe,
152
A.J. Simons, F. Kwesiga, M. Rarieya, eds. Selecting indigenous fruit trees for
domestication in southern Africa: priority setting with farmers in Malawi,
Tanzania, Zambia and Zimbabwe. International Centre for Research in
Agroforestry, Nairobi.
Masamba, C., and J. Ngalande 1997. Inventory data ofbiomass growing stock and
supply. FRIM Technical Report. Forestry Research Institute of Malawi, Zomba.
Martin, G.J. 1995. Ethnobotany; a methods manual. People and Plants Conservation
Manuals, Volume 1. Chapman and Hall, London.
McShane, T.O. 1985. Vwaza Marsh Game Reserve; a baseline ecological survey.
Department of National Parks and Wildlife, Rumphi, Malawi.
Millington, A. Townsend, J., Kennedy, P., Saull, R., Prince, S., and Madams, R.
1989. Biomass assessment; woody biomass in the SADCC region. Earthscan,
London.
Morris, B. 1990. An annotated check-list of the macroflingi of Malawi. Kirkia 13; 323364.
Morris, B. 1998. The power of animals; an ethnography. Berg, Oxford.
Morris, B. and J.D. Msonthi. 1996. Medicinal plants of Malawi. Pages 191-557 in B.
Morris ed., Chewa medical botany; a study of herbalism in Southern Malawi.
International African Institute, London.
Munasinghe, M. 1992. Biodiversity protection policy; Environmental valuation and
distribution issues. Ambio 21;227-236.
153
Munthali, S.M., and D.E. Mughogho. 1992. Economic incentives for conservation; bee­
keeping and Satumiidae caterpillar utilization by rural communities. Biodiversity
and Conservation 1; 143-154.
Ngalande, J.D. 1995. An integrated management plan for Dzalanyama Forest Reserve,
Malawi. M.S. thesis. University of Aberdeen, Scotland.
Ngulube, M.R 1995. Indigenous fruit trees in southern Africa. Agroforestry Today 7; 1718.
Openshaw, K. 1997. Urban biomass fiiels; production, transportation, and trading study.
Alternative Energy Development, Inc/Malawi Ministry of Energy and Mining,
Lilongwe, Malawi.
Orr, B., B. Eiswerth, T. Finan, and L. Malembo. 1998. Malawi public lands utilization
study. Malawi Ministry of Lands, Housing, Physical Planning, and
Surveys/University of Arizona/Forestry Research Institute of Malawi, Lilongwe,
Malawi.
Parent, G., F. Malaisse, and C. Verstraeten. 1978. Les miels dans la foret claire du
Shaba meridional. Bulletin des Recherches Agronomiques de Gembloux 13; 161176.
Paris, S. 1991a. Land Resources Evaluation Project; Land Resources Appraisal of
Blantyre Agricultural Development Division. MLW/85/011 Field Document No.
24. Malawi Govemment/UNDP/FAO, Lilongwe, Malawi.
154
Paris, S. 1991b. Land Resources Evaluation Project; Land Resources Appraisal of Mzuzu
Agricultural Development Division. MLW/85/011 Field Document No. 27. Malawi
Govemment/UNDP/FAO, Lilongwe, Malawi.
Pearce, D.W., and R.K. Turner. 1990. Economics of natural resources and the
environment. Harvester Wheatsheaf, London.
Peham, A.P.K. 1996. Non-timber products as miombo resources: a Malawi case study
with special reference to bark fibre. Diplomarbeit, Institut fur Waldokologie,
Universitat fur Bodenkultur, Vienna.
Pike, J.G., and G.T. Rimmington. 1965. Malawi; a geographical study. Oxford
University Press, London.
Ribbink, A.J., B.A. Marsii, A.C. Marsh, A.C. Ribbink, and B.J. Sharp. 1983. A
preliminary survey of the cichlid fishes of rocky habitats in Lake Malawi. South
Afncan Journal of Zoology 18; 149-310.
Shackleton, C.M. 1993. Fuelwood harvesting and sustainable utilization in a communal
grazing land and protected area of the Eastern Transvaal Lowveld. Biological
Conservation 63;247-254.
Sweeney, R.C.H. 1965. Animal life in Malawi: volume L invertebrates. Institute for the
Publication of Textbooks, Belgrade.
Tweddle D., and N.G. Willoughby. 1979. An annotated checklist of the fish fauna of the
River Shire south ofKapachira Falls. Nyala 5:75-91.
United Nations Population Division. 2000. Infonation: statistical data for member states
of the United Nations (United Nations CyberSchoolBus). Online. Available HTTP:
155
http://www.un.org/Pubs/CyberSchootBus/infonation/e_infonation.htm . November
2000.
Venema, J.H. 1991. Land Resources Evaluation Project: Land Resources Appraisal of
Liwonde Agricultural Development Division. MLW/85/011 Field Document No.
22. Malawi Govemment/UNDP/FAO, Lilongwe, Malawi.
Vitousek, P.M., P.R. Ehrlich, A.H. Ehiiich, and P.A. Matson. 1986. Human
appropriation of the products of photosynthesis. Bioscience 36:368-373.
Wainwright C., and W. Wehmieyer. 1998. Success in integrating conservation and
development? a study from Zambia. World Development 26:933-944.
Wild, R.B., and J. Mutebi. 1997. Bwindi Impenetrable Forest, Uganda: conservation
through collaborative management. Nature & Resources 33:33-51.
Willan, R.G.M. 1947. Annual report of the Forestry Department for the year ended 31"
December, 1946. Nyasland Protectorate, Zomba, Malawi.
Williamson, J. 1974. Useful plants of Malawi: revised and extended edition. Montfort
Press, Limbe, Malawi.
Worid Commission on Environment and Development. 1987. Our common future.
Oxford University Press, New York.
B.9
FIGURES LEGENDS
FIGURE B 1. PER CAPITA PRODUCTION FOR MAJOR CROPS
A = N. Tabacum tabacum L., B = Zea mays L.
FIGURE B.2. PER CAPITA PRODUCTION FOR IMPORTANT MINOR CROPS.
156
A = Ananas comosus (L.) Merr. (pineapple), B = Arachis hypogaea L. (groundnut), C =
Brassica chinensis L. (Chinese cabbage), D = Cucurbita maxima Duch. ex Lam
(pumpkin), E = Ipomea batatas (L.) Lam. (sweet potato), F = Lycopersicon esulentum
Mill, (tomato), G = Mangifera indica L. (mango), H Manihot esculanta Crantz (cassava),
I - Miisa paradisiaca L. M sapientum L. (banana), J = Oryza sativa L. (Gram) (rice), K
= Persea americana Mill, (avocado), L = Pennisetum americatntm (L) K. Schum. (millet),
M = Saccharum qfficinantm L. (sugar cane).
FIGURE B.3. PER CAPITA LIVESTOCK PRODUCTION.
A = Anas spp. (duck), B = Bos indicus Linnaeus var. Malawi Zebu (cattle), C = Capra
spp. (goat), D = Cavia porcellus Linnaeus (guinea pig), E = Gallus domestictis (chicken),
F = Lepus spp. (rabbit), G = Ovis spp. (sheep), H = Snis spp. (pig), I = Terron spp.
(pigeon).
FIGURE B 4. PER CAPITA WOOD PRODUCTION ON AGRICULTURAL LAND.
A = Acacia spp., B = Colophospermum mopane (Kirk ex Benth.) Kirk ex Leonard, C =
Eucalyptus spp., D = Gmelina arborea Roxb., E = Melia azedarach L., F = Piliostigma
thonningii (Schumach.) Milne-Redh., G = Pitms spp., H = Toona ciliata M. Roem..
FIGURE B.5. OVERALL PROTECTED AREA UTILIZATION BY CATEGORY OF
USE (PER CAPITA QUANTES AND RELATIVE PERCENTAGES).
FIGURE B.6. OVERALL PROTECTED AREA UTILIZATION BY CATEGORY OF
USE (PER CAPITA VALUE AND RELATIVE PERCENTAGES).
157
FIGURE B 7 PER CAPITA UTILIZATION OF MAJOR PLANT AND MUSHROOM
SPECIES FOR FOOD.
A = Habenaria walleri Reichb. f., B = mushrooms, C = Oryza longistaminata Chev. and
Roehr., D = Parinari ciiratellifolia Planch, ex Benth., E = Uapaca kirkiana Mull. Arg.
FIGURE B.8. PER CAPITA UTILIZATION OF A SELECTION OF THE MOST
IMPORTANT MINOR PLANT FOOD SPECIES.
A = Adamonia digitata L., B = Amaranthus thunbergii Moq., C = Annona senegalensis
Pers., D = Bidempilosa L., E = Catuhium crassum (Schweinf), F = Dioscorea spp., G =
Momordica foetida Schumach. and Thonn., H = Solanum nigrum L., I = Strychnos
spinosa Lam., J = Syzygium cordatum Hochst. ex Krauss, K = Tamarindus indica L., L =
Thunbergia lancifolia T. Anders., M = Trecidia africana Decne., N = Uapaca robynsii
De Wild., O = Vangueha infausta Burch.
FIGURE B.9. PER CAPITA UTILIZATION OF MAJOR FISH SPECIES.
A = Barbus spp. (small cypriaids), B = Glorias gariepitws Burchell (sharptooth cat fish),
C = Rhamphochromis sp., D = Labeo mesops Giinther, E = Opsaridium microlepis
Giinther (lake salmon), F = Oreochromis spp. (chambo), G = Marcusenius
macrolepidotus Peters (bulldog fish).
FIGURE B. 10. PER CAPITA FOOD UTILIZATION OF WILD ANIMALS,
INCLUDING INSECTS AND HONEY.
A = Aepyceros melampus Lichtenstein (impala), B = Hippotragus niger Harris (sable
antelope), C = honey, D = insects (all species), E = Kobus ellipsiprymmis Ogilby
158
(waterbuck), F = Muridae family (edible rodents known locally as mbewa), G = Numidea
meleagris (Linnaeus) (helmeted guinea fowl), H = Potamochoerus porcus Linnaeus (bush
pig), I = Potomonautes montivagus Chace F. (freshwater crab), J = Procavia capensis
Pallus (rock hyrax), K = Raphicerus sharpei Thomas (Sharpe's grysbok), L = Redunga
arundimtm Boddaert (reedbuck), M = Sigmoceros lichtensteinii Peters (Lichtenstein's
hartebeest), N = Sylvicapra grimmia (grey duiker), O = Syncerus coffer Sparrman
(African buffalo), P = Taurotragus oryx Pallas (eland), Q = Tragelaphus scriptus Pallus
(bushbuck). NB: due to scale, does not include 109.5 kg capita ' of Loxodonta qfricana
Blumenbach in Vwaza.
FIGURE B. 11. PER CAPITA PROTECTED AREA FUELWOOD UTILIZATION.
A = Acacia nigrescens Oliv., B = Aguaria salicifolia (Lam.) Oliv., C = Brachystegia
boehmii Taub., D = Brachystegia floribunda Benth., E = Brachystegia longifolia Benth.,
F = Brachystegia spicifomtis Benth., G = Brachystegia utilis Burtt Davy and Hutch, H =
Burkea africana Hook., I = Colophospermum mopane, J = Combretum apiculatum
Sond., K = Combretum fragrans F. Hoffm., L = Combretum imberbe Wawra, M =
Combretum molle R. Br. ex Don, N = Dalbergia melanoxylon Guill. and Perr., O =
Dichrostachys cinerea (L.) Wight and Am., P = Diospyros mespiliformis Hochst. ex A.
DC., Q = Eucalyptus spp., R = Faurea saligna Harv., S = Faurea speciosa Welw., T =
Julbernardia globiflora (Benth.) Troupin, U = Julbenuirdiapatuculata (Benth.) Troupin,
V = Lonchocarpus capassa Rolfe, W = Newtonia buchananii (Bak.) Gilbert and
Boutique, X = Parinari curatellifolia, Y = Pericopsis atigolensis (Bak.) van Meeuwen, Z
159
= Piliostigma thonningii, a = Pinus spp., P = Syzygium cordatum, y = Terminalia sericea
Burch. ex DC., 5 = Uapaca kirkiana.
FIGURE B. 12. PER CAPITA PROTECTED AREA CONSTRUCTION UTILIZATION
A = Acacia nigrescens B = Aguaria salicifolia, C = Arundinaria alpina K. Schum, D =
Brachystegia longifolia, E = Brachystegia spiciformis, F = Burkea qfricatm, G =
Byrsocarpus orientalis (Baill.) Bak., H = Cassipourea mollis (R.E. Fr.) Alston, I =
Colophospermum mopane, J = Combretum apicidatum, K = Combretum fragrans, L =
Combretum moUe, M = Dalbergia nitudula Welw. ex Bak., N = Dichrostachys cinerea,
O = Diplorhynchtts condylocarpon (Muell. Arg.) Pich., P = Eucalyptus spp., Q =
Friesodielsia obovata (Benth.) Verde., R = Hymenocardia mollis Pax, S =
Hymenodictyon parvifolium Oliv., T = Jatropha curcas L., U = Julbemardia paniculata,
V = Oxytenanthera abyssinica, W = Pericopsis angolensis, X = Syzygium cordatum, Y =
Terminalia sericea.
FIGURE B. 13. PER CAPITA PROTECTED AREA FIBER UTILIZATION.
A = Adenia gummifera (Harv.) Harms, B = Agave sisalana (Engl.) Perrine, C =
Brachystegia boehmii, D = B. bussei, E = B. floribunda, ¥ = B. longifolia, G - B .
spiciformis, H = 5. utilis, I = Cissampelos mucronata A. Rich., J = Colophospermum
mopane, K = Combretum fragrans, L = Dombeya dawei Sprague, M = Julbemardia
paniculata, N = Lannea schimperi (Hochst. ex A. Rich.), O = Piliostigma thonningii, P =
Sterculia africana (Lour.) Fiori.
160
FIGURE B 14 PER CAPITA PROTECTED AREA TOOL AND HANDICRAFT
SPECIES UTILIZATION.
A = Acacia polyacantha Willd., B = Annona senegalensis Pers. C = Anmdmaria alpina,
D = Bauhiniapetersiana Bolle, E = Borassus aethiopum Mart., F = Brachystegia spp., G
= Burkea africana, H = Cassia singueana Del., I = Cissampelos mucronata, J Diospyros squarrosa Klotzsch., K = Diplorhynchus condylocarpon (Muell. Arg.) Pich., L
= Ficiis natalensis Hochst., M = Hyparrhenia rufa (Nees) Stapf, N = Hyparrhenia
nyasica (Rendle) Stapf, O = Julbemardia paniculata, P = Lonchocarpus capassa Rolfe,
Q = Maytenus senegalensis (Lam.) Exell, R = Oxytenanthera abyssinica, S = Pennisetum
purpiirium Schumach., T = Pericopsis angolensis^ U = Phragmites mauritianus Kunth., V
= Psychotria zombamontana (Kuntze) Petit, W = Pteridium aquiUnum (L.) Kuhn, X =
Pterocarpiis angolensis DC., Y = Sida acuta Burm. f., Z = Triumfetta rhomboidea Jacq..,
a = Uapaca kirkiana, p = Widdringtonia whytei Rendle, y = Xeroderris stuhlmannii
(Taub.) Mendonca and E.P. Sousa, 5 = Xerophyta splendens Menezes.
FIGURE B. 15. PER CAPITA PROTECTED AREA THATCH UTILIZATION.
A = Dactyloctenium aegytium (L.) Willd., B = general thatch grass, C = Helictotrichon
elongatiim (Hochst. ex A. Rich.), D = Heteropogon contortus (L.) Beauv. ex Roem. and
Schult., E = Hyparrhenia filipendula (Hochst.) Stapf, F = Hyparrhenia gazensis (Rendle)
Stapf, G = H. nyassae, H = //. rufa, I = Themeda triandra Forsk., J = chambundu grass,
K = Beckeropsis uniseta (Nees) Stapf, L = Brachiaria brizantha (Hochst. ex A. Rich.)
Stapf, M = Chloris gayana Kunth., N = Digitaria diagonalis (Nees) Stapf, O =
161
Echinochloa colona (L.) Link, P = Hyperthelia dissoluta (Nees ex Steud.) W. D. Clayton,
Q = Imperata cylindrica (L.) Beauv., R = Panicum maximum Jacq., S = Pennisetum
purpurium, T = Setaria palustris Stapf, U = Setaria sphacelata (Schumach.) Stapf and
C.E. Hubbard ex M.B. Moss, V = chilambulire grass, W = chisungumbe grass.
FIGURE B. 16. PROJECTED DECLINE IN THE SUSTAINABLE SUPPLY OF WOOD
(MAI ROUNDWOOD EQUIVALENT) AS POPULATION GROWS.
162
B.10 FIGURES
600
t
500
ec
u
in
B fi. 400
o e
P
300
S
200
o .2.
t «
&
o
I.
U
100
• Mulanje
I
71
• Mulanje
• Lfwonde
4
200
• Liwonde
SI Dzalanyama
23
521
S Dzalanyama
• Vwaza
95
369
• Vwaza
FIGURE B.L PER CAPITA PRODUCTION FOR MAJOR CROPS.
A = Tabacum tabacum L., B - Zea mays L.
163
80
oc
_o o
u
U
3
u
S
e
e
e
b
a.
a.
o
TL
H
1
I
K
L
M
N
• Mulanje
23
4
12
9
10
5
38
11
46
8
17
I
7
• Lrwonde
15
1
3
15
1!
22 45
3
1
10
1
0
0
B Dzalanyaim
1
35
15 44 64 37 54
3
7
1
17
• Vwaza
1
33
1
1
36
1
1
8
56
18
10
I
^c
e §•
w
•s U
S a
11
£ S
FII
O
L.
u
• Mulanje
180 60 56 25 54 62 79 34 388 113
1
116 11
• Liwonde
40
10 15 60 278 94 10
8 219
1
50
15
HDzalanyaim 118 513 19 70 346 468 114 11
24^13
1
11
9
117 58 87 15 388 93
1
• Vwaza
7
22 486 4
2
34
5
FIGURE B.2. PER CAPITA PRODUCTION FOR IMPORTANT MINOR CROPS.
A = Ananas comosus (L.) Men*, (pineapple), B = Arachis hypogaea L. (groundnut). C =
Brassica chinensis L. (Chinese cabbage), D = Cucurbita maxima Duch. ex Lam
(pumpkin), E = Ipomea batatas (L.) Lam. (sweet potato), F = Lycopersicon esulentum
Mill, (tomato), G = Mangifera indica L. (mango), H Manihot esculanta Crantz (cassava),
I = Musa paradisiaca L. / M sapientum L. (banana), J = Oryza sativa L. (Gram) (rice), K
= Persea americana Mill, (avocado), L = Pennisetum americanum (L) K. Schum.
(millet), M = Saccharum officinarum L. (sugar cane).
ec
•o
B
es
£
3
S
S
0.0
• Vfulanje
• Liwonde
A
B
0.1
03
B Dzalanyama 0.1
• Vwa2a
C
E
F
G
H
I
1.2 0.9 0.1
0.8
0
0
0
0
0
2.1
0
0.1
0
0.1
1.5
0
0
7.4
0
3.6
0
0.1
E
F
G
H
1
0
0
0
0
0
1.27
0.9 0.1
6.9 2.6
0.1
0
A
B
D
0
0.4 0.1
9.8 0.1
100
80
B
JQ
60
40
9S
s
20
'B
<
0
C
D
• Mulanje
0.68 10.7 11.3 021 S26
• Liwonde
3.83
0
10.7 0J3 21.6 0.05 0.85
B Dzaianyaira 0.08 82.6 32.9 0
• Vwa2a
1.79
0
16.8
0
0
20 0.42
4.49 030 39.3 0.71 0.73 31.6 2.12
FIGURE B.3. PER CAPITA LIVESTOCK PRODUCTION.
A = Anas spp. (duck), B = Bos indicus Linnaeus var. Malawi Zebu (cattle).
C = Capra spp. (goat), D = Cavia porcellus Linnaeus (guinea pig), E =
Gallus domesticus (chicken), F = Lepus spp. (rabbit), G = Ovis spp.
(sheep), H = Suis spp. (pig), I = Terron spp. (pigeon).
165
300
^ 250-
200
w
3
O
150
100
O
O
50
0 JL
A
B
• Mulanje
1
• Liwonde
u
c
D
E
F
G
H
0
83
20
6
0
1
4
20
3
83
0
1
5
0
1
S Dzalanyama
0
0
258
67
0
0
0
2
• Vwaza
0
0
88
28
0
0
0
8
E
F
G
H
8J2 2.04
0.6
0
8.31 0.03
0.1
0.62
0
0.13
30
id
s 25
e
_© 20
w
3
15
ts
e
10
•a
e
o
0
A
B
• Mulanje
0.07
0
• Liwonde
2.48 0.41
U
C
D
0.29 0.39
S Dzalanyama
0
0
25.76 6.68
0
0
0
0.21
• Vwaza
0
0
8.8 2.81
0
0
0
0.82
FIGURE B.4. PER CAPITA WOOD PRODUCTION ON
AGRICULTURAL LAND.
A = Acacia spp., B = Colophospermum mopane (Kirk ex Benth.) Kirk ex
Leonard. C = Eucalyptus spp., D = Gmelina arborea Roxb.. E = Melia
azedarach L., F = Piliostigma thomingii (Schumach.) Milne-Redh., G =
Pinus spp., H = Toona ciliata M. Roem.
166
Mulanje
Liwonde
0
300
600
900
1200
1500
Per Capita UtiUzsatioii (kg)
Mulai^e
Liwonde
0%
20%
40%
60%
80%
100%
Percent of the Total Quantity of Utilization
Splantfbod
Bmeat
Dflielwood
Htool/craft
•construction • thatch
Hfiber
Smedicine
FIGURE B.5. OVERALL PROTECTED AREA UTILIZATION BY
CATEGORY OF USE (PER CAPITA QUANTIES AND RELATIVE PERCENTAGES).
167
Muianje
Liwonde
0
200
400
600
800
Per Capita UtiUzation (MK)
Muianje
Liwonde
•V
DzaJar^ana
Vwa2a
0%
20%
40%
60%
80%
100%
Percent of the Total Value of Utilizatioii
O plant food
Q meat
• flielwood
Htool/craft
•construction • thatch
• fiber
Bmedicine
FIGURE B.6. OVERALL PROTECTED AREA UTILIZATION BY
CATEGORY OF USE (PER CAPITA VALUE AND RELATIVE PERCENTAGES).
168
200
140
120
ec
s ^
100
0 e .a
Ji:E ©. s «
u
S2 « ®" 80
Z S
1 s s 60
ss ?
^ o ^ 40
&•
e
u<
20
0
S
150
•g
3
^
^
c
o«
*s
^
^ 100
^
.2,
a?
5
1s
.mm
I
.2
n
.S
=
•?
3
= •§
I
[iod
A
B
• Mulanje
0
25.4
• Liwonde
0
0.5
D
3.1 8IJ
17
S Dzalanyama 23.8 46.2
0
• Vwaza
0
1.8 11.8
50
0
0
15.5 136
11
9.7
• Mulanje
0
37_28
0
• Liwonde
0
0.5
6.29
0
0
B Dzalanyama 5.95 67.90
0
20.58
181
0.45 1734
0
14.67 1189
• Vwaza
4.17 1083
FIGURE B.7. PER CAPITA UTILIZATION OF MAJOR PLANT AND MUSHROOM
SPECIES FOR FOOD.
A = Habenaria walleri Reichb. f.. B = mushrooms, C = Oryza longistaminata Chev. and
Roehr.. D = Parinari curatellifolia Planch, ex Benth., E = Uapaca kirkiana Mull. Arg.
I
s
e
M
•c .2
e« u
N V
s a
1 .S
C® S
B
e«
• Mulanje
• Liwonde
j
'
i
S Dzalanyama
InVwaza
'
FIGURE B.8. PER CAPITA UTILIZATION OF A SELECTION OF THE MOST IMPORTANT MINOR PLANT
FOOD SPECIES.
A = Adamonia Jigilala L., B = Amaranlhus ihunbergii Moq., C = Annona senegalensis Pers., D = Bidens pilosa L., E =
Canthium crassum (Schweinf.), F = Dioscorea spp., G = MomonHca foetida Schumach. and Thonn., H = Solanum nigrum L.,
1 = Strychnos spinosa Lam., J = Syzygium cordatum Hoclist. ex Krauss, K = Tamarindus indica L., L = Thunhergia lancifolia
T. Anders., M = Treculia africana Decne., N = Uapaca rohymii De Wild., O = Vangueria infausta Burch.
0\
vO
170
35
100
90
30
80
BB
25
70
60
.2 20
A
.s
i 15
40
30
20
I 10
A
B
c
D
I
E
F
G
0
0
0
0
0
0
0
3
0 Jl
IM u lanje
• Uwonde
•fiJL
10
1.5 22.0 13 23 4:2 9.9 3.8
0
• Mulanje
• Uwonde
Jlfl
A
B
0
0
IL
a.
JL
C
D
E
F
G
0
0
0
0
0
6.91 98.5410.1518.90 4.20 44.13 17J21
0
0
0
0
0
S3 Dzalanyama 0.11 0.15
10.8 31.9 0
0
0
3.4
0
• Vwaza
H Dzalanyama 0.2 0.2
• Vwaza
50
7.78 22.95
0
0
0
0
0
0
0
0
2.45
0
FIGURE B.9. PER CAPITA UTILIZATION OF MAJOR FISH SPECIES.
A = Barbus spp. (small cypriaids), B = Clarias gariepinus Burchell (sharptooth cat fish).
C = Rhamphochromis sp., D = Labeo mesops Gunther, E = Opsaridium microlepis
Gunther (lake salmon), F = Oreochromis spp. (chambo), G = Marcusenius
macrolepidotus Peters (bulldog fish).
10
I2
T5
M a
B e
•1
B e>
< .a
ss
1=
0
iUII
A B C D
t
0
• Mulanjc
0
• Liwondc
1.0 0
i
0 I 1.7
i
I j
0 iO.4 0
0
0.1
0
.
i
1
3.2 0.1 1.1
i
-
0 ^ 0
0
0 ' 1.7 1.6 2. l i 7.4 0.5 0.3 7.4 i 0
;
0
I
H
L M N O
•'i
0 ; 0
0.8 0
0
0
0.4
0
0 I 0
0
0
0,9
0
0
0.3
1.2
8.2
1.8;
i
• Vwaza
1
t
f
0
i
0.9 0.7; 0.1 i 0
t
0.8! 0.1 ! 2.5 I 2.0 1.9
0
BDzalanyama
:
J |K
F i GI H I I
1
0
I
i
0.1
0
0 , 1.7 1.8 3.0 3.2 6,2
FIGURE B.IO. PER CAPITA FOOD UTILIZATION OF WILD ANIMALS, INCLUDING INSECTS AND HONEY.
A = Aepyceros melampus Lichtenstein (impala), B = Hippotragus niger Harris (sable antelope), C = honey, D = insects
(all species), E = Kohus ellipsiprymnus Ogilby (waterbuck), F = Muridae family (edible rodents known locally as mbcwa),
G = NumUka meleagris (Linnaeus) (helmeted guinea fowl), H = P()iamochoeru\ parens Linnaeus (bush pig), 1 =
Potomonautes monlivagus Chace F. (freshwater crab), J = Procavia capensis Pallus (rock hyrax), K = Raphicerus sharpei
Thomas (Sharpe's grysbok), L = ReJunga arimdinum Boddaert (reedbuck), M = Sigmoceros lichlensleinii Peters
(Lichtenstein's hartebeest), N = Sylvicapra ghmmia Linnaeus (grey duiker), 0 = Syncerus caffer Sparrman (African
buffalo), P = Taurotragus oryx Pallas (eland), Q = Tragelaphus scriptus Pallus (bushbuck). NB: due to scale, does not
include 109.5 kg capita ' of LoxoJonta africana Blumenbach (elephant) in Vwaza.
100
90
80
70
1
s
•o
M
60
*<0
50
1
40
at
9
U«
30
20
10
0
A
B
C
D
E
F
• Mulanje
0
0
14
12
0
• Liwonde
19
0
0
1
0
BDzalanyama
0
0
4
• Vwaza
0
35 30
u
G
H
1
J
K
L
42
13
11
0
0
1
0
0
1
0
2
55
0
39
11
22 77 31
14
15
0
0
0
1
0
0
10
0
34
0
3
49
M N
R
S
T
U
V
ii
W
X
Y
Z
a
92
11
0
40
0
0
16
10 31
5
42
0
0
0
0
0
0
16
1
0
22
13
0
36 26 50
12 77
0
0
21 57
41
32
0
18 27
0
0
4
O
P
1
4
0
2
17
5
13
0
49 30
0
0
5
FIGURE B.H. PER CAPITA PROTECTED AREA FUEL WOOD UTILIZATION.
J
p
J
7
S
5
9
36
0
0
15
0
13
0
25
9
38
58 42
0
0
20
2
5
FIGURE Q . W - C o f U i n u e d
A = Acacia nigrescens Oliv., B = Aguaria salicifolia (Lam.) Oliv., C = Brachystegia hoehmii Taub., D = Brachystegia
floribunda Benth., E = Brachystegia longifolia Benth., F = Brachystegia spiciformis Benth., G = Brachystegia utilis Burtt
Davy and Hutch, H = Burkea africana Hook., 1 = Colophospermum mopane, J = Comhretum apiculalum Sond., K =
Combretum fragram F. Hoffm., L = Combretum imberbe Wawra, M = Combretum molle R. Br. ex Don, N = Dalbergia
melanoxylon Guill. and Perr., O = Dichrostachys cinerea (L.) Wight and Arn., P = Diospyros mespiliformis Hochst. ex A. DC.,
Q = Eucalyptus spp., R = Faurea saligna Harv,, S = Faurea speciosa Welw., T = Julbernardia globiflora (Benth.) Troupin, U
= Julbernardia paniculata (Benth.) Troupin, V = Lonchocarpus capassa Rolfe, W = Newtonia buchananii (Bak.) Gilbert and
Boutique, X = Parinari curatellifolia, Y = Pericopsis angolensis (Bak.) van Meeuwen, Z = Piliostigma thonningii, a = Pinus
spp., P = Syzygium cordatum, y = Terminalia sericea Burch. ex DC., 8 = Uapaca kirkiana..
25
20
a
e
r
S 15
b(A
e
e
U 10
u
,o
T3
O
e
LiJ
0 J• Mulanje
CD
G I H I I J I K L :M
0 | P
QjR
0
3
0j0 0 0I I
1 ' 0
0; 2
0I 0 0I 0 0
2
0 0I 0
0
0I 0
01 0
0 i 0
0
I
• Liwonde
j
I
1
^
I
0 20' 0 ! I
• Vwaza
I 0
0; 2
8
0
3 ! 1
0 1131 9
0
0
71 7
0
0 i 0
0 17 6
WiX
r
I
•
BDzalanyama j 0
5 i T U
0I0 0
1
i ^
i
1 I
I i 5 0 j2 0 i2 4
I
6 14
6,0
10 0
6
O! 2
I i 0 0i
I 0 I :
3
3
1I
6 0 131
FIGURE B.12. PER CAPITA PROTECTED AREA CONSTRUCTION UTILIZATION.
A = Acacia nigrescens B = Aguaria salicifolia, C = Arundinaria alpina K. Schuni, D = Brachyslegia longifolia, E =
Brachystegia spiciformis, F = Burkea africana, G = Byrsocarpus orienlalis (Baill.) Bak., H = Cassipourea mollis (R.E. Fr.)
Alston, I = Coiophospermum mopane, J = Comhretum apiculalum, K = Comhretum fragnms, L = Combretum molle, M =
Dalbergia nitudula Welw. ex Bak., N = Dichrostachys cinerea, O = Diplorhynchus condylocarpon (Muell. Arg.) Pich., P =
Eucalyptus spp., Q = Friesodielsia obovata (Benth.) Verde., R = Hymenocardia mollis Pax, S = Hymenodictyon parvifoUum
Oliv., T = Jatropha curcas L., U = Julbernardia paniculala, V = Oxytenanlhera abyssinica, W = Pericopsis angolensis, X =
Syzygium cordalum, Y = Terminalia sericea.
175
BC
9N
U9t
£
A
K
• Mulanje
0
• liwonde
1.8 0
BDzalanyaim
0 0.5 0.1
n Vwaza
0.4 0 2.8 0.9 03
0
L M
0 0.5 4.4 1.5
0 5.6
O
0 0.6 0
0 02 1.8 0 0 0
0.4 5j5 0.6
N
03 02
1.1 0.6 0
0 0.6
0
0 0.2 1.8 2.7
FIGURE B.13. PER CAPITA PROTECTED AREA FIBER UTILIZATION.
A = Adenia gummifera (Harv.) Harms, B = Agave sisalana (Engl.) Perrine. C =
Brachystegia boehmii, D = B. bussei, E = 5. floribunda, F = 5. longifolia, G = B.
spiciformis, H = fi. utilis. I = Cissampelos mucronata A. Rich., J = Colophospermum
mopane, K = Combretum fragrans, L = Dombeya dawei Sprague, M = Julbernardia
paniculata, N = Lannea schimperi (Hochst. ex A. Rich.), O = Piliostigma thonningii, P
Sterculia africana (Lour.) Fiori .
20
18 '
16 •
14 •
c
^ .2
II
^ c
6 P
« 2
12 •
>0 =
C«
8
6O !I
4 '!
,!
I!
2 ;;
0
•Mulanjc
• Liwondc
S Dralanyama
• Vwaza
1 lX
A ! B
C
I)
J jtl_Ul
a
H
!• i O H
I
J ; K
0 ;0.4 0.8 0.7 0 7.t;8.7; 0 0.5 0
0.1 1.4
t
1.
M
11Iji
N
0 0.() 0.6 3.1
.
.
>
O
.
0
0
0
18
0 ; 0
0 0.1
0
0
0
0
0
0
0
0
0
0
2.7 0.1
0 0,4
0
0
0
0
0
2.1
2,7 0.1
0
0 0.4 2.7 0 0.8
0
0 1.0 1.7 0 0,8 0
0
l» ! Q
4
0
;
0
R
i
0
1.5U) |0,5
S
.
r 1Ui
V IW
0 8,8' 0 3,8 0
0
0 6.0 0
Z
a
p
y J,
1.8 1,3 1,7 8.7 3.5, 0 8.2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 3.0 0
0
0 1,5 1.2 0
0 0.8
19
Y
0
0 0,5 0,2 0,6 0,5 4.2
0
X
FIGURE B.14. PER CAPITA PROTECTED AREA TOOL AND HANDICRAFT SPECIES UTILIZATION.
FIGURE B.l - Continued.
A = Acacia polyacantha Willd., B = Annona senegalensis Pers. C = Arundinaria alpina, D = Bauhinia petersiana Bolle, E =
Borassus aethiopum Mart., F = Brachystegia spp., G = Burkea africana, H = Cassia singueana Del., I = Cissampelos
mucromta, J = Diospyros squarrosa Klotzsch., K = Diplorhynchus condylocarpon (Muell. Arg.) Pich., L = Ficus natalensis
Hochst., M = Hyparrhenia rufa (Nees) Stapf, N = Hyparrhenia nyasica (Rendle) Stapf, 0 = Julbernardia paniculala, P =
Lonchocarpus capassa Rolfe, Q = Maylenus senegalensis (Lam.) Exell, R = Oxytenanthera ahyssinica, S = Pennisetum
purpurium Schumach., T = Pericopsis angolensis, U = Phragmites mauritianus Kunth., V = Psychotria zomhamontana
(Kuntze) Petit, W = Pteridium aquilinum (L.) Kuhn, X = Pterocarpus angolensis DC., Y = Sida acuta Burm. f., Z Triumfetta
rhomboidea Jacq.., a = Uapaca kirkiana, p = Widdringtonia whytei Rendle, y = Xeroderris stuhlmannii (Taub.) Mendonca and
E.P. Sousa, 8 = Xerophyta splendens Menezes.
-4
178
100
an
80
B V
w
0
is
i^
-I
1^
60
40
20
0
Ju. 1
lifl nlH
A
B
C
D
E
•
G
H
I
J
• Mulanje
0
17.9
0
173
23
)
0.8
13.2
0
15-5
• Uwonde
0
0
0
0
Z5
0
H Dzalanyama
0
35.5 21.1 26.5 2.5
0
0
16.0
0
• Vwaja
2.0
19.6 18.6
20.2 2.4
11.0 3.1
0
64.5 90.0 6.0
15
'3
Ve.
10
K,
0-1
1
1
K
L
M
N
O
• Mulanje
0
0
0
0
0 6.8 1.4 0 2.5 0.5 0 4.6 0
• Liwonde
0 7.8 4.7 0 9.4 0
HDzalanyaim 2.9 0
0
• Vwaaa
0 0.9 0
0
0
11
0
P
Q
0
R
0
S
T
U
0 12 0
V W
0 4.5
0 2J2 1.4 0.6 1.7 23 0
0
0 1.5 0
0
0
0
0 0
FIGURE B.15. PER CAPITA PROTECTED AREA THATCH UTILIZATION.
A = Dactyloctenium aegytium (L.) Willd., B = general thatch grass. C = Helictotrichon
elongatum (Hochst. ex A. Rich.). D = Heteropogon contortus (L.) Beauv. ex Roem. and
Schult.. E = Hyparrhenia filipendula (Hochst.) Stapf, F = Hyparrhenia gazensis (Rendle)
Stapf, G = H. nyassae, H = //. rufa, I = Themeda triandra Forsk., J = chambundu grass, K
= Beckeropsis uniseta (Nees) Stapf, L = Brachiaria brizantha (Hochst. ex A. Rich.)
Stapf, M = Chloris gayana Kunth., N = Digitaria diagonalis (Nees) Stapf, O =
Echinochloa colona (L.) Link, P = Hyperthelia dissoluta (Nees ex Steud.) W. D. Clayton,
Q = Imperata cylindrica (L.) Beauv., R = Panicum maximum Jacq., S = Pennisetum
purpurium, T = Setaria palustris Stapf, U = Setaria sphacelata (Schumach.) Stapf and
C.E. Hubbard ex M.B. Moss, V = chilambulire grass, W = chisungumbe grass.
179
275
i
Q.
S
cw
"9
e 150 H
o
es
w
3
1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
Mularye
Liwonde
Vwaza —Dzalanyaira
FIGURE B.16. PROJECTED DECLINE IN THE SUSTAINABLE SUPPLY OF WOOD
(MAI ROUNDWOOD EQUIVALENT) AS POPULATION GROWS.
C. APPENDIX C: CHOOSING BETWEEN FOREST AND AGRICULTURE
THE EVOLVING ECONOMIC ROLE OF PROTECTED AREAS
Title Page
Authors;
Barron J. Orr
Geospatial Extension Specialist
University of Arizona
1955 E. 6'^ Street
Tucson, AZ 85719
USA
W.M. Kasweswe Mwafongo
Senior Lecturer
PO Box 280
Chancellor College
Zomba
MALAWI
Choosing between Forest and Agriculture:
The Evolving Economic Role of Protected Areas
Number of bytes:
Contact Person:
Barron Orr
University of Arizona
1955 E. 6*^ Street
Tucson, AZ 85719, USA
Tele: +1 520-626-8063
Telefax: +1 520-621-3816
E mail: [email protected]
181
C.l
ABSTRACT
As population densities and the demand for land resources increase, local
inhabitants risk the loss of ecological resources when land is cleared for cultivation. This
dilemma is investigated through a multidimensial socioeconomic and ethnoecological
assessment of427 households in communities adjacent to four protected areas in Malawi.
The research demonstrates that poorer households are more reliant on protected area
proceeds (R^ = 0.84), which also significantly improve equality in the distribution of
income. Ecological resources are shown to meet demand for more people and for a longer
time frame than permitted cultivation on agriculturally suitable land. Yet per hectare
values are 2 to 3.5 time greater for agriculture than for consumptive ecological resource
use. Spatial analysis suggests points of negative land cover change (1984-94) were
associated not with concentrations of population but with the agricultural suitability of the
land. The results suggest the kinds of decisions people will make under extreme stress,
when the consideration of potential impacts is overwhelmed by the need to survive.
182
C.2
INTRODUCTION
Developing nations with predominantly agrarian economies face a difficult natural
resource dilemma. Food and Agriculture Organization (FAO) estimates suggest these
countries lost 14.9 million hectares of forest area annually between 1980 and 1995 (1).
Particularly in Africa, the driving force was generally the expansion of subsistence
agriculture. Thus, as population densities and the demand for land resources increase,
local inhabitants risk the loss of ecological resources when land is cleared for cultivation.
Increasing population densities can also precipitate what Boserup hypothesized as
a spontaneous movement towards agricultural intensification (increased average inputs of
labor or capital in order to increase the value of output) in response to shortened fallow
periods and declining yields (2,3). New technologies demand more farm labor and
investment in land that lead to higher crop yields and household income. In a more
technocentric approach, Simon suggests this process is driven by invention, a function of
population density where more people produce more ideas, thus driving growth (4).
Tiffen, Mortimore, and Gichuki argue that it is the combination of growth in population
and markets that leads to intensification, and that land scarcity is an incentive for
conservation (5).
Indeed, among the numerous studies tying high levels of resource degradation to
population growth, (e.g. 6), Afiican examples of successful agricultural intensification (7),
lower rates of erosion (5), and the improvement of forest resources (8,9) are being
documented more frequently. However, Boserup concedes that agriculture intensification
183
is weak in much of Africa, due in part to poor infrastructure, extension, marketing, and
increased migration (3). Where soil degradation and the loss of soil fertility arising from
more frequent cropping occurs at a rate that overtakes intensification, policy intervention
may be necessary to prevent environmental damage (10). Unfortunately the combination
of fragile soils and government policies that tend to discourage improvements in
smallholder agriculture are all too common in Africa.
Under these conditions rural households may attempt to augment income through
diversification, either to reinvest in agriculture, or simply to survive (11). Often such
efforts to diversify income exhibit production linkages' to local agriculture (12) and
actually help prevent environmental degradation when proceeds are invested in soil
conservation (13). This would suggest the potential for harmony between ecological and
livelihood sustainability, defined as "stable and growing total factor productivity" (14).
However, the ecological and socioeconomic aspects of sustainability are not always
compatible, particularly where perception of risk leads rural households to invest the
proceeds of off-farm earnings away from agriculture (11).
Income diversification occurs when local alternatives to subsistence production are
available at relatively limited risk. The use and sale of protected area natural resource
products (15,16) can provide such an opportunity because the natural resources and the
knowledge necessary for production are locally available. Exploiting protected area
resources is generally viewed as an innovative and relatively low risk livelihood strategy.
However, where population density and agricultural land demand is high, using this
strategy to alleviate poverty may have limited long term potential, whether or not the
184
proceeds are reinvested into agriculture. The dilemma this strategy presents is that the
protected area natural resources must be sacrificed if the expansion of agricultural land
continues.
Where land is scarce and the combination of low soil fertility and limited policy
support impedes the agricultural intensification process, one could assume that
dependence on protected area resources would provide an incentive for conservation,
particularly where local inhabitants "see the relevance of conservation for themselves and
the future of their children" (17). These conditions are prevalent in Malawi (18), where we
investigated the importance of protected area natural resources on the livelihoods of those
living nearby. Our research on the communities surrounding four protected areas in
Malawi demonstrates that, ceteris peribus, local forest utilization is sustainable in the
medium term (~50 years), and that poorer households are more reliant on protected area
proceeds than wealthier households. Indeed, protected area-based income results in an
overall more equitable distribution of income and can actually halve the number of people
who would otherwise sink below the basic needs poverty threshold. Yet in spite of this
reliance, given a choice between land and ecological resources from the protected area,
expansion of cultivation wins out. This presents a dilemma for policy makers who want to
encourage spontaneous strategies for poverty alleviation. The very strategies that support
the poorest households are jeopardized by an overriding need for land.
185
C.3 STUDY AREA
The population of Malawi is predominantly rural, with smallholder farmers
(averaging 1.0 ha of land under cultivation) constituting 90% of the nation's poor
(19).The agroecosystems in Malawi are dominated by a sub-humid tropical, uni-modal
rainfall system (ranging from 700 to 1400 mm annually), with loamy sand soils
characterized as having "low" to "sufficient" nutrient levels (20). High population growth
rates have led to small land holdings (21), and continuous cultivation (limited or no
fallow), whereas in 1938, plots were farmed only 3 to 5 years before extended rest (22).
However, the conditions necessary for autonomous agricultural intensification and
conservation remain weak.
Where agricultural intensification has occurred (mainly burley tobacco), it was a
result of policy change rather than technological innovation, and where it has not (hybrid
maize), transaction costs limiting market access were a major impediment (23). Among
Malawi smallholders, the adoption of better soil management practices is limited not by
socio-economic barriers to participation or perception or risk, but instead by poor access
to information (24). Over 90% of smallholders do not follow government extension
recommendations for improved soil management, in part due to unsuccessful transfer of
knowledge, and perhaps also in part because some of the innovations proposed have been
shown to produce uneconomic gains in yield (25).
Finally, the use of agricultural inputs, free or purchased, is on the decline in
Malawi. Following the removal of farm input subsidies in the early 1990's, the smallholder
sector decreased its use of fertilizer by almost 75%, purchased only enough hybrid seed to
186
plant 7% of the maize area, and reimbursed creditors for only 20% of the inputs purchased
on credit (26).
In 1997, 19% of Malawi's 94,000 km^ of land was under protection in the form of
four wildlife reserves, five national parks, and seventy-seven forest reserves (27). This
percentage is not exceptional in Afnca or elsewhere. However, in the same year, 86% of
the country's 9.65 million people lived in rural areas (28). The associated population
density translates into intense pressure on the 495 km'^of protected areas, primarily in the
demand on new lands for cultivation. However land demand is not the only pressure. In
Malawi, 98% of rural and 94% of urban energy demand is satisfied through fuelwood and
charcoal (29). Land and fuel pressures have contributed to an estimated 50% decline in the
forested area of Malawi between 1946 and 1996. (30,31,32). This has had the effect of
increasing the importance of the remaining forested land area, 53% of which is found in
protected areas (27).
In response to the acknowledged pressure to convert protected areas to
agricultural lands, the Government of Malawi commissioned a Public Lands Utilization
Study (PLUS) in 1996 to study both the environmental risks of conversion and the
importance of the reserves and their resources to adjacent communities (27). The study
was guided by a national Steering Committee on Land, which was made up of 60
government and non-government stakeholder agencies. The Committee selected four
protected areas for intensive study, including Mulanje Forest Reserve, Liwonde National
Park, Dzalanyama Forest Reserve and Ranch, and Vwaza Wildlife Reserve.
187
Mulanje Forest Reserve, located in southeastern Malawi (centered on 15°57'S,
35°39'E, covering 56,314 ha of mostly mountainous terrain) has vegetation ranging from
montane woodland and grassland on the plateau to miombo woodland (mesic-dystrophic
savanna, dominated by Julbemardia and Brachystegia species) on the lower slopes. The
reserve protects a number of catchments from erosion and supplies both hard and
softwood timber. It also shelters considerable biological diversity that includes a greater
variety of wildlife than any other forest reserve in Malawi and over thirty endemic flora
species (33). Mulanje lies in the most densely populated area of the four reserves (211 km'
^in 1997).
Liwonde National Park is located in south central Malawi (centered on 14°50'S,
35°2rE, encompassing 54,633 ha of predominantly flat topography) in an area of high
population density (166 persons km"^). Liwonde protects biodiversity and wildlife in the
Upper Shire and one of the few Malawian examples of mopane woodland (a broad-leafed,
drought deciduous woodland and savanna dominated by Colophospermum mopane) (34).
It is the most important location for ecotourism in Malawi. Enforcement of protection is
stricter in Liwonde than anywhere else in Malawi, due in part to a wildlife fence on the
more densely populated western edge of the Park. It is also the base of operation for the
nation's wildlife scout training program.
Dzalanyama is unique in that it is the largest forest reserve and agricultural scheme
in Malawi (35). The Forest Reserve is in the central, western part of Malawi, (centered on
14°20'S, 33°22'E). It encompasses 98,827 ha of terrain, two thirds of which is comanaged with Dzalanyama Ranch, located in a mostly low lying area of open miombo and
188
dambo (grasslands in seasonally inundated drainage lines). Cattle are excluded from the
remaining western third, which is close-canopy, upland miombo that borders a forested
area of Mozambique that has seen limited use over the past 25 years. The rest of the
reserve is bounded by a population averaging 119 persons km"^. Dzalanyama watersheds,
including two dams, supply 30% of all water needs for Lilongwe, as well as a large
portion of urban fuelwood needs.
Vwaza Marsh Wildlife Reserve (centered on 1 TOO'S, 33°28'E) primarily serves to
protect biodiversity and wildlife, though levels of ecotourism are well below that of
Liwonde. Vwaza also serves to contain the tsetse fly; Trypanosomiasis of cattle is
endemic and an increasing number of sleeping sickness cases among humans have been
reported since 1980 (36). The majority of Vwaza Marsh is composed of miombo
woodland, though some montane woodland, dambo grassland, mopane woodland, and
thicket are also present. The eastern, Zambian side of the reserve is sparsely populated,
while the land surrounding the Malawian boundary averaged 95 persons km~^ in 1996.
This density is greater than much of northern Malawi, due in part to the creation of
numerous tobacco estates over the past 20 years, limiting customary land expansion.
C.4 METHODS
Our study was based on the integration of qualitative and quantitative
socioeconomic data concerning household agricultural production and the use of
protected area species utilization, as well as spatial analysis of biophysical data. Field data
collection was conducted in 1996-97 with a multidimensional approach that combined
189
participatory techniques with a quantitative household questionnaire and an ecological
resource assessment. The approach involved baseline livelihood security data collection
methods developed in the Sahel and Haiti (37) that were adapted to evaluate secondary
forest product utilization (38). The results of the household and species work were then
integrated into a spatial database using a geographic information system (GIS) and the
results of satellite imagery change detection.
To capture household production and protected area resource utilization, we used
a mix of overlapping qualitative and quantitative techniques. These included surveys both
inside the home (where both male and female members participated), and inside the
protected area, where activities could be observed and measured. Results based on
respondent recall were calibrated with an inventory analysis of resource utilization zones.
Other than the formal survey, data collection was participatory in nature: local inhabitants
carried out the investigation, presentation, and preliminary analysis under the guidance and
training of a local research team (39). The field research team was made up of four
Malawian male and female enumerators conversant in the key local languages (primarily
Chichewa, Chiyao, and Chitumbuka), as well as a botanist and two forest mensuration
specialists from the Forestry Research Institute of Malawi (FRIM).
Research began with a rapid appraisal of communities within a 5km zone of
influence around each protected area designed to capture variation among localities and
agroclimatic zones. In all, results from 138 villages provided the patterns of livelihood
systems and natural resource utilization that were subsequently incorporated into
participatory and formal survey instruments specific to each protected area. This included
compilation of local species names and uses, which were later verified by the botanist and
natural resource experts fi^om the village itself during the resource assessment. From the
rapid appraisal, four to five villages were selected for intensive study around each
protected area.
Intensive field study included a participatory mapping exercise to delineate past
and present vegetation and natural resource utilization zones, a protected area resource
assessment (eight 10m x 10m plots per village), key respondent interviews, and a formal,
quantitative survey designed to capture intra-locality variation (Table C. I). The final data
set for the formal survey included 427 households comprised of 2,205 individuals, while
228 key respondents provided qualitative data on specialized resource use (i.e. charcoal
making, healing, etc.).
191
TABLE C 1. SUMMARY OF DATA COLLECTION METHODS
Data Gathering Activity
Primary Objectives
Rapid Appraisal (138 villages)
Interviews with traditional officials
•
Community meetings
• Patterns of livelihood strategies
• Patterns of protected area use
• Crop, livestock, wild resource species list
Focus group interviews
(men and women separately)
• Qualitative specialized use
• Land and resource tenure
• Attitudes towards protection
• Changes in resource availability/access
Village list, locations, resource patterns
Intensive study (17 villages)
Participatory mapping
•
Key respondent interviews
• Quantitative specialized use (223 respondents)
• Local unit volume and weight conversions
• Local market retail prices
Resource assessment
(136 plots)
Formal survey
(427 households)
•
•
•
•
Present and past resource utilization
Vegetation measurements
Local name / Latin name verification
Income (agriculture production, livestock.
remittances, off-farm activities, etc.)
Protected area resource utilization
The socio-economic analysis was intended to capture two key variables. The first
addressed household well being, expressed in terms of both direct and non-direct
household income. The second addressed household protected area natural resource
utilization, broken into seven major use categories; food, fiaelwood, fiber, tools, medicinal
plants, and both wood and thatch for construction.
Income estimates fi'om all sources (direct and indirect) were compiled for each
individual household and converted into per capita values (15 Malawi Kwacha = 1 USD).
192
Direct income was defined as the actual monetary compensation received by household
members for wage labor, sale of agricultural products, remittances, and a variety of offfarm income generating activities, including sales of protected area resource products.
Indirect income was the composite of activities that have utility to the household, each
being assigned an estimated value derived from local retail market prices. The income
from these activities included the value of goods intended for consumption within the
household, such as subsistence maize production and protected area resource utilization.
To create these estimates, it was necessary to develop volumetric and weight conversions
as well as retail prices for all agricultural products and protected area resources (38).
For each protected area, the nature of the resource, pressure on the resource, and
impact on the resource were assessed using the tools of geospatial analysis, and data
captured and analyzed as part of PLUS (27). PLUS maps of the agricultural suitability of
land in all four protected areas were generated from a Land Resources Evaluation Project
(LREP) model and data produced by the Government of Malawi and FAO (40,27). In
order to be considered suitable for agriculture, land had to fall with suitable ranges for all
criteria, including length of growing period, slope, soil depth, surface stoniness, drainage,
and ponding (Figure C. 1). Population pressure was based on a spatial depiction of rural
population in the smallest physical administrative division that was spatially available, the
Environmental Planning Area (41), that was then limited to a 5 km zone of influence
around each protected area. The population figures were augmented to account for urban
influences (i.e. small towns), and then adjusted to 1996 levels using growth rates
determined by the preliminary analysis of the 1998 census (27,28) (Figure C.2). PLUS
193
(27) vegetation classification and change detection between 1984 and 1994 (Figure C.3)
were used to spatially quantify woody biomass and determine the relationship between
negative change (decline in biomass) and agriculture suitability. Of all types of protected
area resource utilization in Malawi, the greatest influence on the overall resource base is
wood extraction. Therefore, estimates of sustainable wood supply were based on the
mean annual increment (MAI, or annual growth) of woody biomass growing stocks (42)
specific to the region and mapped vegetation class. At the protected area level, these were
applied to the vegetation classifications from PLUS. At the national level, the baseline
vegetation classification produced by the Department of Forestry and the Swedish Space
Corporation was used (43), with an annual rate of forest area decline of 1.5% (27).
C.5 RESULTS
The first phase of research on the relationship between protected areas and the
livelihoods of those living in adjacent communities focused on reliance. Reliance was
defined as the share of total per capita income attributed to the direct and indirect value of
protected area resource utilization, as well as associated activities such as healing with
protected area medicinal plants or public land employment. The level of reliance varied
between the four protected areas with respect to total income and land holdings (Table
C.2). In communities around Vwaza and Dzalanyama, where population density was
lower and more land available, income was higher. Reliance was greatest in communities
around Mulanje and Liwonde (despite stricter protection), where land was scarcer and
incomes lower.
194
The pattern of poverty associated with reliance on protected area proceeds was
verified at the household level for the entire data set through a simple linear regression
model (Figure C.4) (38). Results show an inverse income-reliance relationship; for every
100 MK increase in per capita income, the portion of income that is protected area-based
can be expected to decline by 0.1%.
TABLE C.2. RELIANCE (PERCENTAGE OF PROTECTED AREA INCOME
RELATIVE TO TOTAL INCOME), LAND, AND POPULATION.
Reliance Protected Total
Population Relative Level
Mean
Density
of Protection
Area
Income
Land
Income
Holdings
—MK capita ' yr''~ ha capita ' persons km"^
%
Mulanje
20.3
517.76 2552.95
Low
0.146
211
Liwonde
15.8
166
Very
High
384.46 2432.42
0.194
Dzalanyama
10.4
536.18 5167.76
119
Low
0.316
Vwaza
13.3
763.05 5725.91
95
0.329
Medium
Protected area-based income also proved to positively influence the distribution of
income (38). The most common measure of income equality is the Gini coefficient, defined
as the arithmetic average of the absolute values of differences between all pairs of incomes
(44). It can be generally stated that more highly developed countries tend to have lower
differentiation of income (expressed as percentages, a Gini index of 25 to 40%) while
developing nations tend to have higher differentiation (45 to 60%). The Gini index falls
from 56.3 % on income exclusive of protected area proceeds to a more equitable 50.9%
when these proceeds are included (Table C.3).
195
TABLE C 3. MEAN INCOME, INCOME EQUALITY, AND POVERTY WITHOUT
AND WITH PROTECTED AREA-BASED INCOME.
Head Count Poverty Index
Mean
Gini Income
Income
Equality Index
(1700.00 MK
(668.10 MK
MK capita'' yr"'
%
Poverty
Poverty
Threshold)
Threshold)
Without protected
43.6
3646.65
56.3
17.5
area proceeds
With protected
31.9
4320.25
50.9
8.8
area proceeds
A simple and intuitive measure of poverty is the head count index, which is the
portion of the population that falls below the poverty threshold relative to the total
population. Two such thresholds were used in this analysis. The "basic needs" poverty
threshold relates nutritional requirements to the energy provided by a primary diet staple,
which in turn can be converted into a value based on the market price of that staple. This
comes to 668.10 MK in Malawi for the 1996/97 average price of maize. This threshold
understates poverty because our research included all aspects of subsistence production
and protected area resource utilization, the majority of which are not captured in national
studies. We therefore calculated a second "poverty reference threshold" (1700.00 MK) to
include the poorest 32% of all households to reference the national relative prevalence of
absolute poverty reported by the World Bank (19).
It is not surprising that the addition of protected area income reduces the
proportion of the population that falls below the poverty threshold (Table C.3). However,
note this improvement differs, depending on the definition of poverty. The percentage of
196
the population with inadequate income to meet basic needs (668.10 MK) is cut in half by
including protected area proceeds (17.5 to 8.8%), while under the poverty reference
threshold (1700.00 MK), the change impacts only 27% of those considered poor (43.6 to
31.9%). This suggests that in particular, the poorest of the poor use protected area
proceeds as a strategy to rise out of absolute poverty.
The influence of protected area proceeds on the distribution of income and poverty
alleviation is pronounced. Figure C.5 shows that the relative share of income controlled by
each fifth of the population (ranked by increasing income) shifts from rich to poor.
Moreover, the poorest quintile feels the greatest impact: a 40.1% increase in income
share, while the richest quintile experiences a decline of 7.9%.
Though the importance of proceeds from protected area products and related
activities (i.e. healing) is evident, land as a potential agricultural resource was viewed as
more important, at both the village and household level. Of the 138 villages visited during
the rapid appraisal 59% categorized themselves as fully dependent on the protected area
for natural resource products, particularly fiielwood (all year) and wild foods during the
rainy season when food stocks run low. Yet their desire to cultivate at least a portion of
the protected area was more pervasive: 86% of the villages listed this as the most
important issue they faced. When asked to choose between the two, 69% of the villages
put protected area land above their demand for ecological resources. At the household
level, 65% of respondents recommended that management of the protected areas should
be adjusted to permit cultivation, compared with only 31% advocating resource access as
a high priority.
197
These competing demands (expansion of agricultural land versus protection of
ecological resources) were also evaluated in terms of sustainability, at both the national
and protected area level, using population growth rates that have been adjusted to reflect
the preliminary results of the 1998 census (28,45). Demand will exceed the national
sustainable supply of wood will be exceeded by demand in 35 years (Figure C.6). Demand
will exceed the supply of suitable agricultural land will be exceeded by land demand in
only eight years (Figure C.7). The rate of future land demand is projected to flatten as land
becomes more scarce, in part due to urbanization, and in part due to a reduced percentage
of land allocated to infrastructure and public areas in the customary sector (46,47).
Conducting a similar analysis at the protected area level requires spatial
information about agricultural suitability and woody biomass to estimate supply, as well as
population growth rates based on the 1998 census to derive demand (Table C.4). The
density of customary land utilization (ha capita "') was based on the actual 1996 land
occupation figures calculated for the zone of influence, including both cultivated and
uncultivated land (48).
Sustainability analysis at the protected area level shows that in 50 years, the local
demand for protected area wood will exceed sustainable supply in all but Dzalanyama,
where supplies will last another 50 years (Figure C.8). Resource demand was based on the
average per capita wood utilization multiplied by the population in the zone of influence
for each protected area. Each protected area has very different projected rates of
population growth and wood consumption, resulting in varied rates of resource decline.
For example, the sustainable supply of woody biomass in highly protected Liwonde is
198
projected to last a few years longer than Vwaza, despite the much larger resource base
available inside Vwaza Wildlife Reserve.
TABLE C.4. KEY INPUTS FOR THE SUSTAINABBLITY ANALYSIS.
Zone of Influence
Protected Area Land
Rate of Local
Wood
1987- 1998
Total Land Suitable for
1996
Utilization
Population Growth Rate
Agriculture
%
ha
kg capita"' yr"'
%
0.5%
56,314
Mulanje
135,648
2.7%
628
1.8%
Liwonde
54,633
99,604
16.1%
327
98,827
1.6%
Dzalanyama
89.7%
74,868
730
Vwaza
98,214
2.7%
63.0%
43,316
951
The analysis for land at the protected area level (Figure C.9) was based on a
hypothetical scenario, where, in each year after 1999, protected area land that was suitable
for agriculture would be made available only to new population in the zone of influence
around each reserve. The rate of land consumption was a function of the population
growth rate. The results show rapid conversion of land in Mulanje, despite low population
growth rates, because only a small percentage of the mountainous reserve is suitable for
agriculture. Liwonde's suitable land would last only eight years, while much larger Vwaza
and Dzalanyama (which also have much larger per capita land holdings) would last 30, and
58 years, respectively.
Perhaps more important than difference among the reserves is the overall
difference between the sustainability of woody resources noted in Figure C.8 with the
sustainability of suitable agricultural land in Figure C.9. In all cases but Vwaza, ecological
resource utilization can be sustained far longer than conversion of land to agriculture. In
199
Vwaza the difference is less significant because of higher population growth rates and a
higher woody biomass consumption rate. That rate is higher in part due to recent
liberalization of burley tobacco legislation (49), which has resulted in construction of
tobacco-related infrastructure from protected area resources (i.e. pole and fiber-based
drying sheds) that will not require replacement for a number of years.
The results of the sustainability analysis suggest that the demand for agriculture
land in the protected areas is large enough to overshadow consumption patterns for all
other ecological resources. This raised the question of whether the physical evidence
concerning negative change (i.e. declining biomass) identified in the PLUS 1984 to 1994
land cover change detection (27) would prove to be spatially associated with population
pressure from the outside or agricultural suitability on the inside of each protected area. If
land cover changes were due to high fuel demands, proximity to higher population
concentrations would be expected. If land cover changes were linked to agricultural
potential, at minimum the negative change locations would occur more than randomly on
agriculturally suitable sites than not. Results suggest greater changes have occurred on
land suitable for agriculture, irrespective of population concentrations (Table C.5).
Differences in Liwonde National Park land cover between 1984 and 1994 apparent
in Landsat Thematic Mapper (TM) imagery provide a graphic example of this relationship
(Figure C. 10), particularly in the eastern-most portion of the Park (Figure C. 11). Though
this area is suitable for agriculture (Figure C. 13), it is considerably south of the much
higher population concentrations (Figure C.14). The results of the change detection
analysis show this area to have undergone a decline in natural biomass, verified by ground
200
observation as agricultural encroachment (Figure C.15). Moreover, the land adjacent to
higher population concentrations to the north of the encroached area has undergone
minimal negative change.
TABLE C.5. SPATIAL RELATIONSHIP BETWEEN AREAS THAT EXPERIENCED
A DECLINE IN BIOMASS (1984 - 1994) TO BOTH POPULATION OUTSIDE AND
AGRICULTURALLY SUITABLE LAND INSIDE EACH PROTECTED AREA.
Land Area Experiencing
Spatial Association with Areas of
Negative Change
Negative Change
Agriculturally
Areas Adjacent to
Suitable Land
ha
High Population
%
Mulanje
3666
6.5%
limited
high*
Liwonde
5021
9.2%
high*
limited
Dzalanyama
3904
4.0%
high*
limited
Vwaza
9890
10.1%
high*
limited
*all statistically significant using a Chi Squared test.
The final element of analysis involved an economic comparison of agricultural and
consumptive use value of natural resource products on a land area basis in each reserve
(Table C.6). This was accomplished by converting the per capita proceeds derived fi-om
agriculture and those derived from protected-area activities into per hectare values. The
calculation of agricultural values included all crop and livestock production, and like the
sustainability analysis, considered all land in the zone of influence outside the protected
areas, rather than only land under cultivation. This was based on the assumption that
converting protected area land to agriculture would result in the same land use intensity as
currently exhibited in nearby customary land. Protected area values are based on the land
inside each reserve, within 5 km of the boundary, because participatory maps showed the
vast majority of natural resource utilization occurred within that proximity of the each
village surveyed.
201
TABLE C.6. THE CONSUMPTIVE USE VALUE OF AGRICULTURE ON
CUSTOMARY LAND AND NATURAL RESOURCE UTILIZATION ON
PROTECTED AREA LAND *
Agricultural Value of
Protected Area Natural
Resource Utilization
Customary Land
MK ha"' yr"' USD ha"^ yr ' MK ha' yr"^ USD ha' yr-'
Mulanje
67.64
3682.89
245.53
1014.67
Liwonde
49.89
2832.65
748.38
188.84
Dzalanyama
44.78
4743.45
671.69
316.23
Vwaza
45.95
4346.48
289.77
689.29
* all figures are based on gross revenues.
The agricultural value of land in Mulanje and Liwonde is more than 3 .5 times that
of its natural resource consumptive use value. The results are almost double that in
Dzalanyama and Vwaza, where land is less scare and higher value crops or more
prevalent.
C.6 DISCUSSION
These results suggest the kinds of decisions people will make under extreme stress,
when the consideration of potential impacts is overwhelmed by the need to survive. There
is strong indication in both the qualitative and quantitative data that those surveyed were
aware of the longer term risks of converting protected areas to agriculture, and that they
are equally aware of the immediate differences in consumptive use values of both the land
and the ecological resources. Carrasco's (17) position that knowledge of the value of
ecological resources might lead to conservation is unlikely when the difference between
the value of land cultivation and consumptive use of natural resources is so large. The fact
that protected area resources provide a critical income flow that compensates for
202
insufficient land in many cases, does not compensate for short-term risk. The story is more
like what Reardon (II) proposed - that the poor households in this study are thinking
about survival first, that they recognize conditions are not ripe for agricultural
intensification as a solution, that they are aware that protected area resources may be only
a short term stop-gap. The short-term risk in investing in either farming innovation or
conservation overwhelms longer-term considerations.
Unfortunately, using protected area-based income for poverty alleviation
potentially has longer-term limitations as a livelihood strategy. With the exception of the
national analysis depicted in Figures C.6 and C.7, all results assume a system limited to
those living within a 5 km zone of influence around each protected area. Yet the supply
and demand pressures on both land and resources noted in the national-level analysis
underline the notion that the system is not closed. Protected areas supply products not
only to local populations, but society as a whole. Moreover, the total value of protected
area resources to the whole of Malawi goes beyond the economics of consumptive use, to
include genetic diversity, cultural, religious, aesthetic, and intrinsic natural value (50).
Demand for land and the demand for ecological resources will result in difficult
choices at the local level. Reliance on protected area resources can provide much-needed
alternatives to agricultural income in the short run where land is scarce. But in the longer
term, this reliance may expose poorer households to the strong possibility that land
demand will eventually overtake the resource base providing the alternative income
stream. Government and donor environmental policies and their prescriptive interventions
must consider replacing the poverty-alleviation function of protected area resources with
203
income alternatives that are neither land nor forest-based. Furthermore, community
conservation initiated by donor projects that do not address the large financial differences
between the value of land and the value of ecological resources may actually accelerate the
risk faced by the very communities to gain from such measures. As suggested by
Shyamsundar and Kramer (51), proceeds from protection (ecotourism, etc.) need to be
reinvested in local communities to compensate for the value of foregone agricultural lands
and secondary protected area products. In a country like Malawi, where conditions are not
ripe for autonomous agriculture intensification, and income diversification is ofren through
the exploitation of natural resources, investment must be directed towards the
development of alternative income streams that give poorer households options. And it is
essential that these options are not entirely dependent on natural resources, or exacerbate
already critical land shortages.
C.7
REFERENCES AND NOTES
1. FAO. 1997. State of the World's Forests. FAO, Rome.
2. Boserup, E. 1965. The Conditions of Agricultural Growth: The Economics of Agrarian
Change Under Population Pressure. Aldine Publishing Company, New York, USA.
3. Boserup, E. 1981. Population and Technological Change: A Study of Long Term
Trends. University of Chicago Press, Chicago, USA.
4. Simon, J.L. 1986. Theory of Population and Economic Growth. Basil Blackwell,
Oxford, UK.
204
5. TifFea, M., Mortimore M. and Gichuki, F. 1994. More People, Less Erosion:
Environmental Recovery in Kenya. Wiley, Chichester, UK.
6. Hackel, J.D. 1993. Rural Change and Nature Conservation in Africa: A Case Study
from Swaziland. Hum. Ecol. 21, 295-312.
7. Turner II, B.L., Hyden, G. and Kates, R.W. (eds.). 1993. Population Growth and
Agricultural Change in Africa. University Press of Florida, Gainesville, USA.
8. Lindblade, K.A., Carswell, G., and Tumuhairwe, J.K. 1998. Mitigating the Relationship
between Population Growth and Land Degradation; Land-use Change and Farm
Management in Southwestern Uganda. Ambio 27, 565-571.
9. Fairhead, J. and Leach, M. 1996. Reframing Forest History; A Radical Reappraisal of
the Roles of People and Climate in West African Vegetation Change. In; Time-scales and
Environmental Change. Driver, T.S. and Chapman, G.P. (eds.). Routledge, London, UK.
10. Lele, U. and Stone, S. 1989. Population Pressure, the Environment and Agriculture
Intensification: Variations on the Boserup Hypothesis. World Bank, Washington D C.,
USA.
11. Reardon, T. 1998. African Agriculture; Productivity and Sustainability Issues. In;
International Agricultural Development, Third Edition. Eicher, C.K. and Staatz, J.M.
(eds.). Johns Hopkins University Press, Baltimore, USA.
12. Reardon, T., Fall, A.A., Kelly, V., Delgado, C., Matlon, P., Hopkins, J., and Badiane,
O. 1994. Is Income Diversification Agriculture-led in the West African Semi-Arid
205
Tropics? The nature. Causes, Effects, Distribution and Production Linkages ofOff-Farm
Activities. In: Economic Policy in Africa: What Have We Learned? Atsain, A., Wangwe,
S., and Drabek, A.G. (eds.). African Economic Research Consortium, Nairobi, Kenya.
13. Ford, R.E. 1993. Marginal Coping in Extreme Land Pressures: Ruhengeri, Rwanda.
In: Turner II, B.L., Hyden, G. and Kates, R.W. (eds.). Population Growth and
Agricultural Change in Africa. University Press of Florida, Gainesville, USA.
14. Lynam, J.K. and Herdt, R.W. 1989. Sense and Sustainability: Sustainability as an
Objective in International Agricultural Research. Agr. Econ. 3, 381-398.
15. Shackleton, C.M. 1996. Potential stimulation of local rural economies by harvesting
secondary products: a case study of the Central Transvaal Lowveld, South Africa. Ambio
25, 33-38.
16. "Protected area natural resource products" include wild fruits, vegetables, game,
flielwood, timber, thatch, and fiber, as well as such processed goods as handicrafts and
medicines derived from wild species
17. Carrasco, D.A. 1993. Constraints to sustainable soil and water conservation: a
Dominican Republic Example. Ambio 22, 347-350.
18. Mwafongo, W.M.K. and M.L.M. Kapila. 1999. Environmental management in Malawi
- lessons from failure. In: Etrvironmental Planning, Policies and Politics in Eastern and
Southern Africa. Macmillion Press, Houndsmill, Basingstoke, Hampshire, UK.
206
19. World Bank. 1998. Malawi: Human Resources and Poverty. Profile and Priorities for
Action. No. 15437-MAI. World Bank, Washington D.C., USA.
20. Snapp, S.S. 1998. Soil Nutrient Status of Smallholder Farms in Malawi Commun. Soil
Sci. Plant Anal. 29:2571-2588.
21. House, W.J. and Zimalirana, G. 1992. Rapid Population Growth and Poverty
Generation in Malawi. J. Mod Afr. Stud. 30, 141-161.
22. Berry, V. and C. Petty (eds.). 1992. The Nyasaland Survey Papers, 1938-1943,
Agriculture, Food and Health. Academy Books. London, UK.
23. Zeller, M., Diagne, A. and Mataya, C. 1998. Market access by smallholder farmers in
Malawi: implications for technology adoption, agricultural productivity and crop income.
Agr. Econ. 19:219-229.
24. Evans, J., Banda, A. and Seymour, T. 1999. Opportunities for Better soil
Management. European Union/DflD/Danida/USAID, Lilongwe, Malawi.
The authors found in cases where better soil management practices were adopted that it:
"was not the product of spontaneous innovation, but the response to a serious soil-related
production problem catalyzed and supported by external assistance...the result of
fortuitous coming-together of farmers with soil problems and extension workers with an
expanded range of eflfective low-cost solutions."
25. Wiyo, K.A. and Feyen J. Assessment of the effect of tie-ridging on smallholder maize
yields in Malawi. Agr. Water Manage. 41, 21-39.
207
26. Carr, S.J. 1997. A green revolution frustrated: lessons learned from the Malawi
experience. Afr. Crop Sci. J. 5, 93-98.
27. Orr, B., Eiswerth, B., Finan, T. and Malembo, L. 1998. Malawi Public Lands
Utilization Study. Malawi Ministry of Lands, Housing, Physical Planning, and
Surveys/University of Arizona/Forestry Research Institute of Malawi, Lilongwe, Malawi.
28. Malawi Government. 1998. Report of Preliminary Results: 1998 Population and
Housing Census. National Statistics Office, Zomba, Malawi.
29. Arpaillange, J. 1996. Urban Household Energy: Demand Side Strategy.
SEED/Ministry of Energy and Mining, Lilongwe, Malawi.
30. FAO. 1999. State of the World's Forests. FAO, Rome.
31. Openshaw, K. 1997. Urban Biomass Fuels: Production, Transportation, and Trading
Study. Alternative Energy Development, Inc./Malawi Ministry of Energy and Mining,
Lilongwe, Malawi.
32. Willan, R.G.M. 1947. Annual Report of the Forestry Department for the Year Ended
31st December, 1946. Nyasland Protectorate, Zomba, Malawi.
33. Edwards, I. 1985. Conservation of Plants on Mulanje Mountain, Malawi. Oryx 19, 8690.
34. Bhima R., and C.O. Dudley. 1997. The Influence of the Shire River on Liwonde
National Park, Malawi, with Special Reference to Elephant Movements. Koedoe 40, 9-18.
208
35. Ngalande, J.D. 1995. An Integrated Management Plan for Dzalanyama Forest
Reserve, Malawi. M.S. Thesis, University of Aberdeen, Scotland.
36. McShane, T.O. 1985. Vwaza Marsh Game Reserve: A Baseline Ecological Survey.
Department of National Parks and Wildlife, Rumphi, Malawi.
37. Woodson, D.G. 1997. Lamanjay, food security, securite alimentaire: a Lesson in
Communication for BARA's Mixed-Methods Approach to Baseline Research in Haiti,
1994-1996. Cult. Agr. 19, 108-122.
38. Orr, B.J., Chapama, H.T. and Malembo, L.N. nd. The Symbiosis between Protected
Area Resources and Livelihood Security. Econ. Bot. (in preparation).
39. Chambers, R. 1994. The Origins and Practice of Participatory Rural Appraisal. World
Dev. 22, 953-969.
40. Eschweiler, J.A., Paris, S., Venema, J.H., Lorkeers, A.J.M. and Green, R.L 1991.
Methodology for Land Resources Survey and Land Suitability Appraisal. LREP, Field
Doc. No. 30, MOA/UNDP/FAO, Lilongwe, Malawi.
41. Momiere, L., Weiss, E. and Chimwaza S. 1996. A Quest for Causality: Vulnerability
Assessment and Mapping (VAM), Malawi Baseline 1996. WFP/GOM/FEWS, Lilongwe,
Malawi.
42. Masamba, C. and Ngalande J. 1997. Inventory Data of Biomass Growing Stock and
Supply. FRIM Technical Report. Forestry Research Institute of Malawi, Zomba.
209
43. Satellitbild. 1993. Forest Resources Mapping and Biomass Assessment for Malawi.
Satellitbild/Malawian Department of Forestry. Lilongwe, Malawi.
44. Sen, A. 1973. On Economic Inequality. Norton & Company, New York.
45. The preliminary 1998 census results (28) do not as of yet contain an estimate of error.
However, the overall adjustment from the 3.2% national growth rate estimated from the
1977 to 1987 census to 1.9% based on the 1987 to 1998 trend is in line with pre-census
estimates that suggested changes in mortality rates would have a large impact on
population trends.
46. BDPA/AHT. 1998. Customary Latid Utilization Study. BDPA/AHT, Lilongwe,
Malawi.
47. According to the Customary Land Study (46), for every 1 ha of cultivated customary
land in 1996, there were 2.4 ha considered as essential customary land infrastructure (i.e.
fallow or abandoned farm plots, village forests and wood lots, cemeteries, roads, schools,
football grounds, structures, etc.).
48. The percentage of land in the zone of influence that was cultivated ranged between 31
and 37%, and was held constant in the projections of future land use patterns posited for
agriculturally suitable land inside the reserve.
49. The soils in and around Vwaza are highly suitable for tobacco cultivation, an activity
that requires drying sheds. At the time of the study, many farmers were busy constructing
210
these sheds using poles obtained from the reserve - their durability would suggest the
Vwaza wood consumption rates might be overstated for long term projections.
50. Rolston III, H. 1985. Valuing Wildlands. Environ. Ethics 7:23-48.
51 Shyamsundar, P. and Kramer, R. 1997. Biodiversity Conservation—At What Cost? A
Study of Households in the Vicinity of Madagascar's Mantadia National Park. Ambio
26:180-184.
C.8 FIGURES
Length of Growing
Period (LGP) ,
Slope (SI)
Soil Depth (Sd)
LREP Soils &
LREP Agroclimaiic Data
Physiography Dale
LREP Soils &
Physiography Data
Ag Suitable tf
LGP 120-270 days
(LREPcodM 2-tO)
Ag Suitable if
Sl< 13%
Ag Suitable tf
Sd > 50 cm
(IREPcoOn 1-3)
llREPcAdvt 3 5|
Surface Stonlness
or Rockiness (Sr)
Drainage Class
(Dr)
LREP Soils &
Physiography Data
LREP Soils a
Physiography Data
Ag Suitable if
Sr< 15%
(iREPcodM 1-2)
If all yes" - suitable for agriculture.
If any "no" - not suitable.
Agricultural Suitability
FIGURE C.l. AGRICULTURE SUITABILITY MODEL (40,27).
Ag Suitable if
Dr a moderate to well
llRCPcodttM MW W)
Ponding (P)
LREP Soils &
Physiography Data
Ag Suitable if
P s none to sitghtly moderate
(LREPcodMN Sl 6lM)
212
Population Density/
(1987 Census from
FEWS database)
Protected Areas
Map
Population Map / \
EPA Map
(in EPA-based pdygona
within the 5 km zone
around protected area)
FEWS Data
Digrttzed 1:250.000
Survey Sheet
Parks & Reserves
V
I
Grid
Move Population
Density to NonProtected Areas o\
EPAs
(rastenze input maps to
30m cells)
Protected Area
I
P i P ! P
5km Radus
Convert Populatiof
Density to Total
Population
I
Place 5 km Buffer
Around ail
Protected Areas
L = Land Area Cells
inside Protected Area
P = Population
Add an "P" Cells in entire BulTer Zone
divided t}y the area cH ttie reserve
I
Intersect Buffer
with EPAs
(Hits removes aO txjt the 5
icn'population zone*
around the protected areas)
Adda«"P"Cefls
wrthin 5 km Radius
of eacn -L- Cell
z.
Direct
Population
Pressure
Potential
Population
Pressure
I
Classified as;
- Beiow National Average
- Above National Average
Population
Pressure
FIGURE C.2. POPULATION PRESSURE SPATL\L MODEL (ADAPTED; 27).
213
Satelite Imagery
Satelite Imeigery
1984 TM Data
(30fn cells)
1994 TM OataGnd
(30m cells)
1. Miombo Woodland
2. Open Mtombo WootSand
3. Broadteaf Oeoduous Forest
Forest = F
4. Broadteaf Evergreen Forest
Image Processing
7. OamCx>
e. Grassland Hert)aceous = H
9. Marsh
10. Pine
11. Eucalyptus
Plantation = P
12. Outcrop
13. Shadow
f4. WaAcr
Classification
Benign = B
Agriculture = A
17. Soils and Clay
1984 Land Cover
Soils
S
Categorize Land
Cover Classes
1994 Land Cover
Environmental impact of Land Cover Change
Between 1984 ar>d 1994:
7
7
14
14
7
7
1
7
7
14
14
7
1
1
7
7
14
14
1
1
1
\
1
1
10
10
7
7
10
15
15
10
15
15
From:
s
Miornbo
Pine
Agriculture
10
Change Detection Matnx
Classified as:
A. Positive
B. Negative
C. Neutral
tf A,F.H.P »> F
lfS»>P
If A,F.H.P.S »> S
IfF.H.P >» A
If A.F.H »> P
If F»> H
IfS.B »>S
If A,B.S »>A
If B >» F.H
If BP »> P
If A,B.F.H.P.S »>B
positive change
negative ctiange
neutral or no ctiange
Impact
(land cover change)
FIGURE C.3. VEGETATION CLASSIFICATION AND CHANGE DETECTION
MODEL (27).
214
40%
35%
X
30%
y = 0.2949e
R- = 0.8444
••
V
u
"2 25%
20%
F =151.899
(Sig. .000)
••
Z 15%
t = 22302
(Sig. .000)
E
e
e
^ 10%
e
5%
0%
0
5.000
15,000
Per Capita Income(MK)
• reliance —— predicted reliance
FIGURE C.4. REGRESSION FOR PER CAPITA INCOME AS A PREDICTOR FOR
RELIANCE ON PROTECTED AREA RESOURCES.
215
100%
80%
u.
se
60%
V
E
o
w
s
40.1%
40%
V
>
20%
S
0%
OC
-20%
21.9%
12.5%
5.0%
-7.9%
,CN
/
>&
</
Income Groups (fifths ofthe population)
•change in share of income
FIGURE C.5 CHANGE IN RELATIVE INCOME SFL\RE RESULTING FROM THE
ADDITION OF PROTECTED AREA-BASED PROCEEDS TO HOUSEHOLD
f^POME
216
35,000
30,000
25,000
• •
^^
i* 20,000
ts
>
's
O" 15,000
E&
0
c
10,000
•s
B
1
sustainable wood supply
wood demand
5,000
1940
I960
1980
2000
2020
2040
NB: "sustainable supply" refers to the mean annual increment
— each year's new growth measured in m^ ha ' y r ' .
FIGURE C.6. WOOD DEMAND VERSUS SUSTAINABLE SUPPLY IN MALAWI.
217
100.000
Total Land
90.000
(94,000 km")
80.000
70.000
£
60.000
Agriculturally
Suitable Land
X
u
50.000
(57J00 km")
•a
e
sa
40.000
30,000
20.000
land demand
10.000
0
1900
1950
2000
2050
NB: Land demand (ha capita ) is based on historic, actual and projected resident population
figures multiplied by the relative land area per capita represented in the customary (0^0 ha
capita '), urban (0.07 ha capita ') and estate (13 ha capita ') sectors. Added together, this
amounts to land demand overtime.
FIGURE C.7. LAND DEMAND VERSUS AVAILABILITY IN MALAWI.
218
250 4
a.
s
c/i
e
o
z
ee
<n
3
(/J
1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
•Mulaig'e —Liwonde -*—Vwaza —Etolanyama
FIGURE C.8. PROJECTED DECLINE IN THE SUSTAINABLE SUPPLY OF WOOD
(MAI ROUNDWOOD EQUIVALENT) AS POPULATION GROWS.
219
800 -
700 -
200 -
1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
Mulai^e —^ Liwonde —Vwaza —Dzalanyama
FIGURE C 9. HYPOTHETICAL DECLINE IN AGRICULTURALLY SUITABLE
LAND INSIDE THE PROTECTED AREAS AS POPULATION GROWS.
220
Landsat Thematic Mapper Images:
1984 Compared to 1994
Liwonde National Park
Landsat TM Bands 4, 3,2 shown as R, G, B
10
15 Kilometers
Arizona Remote Sensing Center
Office of Arid Land Studies
University of Arizona
FIGURE C.IO. LANDSAT THEMATIC MAPPER FALSE COLOR IMAGES OF
LIWONDE NATIONAL PARK IN 1984 AND 1994 (27).
221
Zoom of 1984 and 1994 Landsat TM Images
Liwonde National Park
1984 Fdse Color Conqwsite
1994 False Color Couqwsite
AriaoiuRcmot* S*B5iac Center
Officc of And L«iid S todies
Uiiiveiiiiy of Ariaoiu
FIGURE C.l I. EASTERN-MOST PORTION OF LIWONDE NATIONAL PARK IN
1984 AND 1994.
222
Nature of the Resource:
Agricultural Suitability Model Results
Liwonde National Park
Agricultural Suitability
HHl Suitable (17.0%)
Not Suitable (83.0%)
10
10 Kilometers
Arizona Remote Sensing Center
Office of Arid Land Studies
University of Arizona
FIGURE C.12. AGRICULTURAL SUITABILITY OF LAND IN LIWONDE
NATIONAL PARK (27).
223
Pressure on the Resource:
Potential Population Pressure
Livvonde National Park
Population Pressure
Relative to
National Average
Below Average
Above Average
Population Density
Surrounding Site
0
I
10
15
20 Kilomciers
people per km
1-50
50-100
100-150
150 - 200
200 - 250
250 - 300
Arizona Remote Sensing Center
Office of Arid Land Studies
University of Arizona
FIGURE C.I3. POPULATION PRESSURE AROUND LIWONDE NATIONAL PARK
(27).
Impact on the Resource:
Land Cover Mapping and Change Evaluation
Liwonde National Park
1984 Land Cover Map
84 l.and Cover
I
Mopanc Woodland
Riverine forest
lliickct
Marsh Vegetation
(^Grassland
H
Water
Dambo
I—I Mixed
litnd cover interpreted from
unsupervised classification of
I'M Bands 1,2,3,4,5,7,
MSAVI, and K-TGn;eness
1994 Land Cover Map
94 Land Cover
Evaluation of Change (84-94)
lijnd Cover Change
I'lvaluation
Miombo Woodland
Mopanc Woodland
Riverine forest
lliickct
Marsti Vegetation
Gmssland-Agriculturc
Water
Dambo
n Mixed l'(
I
H
10
15 Kilometers
I neutral or no change
negative
posiliw
Arizona Remote Sensing Center
OlTiee of Arid land Studies
University of Arizona
FIGURE C.14. LAND COVER CLASSIFICATION AND CHANGE DETECTION IN LIWONDE FROM 1984 TO 1994.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement