Farming systems research in three counties in Liberia

Material Information

Farming systems research in three counties in Liberia a reconnaissance survey on Grand Gedeh, Nimba, and Bong counties
Series Title:
Applied Anthropology Documentation Project
Frankenberger, Timothy R
Farming Systems Support Project
Place of Publication:
Gainesville Fla
Publication Date:
Physical Description:
xv, 178 p. : ; 28 cm.


Subjects / Keywords:
Agriculture -- Liberia -- Grand Gedeh County ( lcsh )
Agriculture -- Liberia -- Nimba County ( lcsh )
Agriculture -- Liberia -- Bong County ( lcsh )
non-fiction ( marcgt )


General Note:
Supported by the Farming Systems Support Project, University of Florida with cooperation from : the University of Kentucky, USAID/Monrovia, and the Central Agricultural Research Institute, Suakoko, Liberia.
Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
Statement of Responsibility:
by Timothy R. Frankenberger [et al.]

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services ( with any additional information they can provide.
Resource Identifier:
022153595 ( ALEPH )
14208881 ( OCLC )
ACY8138 ( NOTIS )

Full Text
2Z. 0o7
Farming Systems Support Project
International Programs Office of Agriculture and
Institute of Food and Office of Multisectoral Development
Agricultural Sciences Bureau for Science and Technology
University of Florida Agency for International Development
Gainesville, Florida 32611 Washington, D.C. 20523

Timothy R. Frankenberger
John A. Lichte
Arthur S. Gedeo
John Kpakolo Jallah
Maran J. Sherman
April 1985
Supported by the Farming Systems Support
Project, University of Florida, with
Cooperation from: the University of Kentucky, USAID/Monrovia, and the Central Agricultural
Research Institute, Suakoko, Liberia

A farming systems reconnaissance survey was conducted in Grand Gedeh, Nimba, and Bong counties in Liberia. The research team consisted of an anthropologist, two agricultural economists, a soil scientist, and a tree crop researcher. The survey was conducted over a three week period in 19 villages. One hundred and fourteen farm families were interviewed. A detailed topical outline was used to guide interviews, and each interview was conducted on the farm family's field with both the husband and wife present. The major findings of this study are presented below.
Two basic farming systems are found in the three counties surveyed. The key factors which distinguish the systems are the length of fallow, the rice planting method, and the degree of diversity. The rice-cassava intercrop pattern is also important in distinguishing sub-system differences in Bong and Nimba.
Grand Gedeh Farming System
Grand Gedeh farmers typically make their upland fields in high bush which has been in fallow for up to 20 years or more. The presence of numerous large trees makes tree felling the most difficult and constraining field preparation activity. Groundhogs, weeds, and birds are less of a problem than where the fallow period is normally shorter. Rice is planted using the dibble method and no soil tillage is performed. If a reasonable burn is achieved, farmers do not attempt to clear the field of remaining tree trunks and debris.
Cassava is planted on a portion of the rice field about one month after the rice is planted. This delay prevents cassava from shading the rice and minimizes competition. Bananas and plantain are intercropped with the rice on a different portion of field. A number of tubers and other vegetables are intercropped with the rice throughout the entire field. These include: cocoyams. yams. sweet potatoes, corn, okra, peppers, bitterball, pumpkin, squash, plato, watergreens, tomatoes, and sesame.
Both early and late maturing rice varieties are usually planted. The early variety is planted first so that rice harvesting can begin as early as July or August.
Farmers tend not to build fences to protect their rice from groundhogs. Instead. they try to increase the size of their field and/or locate their field far from areas of secondary bush where groundhogs are likely to be numerous.
In high bush, the dense shade of the mature forest limits the presence of weeds. With little weed seed in the soil to germinate, rice gets a head start and dominates weed competition. Farmers generally do not expect to weed their high bush fields.
Few farmers have either traditional or improved swamp rice farms. A portion of the upland rice field often descends into a swampy area. However, the swamp is considered part of the upland

rice field and is cultivated in the same manner.
Women often plant a short maturing rice and intercrops in a small "hungry farm" separate from the main field. It is usually placed on a portion of the previous year's rice f ield or some other area of secondary bu s h. This f iel1d assures the family's rice needs until the main field is ready for harvest. The remaining produce of the "hungry f arm" is used to meet the personal needs of the woman.
The reciprocal kuu system is not generally practiced in Grand Gedeh villages and farmers rely more on family labor. Although some labor is hired to help with many farming activities, hired labor is more expensive and less frequently employed than in the other counties surveyed.
Cocoa and coffee are the dominant cash crops in Grand Gedeh. In addition, rice, cassava, and minor crops are sometimes sold. The sale of wild meat and produce collected from the forest (e.g. wild palm oil fruit) are other important sources of income. Farmers own and produce few animals even though they are important in meeting social obligations and as a source of ready cash for emergencies.
Nimba and Bong
Many of the characteristics of the farming system in Nimba and
Bong Counties are similar. Both counties are experiencing some degree of land pressure, which has shortened fallow periods, and compels farmers to make their farms on secondary bush. In Nimba, t h is pressure comes from a high population density, while in Bong, it comes f rom the presence of concessions and from the private ownership of large estates. Problems closely associated
with farming secondary bush, such as groundhogs, weeds, and rice b ir d s, have an important impact on the characteristics of this farming system.
Farms appear to be slightly larger than in Grand Gedeh, perhaps due to some combination of the following factors: 1)
felling trees is less of a constraint; 2) when large kuus do the brushing, more land may be cleared than would be otherwise by family labor; 3) poor fertility and pests associated with
secondary bush may oblige families to farm more extensively to meet food requirements; 4) the absence of "hungry farms" may both
permit and oblige the cultivation of a larger main upland rice f ielId .
Few large trees are present, so the major land preparation a ct iv it y is brushing rather than felling trees. More e f for t is also devoted to clearing small tree trunks and debris after
burning so that it will be easier to hoe the field. Hoeing (or scratching) the field does not begin until 2 to 8 weeks af ter burning, in or d er to give weed s e ed in the s oilI a chance to germinate. Broadcast planting is done at the same time as the scratching in order that this single hoeing might both eliminate

the weeds and cover the seeds. After hoeing, the weeds are
removed from the soil and piled so they will not have a chance to grow back. A second weeding often seems to be necessary, at least in years of high rainfall. This may also be related to the severely reduced fallow period.
In Nimba, cassava is planted at low density at the same time as the rice. It is typically intercropped on the entire rice field, eliminating bananas and plantain as an important rice intercrop. If the cassava gets too tall relative to the rice, the lower branches and leaves are stripped to reduce shading. In Bong, the cassava density is perhaps even lower than in Nimba and it is planted 2 to 4 weeks before the rice. It is often planted on only a portion of the rice field and stripping the cassava leaves to prevent shading is less frequently practiced. Cassava, in Nimba seems to have a more important role in both production and consumption than in the other two counties.
In Bong, some of the corn is planted with cassava before the rice is planted. Some of the vegetable intercrops tend to be planted after the rice in both counties.
Early and late maturing rice varieties are usually planted, but the pattern is not consistent. Where birds are a serious problem, the late maturing varieties tend to be planted first. The fear is that an early maturing field will be decimated by the birds, but the attack will be spread across more fields later in the season. This strategy may lengthen the "hungry season" by delaying the new harvest by 2 to 3 months.
Farmers regularly build fences in Bong and Nimba to protect rice from groundhogs. This will occupy the men for a month or more while women finish the planting and weed. Bird-watching is also common for a week after planting and for about a month while the rice is heading.
A few farmers have swamp rice fields. Women are more often responsible for the traditional swamp rice fields, but- men tend to be more involved in the improved rice fields sponsored by projects. "Hungry farms" are not common, but in some cases the swamp rice fields will take their place as well as provide a source of income for personal needs.
The diversity of farm enterprises is one of the biggest differences between the farming system in these two counties and that of Grand Gedeh. Farmers are often involved in both cocoa and coffee and perhaps sugar cane or groundnuts, as well as in their upland field and occasionally swamp rice. Rubber, citrus, or cultivated oil palm may also be present. Farmers are also
more likely to be involved in some form of off-farm employment. These numerous activities in any one family tend to strain family labor resources and management capabilities. For these reasons, hiring labor and hiring kuus as well as using reciprocal kuus are very prevalent in these two counties. Farmers in Nimba often raise pigs with the intention of using them to feed kuus. The use of hired labor is further encouraged by lower day wages than those prevalent in Grand Gedeh.

The survey has identified a number of general and crop specific constraints which are impinging upon the existing farming systems. The major constraints are summarized below along with possible strategies to deal with each type of constraint.
I. General Constraints
A) Access to land
I) Identify the most appropriate crop rotation system for
specific areas.
2) Investigate how tree crops can be combined with food
crops in the same field.
3) Emphasize improved swamp rice farming in areas which
are experiencing severe land pressure.
4) Investigate the possibility of integrating chemical
inputs into crop rotation systems and improved swamp
rice farming.
B) Access to labor
1) Projects should refrain from introducing too many
interventions to a given farmer at one time.
2) Consider use of intercropping and cover-cropping
strategies to reduce labor requirements.
3) Investigate the economics of existing mechanical
interventions and new types of appropriate
C) Access to capital
1) Consider introducing alternative cash crops into the
2) Improve pricing p oIi c ies, processing techniques, and
marketing channels for existing cash crops.
3) Explore replication of successful village credit
II. Upland Rice Constraints
A) Pests and diseases
1) Investigate alternative fencing techniques and/or
chemical repellants and poisons as deterrents to
2) Study the i f e cy cl1e and breeding habits of
-3) Incorporate bird resistant t r a its into improved rice
v ar ie t ies .
4) Consider the adverse consequences of using poisons to
control birds and rats.
5) Investigate alternative means f or controlling
6) Continue research on designing inexpensive r ic e

kitchens that control rats.
7) Continue efforts of selecting for disease resistant
rice varieties.
8) Encourage project and extension personnel to educate
farmers on how to recognize rice diseases and how to
deal with these when they are identified.
B) Weeds
1) Initiate studies on the effectiveness and costs of
using herbicides.
C) Soil fertility
1) Promote crop rotation systems which integrate
nitrogen-fixing legumes as rotation crops or
2) Continue experiments on minimum tillage practices and
other erosion control measures.
3) Investigate the use of composts, manures, and other
indigenous materials like rock phosphate for restoring
4) Consider managed fallow using legumes as a substitute
for natural fallow.
D) Other areas of investigation for rice
I) Compare dibbling and scratching methods of planting to
determine effect on plant density, weed population, soil degradation, labor demand, ease of intercropping,
and yield.
2) Investigate the advantages and disadvantages of
intercropping other crops with rice.
3) Screen and compare local rice varieties with
recommended varieties through on-farm trials.
II. Cassava Constraints
A) Pests and diseases
1) Continue multi-locational trials of disease resistant
varieties through on-farm trials.
2) Encourage project and extension personnel to educate
farmers on how to recognize cassava diseases and what
to do about them.
B) Other areas of investigation for cassava
1) Study the effects of different spacing and timing
strategies of planting cassava in relation to rice.
2) Determine the effects of striping the leaves and
lower branches of young cassava plants.
3) Identify potential crops that can be intercropped
with cassava which could serve as cover-crops.

IV. Swamp Rice Constraints
A) Lab-rav'ailability
1 ) Research should focus on the economics of combining
improved swamp rice with upland rice and other farming
2) Identify cost-effective, labor-saving techniques to
reduce labor demands of swamp farming.
3) Encourage farmers to delay swamp-farming activities
until upland activities are completed.
4) Study areas where improved swamp rice farming has been
successfully adopted.
B) Other areas of investigation for swamp rice
1) Consider the use of nitrogen fixing aquatic plants.
2) Investigate the effects of cutting back rice to deal
with lodging.
V. Cocoa and Coffee Constraints
A) Pestsan diseases
1) Encourage farmers to follow practices which help
control stem borers such as pruning, destroying
infested branches, and timely underbrushing.
2) identify potential hosts which may harbor stem borers.
3) Investigate the economic feasibility of using
insecticides f or prevention and/or treatment of
infested trees.
4) Consider the ecological effects of eliminating red ants
before proposing control methods.
5). Explore alternative means of preventing wild a n im aIs
from eating farmers' cocoa.
6) Identify or develop black pod resistant varieties of
cocoa and/or identify other means to control the
B) Underbrushing
1) Explore the possibility of using cover-crops for cocoa
and coffee. These could possibly be food crops.
C) Site selection
1) Determine the suitability of different soil types for
cocoa and coffee.
D) Other areas of investigation for cocoa and coffee
1) Research should determine the appropriate spacing for
cocoa and coffee. Explore the economic trade-offs between labor expenditure and potential yield loss when
considering spacing.

VI. Other Tree Crop Constraints
1) Continue to identify possible varieties of fruits
which bear year round or at different times of the
year to deal with the problem of seasonality.
VII. Animal Constraints
1) Encourage projects and extension services to provide
better veterinary services to diagnose and treat
infected animals.
2) Determine the extent to which i d e nt if ied diseases
have spread.
B) Other areas of investigation for animals
1) Research should focus on why traditional farmers are
not investing more in animals.
VIII., Marketing Constraints
A) Access to market information
1) Develop a low co0s t method f or disseminating
marketing information, i.e., radio broadcasts in
local languages.
B) Grading of cocoa and coffee
1) Projects and extension services should initiate
efforts to train farmers how to grade their own
cocoa and coffee.
2) Instruct farmers as to the proper processing
techniques for improving the grades of cocoa and
IX. Food Consumption Constraints
A) Areas of investigation for consumption
1) Conduct studies of food preferences, preparation
techniques, and food taboos to determine acceptability
of proposed interventions.
2) Varietal work on any crop should consider taste
preferences .
3) Encourage the introduction of other beans and pulses
into the farming systems such as mung beans and
4) Conduct limited consumption surveys to determine the
nutritional effects of seasonal shortages of food during peak labor periods. This information can guide
crop choices for intervention.

We would like to take this opportunity to thank all of the. staff at the Central Agricultural Research Institute as well as the LSU project staff for their contributions and assistance they provided us while carrying out our research. Special thanks are due to Mr. Andrew Paye, Director of CAR1; Dr. Harold Young, LSU Team Leader; Dr. Charles Mulbah, Head of Land and Water Resources Department at CAR1; Dr. Joe Subah, Research Coordinator of CAR1; and Mr. Sam Hooke, Administrative Officer of CAR1. They provided the staff, vehicle, and backup support which enabled this mission to accomplish its task. We are grateful for their support.
We also wish to thank the staff at the AID Mission in Monrovia. We are grateful for the help and support provided by Dr. John Flynn, Mr. Myron Smith, Mr. Mark Smith, and Ms. Lois Richards. John Flynn was especially instrumental in coordinating and organizing the initial efforts of this study. John provided the team with valuable background information on the study area, introduced us to important contacts, and arranged the meetings between the CART staff and the research team. To John and the rest of the AID staff, we extend our gratitude.
A third group of individuals we are grateful to are the various government officials and project staff in the counties we worked in who aided us in our endeavors. Their contributions were very important to this study. They provided us with background information on the county as well as helped us select villages for our survey. Many of their staff also accompanied us as translators and/or participated in the survey. Special thanks
are extended to Mr. Alonzo N. Munyeneh, Station Manager, SayeDube Sub-Research Station as well as the Regional Agricultural Office in Grand Gedeh County. Alonzo not only contributed his time and energy by participating in the survey, he also opened his home to the research team during the course of the study. To him and his wife we are extremely grateful. We also wish to thank Dr. Heiko Dekena, Mr. Karl Kirsch, Mr. Mah, and the Extension
Personnel of NCRDP for their support and assistance. In addition to helping select villages and assisting us with their staff, they provided us with excellent lodging, warm hospitality, and an opportunity to exchange ideas. To them we express our gratitude. Thanks are also due to Jerry Mason and his field staff of BCADP. Jerry was instrumental in coordinating our survey efforts in Bong. To him and his staff we are grateful.
Thanks should also be extended to U.S.AID, Washington; the Farming Systems Support Project at Florida; and the University of Kentucky for providing cooperative support for this project. We are especially grateful to Ms. Gloria Steele, U.S.AID, Washington; Dr. Susan Poats, University of Florida; Mr. Wendell Morse and Ms Jo Albert, U.S.AID, Washington; and Dr. Herbert Massey, University of Kentucky for their efforts in initiating this project.
Finally, we are sincerely grateful to all the Liberian farmers who participated in this survey. The patience and kindness they

showed us in the face of persistent questioning was truly noteworthy. We are indebted to them for their hospitality. We hope that someday the outcomes of our research efforts will repay this kindness.

I. Introduction ......... ......................... I
A. Methodology ......... ....................... I
B. Content of Report ......... ................... 8
II. General Farming Sysytems Characteristics .... ........ 9
A. Access to Land-Fallow Rotation System .... ........ 9
B. Spatial Arrangements of Farmers' Fields ......... ..10
C. General Labor Patterns ..... ............... ...11
C.1 Sources of Labor ...... ................ ...12
D. Upland Rice ........ ..................... ..12
D.1 Upland Rice: High Bush Compared to Secondary Bush 12
Strategies for Use of High vs. Secondary Bush 13
Site Selection ...... ................. ...13
Land Preparation ...... ................ ...14
Importance of Weeding ..... .............. .14
Farm Size ....... .... ............ . 15
Planting Methods . ..... .............. 15
Vareities Grown and Time of Planting ....... ..17 Intercropping Strategies .... ............ 18
Harvest and Post Harvest Methods .......... ...19
Pests and Diseases ..... ............... ...20
Soil Fertility Considerations .... ......... 21
Hungry Farms ....... .................. ...22
E. Cassava .. ......... ....... .............. ..22
E.1 Cassava Intercropped with Rice ... .......... .23
Cassava Planted After the Rice ... .......... .23
Planting Cassava with the Rice .. ........... ..23
Cassava Before Rice ..... ............... ...23
Rice-Cassava Double-Cropping ... ........... ..24
E.2"Separate Cassava Fields ......... ......... 24
E.3 Cassava Varieties ...... ................ ...24
E.A Number of Cuttings ..... ............... ... 25
E.5 Pests and Diseases .... .............. 25
E.6 Other Crops Intercropped with Cassava....... 26
F. Swamp Rice Cultivation (General) .... ........... .26
F.1 Traditional Swamp Rice ..... .............. .26
Site Selection .............................. 26
Methods of Cultivation ..... .............. .27
Varieties Grown ...... .................. 27
Post-Planting, Harvest and Post Harvest Methods 27
F.2 Improved Swamp Rice ..... ............... ...28
Location of Improved Swamps ... ........... ...29
Varieties Grown ...... ................. ...29
Methods of Cultivation .... ............. ... 29

G. Other Major Field Crops .. ............ ..31
G.1 Sugar Cane. .. ..................31
Site Selection. .. ................31
Land Preparation and Methods of Cultivation . .31 Varieties Grown .. ...........................32
Harvesting and Processing Techniques. .. ....32 Sugar Cane Growing Strategies .. .........33
G.2 Groundnuts. .. ..................33
Site Selection. .. ................33
Land Preparation and Methods of Cultivation 34
H. Tree Crops. .. ....................34
H.1 Cocoa .. ....................................35
Years of Experience .. ......................35
Sources of Planting Material. .. ........35
Site Selection...................................35
Methods of Cultivation. .. ............36
Pests and Diseases. .. ..............39
H.2 Coffee .. .......... ..........40
Years of Experience .. ........... ...40
Sources of Planting Material. .. ........40
Site Selection. .. ................41
Methods of Cultivation. .. ...........41
Diseases and Pests. .. ..............43
H.3 Other Tree Crops. .. ...............43
Banana and Plantain .. .......... ....43
Oil Palm .. .......... .........44
Rubber .. .......... ..........45
1. Livestock .....................................46
11.1 General Husbandry Patterns. .. .........46
1.2 Goats .. ............. ........47
1.3 Sheep .. ............. ........47
1.4 pigs. .. .....................47
1.5 Cattle. .. ....................48
A.6 Poultry .. ............. .......48
J. Marketing .. ............... ......48
J.l Marketing Channels. .. .............48
J.2 Marketing of Cocoa and Coffee .. .........49
J-3 Marketing of Rice .. ........... ....51
J.4 Marketing of Cassava. .. .............51
J-5 Sugar Cane (Cane Juice) .. ............51
J.6 Other Crops .. ............. .....52
K. Other Sources of Income .. .......... ....52
K.1 Off Farm Employment .. ........... ...53
K.2 Farm Laborers .. ............ .....53
K-3 Arts and Crafts .. ............ ....54
K.4 Wild Game .. ............ .......54
K.5 Palm Wine .. .......... .........54
K.6 Fishing .. ............ ........
K.7 Money Sent From Relatives.............54

L. Cred it .. ....................55
L.1 Credit from Government Organizations . . 56
M. Food Consumption .. ...............56
M.1 Food Preferences .. .............56
M.2 Seasonality. .. ..............56
M.3 Sources of Meat. .. ............57
MAL Dietary Patterns .. .............57
M-5 Food Taboos .. ........ ........58
M.6 Other Culturally Prescribed Foods .. .....58
M.7 Nutritional Implications of Current Dietary
Patterns .. .................59
N. Community (Communal) Farms .. ..........60
0. Government Interventions .. ...........60
0.1 Project Activities .. .................6o
0.2 Constraints Imposed on Farmers by Project
Activities .. ................61
III. Farming Systems Characteristics in Grand Gedeh . .63
A. Access to Land .. .......... ..........63
B. Upland Rice. .. ...................63
C. Cassava. .. .....................64
D. Swamp Rice .. ............... .....65
E. Other Field Crops. .. ...............65
F. Tree Crops .. .................. ..65
G. Livestock. .. ....................66
H. .Marketing in Grand Gedeh County. .. ........67
1. Other Sources of Income. .. ............67
J. Credit in Grand Gedeh County .. ..........67
K. Food Consumption .. .................68
L. Government Interventions in Grand Gedeh County 68
IV. Farming Systems Characteristics in Nimba County . .70
A. Access to Land. .. ..................70
B. Upland Rice .. ................ ....70
C. Cassava .. ................. .....71
D. Swamp Rice. .. ....................72
E. Other Field Crops .. .................73
F. Tree Crops. .. ....................73
G. Livestock .. ............... ......74
H. Other Sources of Income .. ..............74
I. Marketing in Nimba County .. .............74
J. Credit in Nimba County. .. .............74
K. Food Consumption. .. .................75
L. Government Interventions in Nimba County. .. ....75
V. Farming Systems Characteristics in Bong County. .. ...76
A. Access to Land .. ................ ...76
B. Upland Rice ... .. ...................76
C. Cassava. .. ......................77
D. Swamp Rice .. ................ .....77
E. Other Field Crops. .. ................78

F Tree Crops 78
G. Livestock 79 H. Marketing in Bong County . . . . . . . . 79
1. Other Sources of Income . . . . . . . . 79
J. Credit in Bong County 79
K. Food Consumption 80
L. Government Interventions in Bong County . . . . 80
VI. Summary of County Specific Farming Systems Characteristics 81
1. Grand Gedeh 81
11. Nimba and Bong 82
VII. Constraints, Areas of Investigation and Recommendations 84
1. General Production Constraints . . . . . . 84
A. Access to Land 84
B. Access to Labor 86
C. Access to Capital 87
11. Crop Specific Constraints and Areas of Investigation 88
A. Upland Rice 88
1. Pests and Diseases 88
2. Weeding 92
3. Soil Fertility .. i i 93
4. Other Areas of Investigation for Upland Rice 93
B. Cassava 94
1. Pests and Diseases 94
2. Other Areas of Investigation for Cassava . . 95
C. Swamp Rice 96
1. Labor Availability 96
2. Other Constraints to Improved Swamp Rice
Farming 97
3. Other Areas of Investigation for Swamp Rice . 98
D. Tree Crops 99
1. Cocoa and Coffee . . . . . . . . 99
2. Other Tree Crops . . . . . . . 103
E. Other Field Crops 104
1. Sugar Cane 104
2 Peanuts 104
111. Animals (Goats, Sheep, Pigs, Cattle, Chickens) . 104
A. Diseases i i 104
B. Other Areas of Investigation for Animals . . 105
IV. Marketing 105
A. Access to Markets . . . . . . . . 105
B. Access to Market Information . . . . . 105
C. Grading of Cocoa and Coffee . . . . . . 106
D. Lack of Processing Facilities for Citrus . . 106

V. Food Consumption .......... .................... 106
A. Areas of Investigation for Food Consumption .. ..... 106
1. Food Preferences, Preparation Techniques and Food
Taboos .*.......*. ....................... 106
2. Beans and Pulses ......... ............... 107
3. Seasonal Food Shortages .... ............. ...107
A. General and County Specific Farming Systems Hypotheses 109
B. Summary of Results for Farming Systems Reconnaissance
Survey for Grand Gedeh County ..... .............. .127
C. Summary of Res.ults for Farming Systems Reconnaissance
Survey for Nimba County ..... ............... .... 136
D. Summary of Results for Farming Systems Reconnaissance
Survey for Bong County ......... ............... 145
E. Timing of Cropping Activities for Grand Gedeh ........ .153 F. Timing of Cropping Activities for Nimba County .. ..... 154 G. Timing of Cropping Activities for Bong County ........ .155
H. Crop Calendar: Months of Main Farming Activity, by County,
Liberia .......... ......................... ...156
I. Rice Varieties by County ...... ................ ...157
J. Cassava Varieties by County ..... ............... ..164
K. Primary Sources Consulted ..... ............... ...166
L. List of People and Institutions Contacted .......... .168
M. List of Acronyms ...... .................... 170
N. Topics of Inquiry for Farming Systems Reconnaissance
Survey for Grand Gedeh, Nimba and Bong Counties .. ..... 171

Map 1 Location of Counties in Liberia ..... ............. 2
Map 2 Location of Villages Surveyed in Grand Gedeh County 4 Map 3 Location of Villages Surveyed in Nimba County .. ..... 6 Map 4 Location of Villages Surveyed in Bong County 7

This research report presents the findings of a farming
systems reconnaissance survey that was conducted in Liberia in Grand Gedeh, Nimba, and Bong Counties (see Map 1). The survey was conducted over a three week period in July, 1984, in 19 villages. One hundred and fourteen farm families were interviewed. The study was supported by the U.S. Agency for International Development Farming Systems Support P ro je c t, University of Florida with cooperation from the University of Kentucky; the U.S. AID Mission, Monrovia; and the Central Agricultural Research Institute (CART), Suakoko, Liberia. The primary objective of this study was to provide information on the rice-based farming
systems that are found in Grand Gedeh, Nimba, and Bong Counties to help establish research priorities at CARl. In addition to providing baseline data on cropping patterns, animal husbandry. off-farm economic activities, marketing and consumption, the
study helps identify some of the key constraints facing farmers within these counties. Although this information may need to be verified through formal surveys and on-farm testing, it serves as a starting point for orienting research to the needs of farmers.
A. Methodology:
To facilitate data collection on the various factors which make up a farming system, a multi-disciplinary team was used. The team members included: Timothy R. Frankenberger (anthropologist); John A. Lichte (agricultural economist); Arthur S. Gedeo (agricultural economist); John Kpakolo Jallah ( soil scientist) ; and Maran J Sherman (t ree crop researcher). The f ir st two researchers mentioned were Americans while the other three were Liberians from CART.
To help guide interviews, t he team constructed a d et a iIe d topical outline prior to going to the field (see Appendix N). A structured interview format was not employed to avoid collecting biased information. In order to develop a preliminary
understanding of how farmers express themselves on a variety of issues, an open ended topical list was used. The topics included in this outline were obtained from three major sources. First,
secondary data sources were reviewed to obtain a list of important variables which were identified in past studies. Second, each department head at CARI or one of their representatives was consulted to identify the various kinds of information needed that would be directly relevant to ongoing research at the station. Third, the team members drew upon their own knowledge and past research experience in devising topics. Once the team reached total agreement on the items included in the list, the survey was initiated.
As stated earlier. the study focused on three counties. Grand Gedeh County was the first county studied. This county was

Location of Counties in Liberia
I.* 2I 9.B.
..%% CAve' Sanoliole
-- %
0 Do 5 75 100 KILOMETRES
0 25 s 75 100 MILES
SCALE 1:30O0000

selected because i t tlIe information existed on the region and CARI was planning on opening a sub-station there at Saye-Dube. Prior to initiating any village surveys, meetings were held with county extension personnel for a briefing on the county and to enlist their aid in selecting villages for the survey. Factors taken into account in village selection included location, size, access to roads, and institutional complexity. In addition,
attempts were made to have a village representing each of the sub-tribes in the vicinity. The seven villages selected were geographically distributed around Zwedru and the Saye-Dube SubStation (see Map 2) Each village was located within a 40 mile radius of Zwedru to ensure that they were reasonably accessible to the Sub-Station for future on-farm experiments.
Once the villages were selected, the team visited one village each day to conduct interviews with farm families. Upon arrival in a particular village, the team would f ir st meet with the
village chief and other villagers present to explain their mission and why so many questions would be asked. General inquiries were directed to the group assembled concerning village infrastructure, land tenure, community farming, sources of credit, project intervention, marketing, and typical labor
arrangements. Aftef these initial inquiries, the team split up into three groups. Each group was accompanied by a translator 2or one of the team members who understood the local language. A
farm family with a field within a 40 minute walk of the village was then selected by each group to be interviewed. Interviews were conducted in the fields away from the village so that team members could see the fields they were inquiring about and to prevent other farmers from biasing the farmers' responses. In most cases, both the husband and wife were present for the
interview. The team felt it was necessary to get both of their inputs because women were responsible for a considerable amount of the labor performed on the rice fields. Usually each group could do two complete interviews in a day. In some cases, it was only possible to do partial interviews and a group could interview more farmers in a day.
Team members did not work in the same group every day. Thus the groups did not remain constant. This was done to give every team member an opportunity to work with and learn from the other team members. This greatly facilitated an exchange of ideas and helped establish better communication between team members.
After interviews were completed for selected villages in Grand Gedeh, the team members got together and f ormulated hypotheses about the farming system which characterized that region (see Appendix A). This procedure helped summarize the important
1Alonzo Munyeneh, Manager of the Saye-Dube Research Sub-Station,
participated in the study and joined the team for the duration
2 of the Grand Gedeh survey.
Two of these translators were local extension agents.

Map 2
Location of Villages Surveyed in Grand Gedeh County
3 7
Village Names:
1. Gaye Town 2. Jarwodee 3. Toyelabli
4. Bawaydi 5. Toozon 6. Gbolue 7'. Gleplay

attributes of the farming system and provided a basis for comparison when survey work started in the other two counties.
Upon completion of the work in Grand Gedeh, survey work was initiated in Nimba county. The study concentrated in the Saclepea area in order to establish collaborative links with the Nimba County Rural Development Project (NCRDP). This cooperation was considered important because NCRDP had been cited as a good model for farmer oriented development efforts. It was felt that the information provided by the study could be of use to NCRDP as well as CART, and that future collaboration could be beneficial to both organizations. With the help of the NCRDP staff, an initial selection of villages was made. Selection criteria were similar to those used in Grand Gedeh. This selection was somewhat restricted since many of the roads were impassable due to rain. This list of villages was later modified to include some nonproject villages because the initial selection strongly emphasized project villages. Interviews were conducted with
farmers in seven villages in total (see Map 3). NCRDP extension staff usually accompanied the team to each of the villages surveyed.
The survey procedures followed in Nimba were similar to those followed in Grand Gedeh. Following the survey, hypotheses were also formulated about the farming system which characterized Nimba County (see Appendix A).
The last county surveyed was Bong. Bong was a logical choice for this study since CART was located in this county. If on-farm experiments are to be carried out in villages located within a reasonable distance of the experiment station, then information is needed on the existing farming systems in the area. To facilitate data collection efforts, collaborative links were also established with the Bong County Agricultural Development Project (BCADP). As was the case in Nimba County, it was hoped that such links between BCADP and CAR1 could be beneficial to both
organizations in future development efforts. The BCADP staff gave the team the names of extension officers in the different regions of the county to contact. These extension officers helped the team identify a variety of villages including some which had relatively little project influence. Again village selection criteria were similar to those used in Grand Gedeh and Nimba. Interviews were conducted in five villages in Bong (see Map 4). In some cases, the BCADP extension staff accompanied the team to the villages. Similar survey procedures were followed in Bong as in the other two counties. Likewise, hypotheses characterizing the farming system found in Bong County were also formulated after the completion of the survey (see Appendix A).
Once the survey was completed, hypotheses were generated which applied to the farming systems found in all three counties. Following these formulations, a series of possible recommendations were derived which could help alleviate some of the constraints which were identified in the survey. After these tasks were completed, the results of the study were written up. Each team member was given responsibility for a certain section

Map 3 Location of Villages Surveyed in Nimba County
Village Names:
1. Boitain
2. Dohn
3. Zayglay
4. Boweh
5. Zehplay 6. Gwehrlay
7. Lampa

Map 4 Location of Villages Surveyed in Bong County
Village Names:
1. Gbarna 2. Seata 3. Santa
4. Kollieta
5. Janniepeleta

of the report.
B Content of the Report:
The information provided by this report is presented in three major sections. First, the general farming systems characteristics of the three counties taken together are presented. This section addresses such topics as access to land, spatial arrangements of farmers f ielId s, labor patterns, upland rice cropping patterns, cassava cropping patterns, sugar cane
cropping patterns, groundnut cropping patterns, swamp rice cropping patterns, animal husbandry patterns, marketing patterns,
other sources of income, access to credit. consumption patterns, community farms, and government interventions.
The second section of the report presents *the specific
findings for each county on the various topics addressed in the previous section. if farm families in all three counties follow similar patterns for a given practice. then that topic is not presented in this section.
The third section of the report presents a summary of the farming systems found in the three counties surveyed as well as the teams recommendations. The recommendation subsection is
structured in such a way that: 1) a constraint is identified; 2) the farmers' present compensating strategies to deal with the constraint are outlined; and 3) recommendations are proposed that take these strategies into consideration. The constraints
addressed in this section include general economic constraints, crop specific constraints, animal husbandry constraints, marketing constraints, and consumption related constraints.
The f inalI se c t ion of this report contains the appendices. These appendices present d e t a iled information that does not appear in the main body of the report which should be useful to researchers and administrators.

A. Access to Land-Fallow Rotation System:
In the past, usufruct land tenure arrangements were the predominant tenure patterns found throughout Liberia. Land
availability was not a constraint due to a low rural populationto-land r a t io. Presently, land access is a limiting factor in areas of high population concentration and/or where concessions and private estate ownership are prevalent. Land purchases and/or land r en talI are not uncommon in such areas. All three tenure arrangements are found in the study area.
In Grand Gedeh, usufruct arrangements predominate in all of the villages surveyed. Land is not being purchased nor is land actually being rented. Due to the low population densities characterizing the county, a considerable proportion of the land
is still in high bush. Individuals gain rights to land by being the first to clear high bush from an area. Anyone else wishing
to make a farm on this land must obtain the permission of the original farmer.
When all land in the vicinity of a village has been farmed, farmers will make farms farther away by seeking out high bush or older secondary bush to clear. If these farms are far from the v illa ge s at elli t e vi IlIa g es may be created. This pattern is possible in Grand Gedeh because of the low population density and availability of land.
In Nimba, the tenure situation is more complex. Many of the villages surveyed in this county are located in an area of high population concentration.
Land pressure seems considerable in such areas. in addition
to usufruct rights similar to Grand Gedeh, cases of land purchases and land rentals are found. Besides individual land purchases, sometimes a whole village bands together and purchases
the land they are farming. As for land rentals, two forms are found. One is the more traditional quasi-rental pattern of giving a portion of rice at harvest for the right to farm an area. The
second form is an actual cash payment of ten to twenty dollars for one year's access.
One consequence of the high population density in Nimba has been the removal of a considerable amount of the high bush in the area. Presently, most of the land accessible to farmers is
secondary bush. In addition to this, land pressure in the area has forced many farmers to shorten the fallowing period on their land. Both th11e prevalence o f secondary bush and the shorter fallow period have had a significant impact on the farming system
3The amount of rice given now for land access is much greater than the traditional token. In the past, a wash pan or bucket of rice may have sufficed. Presently, 2 bags of paddy rice is
not uncommon.

found in the area.
In Bong County, the tenure situation is similar to t h at in Nimba but for different reasons. Farmers in this county are experiencing land pressure due to the prevalence of concessions (mostly rubber) and private estate ownership. As in Nimba, usufruct tenure arrangements a re s t ill common; however, land purchases and land rentals are also found. The traditional rental
arrangement of giving a quantity of rice at harvest is the only pattern identified.
Farmers indicated that cash rental arrangements are not common
because of the fear that renters may later claim that they have purchased the land. Possibly to compensate for this tendency of not accepting cash rental payments, the quantity of rice given at harvest for access to land is considerably higher than that- in Nimba. For instance, one farmer had to agree to give two bags, of cl1e an r ic e to g a in access to the land he was farming. In addition,' it appears t h at the quantity of rice charged from a stranger is more than is charged from a relative or fellow> v illIag er .
Because of the rubber plantations, private estates and government farms located throughout Bong, high bush is becoming less available to farmers. Similar to Nimba, farmers are being forced to' farm in secondary bush with shorter fallowing intervals due, to t h is land pressure. Again, t h is has had a significant impact od the farming system found in this county.
B. Spatial Arrangements of Farmers' Fields:
Regarding the location of f iel1d s, a similar sp at ialI arrangement is f ound in all1 three counties surveyed, de sp it e differences in topography. Upland rice fields tend to be located some distance f rom the village. This is especially true where domestic animals are prevalent and/or good farmland is a
considerable distance away or not separated from the village by streams or swamps which animals are not likely to cross. If no such natural barriers exist, fences are constructed along paths or roads to keep animals out. The cocoa and coffee fields tend to
be located closer to town, often encircling the town itself. In
ar e as where traditional swamp fa r m ing is practiced ( Nimba and Bong) the preferred pattern is to place swamp farms in close proximity to upland farms to facilitate easy access. If this is not possible, they may locate the swamp farms elsewhere.
Decisions regarding the location of upland rice farms are strongly influenced by the availability of land. If high bush is accessible and within a reasonable distance f rom the v illIa ge farmers will opt to plant their upland farms there. This tendency often results in upland farms being 1 to 2 hours walking distance f ro~m the village Because of these di stances f armers may opt to spend most of their time living on the farm during the cropping season. 'Satellite villages may be created through this process.
In villages where h ig h b u sh is not accessible and land

pressure is more prevalent, l es s flexibility exists as to where upland f ar ms can b e located. The tendency i s t o p lan t upland farms on secondary bush with short [allow periods.
Topographic considerations do not seem to play a critical role i n farmers' de c is io ns regarding upland farm location. Although some farmers indicate that they prefer flat areas to make 4th e ir farms, they also say that rice does just as well on slopes.4
C. General Labor Patterns:
Two major points can be drawn about the labor patterns found in the three counties surveyed. First, most labor activities seem to be adjusted around the upland rice cycle. Second, though not absolute, a division of labor exists for the performance of agricultural operations.
Upland rice farming is the primary agricultural activity performed by the farm family. A major portion of the f ami ly 's time is spent on the upland farm in order to secure an adequate supply of rice. All other agricultural activities not tied to the upland farm are of secondary importance as far as labor all Ioc a t ion is concerned. The major labor operations associated with upland rice farming include brushing, tree felling, burning, clearing, planting. weeding, fence building, bird watching and harvesting (see Appendices E through H for labor calender).
C.l. Sources of Labor:
To perform their agricultural activities, f armers will draw from three primary sources of labor. First, farmers may rely on family labor to do the activity. In the past, having large
families was a way to compensate for the labor constraints associated with slash and burn farming. Recently, family labor
supplies are decreasing due to out-migration of c h iId r en fo r
school or wage employment. S t illI, the number of wives a male farmer has basically represents his core family labor assets. if enough family labor is not available to perform the task, farmers have to resort to other sources of labor to get the job ac c om pli s h ed.
Kuus are a second source of l ab o r. A kuu is a reciprocal communal labor arrangement which involves a group of laborers working together on a particular task, rotating from one member's f ielId to another. Food and drink are provided by the farmer having the kuu on his/her field that particular day. Kuu labor
groups for most operations are quite common in Nimba and Bong but less common in Grand Gedeh.
4in f act, some f arms in Bong County are located on 45 degree sl1op es .

A third source of labor is hired labor. In all three counties, individual laborers may be hired on a daily or contractual basis. Daily wage rates vary from one county to another. In Grand Gedeh, a common daily wage rate is $2.50, while in Nimba and Bong, laborers would be paid $1.00 to $2.00 a day. Most hired laborers come from within the village or in the near vicinity. Occasionally, outside laborers from Guinea or Ivory Coast may be hired. In addition, students home on vacation will hire their labor out to earn money for school.
Besides individually hired laborers, farmers may also purchase the rights to a kuu from one of the kuu participants. Normally, the farmer pays an agreed upon sum to the kuu member who is foregoing the kuu on his/her own field, plus the farmer provides food and drink to all the participants. Hiring kuus is a common practice in Nimba and Bong but not in Grand Gedeh.
D. Upland Rice:
The upland rice field with its many intercrops, is the core farming enterprise around which family labor and other farm enterprises are organized. The traditional method of slash and burn cultivation minimizes the risk of crop failure by effectively dealing with problems like low soil fertility, soil erosion, and weeds. In some areas, this traditional system is coming under pressure from increasing population density or the sale of tribal farm land. Fallow periods have been severely reduced and certain ecological imbalances have occurred. These
changes, and the farming systems which have developed over time to dealwith them, are characterized by the present situation in Bong and Nimba counties in contrast to that of Grand Gedeh. Many of the major differences between the farming systems are related to the continuing existence of high bush or mature forest in Grand Gedeh and the necessity of farming on secondary bush or immature forest in Nimba and Bong.
D.I. Upland Rice: High Bush Compared to Secondary Bush
Under normal circumstances, farmers prefer to farm in high bush. The longer fallow period tends to improve soil fertility, and the dense shade eliminates many of the grasses and other
-----------------------------5 In Nimba and Bong, the equivalent to one dollar a day is paid
for each member of a kuu hired to work a field, while $2.00 a
6 day is a common wage rate for individually hired laborers.
The groundhog population lias increased rapidly as its favored secondary bush environment expanded and its main predator, the
leopard, has been eliminated.

weeds. The major disadvantage is the amount of work required to fell the many large trees. A few farmers rent chainsaws for $25 a
day, but most farmers do not have access or do not feel they can afford this luxury. However, many farmers in high bush areas state a desire to use chainsaws. once high bush is cut, it will
be farmed again once or twice in more rapid succession, i.e., a second time after 4 to 8 years and perhaps a third time a fter an additional 8 to 12 years. Then, the field is again allowed to lie
fallow for an extended period. This succession of short fallows comparable to secondary bush makes data on high bush fallows quite confusing.
With secondary bush less time is required to fell trees,
although this is partially offset by the need to spend more time brushing. Secondary bush has the additional disadvantage of being
a better environment for groundhogs, termites, and a number of the rice-eating birds.
Strategies for use of high vs. secondary bush:
The availability of maTle lab-rand -the distance one has to go to find high bush are the considerations which dominate
strategies concerning which type of bush to use. Felling trees with an axe is primarily a male activity. A secondary bush site is likely to be chosen if the head of the family is elderly, sick, or involved in off-farm employment. A woman alone would likely choose secondary bush in fallow only one or two years so that brushing (using a c u tla ss) alone would be sufficient to prepare the land.
Distance is also a constraint, particularly when a family has other important enterprises at fixed locations. For some reasons,
a farmer may move his family to a distant field location during the farming season. This may be impractical, however, if the
farmer: 1) is elId e rly ; 2) is planting tree c ro p s; 3) has an important tree crop or sugar cane activity near the village; or 4) has important social responsibilities. Under such
circumstances a farmer may decide to use secondary bush to remain near the village.
Site selection:
* In areas where the bush fallow may last up to 20 years or more, farmers have little experience with any particular piece of
ground. They therefore rely on various tests and indicators to find a field site which will grow good rice. Some farmers will take a handful of soil and make a ball, checking for soil
texture. Others test to see how easily plants can be pulled from the soil. Another test is to hold the soil to one's chin to see if it is "warm". One farmer tests the soil by checking if his feet left imprints. Many farmers rely more heavily on the
presence of certain plants such as palm trees, "sidrew-s igbor,'

'pohseken," o r a type o f f e rn c al Ie d "to-tay. Where f al Io w periods are shorter, some farmers u s ed indicators, but mos t s ta ted that the f ie ld had previous ly produced good rice f or them or their fathers.
Land preparation:
The critical aspect of land preparation is obtaining a 'good burn. If the f ield burns well, there is more open space which is conducive to growing r ic e, and the la t er field operations are easier to perform. A good burn also produces more ash. This ash is the major source of fertilization, providing some additional potash, phosphate, and micro-nutrients, as well as improving the pH so that more phosphate is available to plants.
To get a good burn, farmers must remove the branches from trees and pile together the branches and other brush. If this is not done, then some of the remaining debris will have to be piled and burned again at a later date. Farmers using dibble planting
want to burn as late as possible before planting. so that weeds do not get a head start on the rice. But this means waiting until the rains have started and hoping to find a day when the brush is
dry enough to burn properly. In 1984, the rains came early and there has been more rain than normal. Many farmers have had to spend several months clearing their farm after the original burn. Others have had portions of th e ir f iel1d which could not be planted at all1 due to the quantity of unburned brush. A few farmers~try to finish brushing and felling their field before the rains start to avoid this kind of problem. They feel that weeds can be dealt with later if necessary but that a good burn and 'planting early are critical.
importance of weeding:
In secondary bush areas, weeds are a major constraint.
Competition from weeds can seriously reduce rice yields directly and al1s o d eIa y the time of pl1a n t ing when the f iel1d is hoed (scratched) prior to planting. After burning the field, farmers will often wait 2 to 8 weeks before starting to hoe and plant. This allows the weed seeds a chance to germinate so that the hoeing* will be effective. In years of greater than normal rainfall, when weeding takes extra time, planting may be seriously delayed. The scratch-planting does save on labor by combining the hoeing and planting into a single activity. The hoeing is faster and more effective than trying to weed rice that has been broadcast sown by hand. As such, the hoeing does reduce
the total labor time required for weeding during the year. Where the dibble method of planting is used, planting is done directly after burning or following the initial early rains. On secondary bush, weeding will normally begin about 6 weeks after planting. On high bush, farmers hope that weeding will not be necessary

except perhaps in the lowland or swampy portions of the field. Some farmers are prepared to weed late in the season if necessary while others state that no weeding is done.
Farm size:
The farm size in Grand Gedeh seems to be constrained by the labor necessary to f ell trees on the predominately high bush
fields. Even though farmers state that their strategy is to
increase farm size rather than construct groundhog fences, their fields appear to be a little smaller than the fields brushed in Nimba and Bong. A few farmers in Nimba and Bong are only about halfway through with clearing, weeding. and planting when visited between July 1.5 and July 29, 1984. In such cases, it is not clear whether or not they can finish in time to get a decent crop or if some portion of the field will be abandoned. This delay may be characteristic of years of greater than average rainfall or may indicate a willingness to brush an extensive area and later see how much the family can actually manage. In areas where most brushing is done by kuu labor, farmers must determine field size on the basis of the number of days a kuu should work. It is safer to have too much land cleared than too 'little. Given the problems of low fertility, groundhogs, weeds rice birds, and termites that tend to be associated with secondary bush, it seems logical that farmers might try to be more extensive if not constrained by the availability of land.
BCADP estimates that average farm size is 3 to 3.3 acres in Bong. and NCRDP estimates 2 acres in Nimba. We did not measure fields, but our visual estimates of rice fields are often larger
than this, without taking into consideration that some families may have several rice fields. Furthermore, farm size should take into account the areas that a family has devoted to other t h an r ice .
Planting methods:
Two methods are f ound in the three counties surveyed. The
dibble method is used throughout Grand Gedeh, while the scratchplanting method is used in Nimba and Bong. The dibble method employs a dibble stick to open a shallow hole in the soil, and a snail shell f~om which the seed is poured in spurts using a snap of the wrist. This action determines how many seeds are planted
7A dibble stick has a short stiff metal blade about one and a half inches wide and a straight wooden handle. The total length is usually 12 18 inches. Local names include pla or squagba

in each hole (e.g. usually from 5 to 20). The scratch-planting method involves broadcasting the seed by hand and scratching the soil with a. short handled hoe to cover the seeds and remove any weeds in a single action. Broadcasting is a critical operation learned only by experience. To be efficient, the individual who is broadcasting must maintain not only a uniform seed distribution but also a uniform seeding rate.
Dibble planting seems well adapted to the farming conditions where high bush is prevalent and weeds are not a serious constraint. Such conditions exist in Grand Gedeh. Some of the advantages of dibble planting are:
(1) The absence of soil tillage which might promote erosion
loss and soil degradation.
(2) The presence of tree trunks and residual debris to reduce
erosion. The field is not completely cleared after burning since this interferes little with planting and
there is no tillage which it might impede.
(3) Labor requirements are low and planting is rapid, allowing
early completion. Loss of potential yield due to late
planting is avoided.
(4) Early planting (of short maturing varieties) allows an
early harvest and reduces the length of the hungry season.
(5) Open spaces between the hills of rice permit the
i ntercropping of cassava after the rice is planted without disturbing it. Planting cassava a month after the rice
minimizes competition between the two crops.
Dibble planting (as practiced in Grand Gedeh) is ecologically sound and stable; it has low labor requirements and facilitates the rice-cassava intercropping. However, it would not work as well in areas where secondary bush is prevalent, and, hence, where weeds are a serious constraint, i.e., Nimba and Bong. Under such conditions, the scratch-planting method seems more appropriate because of the weed control that it provides.
Farmers using scratch-planting wait 2 to 8 weeks after burning and/or the early rains before starting. This eliminates the advantage of early planting but allows the weed seed in the soil to germinate so that hoeing will provide effective weed control. This reduces labor expenditure for later weedings. The delay
also allows families with different resources and strategies of land preparation to coordinate planting and harvesting to avoid the risk of serious bird damage. Broadcast planting provides a more uniform, and probably a higher plant population than dibble planting. In addition, the hoeing should improve germination and perhaps enhance soil structure in the short run (even though the long term effect may be to increase erosion and 'soil degradation). These factors may help offset the yield-reducing
-----------------------------in Garbo Krahn and blobla in Tchien Krahn. The snail shell is about the size of a fist and has a hole drilled in the closed

effects of the lower expected fertility of secondary bush and the late planting.
Varieties grown and time of planting:
Because traditional farmers tend to be risk-aversive, they diversify as much as possible. It is, therefore, not surprising that most of them grow more than one variety of rice. In the three counties surveyed, these tend to be both long-maturing and short- to medium-maturing varieties. (See Appendix I).. Several farmers, especially in the study areas of Bong and Nimba, are growing LAC 23, both red and white. Most farmers growing this variety consider it to be high yielding and palatable. One
farmer claimed that it is somewhat less favored by groundhogs than the local varieties he has been growing. However, the red LAC 23 is said to be very difficult to mill, since consumers insist that prepared rice must be white.
Of the many8 local varieties found, several appear to be
commonly grown. According to farmers, some of these may be equally or more productive and/or palatable than the recommended varieties. Thus, although several germ plasm collection efforts have been undertaken in the past, it would be worthwhile to repeat this effort, at least in certain areas, and/or re-evaluate the germ plasm already accumulated.
In the three counties surveyed, farmers appear to be employing various strategies concerning the order of planting of long- and short-maturing varieties. In Bong and Nimba where birds are a common problem, farmers are planting the long-maturing varieties first to ensure that both short- and long-maturing varieties ripen at about the same time. The reasoning here seems to be that planting the short-maturing varieties first means they mature before most nearby fields, making severe damage by birds likely. In contrast, in Grand Gedeh where bird problems are less common, farmers seem to be planting the short-maturing varieties first with the hope that these can be harvested early enough during the season to meet some of the hungry season food requirements. In certain areas of Bong and Nimba where birds seem less of a problem, farmers also practice the Grand Gedeh strategy of
planting the early-maturing variety first.
Where dibble planting is practiced, the planting season coincides with the beginning of the rainy season. However, since the operations of land preparation must precede planting and since these may be delayed by the unavailability of labor and/or inclement weather, the actual time that planting commences may be delayed until sometime after the rains begin. In Grand Gedeh,
Rice varieties are listed in the appendix by county, village, and approximate length of maturity. Desemah in Grand Gedeh and
Nakatua in Nimba stand out as popular varieties.

March, April, May, and June are common planting months for farmers, while April, May, June, July, and early August (for late planters) are common planting months i.n Bong and Nimba. Due to yearly weather changes the ideal period for planting may vary somewhat. There is, therefore, a need to intensify agrometeorological efforts to monitor not only weather differences across the country but also year-to-year weather fluctuations. Information gathered from such a study would be useful in guiding farmers regarding the risk or expected results of a given planting date. However, only if land preparation is finished in advance can farmers take advantage of early planting opportunities in years when the rains begin early.
Planting has historically been considered a woman's job. In Grand Gedeh this is still largely true. In contrast, in the surveyed areas of Bong and Nimba, women continue to play a major role in planting, but men are also involved, especially where kuu labor is employed. The broadcasting operation associated with planting continues to be done by women.
Intercropping strategies:
The fact that farmers practice subsistence farming suggests that they would tend to diversify for food crop production. Accordingly, other crops are cultivated in association with upland rice in a mixed cropping pattern. Thus, practically all upland rice farmers interviewed in the three counties were found intercropping one or more crops with rice. These included: cassava, corn, bitterball, hot pepper, okra, tomato, eggplant, cucumber, watermelon, pumpkin, sesame, bean, water green, plato, yam, cocoyam (eddo), sweet potato, banana, plantain, sugar cane, and others.
Of the above intercrops, all but cassava, banana, plantain, yam, eddo, sweet potato, and sugar cane can be mixed with the rice seed and planted at the same time as the rice. However, not all farmers in the study areas are mixing other seeds with rice. In a number of cases, crops such as corn, beans, bitterball, pepper, and eggplant were planted separately by farmers before, during, or after planting rice. The other crops (cassava, yam, eddo, sweet potato, banana, plantain, and sugar cane) can only be planted by stem, tuber, or sucker and are thus not mixed with the rice seed. The timing of planting these crops also varies.
Intercropping is an attempt to meet multiple objectives, to reduce risk through diversification, and to maximize total production with limited resources. Competition between crops typically reduces the yield of any given crop from what it would be in a pure stand, but total production is oftgn greater than it would be from the same acreage in pure stands. Typically, crops
9 The Land Equivalent Ratio is often used to measure the

grown together will1 ha ve different nutrient requirements and growth patterns, so there are important complementary effects as well. Where crops differ in their requirements for moisture, such as rice and cassava. intercropping can greatly reduce the risk of total crop failure due to variations in rainfall. Certain crops particularly legumes, may both compete with and benefit from a cereal intercrop. Research efforts should be initiated to learn the effects of the common intercropping patterns practiced as well as to determine how cropping systems might be improved.
Harvest and post harvest methods:
Harvesting rice is a long and labor intensive operation. It consists of manually cutting the panicles from rice plants one at a time, using a locally made penknife. This same method is
employed in all three counties. Harvest may begin as early as late July or August in Grand Gedeh where the rains start early and short-maturing varieties are planted first. In Bong and Nimba, harvest usually does not begin until October. Rice harvest continues through December in all three counties. The entire family participates in harvesting, and kuus or hired labor may be employed as well. Fields that mature earlier than other fields in
the village face the threat of serious bird damage where birds are a problem. The same threat did not seem to apply to fields harvested later. Farmers prefer to harvest quickly after the rice reaches maturity, since dry heads tend to shatter over time and the grain falls on the ground and is l o st. This is easier to accomplish when birds are not a problem and harvest can be spread over a number of months.
The mechanization of harvesting and threshing using s ic kle s and pedal threshers has not been adopted. Farmers complain that to harvest with a sickle would be harder because one must bend lower to cut the straw. Farmers use the same justification for not liking short straw rice. However, the major problem appears to be the need to change the entire storage system to accommodate threshed grain. It is likely that farmers will only be willing to
adopt mechanized harvest techniques when a practical method for storing rice grain adopted to humid and high rainfall conditions has been developed, proven by farmers, and extended on a wide s c aIe .
After harvest, the rice is taken direct ly to the kitchen or left to dry in the field on stumps, logs. or in heaps on scaffolds for a number days before transporting it to the rice kitchen. The farmers interviewed seem to be practicing whichever
method proves to be reasonable at the time of harvest, depending on the distance between the rice field and kitchen or the frequency of fires under the rice kitchen.
(increased) productivity of intercropping.

Post-harvest operations a re alIs o f o und t o b e similar among farmers from the three counties surveyed. Storage methods seem' be closely related to the harvesting technique and involve
storing (in heaps) bunches of panicles in specialty constructed kitchens made of wood, thatch, and leaves. The family cooking is done in these kitchens, and frequent fires help to ensure proper drying. This also indirectly controls insects. The rice is stored in these kitchens for periods up to a year. The amount to be consumed is periodically removed, threshed, dried, and milled, often using a mortar and pestle. The milling is usually done by females.
Pests and diseases:
The major pests reported bTy farmers in the surveyed areas of the three counties are groundhogs, r ic e b ir d s, termites, s te m borers, and other small animals. The severity of damage done by these, however, varies not only within counties but also between counties. Groundhogs and birds are reported to do the most harm.
Within counties, groundhogs and birds seem to be less problematic in high bush areas, provided such high bush areas are isolated from secondary bushes, (especially for groundhogs). This
implies that groundhogs and birds are less of a problem in Grand Gedeh than they are in Bong and Nimba because of the prevalence of secondary bush in the latter two counties.
Farmers appear to employ similar methods to control pests in all of the counties surveyed. They do practically nothing about termites, stem borers, and small animals. They control groundhogs by making fences or traps, or by using guns. Birds are controlled through the use of slings, scaffolds, and rattles on ropes. Birdwatching is done by women and children and assisted by men in some cases.
Other minor control measures or compensating strategies fo r groundhog and bird control include: taking larger farms, using nets, and employing the use of poison In Bong, the practice of not intercropping rice with cassava is reported to be another control measure for groundhogs. The practice, according to farmers, is based on their observation that groundhogs are attracted more to cassava than rice.
The farmers in all three counties strongly feel that groundhogs are the most s er io us p es t s. Estimates of groundhog damage range as high as thirty percent on some fields. Control measures or compensating strategies presently employed require
10 A few farmers in Bong county boil rice with an insecticide and
throw the poisoned rice on the field for birds to eat. Farmers say that the poisoned rice is the most effective control, but it represents a serious long term as well as short term danger,
given the use of products like aldrin.

much labor and/or are not very effective. Research projects to study the feeding habits and other characteristics of these pests should be initiated to facilitate the development of alternative, less costly, and ecologically sound control measures. The use of
leopard scent as a repellent is one possibility that should be explored.
Although numerous diseases are known to af fect rice, farmers in the three counties have no idea as to the extent of damage caused by these diseases. In fact, the only major disease
identified by farmers is false smut. Most farmers do not regard this as a very serious problem, and no attempt seemed to be made to eliminate panicles affected by false smut. Therefore, there is
a need for a pathological survey to investigate the incidence of diseases associated with upland rice and how extensive these diseases are.
Soil fertility considerations:
Soil fertility depletion is a well known phenomenon associated
with traditional upland rice cultivation which forces farmers to shift periodically. In the predominantly high bush study areas of
Grand Gedeh, it appears that two crops of rice are possible with practically no fertility improvement. This may account for why farmers can a fford to plant a second year when the previous year's burn is bad or spotty. However, in the secondary bush
areas of Bong and Nimba,even the first year's crop does not seem
to be adequately supported with respect to nutrient supply. Many rice fields in these study areas are experiencing nutrient deficiencies, especially nitrogen deficiency. it is therefore
suspected that rice yields per unit area in these counties are of ten lower than those obtained under the conditions found in Grand Gedeh. Farmers in Bong and Nimba seem to be compensating f or this by making larger farms. This pattern appears to be exacerbating the land pressure that is already evident in some areas.
At present. chemical fertilizers are not commonly used by farmers and the limited amount being distributed in project areas seems to be used for tree crops and improved swamp rice. It would
be interesting to investigate the economics of minimal chemical fertilizer applications for upland rice, at least in those areas initially suspected of being low in inherent fertility. Research on 3 5 year rotation schemes involving f ood and cash crops would also be worthwhile to conduct. Such research might investigate various intercropping strategies.

Hungry farms:
Although most farmers succeed in producing a major portion of the foods consumed, they do not always produce enough to feed the
family at a sufficient quantity for the entire year. The period between planting and the main harvest is often called the "hungry season." To meet hungry season food requirements, farmers employ various strategies. in Grand Gedeh, farmers with more than one wife oft 'en have several upland r ic e f iel1d s including a major field and one or more minor fields. These smaller fields serve
as "hungry farms." Early-maturing varieties are planted on these minor fields and are harvested some time during the hungry season. For those farmers having only one wife, one large rice farm is usually planted with both short- and long-maturing varieties. The short-maturing varieties tend to be planted first so that they mature at some point during the hungry season. Some families may have both hungry fields and short cycle varieties in
the major field(s). Other families eat cassava as a substitute for rice during the hungry season.
The strategies of planting hungry farms and/or the early
planting of short-maturing rice varieties in major rice fields are not common in Bong and Nimba. The prevalence of birds makes it impractical to have small quantities of rice mature prior to the main harvest. In Nimba, cassava and/or local rice purchases seem to be the main compensating strategies farmers use that deal
with seasonal food shortages. In Bong, where less cassava seems to be grown relative to Nimba, rice purchases appear to be the major strategy. In a limited number of cases, swamp rice is reported to be an important source of hungry season f ood in
certain areas of Bong and Nimba.
E. Cassava:
Cassava is the second most important food crop in the areas surveyed and a major food source that people rely on during ;the hungry season prior to the new rice harvest. It can be left in
the ground a year or longer once it is mature. Although somewhat demanding of phosphate and potash, it requires little nitrogen and grows well even in soils considered to be low in fertility. Cassava is also very resistant to drought. For these reasons the rice-cassava intercrop is nearly an ideal combination.
E.l. Cassava Intercropped with Rice:
Rice is demanding of nitrogen, while cassava grows w ell even after the nitrogen has been depleted. Rice is sensitive to
drought but tolerant of excess moisture, while cassava tolerates most conditions except poorly drained soil. Intercropping
cassava 'with rice greatly reduces the risk of total crop failure and helps stabilize both seasonal and annual production of both crops. A different rice-cassava intercrop pattern is dominant in

each of the three counties surveyed and a few farmers in each county plant their cassava after the rice harvest.
Cassava planted after the rice:
In Grand Gedeh, c as s ava is planted 4 to 6 weeks a fter the r ic e. Since the rice is dibble planted, it is not difficult to plant the cassava cuttings. Because of its earlier planting, rice
shades the field and dominates the cassava, limiting the effect of competition from the cassava. Cassava is usually only planted on a portion of the f ield and is often intercropped with the first rice planted. The fact that the earliest maturing rice varieties are usually planted first helps assure that the cassava
will not shade the rice. Once the rice is harvested, the cassava can grow and mature normally. Since the cassava does not shade the rice, cassava can be planted at a fairly heavy density.
Because of the limited competition from cassava, this is probably the most efficient of the three intercropping strategies
in terms of Land Equivalent Ratio. However, it is probably not
practical in areas where broadcast planting is practiced. In such
areas, planting cassava after rice would probably result in the removal or destruction of many rice plants. It is doubtful that farmers would accept this.
Planting cassava with the rice:
In Nimba, farmers plant the cassava cuttings at the same time that they are planting the rice. In this manner, the broadcast rice seed is moved to the side, but no plants are destroyed. The cassava does tend to grow faster and at some point begins to shade the rice. To reduce this competition, the cassava is planted at low density and the lower half of the branches and leaves are removed when it starts shading the rice. Farmers say that the leaves of immature cassava plants are not good to ea t but may be fed to livestock. Cassava is intercropped throughout the entire rice field except for swampy areas. This larger area of cassava reflects its importance in consumption. Intercropping bananas and plantain in the rice f ielId is largely eliminated
since farmers believe they are too competitive with cassava.
Cassava before rice:
In Bong County, the common intercropping pattern is to plant the cassava shortly after burning the field, which may be any time f rom 2 weeks to 2 months before the rice is broadcast. The cassava will have germinated and will be growing before hoeing begins. In this manner it can be identified so that it is not hoed out and does not interfere with the broadcasting of rice seed. The stripping of leaves and branches to prevent shading of

the rice is not as commonly practiced as in Nimba, so farmers often rely on the low cassava density alone to reduce
competition. In contrast to Nimba, cassava is frequently planted on only a portion of the field.
Rice-cassava double-cropping:
In all three counties, a few farmers plant their cassava as a second crop, after the rice is harvested. In some cases, particularly in Bong county, farmers state that this 'is because they believe ground hogs like cassava even better than rice. In such cases, planting cassava in this manner is a strategy to help reduce the groundhog damage to the rice field and yet take advantage of the groundhog fence, if one has been built. In many other cases, this cassava planting strategy is reaction to problems encountered earlier in the year. If they did not have cuttings available at the normal time of planting, or if the initial planting has not done well, farmers may compensate by pl-Anting cassava after the rice harvest.
E.2. Separate Cassava Fields:
Some farmers in each county also plant a cassava field entirely separate from the rice field. In some cases, this is in
addition to intercropping cassava with rice, but in other cases it represents an alternative strategy. These f ielId s are of ten made on a portion of the rice field from the previous 1 to 3 y ea r s, so land preparation is not too difficult. The time of
planting varies throughout the year but one important strategy is to plant a fast-maturing variety in March or April. If necessary, farmers could begin to eat t h is cassava 4 to 5 months later during the hungry season.
Farmers who do not build a groundhog fence for the rice field will sometimes plant a small field of cassava separately, and build a groundhog fence for it alone. As stated earlier, this separation is often an attempt to protect the rice because farmers believe that groundhogs are more attracted to cassava than rice. This belief is not universal.
E.3. Cassava Varieties:
A number of cassava varieties are found in both Grand Gedeh and Bong Counties (see Appendix J), but only a single variety is identified in Nimba. "Matadi," an introduced "sweet" cassava, is universally the choice of Nimba farmers. This variety has replaced all of the traditional local va r ie t ies in this area
where cassava is eaten practically every day. In areas where cassava is less important in the diet, farmers still use many of their local varieties. Some of these are "bitter" varieties which

contain hydrocyanic acid (HCN) and must be processed before eating to prevent cyanide poisoning. Farmers say that the "bitter" varieties yield a larger tuber and mature faster than the introduced sweet varieties like "Matadi." Some of the local varieties mentioned by farmers are: mornfo. boutoh, banweh and coco in Grand Gedeh; with fusan, behuna, gbarkpalin, kpelemana, two cents, and awakana mentioned in Bong. Say-ton-pon and
belaminah appear to be other names of "sweet" varieties in Grand Gedeh and Bong, respectively. One additional variety in Bong called gorbu is used for leaf production only.
One field trial of cassava varieties in Bong was visited. It included CARICASS I, II and III, NUCASS I and II, and ROCASS I and III. The farmer planted them without any specific instructions and later the extension agent made him dig them up and replant them by variety. About 700 of the 2100 plants died in the process, but the rest looked good at the time of observation. The CARICASS and NUCASS varieties did not show any cassava mosaic, but the control showed a severe attack. These varieties need to be tested in on-farm agronomic trials, particularly intercrop trials with rice. They are bushy and should produce a good leaf yield as well as tubers; however, the additional bushiness may cause shading problems as a rice intercrop.
E.4. Number of Cuttings:
Throughout the 3 counties surveyed, farmers plant cassava by digging a shallow trench and laying 3 or 4 cuttings parallel in the trench. If people are short of cuttings, the number might be reduced to two. The only major exception to this pattern was a woman in Bawaydi, Grand Gedeh, who planted 1 cutting at about a 60 degree angle in areas where there was a lot of grass. Except in grass, she also used 3 or 4 cuttings in a shallow trench.
E.5. Pests and Diseases:
Farmers do not seem to recognize cassava diseases. When asked, they consistently state that they have no disease problems with cassava, even though cassava mosaic is present and serious in every cassava field observed. Women do not even discriminate against leaves affected by cassava mosaic when picking them for eating. Cassava bacterial blight has been identified in some fields, but it is not nearly as prevalent as cassava mosaic.
The common pests identified by farmers as a problem with regard to cassava are: groundhogs, porcupine, and ground squirrel. On occasion, grasshoppers are also cited for attacking the leaves. As mentioned earlier, groundhogs are a serious
problem. Stake fences are often built, but even this is often not sufficient to keep the groundhogs out. Controlling groundhogs is the primary problem cited by farmers for cassava as well as for rice. Porcupines are a problem, but they are less numerous than

groundhogs. Porcupines are heavily hunted since they are prized for their meat. Ground squirrels seem to be more of a problem in Grand Gedeh than in Nimba and Bong. It appears that Grand Gedeh hunters are less willing to waste a shell on them, given the abundance of larger game.
E.6 Other Crops Intercropped with Cassava:
When cassava is planted separately from rice, it is not usually intercropped except for a few vegetable and pineapples. However, several exceptions were observed. One farmer in Grand Gedeh intercrops cassava and sweet potatoes, and states that the sweet potatoes help to reduce underbrushing. Several farmers, particularly in Bong county, are growing groundnuts with cassava. One farmer admits that it is difficult to dig the groundnuts without damaging the cassava tubers. One tree crop specialist in Bong (a Bassa man) plants all his coffee seedlings by intercropping them with cassava. He uses his preferred food crop to provide shade for the seedlings. One other man in Grand Gedeh plants cocoa seedlings with his cassava.
F. Swamp Rice Cultivation (General)
F.I. Traditional Swamp Rice
Traditional swamp rice farming, as found in the study areas of the three counties surveyed, does not adversely compete with upland farming. This is because it is either planted as part of the upland rice field, located close to the upland farm, and/or operated in a manner that creates little conflict with the upland rice farm.
In Grand Gedeh, traditional swamp rice farming is usually part of a traditional upland rice farm. Women tend not to have individual swamp rice farms. This contrasts somewhat with the practices found in Bong and Nimba. There, a number of women do have individual swamp farms which are located, whenever possible, near the upland rice farm for their convenience. Traditional
swamp rice farming may not be as prevalent in Grand Gedeh due to the availability of upland farmland in the county.
Site selection:
In those cases where the traditional swamp is associated with the upland rice, site selection is dictated by the location of the upland farm. In other cases, proximity to the upland is a major consideration. Thus, site suitability in terms of nutrient supply and water availability and control are seldom given
consideration. The association of the swamp to the upland farm also suggests that continuous cropping is seldom practiced.

Methods of cultivation:
When associated with the upland rice farm, a number of preplanting operations (brushing, felling, and sometimes burning) are done at the same time and in a similar manner as those of the upland rice farm. In some cases, even when associated with the upland farm, the brushing and burning may be delayed because of involvement with the upland field. Besides being delayed, the burning may be made impossible if the swamp is too wet. In this case, the swamp must be cleared by hand before planting. These factors, necessarily, also delay planting which is then completed after the activities in the upland field have subsided. In those cases where the swamp is located a distance away from the upland farm, a delay in the completion of pre-planting operations and, consequently, a delay in planting is a likely occurrence.
A number of examples of swamp farms located a distance away from the upland farm are found in Bong and Nimba. In one instance, the woman brushes the swamp in March (using kuu labor) while the husband fells trees in the upland After completion of the upland, the husband then helps fell trees in the swamp. In another case, the woman brushes the swamp in June with the help of a kuu consisting of eight women, but does not have to fell trees snce the swamp had been cultivated the previous three years. Most traditional swamps are planted by broadcasting the rice seeds on the wet soil surface. In a few cases, broadcasting is followed by scratching if the soil surface is not wet enough.
The few swamp farmers interviewed in the study areas reported that they broadcast on wet soil surface or in standing water.
Varieties grown:
Gissi 27 is the most common variety grown in swamps. Sogada and Gbokala are also reported to be grown in Nimba. Menonkor, an upland variety grown in Grand Gedeh, is also reported to be a good swamp variety.
Post-planting, harvest, and post-harvest methods:
Some of the post-planting operations (weeding, bird watching, fencing) and practically all of the harvest and post-harvest operations are similar to those practiced in the upland. Weeding is performed by the women using the hands; fencing (if and when required) is done by the men. Harvesting is done mostly by the women who are sometimes assisted by men.
Most farmers interviewed report that they either sell the rice and/or use it to supplement the rice obtained from the upland.
The swamp had been cultivated two years by other farmers and
the third year by the farmer interviewed.

F.2 Improved Swamp Rice:
The term "improved, as used to distinguish certain kinds of swamp rice f arming from traditional swamp rice farming, is very r ela t ive This term refers to the method of swamp r ic e farming where production on the same location is anticipated for a number of ye a rs and some degree of water control is attempted. This method involves a lot of labor put into land preparation, such as
the construction of bunds and canals and the leveling of swamp fields to ensure water control. In some cases, the construction of reservoirs to ensure year-round water availability is alIs o implemented. Continuous production also suggests that chemical inputs may have to be introduced at some point to maintain f er t iIi t y. It is this kind of swamp farming, that traditional farmers have shunned for so many years and still continue to be suspicious of today.
A number of reasons have been advanced to explain the indifferent attitude of many traditional farmers towards improved swamp f a rm ing. Jenne (1982) has c it ed some of the following reasons:
1. The traditional view that the upland rice farm is the core
of the farmer's lifestyle and must not be abandoned for any
new, unfamiliar and risky enterprises;
2. The fear on the part of farmers that diseases (especially
schistosomiasis) abound in swamps and may pose serious.
health problems with prolonged exposure of farmers to swamp
3. The impossibility of simultaneously growing the many
intercrops farmers are accustomed to planting in the
u pla nd ;
4. Farmer's lack of experience and understanding of the
importance of water control;
5. The general belief that swamp farming is the responsibility
of women; and
6. Farmer's taste preference for upland rice relative to swamp
r ice .
Another major factor i de n t if ied through this survey is the heavy labor demand for not only building but also yearly maintenance of the bunds and canals. This work generally needs to be done at -the same time that the upland rice field is being prepared, creating direct competition between the two. Since the upland field is considered the core of farm activity and production, it takes priority. For many farm families, this means t h at the swamp infrastructure is not well maintained or labor must be hired to do the work. During the years in which the original development loan is being repaid, the repayment plus the
cost of labor for maintenance may approach the sale value of the rice produced. When less labor is hired for maintenance, the work
is often not completed, and yields tend to fall correspondingly. Swamp rice farming is only economically attractive to families who have a surplus of labor to do the work, or who lack access to
land and do not have the alternative of making an upland rice

farm. Like many capital investments proposed to small farmers, developing a swamp appears to be economical only if the cost of the investment is ignored, or it is assumed that labor has no valuable alternative employment during the period in question.
Despite all of the obstacles to the adoption of improved swamp farming, a number of farmers are found practicing this new method. Most of these farmers participate in groups encouraged by projects such as NCRDP in Nimba and BCADP in Bong. Besides the project-motivated swamps, other developed swamps operated by private farmers are found in urban areas. Typical examples of
this are found around the cities of Zwedru and Saclepea in Grand Gedeh and Nimba, respectively.
Location of improved swamps:
The location of improved swamps, when project-motivated, tends to be determined by project personnel and approved by farmers. In some cases, farmers identified the location of swamps which were then approved by projects after some investigation. Especially important in this light are those swamps being operated by working groups call "Farmers' Development Associations" (FDA's) and motivated by the Nimba County Rural Development Project (NCRDP). Non-project-motivated swamps tend to be located in urban areas or areas that were once government project swamps.
Varieties grown:
Many farmers are found growing improved swamp rice varieties. These include Gissi 27, IR5, Suakoko 8, 10,and 12, and BG90-2. No local varieties are found being grown in improved swamps.
Methods of cultivation:
The pre-planting operations involved in the development of improved swamps consist of brushing, felling, burning, clearing, and destumping, construction of bunds and canals, and leveling operations. In some cases, the construction of reservoirs to ensure the availability of water year-round is also a major preplanting operation. However, not all of these operations have been implemented in the study areas. The construction of reservoirs is certainly not common. Also, it is only in new swamps being cultivated for the first time that most of the other operations are reported to be implemented.
Unlike traditional swamps that appear to be the responsibility of women, men and women are equally involved in improved swamp rice farming. This is especially true of the Nimba study area, where the FDA's consisting of both men and women are the active improved swamp farming groups. In one study area of Nimba,
working groups consisting exclusively of men as well as those

consisting exclusively o f women are f ound. The swamp being operated by the women has, however, not been laid out. An interview with the leader of the women revealed that the men have
assumed the responsibility of developing the site being operated by the women after completion of their own site. In another area, working groups consisting of both men and women have been organized, each operating in a separate swamp To meet labor requirements, farmers in these working groups report tha t they each sacrifice a day or two per week during peak working seasons to'participate in the FDA swamp farm activities. in non-project swamps the source of labor is the family and/or hired labor.
Planting is done by broadcasting or transplanting. Farmers are found practicing both techniques. At least one farmer claims that
broadcasting gives a greater yield per acre than transplanting, although h is transplanted r ic e appears to have a low plant popula tion.
Some post-planting operations implemented by farmers include weeding, birdwatching, fencing, and limited application of chemical fertilizers and insecticides. The latter operations are restricted to the study areas of Bong, where Urea, Triple Superphosphate and Diazinon are distributed to farmers. i An interesting post-planting operation that is being practiced in Grand Gedeh by one farmer is the brushing back of rice to control lodging of Gissi 27 in very rich soil environments. According to the farmer, bushing back the rice 3 months after planting results in reduced plant height without any delay in maturity, thereby prevent ing lodging.
The many operations discussed above give an indication of the
high labor demand of swamp rice farming. A number of farmers, es pe c iall1y in the Bong and Nimba study a re a s, appear to be ge t t ing involved in improved swamp rice cultivation. Some are urban dw'ell Ie rs lacking access to land for upland rice near the city. Many others are attempting to maintain their traditional f iel1d s and cultivate swamp rice in addition. Often, this can
only be accomplished at the expense of other enterprises.
Neglecting the underbrushing of cocoa and/or coffee is a common result of being over extended. This neglect is intentional to the extent that farmers admit they will not be able to underbrush their tree crops this year because of the amount of time spent on the swamp rice. in most cases, farmers believe this labor constraint will ease when the infrastructure is in place.. A
second compensating strategy is to simply limit the size of the swamp field even though this may conflict with the recommendations of project staff (and project targets). Many more farmers seem to be motivated by the FDA group approach encouraged
in Nimba by the NCRDP than by private involvement. Many farmers continue to participate, even though it means that they must
either reduce or neglect some of their family farm enterprises. it is unlikely that FDA labor requirements can continue at the present level without discouraging farmers unless substantial compensation of some form is forthcoming. it is unclear whether village improvements will continue to be sufficient compensation

to motivate them.
G. Other Major Field Crops
G.1 Sugar Cane:
Although considered a capitalI-intens ive crop, many farmers in the study areas of Bong and Nimba were found growing sugar cane as a cash crop. However, in most cases, the high initial capital investment required is not found to be an obstacle. The majority of farmers have access to a wealthy farmer's processing facilities for a fee or a share of the product. Of course,
accessibility is not only a question of availability, but it is also a question of the distance between sugar cane f ielIds and processing facilities. Thus, farmers also have to absorb the cost of transport to the processing site which can greatly reduce profitability. Most farmers in this situation hoped to obtain their own processing facilities over time.
Site selection:
Most farmers interviewed feel that alluvial soils along river
banks and deep, somewhat poorly- to well-drained soils are the best for sugar cane cultivation. Other sites with less favorable characteristics are also being used, e sp e c iallIy f or the black Chinese variety that appears to be more adaptable to a number of soil conditions.
Other factors that farmers claim they consider in site
selection are access to water for processing or proximity to the nearest processing facility. To save on labor, some farmers are using old rice fields, as long as these have characteristics favorable for sugar cane growth..
Land preparation and methods of cultivation:
After site selection, the next step in sugar cane cultivation is land preparation. This involves brushing, felling, burning, and clearing. In the event that an old rice f ield is used, only brushing and burning need be done. The planting period depends on the location. En river bottoms and very swampy areas, planting is
done about January to avoid standing water. In drier locations, planting should coincide with the beginning of the rains.
However, this places the land preparation and planting of sugar ca ne in conflict with those same operations for upland rice, the Priority crop. This suggests sugar cane operations would be d el1ay e d, which farmers agreed often happened in reality. The methods of planting ag a in seem common throughout those study
areas growing sugar cane. It consists of first preparing holes about a foot wide and an inch or two deep, using a hoe. Sugar

cane cuttings (usually taken from the top 2 to 3 feet of the cane) are then placed horizontally in these holes (one or two cuttings per hole) and covered with previously dug soil, using hoes or hands.
Af ter planting, frequent underbrushing is carried out until harvest time. Few farmers make fences during the growing season to protect the cane from groundhogs. Due to the perennial nature of sugar cane and the susceptibility of such fences to termite damage, their use is limited to the first cane crop. Thereafter, a new fence has to be made each year. This is probably one reason why most farmers ignore the fencing operation.
.Varieties grown:
All farmers interviewed growing sugar cane recognize only, two varieties: the local yellow type and the black Chinese variety. Farmers report that the local variety takes about a year to reach maturity after the first planting, produces more sugar juice (commonly c al Ie d "beer" by local farmers), but d o es not necessarily produce more alcohol (commonly called "cane juice"). However, farmers claim that the local variety is difficult to harvest, and can be harvested only once a year. The Chinese
v ar ie t y, according to farmers, is ready for harvest in 7 8 months. 'Farmers reported that it has the additional advantage of being more adaptable to a variety of soil conditions, is less difficult to harvest because of its erect growing habits and can be harvested twice a year. It also appears that it is less favored by groundhogs s in ce it produces harder stems than the local variety. Some farmers. however, do not hold this view. At least two farmers interviewed also claim that the Chinese variety yields more cane juice from each drum of beer distilled. They attribute this to their observation that a drum full of beer from the local variety has a higher percentage of water, particularly during the wet season, than the Chinese type.
Harvesting and processing techniques:
According to the sugar cane farmers interviewed, harvests from
the ratoon crop may continue for as long as ten years or more. However, because farmers make no attempt to maintain soil
fertility, yields tend to decline progressively from the first ye a r. Most farmers are harvesting only 3 or 4 times from the ratoon 'crop after which they have to find a new site. Frequent interplantings within the first few years are also reported.
The harvesting method practiced by farmers consists of cutting
each cane as close as possible to the base using a cutlass and then removing the top 2 to 3 feet for subsequent planting. The lower sugary stem is then tied into bunches and transported to the milling site where juice extraction by crushing begins as soon as is feasible.

The processing technique for alcohol, the major reason fo r which farmers grow sugar cane, consists essentially of three
s te p s: crushing the cane to remove the j u ic e, fermenting the juice, and distilling for alcohol. The first operation is done by a hand or a motor-operated mill. Many farmers are using the former hand-operated type. The fermenting process is accomplished by souring the crushed juice in drums made partially air-tight by covering with leaves, thatch,, or other appropriate material. Farmers report that they use a minimal amount of yeast to speed up the fermentation process. After fermentation, distillation follows. This involves heating the fermented juice in a
distilling pot and gradually cooling the volatile vapor using long pipes that meander from the distilling pot through a cooling apparatus (often a drum full of water) to a collecting container.
The liquefied volatile material is ethyl alcohol (CH3OH), which is a favorite intoxicating drink for many Liberians.
Farmers interviewed report that for every drum of fermented juice, two to two and one half demijohns (5 gallon container) of alcohol are obtained. For farmers renting the processing facilities of another, one-fourth to two-fifths of t h is yield must be given to the owner of the processing facilities. Some farmers pay this fee on the basis of fermented juice, for which one drum for every three is a common fee.
Sugar cane growing strategies:
In all surveyed a re a s, two primary strategies f or growing sugar cane seem to be employed. First cane can be planted in separate fields sparsely intercropped with corn, banana, plantain, and other crops. Second, sugar cane can be intercropped in the rice field with other crops as well. Another labor-saving
strategy also practiced is the use of old rice fields for sugar cane cultivation. This eliminates a number of steps associated with pre-planting operations. The strategy practiced depends on a
number of factors. No strategy seems to be exclusively practiced in any one area.
G.2. Groundnuts:
The cultivation of groundnuts is found to be restricted to the study areas of Bong and Nimba, and cultivation methods and strategies seem similar for both counties.
Site selection:
The common tradition found among farmers growing groundnuts is
to cultivate it as a second crop after rice. This seems logical since groundnuts, being a legume, are not too demanding of nitrogen, the nutrient most required and easily depleted under

rice cultivation. The choice of site also saves on labor requirements, since very i t tlIe effort needs b e put into preplant operations. This great savings on labor and the ease with which other planting and post-planting operations are done are probably the reasons why groundnut cultivation is considered a woman's job.
Land preparation and methods of cultivation:
Brushing and burning the stubble from the old rice field are the major pre-planting operations. Two methods of planting are employed: 1.) broadcasting and covering the seed by scratching the soil with a hoe; and 2.) dibbling, which involves hoeing and leveling the top 1 or 2 inches of soil, then digging holes an inch or two deep into which 1 to 2 seeds are placed and covered with soil. The latter method seems common in the areas surveyed.
Planting is usually done in March or April and the groundnuts mature during June or July. This facilitates harvesting since the
soil is moist and loose and the groundnuts can be removed from the soil with relative ease. Weeding can be done either before or after flowering, but when groundnuts are densely planted, little weeding is required.
H. Tree Crops:
Cocoa and coffee are important components of the farming
systems in the three counties surveyed. Most farmers interviewed are growing one or both because of their complementariness to the upland rice-cassava field. Once established, the labor
requirements for these crops do not fall within the periods of. peak labor demand for rice. The risk of growing these tree crops
is low since maintenance is flexible and underbrushing may be omitted for several years, once the trees are established,
without destroying the investment. The cost of establishing the trees is r elIat iv elIy low, and they are a good source of cash income once mature. The relative prices at the time a farmer plants are probably a key factor in determining which crop he selects. Access to land may be a constraint in areas of severe land pressure for newcomers to a village because of the relationship between tree planting and land rights.
Cocoa is the tree crop most of ten grown by farmers in the three counties surveyed. It has the advantage of requiring less underbrushing than coffee, once the trees are mature. In addition, for many years farmers received a much higher price for cocoa than for coffee. Recent cocoa price instability is probably encouraging diversification of tree crops.

H.l. Cocoa
Years of experience:
Some farm families interviewed have been growing cocoa for as much as forty years or longer. A number of families have 20 year old trees in need of rejuvenation. Frequently, the cocoa farmer either has considerable experience with cocoa or represents a new generation who had grown up with cocoa. On the other hand, there are a considerable number of relatively new cocoa growers in Nimba and Bong who have been encouraged to grow cocoa by the county development projects.
Sources of planting material:
As might be expected, the source of planting material depends greatly on the period when a farmer acquired his trees. The early source seems to have been seeds that were brought back from Ivory Coast. The few original growers gave seeds to neighbors and friends in a process of spontaneous adoption. In the 1950's, the Cocopau Plantation in Nimba, and perhaps several others, became heavily involved in cocoa production. Directly or indirectly
these plantations provided the seeds, and often the experience, which encouraged the spread of cocoa production.
At present, sources are more diversified. Farmers still get cocoa seeds from neighbors and friends. LCCC and county agricultural development projects not only provide seedlings but also give loans which cover the cost of seedlings, tools, and money to hire laborers for underbrushing for the first two years. These organizations also provide some extension service and may help organize marketing arrangements as well as provide guidance in selecting a site.
Site selection:
Farmers have a number of different means which they use to determine the suitability of a site for cocoa. These fall into 4 categories: 1) plants which indicate good soil; 2) traditional soil tests; 3) folk wisdom; and 4) experimentation. However, it should be recognized that the areas to which an individual farmer has land rights may not include the most appropriate environments for cocoa.
As mentioned in the rice section, there are a number of plants which farmers in different areas recognize as indicators of good fertility. These include palm trees, a fern called "to-ay", "sidrew-sighbor' and "poseken." Some farmers will look for an area where one of these is common to plant cocoa.
Traditional soil tests are also commonly used. These include looking for black soil which is considered to be rich; checking

to see how easily plants can be pulled from the soil; or squeezing a fistful of soil to test for texture.
Folk wisdom is also a common guide. The folk wisdom concer .ning where to plant cocoa suggested one of three possibilities: near a river; along a swamp where there is moisture but not standing water; and on an old town site. It is interesting that this folk wisdom recognizes the importance of alluvial soils or soils made rich by composting household refuse but it may lead farmers to plant in areas that are wetter than is desirable.
IFinally, some farmers use different forms of experimentation. One common practice is to plant cocoa on an old rice field, using the performance of the rice to indicate if the s oil1 is rich enough for cocoa. This process has the advantage of avoiding the need to clear a separate s it e. However, unless a farmer is
w i Ii n g to. wait about 2 y ea r s, sufficient shade may be a constraint. A few farmers actually plant a few seeds or seedlings on a promising site and watch their development for about a year; only if they do well does the farmer go ahead and plant additional trees there.
P r o ject personnel often pick s it es f or farmers involved in county development projects. Unfortunately, the sites selected in
the past have not always been appropriate. The introduction of soil testing of potential sites will hopefully rectify this prob lem.
Method of cultivation:
Planting: Most farmers who have been growing cocoa for a long
time began by planting seeds. It was probably in the 1950's 'when plantations became involved that people began to recognize the u t iIi ty of nurseries and transplanting seedlings. Most farmers still only had access to seeds, but a few began making their' own nursery rather than planting 2 or 3 seeds per hole. At present, seedlings are available from projects and LCCC, but some farmers are unwilling to pay their price of .15 cents per seedling and continue to procure seeds from other villagers.
Intercrops: Cocoa seedlings are sometimes intercropped. Usually, this appears to be related to the strategy of using a field which has already been cleared and sometimes for overcoming the lack of shade on such fields. Intercrops with cocoa seedlings include rice, cassava, sugar cane (when planted in a swamp) bananas, and plantain. Cola (nut) and palm trees are somefimes in tercropped with mature trees. There is a suspicion that cola trees may be an alternative host for some of the pests attacking cocoa. This should be investigated.
Shad ing: Cocoa seedlings should be planted in shaded areas. Farmers provide shade in different manners. Some will cut part of

the trees when brushing an area for planting cocoa, leaving a
general but somewhat thin canopy. Others will leave all the
large trees but girdle a portion of them so they will die slowly. Farmers who plant on a previous rice f ield sometimes have a
problem with lack of shade unless they wait several years and brush again, leaving some of the regrowth to shade the seedlings.
One technique observed on a few farms which seems to overcome this lack of shade is to plant the area to bananas and plantain a few months before planting the cocoa seedlings. In this manner,
the field also remains productive during the 4 to 7 years which cocoa requires to mature.
Once cocoa trees are mature, they develop a shade canopy which provides most or all of the shade necessary. This is especially true given the way farmers typically plant them. The extension services recommend spacings that vary from 10 x 10 feet to 12 x. 12 fe e t. Most of the older f ields were planted randomly, so
spacing varies, but the spacing is approximately 8 x 8 feet or sometimes less. One farmer observed planting cocoa seedlings has
two 10 f t. lengths of bamboo so he can use the triangulation method of spacing, but he will not use the full length of the two poles. He said this is because some of the seedlings will die and
the "World Bank" (BCADP) will not provide replacements. However, in talking to farmers it is evident that they are very concerned with underbrushing. it appears that some farmers are
intentionally planting more densely than recommended so that a heavy shade canopy will develop which reduces or even eliminates the need for underbrushing. Branches of trees in the older fields
do interlock and provide a dense shade. This leaves the lower branches shaded and competing for sunlight. It can be expected that roots will extend as far as the branches and thus will also be intertwined and competing for nutrients. For these reasons, it is expected that the farmers' higher density will reduce
potential yields. In fact, the number of pods observed on trees was very small in most cases: 5 to 10 pods per tree. .Research and extension recommendations on spacing should reconsider this trade-off between yield and underbrushing, taking into account
the risk that wide spacings have for farmers who cannot afford to underbrush.
Underbrushing: Cutting back the undergrowth is the only maintenance activity commonly practiced and the main labor activity concerning cocoa. For this reason, it is also the
primary constraint. As mentioned earlier, farmers appear to be more concerned with the problem of underbrushing than with the yield per tree they receive. Young trees are commonly underbrushed twice a year in December-January and in July-August.
When the trees are mature, this is reduced to only once a year, just prior to harvest if any underbrushing is done at all. This is commonly in July or August, but may be later depending on the harvest cycle of the trees in question.

Thinning and pruning: Except for one or two tree crop speciali-sts interviewed, farmers rarely do any thinning or pruning o f t re e c ro p s. Many farmers do not want sunlight to penetrate so the stand is not thinned, and the canopy is not pruned to allow sunlight to reach the to lower branches. Even dead, diseased, or insect-infested branches are not removed. Very old trees are left to decline naturally, though they produce very
few pods, rather than cutting them back to generate a new tree. Farmers' reluctance to thin and prune may be attributed to the following:
1) Fear that thinning and pruning will reduce shade and
thereby increase the need for underbrushing.
2) Time and labor constraints which prevent a farmer from
bothering with activities which do not have a direct effect
on production.
3) Hesitancy to remove a tree that is producing, given ,the
long delay before a new tree will provide any income.
4) Lack of information on the importance of pruning,
particularly of removing and destroying diseased or
infested branches and trees.
Harvest: Harvesting must be done periodically as pods ripen, but it does not require a lot of time, particularly given the low yields commonly observed. A number of farmers have indicated that
harvesting is done about every third day once some of the pods are ripe. This usually begins in August and continues for one to two months. A few farmers have stated that their trees do not begin to produce until November or December, perhaps due to varietal difference. Some, but not all farmers use a knife to remove the pods so that the tree will not be injured.
Post-harvest techniques: The adoption of recommended postharvest techniques varies greatly. Farmers in counties with
projects and well-developed extension services tend to be more aware of these techniques, but most farmers still have little incentive to practice them. With the exception of cocoa purchased by some project-sponsored cooperatives or FDA's, there is no grading of cocoa at the farmgate. Since there is no price differential for quality, farmers have no reason to follow recommended practices.
Once pods are harvested, they are lef t in a pile or in a basket for about 3 days before attempting to break them. Newly harvested pods are difficult to break and hurt one's fingers when
they crack. The delay softens the pods, making the task easier and less painful Following harvest, the next two post harvest techniques followed by most farmers include the following:
1) Fermentation
Some degree of fermentation is practiced by many farmers in all 3 counties. Cocoa beans are piled into a box or basket
lined with banana leaves and are covered with more banana

1 12
A weight is placed on top to keep the leaves in place and perhaps
to compress the cocoa as well. The amount of time the beans are allowed to ferment ranges from 3 to 7 days.
2) Drying
The beans are removed from the box or basket and placed in the sun to dry. Drying time varies f rom 3 to 4 days to several weeks. The time required varies according to the weather and whether farmers start the drying gradually, a few hours at a time. Some farmers place the beans on a mat on a raised platform of small wood logs or bamboo sticks to help keep the beans clean. Others simply place the mats on the ground. Most farmers periodically stir the beans as they are drying on the mat to allow for more even drying.
Farmers can tell when the beans are dry by cracking open
one of the beans.
Pests and diseases: Black Pod disease is a serious problem for cocoa throughout the three counties. The disease is evident on every cocoa field visited. Pods have been observed in various stages of the disease's development, from a few black spots to entirely black and empty husks. Immature pods show signs of attack less often than mature pods. More information is needed on
this disease. The development of Black Pod resistant varieties would be highly desirable.
Stem borers appear to be the most serious cocoa pest. Some cocoa fields have dead branches on a majority of their trees which appear to be caused by stem borers. Farmers also have pointed out holes in the trunk of some trees going all the way to
the pith, which caused the tree to die. Farmers mention that the dry season is the primary time of attack and that trees with a lot of sap are resistant to stem borers. One tree with a lot of sap had 6 or 8 holes, but the stem borer had not reached the pith. It appears that there may be two types of stem borers involved. one is a completely white beetle about the size of a large fly. Larvae with the appearance of small maggots have also been observed, but farmers say these are not the larvae of the white beetle.
The severity of stem borer attacks seems to be associated with bush overgrowth. i.e., lack of underbrushing. The fields with the
most severe stem borer attack appear to be those which have not been underbrushed for several years or have been partially abandoned.
Termites are a, problem in many cocoa fields observed. However, it appears that they usually start as a secondary parasite
feeding on dead and decayed branches and only later spread to
-------------------12 In some cases, farmers will place the beans in a sack to let
them ferment.

healthy plants. If true, this strongly supports the need for pruning and destroying dead and infested branches.
Squirrels, domestic livestock, wild animals, and even some birds are additional pests to cocoa that are cited by farmers. Planting cocoa near the village may afford some protection from wild animals, but it increases the likelihood of damage by domestic animals unless the field is fenced or protected by a natural barrier such as a river or a stream.
H.2. Coffee:
Coffee is the second most popular tree crop. It has the advantage of performing well in somewhat less fertile soils than cocoa, but it does not usually develop a dense shade canopy like cocoa and requires more labor for underbrushing.
Years of experience:
In Grand Gedeh and Nimba counties, which border on the Ivory Coast, farmers have about the same experience growing coffee as cocoa. Some farmers have grown coffee for as many as 15 to 20 years, while perhaps one-third have started growing it in the last six years. Growing coffee seems to be a more recent phenomenon in Bong County. Recently, many farmers have begun diversifying by planting coffee as well as cocoa. However, older coffee fields tend to be abandoned more readily than old cocoa fields, probably because of the greater labor requirements
associated with underbrushing coffee.
Source of planting material:
Until recently, neighbors or friends have been the most common source of seeds or seedlings. It appears that, as with cocoa, a few farmers originally obtained seeds from the Ivory Coast.
Later, Cocopau and perhaps other major plantations became an important source of seed and experience. More recently, LCCC,
LPMC, and county development projects have become an important source of seedlings, credit, and technical advice. Most of the coffee grown in the three counties is Robusta, with only one or two plantings of Liberica or Excelsa observed. Seedlings available within the village are frequently from coffee beans that fall and germinate.

Site selection:
Opinion is divided as to the proper soil for planting coffee. One opinion contends that coffee, like cocoa, should be planted in rich loam soils, but unlike cocoa, it should not be planted in swampy areas. These sites can be identified by the same plant indicators and traditional soil tests as used for cocoa. The second opinion maintains that coffee can grow anywhere that is well drained and that gravelly soils in particular are good for growing coffee. Coffee does seem to adapt to a wider range of soil types than cocoa, and this is advantageous to those farmers who lack access to lowland alluvial soils that are fairly well drained. On several occasions, farmers have commented that they have previously tried growing cocoa on their coffee site, but the cocoa died.
Methods of cultivation:
Planting: Coffee is more likely to be planted first in a nursery and transplanted in a field than is cocoa. Even when farmers get seeds from a neighbor, they may well make their own nursery. Farmers do not seem to believe that coffee seed is as viable as cocoa seed. This perception is perhaps related to the smaller seed size.
Intercrop: Coffee adapts to a wider range of intercrops than cocoa because coffee seedlings tolerate but do not require much shade. Coffee seedlings grow well with other tree crops like bananas and plantain, but they can also be grown with field crops such as rice and cassava. The potential for using coffee in alley cropping combinations should be explored.
Shading-underbrushing: Providing shade is not as important a consideration with coffee seedlings as it is with cocoa seedlings. However, farmers still consider shade an important aspect in controlling underbrush growth. Extension services
recommend a spacing of 9 x 9 feet or more with pruning when the tree gets too tall for easy picking. Farmers are more likely to plant with an average spacing of 6 x 6 feet or 7 x 7 feet, and they often plant several seeds or seedlings together. This close spacing combined with no pruning provides enough shade that the number of underbrushings can be reduced from the two or three that are recommended. Underbrushing is most commonly done only once a year prior to harvest, even though many farmers would prefer to underbrush twice. The yield effects of this close spacing and lack of pruning should be studied and compared to the labor or cost savings from fewer underbrushings. As an alternative to close spacing, cover crops and alley cropping techniques should be investigated.

Thinning and pruning: Very little thinning or pruning of
coffee has been observed except in the case of one or two treecrop specialists. As with cocoa, the reluctance to thin and prune
can be attributed to the desire to control underbrush with shade to deal with labor constraints, as well as to lack of information about the importance of pruning in controlling disease and pests.
The effects of pruning as a means of controlling stem borer and termites in particular should be investigated, and the results made available for extension activities.
Harvest: Harvest recommendations call for the selective
picking of only ripe red cherries (coffee beans). Using this procedure, farmers must return periodically over a period of about three months to complete the harvest. This assures high quality coffee but is labor intensive over an extended period (either from November through February or July through September,
depending on the variety). This could be in conflict with labor requirements to brush and fell trees on the upland rice field which is primarily done in January and February. Furthermore, except for several project-sponsored coops, buyers do not grade coffee or pay a higher price for quality. Without such incentives, there is no reason for a farmer to use a procedure requiring additional labor or to upset the work schedule on his main subsistence crop. Instead, most farmers wait until a good portion of the beans are ripe and then harvest all of them at once. This is done either by rubbing a bunch of beans with both hands or vigorously stripping the length of the branch. The presence of red ants on the trees is also a serious harvest constraint, and this rapid procedure minimizes contact with them.
Unfortunately, some farmers, in a hurry to complete the harvest so they can prepare their upland rice field, do not even wait until most of the coffee beans are ripe. Grading coffee should be
tested as a means of inducing farmers to produce better quality coffee and as a means of potentially improving the price paid to farmers.
Post harvest: After harvesting coffee, most farmers simply dry the in the sun for 3 to 14 days. One farmer had a
cement slab on which to dry his coffee but most farmers place the
beans on mats. The mats may be placed directly on the ground or on a short scaffold which helps minimize the amount of foreign matter that gets mixed in with the coffee. A few farmers mentioned that they mill their coffee either by hand or at a local mill in order to sell clean coffee. However, in some areas buyers do not seem to offer a price differential between cherry and clean coffee so many farmers have no reason to bother with mi .11ing. Information should be gathered or synthesized to establish the effect of the different milling processes on coffee quality and quantity.

Diseases and pests: Farmers did not mention any diseases
attack ing the-Tr-rcoffee and are probably not able to identify any particular coffee diseases. A few farmers did mention that a few trees died back, starting with the apex, but that they did not know why. Stem borer and termites are the major pests attacking coffee. As in cocoa, Stem borer seems to be worse in fields where
underbrushing has been neglected. Termites seem to be attracted by ,the presence of dead and decaying plant material but later attack healthy portions of the tree as well. Certain caterpillars and grasshoppers do attack the leaves but are observed to pose a
threat only to seedlings in a few isolated cases. Red ants are a serious harvest constraint but do not appear to attack the tree itself except for the damage done by making a nest. The
ecological role of these ants should be closely studied before any program to exterminate them is considered. It is possible that they help protect the tree from other insects.
H.3. Other Tree Crops:
Banana and plantain:
Banana and plantain are- important secondary crops on practically every farm visited in the three counties. Most farmers grow banana and plantain with vegetables in an area
around their house and/or rice kitchen. In addition, they are often intercropped in one or several farm fields. Most of ten, bananas and plantain are found on some portion of the upland rice field. If cassava is planted on only a portion of the rice field,
bananas and plantain are usually found on a portion not: used for cassava. The pattern in which they are found varies from dense plantings on a small portion of the field to widely spaced
plantings on a major portion of the field. Bananas and plantain are also intercropped with coffee and cocoa seedlings and have been observed on perhaps 20 percent of the f ields of immature coffee and cocoa. They also continue to be present in and around
some mature cocoa and cof fee f ields and may be used to f ill the space where a cocoa or cof fee seedling has died. A f ew f armers intercrop bananas and plantain in their sugar cane.
Bananas and plantain are picked slightly green if they are to
be marketed so that they may be transported without bruising. Because of their perishable nature, farmers have little control over when they are sold. The time of harvest depends on the time of planting. It takes about a year for the plant to flower and three to five months for the fruit to mature, depending on the variety.
No pests or diseases are specifically mentioned for banana and
plantain but browning and yellowing of the lower leaves is often observed.

The importance of citrus varies from county to county, but only a small percentage of farmers grow citrus as a cash crop. A
number of farmers have 3 or 4 trees for home consumption, but very few have 20 trees or more. Farmers complain that the flood of fruit during the short harvest period prevents prices f rom being favorable.
,Oranges are by far the most common citrus grown followed by grapefruit and tangerines. A few rough lemon and lime trees are
also grown. Most families get their seeds or seedlings locally from neighbors, relatives, or friends. No farmer interviewed has
obtained any of the improved budded seedlings being distributed byCARl. Most of the citrus observed are planted around the house and/or rice kitchen. Farmers with a number of citrus trees tend to plant their citrus on the black or alluvial soils and old town
sites popular for cocoa and coffee production. Underbrushing is done one or two times a year, depending on the availability of labor, and cash. As with other tree crops, thinning and pruning are not generally practiced.
Mistletoe is the only pest specifically mentioned by farmers other than the red ants which hamper harvest activities. Mistletoe is not widely distributed and is not recognized as being a potential yield reducing agent. Farmers did mention that trees occasionally die back for no apparent reason and that the black leaf condition is also prevalent.
Research should investigate varieties that would allow the citrus production to be distributed throughout the year. The feasibility of processing citrus and marketing the produce should also be studied.
Oil palm:
Oil palm cultivation is primarily concentrated on large-scale corporate or parastatal plantations. With rare exception, small farmers in the three counties have not traditionally been
involved in cultivating oil palm. This is changing at least, in Nimba county, where the NCRDP has recently begun a program providing oil palm seedlings to farmers and encouraging their cultivation as an alternative tree crop. most small farm families throughout the three counties use oil palm products, but exploit wild oil palms to get them. Oil palm nuts are processed into an
oil which is used for cooking throughout the country. Many farm families produce at least enough palm oil for their own consumption, and many have a surplus of nuts or oil which is sold as well. Palm wine is the most common alcoholic beverage produced in, the village and plays an important role in social occasions, greeting visitors, and even the daily nutrition of many people. Palm leaves are also sometimes used as a source of thatch for roof ing other oil palm products used in the home or sold as a source of cash include the palm kernels, palm cabbage, cooking soda, and brooms. Given the many uses of the oil palm, the

improved productivity of the dwarf species, and the fact that one project is already promoting their inclusion in the family farm, research should investigate the best manner for incorporating oil palm using valley cropping techniques. Since oil palm is not cultivated, farmers seem to pay little attention to disease and pest problems. Information on oil palm pests and diseases is probably available from LCCC and LPMC or at least from the
individual plantations, but it may need to be synthesized and made available to farmers through the extension services.
Rubber is the most important commercial crop in Liberia. When traveling along main highways, one drives through stands of rubber trees for miles and miles. However, most of the rubber is grown on large concessions or estates rather than on small farms. Those villagers growing rubber tend to be wealthier than the typical farmer. They of ten are the direct descendents of the village founder and can make labor demands on other villagers. Many of these families have been growing rubber for 20-25 years,
and they tend to be located in close proximity to one or more rubber estates. In the past, rubber seed was usually obtained from neighboring estates or by working on one of the large rubber concessions. Some family member usually worked on a rubber estate or concession and learned the techniques for growing rubber during that experience.
Small farmers prefer high bush locations for growing rubber but, except for avoiding swamps, show little concern for site selection. Some farmers are able to plant a large tract all at one time. Others progressively clear land or their rice field and intercrop the rubber in July or August when the crops are planted and more labor is available. Farmers do not know the recommended
spacing for rubber trees, but many say they measured an 'd cut sticks or paced off the distance between trees on the estate or concession where they had worked. If labor or cash are available, rubber is underbrushed twice a year in July or August and during/af ter the rice harvest. Farmers complain that underbrushing is a serious problem but do not seem to use a cover
crop like kudzu as many of the estates and concessions do. No thinning or pruning was observed.
Rubber fields are not generally tapped continuously but tapping increases as the price rises. of ten a portion of the field is tapped for about a month and then left alone while another portion is tapped. Tapping is only done for about 3
months during the rainy season. Bark is stripped from a portion of one side of a tree and tapped for about 5 years. Then the bark
is stripped from the other side, and tapping continues there for another 5 years.
Farmers usually sell their rubber as cup-lump (coagulated) to
the nearest estate willing to buy. Because of transportation and storage problems, many small farmers are not able to sell their

rubber in the more valuable latex form. They may not be knowledgeable about rubber grading and subsequent pricing, but u ntilI recently. most have had i t tl1e or no alternative as to where they may sell. Farmers are very suspicious about the p r ic es they receive and tend to feel that the buyer is taking advantage of them. A new rubber marketing agency program called the Rubber Corporation of Liberia (RCL) has been established to collect rubber from small producers. This program will hopefully help small farmers receive better prices.
The only disease or pest of mature rubber trees mentioned by farmers is a type of borer described a "the rubber worm** Termites and other animals (possibly groundhogs) do attack seedlings.
The possibility of small farmers using a cover crop like kudzu should be explored to reduce underbrushing. In addition, the potential f or incorporating rubber in cropping patterns using alley cropping techniques should be investigated.
I. Livestock:
Raising livestock does not appear to be an important part of the farming systems in the areas surveyed. Farmers claim that they own few animals and few animals were observed around the v villages Why so few animals are present is really not clear. Certainly d ise a se and the lack of veterinary services is a serious problem. But if health problems could be controlled the potential for raising livestock would appear to be quite good. Several social factors may also be important constraints. First, farmers do not have a strong husbandry tradition so that the risk of livestock dying is high. Second, it would appear that private ownership of livestock is not completely respected. on occasion, a villager may grab any animal available to meet an important social obligation. Such factors may discourage investment in livestock.
1.1. General Husbandry Patterns:
Farmers tend to leave animals alone to fend for themselves. They roam the village uncontrolled, except for the use of fences
and gates to prevent them from wandering too far down the road. Some f armers also put a gate on the path to their f ields to discourage animals from wandering in that direction. Given this lack of control, farmers may not see their animals from one day to the next, or even every week in some cases. This is probably the reason that farmers often do not know why an animal died and probably why farmers do not recognize many disease symptoms. A few farmers have been observed feeding livestock cassava or
leaves as a means to induce them to return to the house. This reduces the risk of their wandering off and allows the farmer a chance to observe them.

The freedom of livestock to roam does have an important effect
on the spatial pattern of crops. Farmers are unwilling to plant field crops near the village because they would be destroyed by the livestock. Thus, even where land is not a constraint, most f armers walk 30 minutes or longer to get to their rice f ield. Tree crops like cocoa and coffee are often planted near the village, but cocoa in particular is subject to livestock damage. Pods near the ground are often eaten, and even higher pods may be
endangered if cattle are present. Livestock problems tend to be more noticeable around large villages, because they have more livestock and are more likely to have cattle.
Animals are used primarily as a ready source of income, a source of meat for feeding kuus, and as a means of meeting social obligations. For example, in Grand Gedeh, the bride price is
still of ten paid at least partially in animals. Animals might al1s o be killed on other important social occasions such as weddings, funerals, or the visit of a son or daughter from the city.
1.2. Goats:
Other than chickens, goats are the animals most consistently found in villages. Except in Nimba, farmers usually owned more goats than sheep, pigs, or cattle. One of the diseases observed in goats is a form of blindness in which the center of the eyes turns white (perhaps contagious conjunct ivi tis) A skin disease called "zeh" affects sheep and goats where animals lose hair and develop warty growths around the mouth and on other parts of the body (perhaps mange or scabies).
1.3. Sheep
Sheep are commonly found in villages except in part of Grand Gedeh where ze h "i s endemic. Farmers say that the d ise as e affects sheep more than goats and is so bad that they do not try to raise sheep. In one village, farmers inspect all sheep brought
to the village and will not allow any sheep infected with 'zeh" to enter.
1.4. Pigs:
Pigs are found throughout the three counties but are especially numerous in Nimba. Farmers say that they are
particularly useful for feeding kuus since there is more meat on a pig than on a sheep or goat. The only pig disease described is diarrhea which kills piglets.

1.5. Cattle
Cattle are less prevalent in all the villages than are goats, sheep, and pigs Some of the villages do not appear to have any cattle. Among the cattle observed, Matura are perhaps more common
than N'Dama. A small group of Zebu have also been seen in Nimba on the outskirts of a city, but they are probably the inventory f or a local butcher. Cattle are s till1 sometimes used in Grand Gedeh as part of the bride price. Because no farmer interviewed admitted to owning cattle, data are not available on cattle diseases.
1.6. Poultry:
Chickens are. owned by practically every family interviewed. Numbers vary from a few to 40 or more. Some farmers have mentioned that they had many chickens at one time, but an
epidemic killed many of them. Symptoms mentioned are diarrhea, drooping head and salivating, and the appearance of being wet. Because of their lower value, chickens may be used for both major and minor social occasions.
Ducks are also found on a number of farms, but unlike chickens, they are usually found only on farms which had other livestock as well. Guinea fowl are also raised in a few villa ges .
J. Marketing:
The Liberian Produce Marketing Corporation (LPMC) was established in 1962 as a joint venture between a Danish Company, the East Asiatic Company (E.A.C.) and the Liberian Government. It is a monopoly buyer and sole exporter of cocoa, coffee, and palm
kernels. It is also mandated to purchase locally produced rice and to provide extension services for tree crop (cocoa and
coffee) development. Since 1975, the corporation has been solely government owned. Despite its relatively long existence, it appears that traditional* farmers in the areas surveyed are
experiencing difficulties in marketing their produce and are not receiving prices comparable to those LPMC pays for high quality produce.
J.1. Marketing Channels:
In order to carry out its marketing activities, LPMC operates
two primary buying centers: one in Gbarnga, Bong county, and the other in Voinjama, Lofa county (and a sub-center in Ganta, Nimba county). In addition to these centers, LPMC licenses private traders and cooperative societies throughout the country to act as buying agents on its behalf.

The marketing cooperatives were established in the early 1970's to provide economic and social services such as credit and input supplies to farmers and to serve as catalysts to stimulate agricultural and economic development. In this connection, they are licensed by LPMC as buying agents and are required to pay farmers prevailing LPMC prices.
In the areas surveyed, only certain villages have an active cooperative. Where they are found, cooperatives merely serve as buying agents for LPMC, engaging in the buying and selling of produce for commissions. They have not been able to play the expected role of providing adequate and satisfactory marketing services to farmers. The poor performance and subsequent dissolution of some of the cooperatives may be attributed to poor management, the lack of "ready cash" to buy produce on delivery, inability of the cooperatives to loan farmers needed cash to pay for underbrushing or harvesting of tree crops, and in some cases, to the lack of logistics.
Because most villages are not served by an active cooperative, private licensed traders and local merchants (mostly Mandingo) dominate the produce trade. Private buying agents are more active in areas where: 1) roads and access to markets are major
constraints; 2) LPMC buying centers are not present; and 3)cooperatives are irregular and inefficient. In these areas, licensed agents (Mandingo) go to the villages or hire sub-agents to go to the villages to buy produce. The buying agents provide the instruments of measurement, usually buckets or pans instead of scales. Where scales are provided, it is uncertain whether correct readings are given, because most farmers do no know how to read or use the scales. Farmers have little choice but to accept the weight and payment given by the buying agent.
J.2. Marketing of Cocoa and Coffee:
The marketing of cocoa and coffee in the area studied is constrained or enhanced by several factors. Some of the major factors are:
1) The presence or absence of LPMC buying centers;
2) The presence of active functioning cooperatives or their
non-existence in a given area;
3) Price structure and the timely dissemination of LPMC's
prevailing price information to farmers;
4) Access to and conditions of roads;
5) The lack of grading by buying agents at the farm gate.
The presence of an LPMC buying center and/or an active cooperative are important factors in establishing the price
farmers receive for their cocoa and coffee. Cooperatives pay members the prevailing LPMC prices but often pay less to non-

members. 13Farmers typically know the price paid by the center or cooperative. Even if they choose not to sell to the center or. cooperative, they are in a position to negotiate a favorable price from traders because they have a known alternative. Farmers may still sell to traders to avoid transporting produce to the local center or coop, or because cash needs force them to sell at the first opportunity, even at disadvantageous prices. LPMC's tendency to not announce prices until well after the
beginning of the harvest season exacerbates the disadvantage of farmers who need cash quickly. Traders or their agents go to the villages to engage clients at the beginning of the harvest season
before LPMC prices are announced. Often they insist that cocoa and coffee prices for the coming season will be low. Farmers in need of cash who have no current price information often sell for whatever price is offered. In addition, farmers who need to borrow money may obligate a portion of their produce substantially lower prices even before harvest season arrives. Given these circumstances, the timely dissemination of price. information might help farmers negotiate more advantageous p r ic es.
Good roads are also an essential factor in farmers' access to markets. When villages are not accessible by truck or pick-up, transportation costs are high and farmers have few marketing
opportunities. This places them at the mercy of the rare trader who does show up to buy their produce.
The fact that traders and buying agents don't grade the cocoa and coffee purchased from farmers tends to keep prices low.
Buying agents and traders are paid according to the grade when selling their cocoa and coffee to LPMC. Because of the risk that the produce they buy may not receive a good grade, they are only
willing to pay a price comparable to those paid for the lower grades by LPMC. Farmers receive the same price, irrespective of the quality of their produce, so there is no incentive to improve qufiity by following recommended harvest and post-harvest practices.
Mandingo seem to dominate the produce trade in the study areas for the following reasons:
1) They are willing to go to small villages and buy produce in
small quantities;
2) They are willing to transport the produce out of the
village (even small quantities) which might otherwise cost the farmer more than the difference between what the trader
pays and what he could get elsewhere;
13 Non-members get around this problem by having members sell the
14 produce in their name.
Some farmers in Grand Gedeh said they intentionally grow cocoa 15and coffee so that they can borrow money from merchants. 1Most farmers do ferment and dry cocoa, but the procedures
followed vary, and may not be those recommended.

3) Traders are present and willing to buy produce early in the
season which helps those farmers in need of cash;
4) Traders are one of few sources of credit available to
J.3. Marketing of Rice:
Rice i s the staple f ood in Liberia and i s grown by all1 traditional farmers. Therefore, the primary reason for growing it is for home consumption rather than for sale. However, rice sales
are not uncommon in the areas studied, especially when there is surplus production.
Rice is sometimes sold through the same marketing channels as coffee and cocoa. LPMC is also responsible f or the buying of locally produced rice. It uses the marketing cooperatives and licensed traders as buying agents. Unlike coffee and cocoa however, it appears that speculators, (Mandingo and local merchants) buy rice at low prices (7 to 12 cents per pound)
during harvest seasons when rice is plentiful. Later in the year
during the hungry season, these traders can often sell it at the government set price of 18 cents.
In most of the villages studied, the sale of rice is the
responsibility of the women. During harvest, women sell small quantities of rice at 25 cents per cup to purchase basic necessities such as salt, chicken soup, soap, meat, etc.
J.4. Marketing of Cassava:
Cassava is the second most important s t a ple. In addition to being produced for consumption, cassava is often sold to purchase
rice and other basic necessities, especially during the hungry season. Unlike the three cash crops previously discussed, it is sold in piles at 25 ce n ts per pile or in 100-pound bags in
village, town, and city markets. Usually, it is retailed by
producers/sellers in the village market or bought by market women
who sell it in major towns and cities. Less cassava is sold in Grand Gedeh relative to Nimba and Bong counties. It appears that more cassava is consumed in Nimba than in Bong. However. because of i ts proximity to Monrovia, the market f or cassava in Bong seems favorable.
J.5 Sugar Cane (Cane Juice):
The amount of fresh sugar cane marketed is small compared to the amount processed for "cane juice." Most of the cane produced

is milled. 16
Because cane juice (commonly called C.J. ) is a popular local rum, a significant portion of it is sold in the village. Sales outside the vi 1 lage usually occur when the farmer has sat is f ied village demand. The marketing of cane juice is restricted by government regulations to licensed traders, shops. and bar owners. Therefore, farmers usually sell their cane juice wholesale in local markets. Of course, farmers in the same
villages sell at retail. The wholesale price ranges from about five to seven dollars per gallon, depending on the supply at the time of sale. The principal buyers are market women, shop owners, and other businessmen in the local major towns and urban centers. A few farmers transport their cane juice to Monrovia where prices are relatively higher.
J.6. Other Crops:
Minor crops such as peanuts, vegetables, bananas, plantains, and palm oil are marketed mainly by women. Besi-des being produced for home consumption, production over and above family subsistence requirements is sold to meet other basic needs.
While these c r o p s are considered minor relative to coffee, cocoa, rice, and cassava, they serve two useful purposes:
1) They are an important source of income for some farmers. It
is usually from the sale of these crops that farming wives are able to maintain their homes. According, to discussions held with farm families during the survey, women receive all proceeds from the sale of minor crops except where the amount involved is substantial. As such, they are not expected to ask their husbands for money to purchase
household basic needs.
2) They serve as an important source of fruits and vegetables
for city dwellers.
K. Other Sources of Income:
The most common sources of income are from the sale of cash crops like cocoa, coffee, and sugar cane (cane juice) and/or from the s a I e of a portion of the f ood crops like r i c e cassava, bananas, and plantain. In addition, some villagers are participating in a number of other income generating activities.
-------------------16 The milling and processing arrangements are discussed under
"Other Major Field Crops" section of this report.

K.1. Off Farm Employment:
Several forms o f of f-f arm employment are available. A few people in many villages are involved in retail trading. This may be limited to a single commodity like kerosene or cane juice, or several commodities such as cigarettes, matches, and snuff. In a
few large villages or along a highway, some villagers have a small shop in the front portion of their houses where they sell a
variety of articles. In a few exceptional cases, villagers near important centers may be employed by government agencies. In one village, employment is generated by a nearby government oil palm plantation. In another village on a highway near an important city, villagers are employed by a range of government agencies.
Migration provides a second type of off-farm employment. Two types of migration have been identified. Many young men leave the village for several years to work in the mines, on rubber plantations, or in Monrovia. Many will find permanent employment and remain. Others will return to the village, perhaps to get married. Their objective seems to be to make enough money to construct a house and to help pay the bride price. In addition. a number of farmers are involved in seasonal or periodic migration. Frequently, these people find jobs tapping rubber or
underbrushing tree crops for a concession or large plantation. Several farmers also mention they are involved in processing and gravel from river beds in search of gold. Such migrants may or
may not participate in this activity every year, depending on their cash needs, family labor situation, etc.
K.2. Farm Laborers:
Working as a farm laborer may be occasional, seasonal, or in the case of one man interviewed, fairly constant. People willing to work f or cash seemed to be available in every village. Employment may be organized on either a day wage basis (common f or food crops) or a contract basis (common for underbrushing tree crops). If a village has a "big man" with rubber or other extensive tree crops, some of the villagers will typically spend part of their time working for him. Kuus may also work for money once the reciprocal obligations are met, or a participant who does not need the kuu's services may sell them to someone who does. Kuu wages run about $1 .00 per person plus a good meal which usually includes some palm wine or cane juice. Student kuus
are organized specifically to work for money when students are home on vacation. Another quasi- wage labor activity is the
practice of helping someone harvest rice and receiving a basket or a pan of rice for each day of labor.
One old man encountered works mostly as a laborer. He is not married and says he cannot make a farm alone. He lives with his brother and helps him when he was not working on contract for day wages.

K.3. Arts and Crafts:
A number of men in each village are involved in local crafts. Those commonly encountered include mats, fish traps, cane chairs, baskets, and fanners (a winnowing device). it is difficult to ascertain how much money is generated by these activities, but one man does admit to making $60 to $70 from mats and fish traps.
A few people also have some income from artistic activities like playing drums, sasa (a gourd enveloped in a network of strung beads) or, dancing.
K.4 Wild Game:
The sale of wild meat is an important source of income cited by many farmers, particularly in Grand Gedeh. Fresh meat is retailed locally, and dried meat may be sold locally or to traders who take it to Monrovia. Other animal products suchas the hides may also be sold. Deer quarters may sell for about. $2 to $5, depending on size. A quarter of bush hog may bring $12.
K.5. Palm Wine:
Palm wine is a common beverage in most villages. Many farmers
tap only a small amount for personal use or occasionally sell a gallon for pocket money. A few specialize in collecting palm wine for sale, especially if they are near a highway or city. Men have
been observed selling as many as 5 or 6 gallons. Palm wine sells for about $1.25 a gallon.
K.6. Fishing:
Most families fish primarily for household consumption, but if
the catch is good, they may sell some as well. An excess amount is most likely when women fish out the pools remaining in the swamps during the dry season. However, like most activities, a few people do it more than others and may regularly have some surplus for sale.
K.7. Money Sent From Relatives:
Many families receive money or goods from relatives who have migrated out of the village. The most common situation is to send
money home to parents or a brother or sister. The amounts may vary from $10 or a few articles of clothing to f financing the construction of a new house.

L. C re d it
Credit is one of the major production constraints for Liberian farmers in general. Farmers need credit to:
1) Hire additional labor to work on swamp or upland rice
fields where family labor is inadequate to carry out all
farming activities;
2) Underbrush tree crops, cocoa and coffee in particular;
3) Purchase ba s ic farm to ol1s and inputs such a s cutlasses,
axes, hoes, and chemical inputs where they are used;
4) Meet social obligations such as funerals, paying children's
school fees, and bride price.
In the areas studied, it appears that credit is a problem for small farmers. The credit services available to them are limited to village credit associations (susus), Mandingo traders, friends and relatives, and occasionally government institutions.
The village credit associations seem to be the most common sources of credit. The organization and management of these
associations varies slightly from village to village. Basically, however, they are organized and operated on the same principle, which is to provide credit for their members in Limes of need. They alIs o serve as "mini banks" which are used to m ob iIi z e
village savings. To join a susu in some villages, farmers are required to pay registration fees. In others, registration fees are not required. Members of the village credit association make weekly/monthly payments to the club. In some susus, weekly payments are standardized, while in others, members pay according to their ability. In the entire area surveyed, weekly payments into the susus range from as low as 10 cents to as much as $10. The amount a member is allowed to borrow is associated with his savings in the credit association. No one is allowed to borrow more than his savings. Non-members must go through club members to borrow.
In villages where susus do not exist, village money lenders, friends, and relatives are the credit sources. Village interest rates are relatively high; usually the repayment is 125 percent to 150 percent of the loan amount for a three-month period. If computed on a per annum basis, these rates will be 100 percent and above.
Madigotraders sometimes prvd rdtfrfarmers. Most
often, they provide credit for cocoa and coffee farmers to hire labor for underbrushing and harvesting. Though the Mandingo loans
are usually interest free, they are issued to secure clients. In a re as where access to markets is i m it e d, farmers who r ece iv e
such loans are paid lower prices for their produce.

Credit from government organizations:
The agricultural development projects ( ADPs) and some of *the public corporations such as LPMC, L P PC, and the Agricultural
Cooperative Development Bank (ACDB) do provide some limited
credit for farmers where they exist. Most of the credit extended by the ADPs and the public corporations is channeled through the farmers' cooperatives which serve as the links between farmers and these institutions. The presence and the role of each of the credit institutions in providing credit services vary from county to-county. These variations will1 be highlighted under countyspecific discussions on credit.
M. Food Consumption:
Although numerous similarities exist between the three counties regarding food consumption, some significant differences
have also been observed. Some of the topics investigated in this study include food preferences, seasonality of foods, sources of meat, die ta ry pat terns food taboos, and culturally prescribed f oo ds.
M.1 Food Preferences:
Throughout most of the areas surveyed, rice is the preferred staple. However, one exception to this is found in Nimba County among th-e Gio tribe. The Gio farmers who have been interviewed consistently indicate that they prefer cassava prepared as
gigbah" (pounded into a dumboy-like substance) more than rice. In fact, the amount of cassava they are growing and the number of
times a. day they consume it reflects this preference. (They eat it at least once a day.) This preference seems to be distinctly tribal, because the Mano living in close proximity do not grow or eat as much cassava as the Gio. The Mano farmers will often refer
to the Gio as cassava eaters. In the rest of the area surveyed, cassava -is considered the second most important staple. Almost every farmer interviewed has some intercropped with the rice or in a separate stand.
M.2. Seasonality:
In all three counties, a marked seasonal difference in access to food resources is found. Although large quantities of rice and a wide variety of food stuffs are a v a iIa bIe after harvest, supplies of rice begin to dwindle during the next year's cropping season. many farmers run out of rice during July and August in Grand Gedeh, and August and September in Nimba and Bong. This season is referred to as the "hungry season". To deal with this problem, farmers are obliged to purchase rice, reduce their food intake, substitute cassava as the main staple or some combination of these courses of action. The period of time that this hungry

time normally occurs coincides with some of the peak labor periods in the upland rice cycle. During this time, women are usually doing the planting and weeding, and men are building groundhog fences and underbrushing their tree crops. These seasonal shortages may have significant nutritional effects on farm families in these counties, adversely affecting their labor productivity and health status.
M.3. Sources of Meat:
Some interesting differences are found among the three
counties regarding their access to meat sources. In Grand Gedeh, probably because of low population densities and the prevalence of high bush, farmers have more access to wild game than the other two counties. Wild _Pe t is used both for home consumption and as a source of income.1 Such wild meat is also used to feed kuu labor groups when such labor patterns exist. Freshwater fish is also consumed when it is in season.
In Nimba, less wild meat is available because of the secondary bush and high population densities. As a consequence, farmers rely more on freshwater f ishing, market-purchased meat (especially marine f ish) and domestic animals for their sources of meat. In addition, domestic pigs are commonly used to feed kuu labor groups. Pigs are preferable to goats and sheep for this because they contain more meat per animal than the other domesticates and can feed more laborers. Thus, it appears that pig raising is an animal husbandry pattern specifically adapted to prevalent labor patterns.
In Bong, f ish is the most common meat source for most farm families. In addition to fresh water fish, dried marine fish from Monrovia is often purchased in local markets. Fish also is a common meat used to feed kuus. Some wild meat, market purchased meat, and domestic animal meat is also consumed, but less frequently than fish.
M.4. Dietary Patterns:
It is difficult to generalize about dietary patterns in the areas studied primarily because a considerable amount of
variation exists from one family to another. One common pattern found in all three counties is the tendency for husbands and wives to eat separately. However, no consistent pattern exists as to the order of eating; sometimes a women will eat 1f rst followed by the man, other time's the men will eat first. Frequently,
1Buyers come up from Monrovia to purchase wild dried meat from
18 farmers to sell in Monrovia and other urban centers.
These findings do not support the commonly held belief that

both eat at the same time. S im ila rlIy, no regular patterns are found for the other family members. Sometimes children will eat separately, other times male children will eat with the father and female children will eat with their mother.
As for the number of meals consumed a day, again, a considerable amount of variation exists acrc1w families. Anywhere from one to th-ree meals a day are consumed. It Iappears that the
evening meal is the main meal for most families. It also appears that fewer meals are consumed during the hungry season than during other times of the year. This pattern could have
significant nutritional consequences, since the energy expended on labor is quite high.
M.5. Food taboos:
Although food taboos are not always strictly adhered to by everyone (especially young p e o pIe ), they appear to be quite common in all three counties. Most food taboos concern
consumption of certain kinds of wild animals which the farmer is not supposed to eat. For example, some types of deer, monkey, or freshwater f ish might be forbidden f or a farmer to eat. The traditions that of ten surround these f oo d taboos concern some r elIationship or experience that a family ancestor had with the animal in question. Food taboos can be tribal and/or specific to one. family.
In addition to wild animal food taboos, cases are found where some domestic animals and certain food crops are also considered taboo. For instance, one farmer will not eat sheep, while another
farmer will not eat bitterball (a common vegetable). A third
farmer could not eat cassava prepared in certain ways.
These f indings on f o od taboos have important implications. They indicate that many farmers are excluding certain foods from their diet. If such foods are common sources of protein for the rest of the community (e.g., f ish) this can have detrimental effects on the farmers' nutritional status as well as on the well-being of his/her family.
m.6. Other Culturally Prescribed Foods:
In the area surveyed, certain foods appear to be prescribed
men always eat first and receive the prime share of the high 19 quality food served during the meal.
Although some families indicate that only one or two meals are
consumed a day, a considerable amount of snacking goes on. Not
only will left-overs from the night before be eaten the next* morning, but a lot of raw and roasted cassava is eaten
throughout the day.

f or c e r t a i n occasions and f o r pregnant and lactating women. However, not all farmers follow these prescriptions. In Grand Gedeh, funerals appear to have certain food restrictions associated with them. For instance, if a family member dies, other relatives and friends will only bring cassava to. eat. The belief is that rice will make the bereaved individual happy, and
this is culturally inappropriate. A certain mourning period is often expected. The amount of time the bereaved individual must r e f r a i n from eating rice varies from one village to another. Typically, the mourning period is longer for the death of a male member of the family than a female member.
Domestic animals such as goats, sheep, or chickens are often consumed during weddings. This pattern holds for all three counties.
There are also certain cultural prescriptions that prohibit women from eating certain foods. They are often restricted from eating animals which have behavioral characteristics which are not desirable in a child. For instance, they don't eat turtles and ant eaters because these animals tend to be shy. Another food
which they are not supposed to eat is the hippo. This is because it is believed that hippo meat will prevent the mother f rom producing breast milk. Other common restricted foods are groundhogs and water deer.
Although not common throughout the area, sometimes there are also restrictions on what lactating women can eat. For example, in Grand Gedeh, lactating women are not supposed to eat plato and
okra for the first three months after childbirth because it is believed that such foods will hurt the mother's stomach. Food restrictions for lactating women in Bong and Nimba are less
evident .
As for weaning f o o d s most families appear to feed infants from the family pot when they stop breast feeding. The timing of this varies, but usually this is done when the baby starts crawling and has grown some teeth.
M.7 Nutritional Implications of Current Dietary Patterns:
Given the array of foods consumed by most households in the study area, it appears that farm family diets are quite diversified (see section on upland rice). Along with such staples as rice and cassava, farmers are consuming a wide range of vegetables, tubers, legumes, and f r u i t s in addition, they are supplementing the crops they grow with wild foods such as leafy greens, fruits, roots, etc.
Despite this dietary diversity, a number of consumption patterns have been identified which could have some adverse ef fects on the nutritional status of households. To summarize, these include:
1) The seasonal shortages of food which occur prior to harvest
during peak labor periods in the upland rice cycle;
2) The current practice followed by some households of selling

wild meat and domestic animals to urban buyers rather than consuming it themselves. This practice may lower f ami ly
protein intake levels considerably;
3) Food taboos which restrict consumption o f regularly
available sources of protein such as fish and other
animals ;
4) The current weaning practices followed by farm households
which may be detrimental to children's nutritional status.
The shift from breast feeding to family pot may result in
insufficient absorption of nutrients by young children.
Although these findings are preliminary, they point to the need for more thorough investigations into consumption patterns and their implications for the nutritional status of households. Limited consumption surveys could be incorporated into investigations of farming practices to determine the linkages
between production activities and consumption. Such surveys may
help identify interventions which would have a positive nutritional impact. For instance. interventions could be oriented
towards 'alleviating the problems associated with seasonal f ood shortages (e.g. improvements in storage, in the timing of production etc.) as well as towards increasing production.
N. Community (Communal) Farms:
The community farm concept was introduced by the PRC Government in October o f 1981 as a means of increasing f o od
( r ic e) production and generating fu n ds to f inance "self help" development projects in the political sub-divisions. The program is administered by the Ministry of Internal Affairs.
Community farms are found in most of the villages visited during this study. In Grand Gedeh, there appears to be a shif t from clan organized to individual village based community farms. Community farms in Bong and Nimba appear to be individual village organized.
0. Government Interventions:
The Ministry of Agriculture is responsible for providing extension and other support services to farmers throughout Liberia. Because of institutional and other constraints, the Ministry adopted a strategy of establishing area-specific Agricultural Development Projects (ADPs) and other parastatal
agencies in the late 1970s as its extension arms. In areas where ADPs and the parastatal bodies are not present, the Ministry is directly responsible to provide extension services.
0.1. Project Activities:
There are two ADPs in the area studied; the Nimba County Rural

Development Project (NCRDP) and the Bong County Agricultural
Development Project (BCADP). Unlike the BCADP. the NCRDP has adopted a low-key and less costly strategy which encourages 'self-help" rather than direct financial aid to farmers. It encourages the development of Farmers' Development Associations (FDAs) as village working groups and provides a range of services through them. The FDAs are multi-purpose, village-based organizations which are intended to encourage cooperation among farmers and the use of communal labor in various farming and nonfarming activities. They are involved in the making of group swamp rice farms, tree crop development, fish pond construction and management, and the organization of credit associations
(banks). Although it appears that NCRDP does not provide direct credit to individual farmers, it assists in arranging credit for improved seeds and seedlings (cocoa, coffee, and rice), oil palm planting materials, fingerlings, rice mills, and rototiller services. It does not encourage the use of chemical inputs.
The Bong County Agricultural Development Project operates somewhat differently from NCRDP. It provides direct credit to farmers and encourages the use of chemical inputs. Like NCRDP, BCADP provides improved rice seeds (swamp and upland) and improved cocoa and coffee seedlings. Both projects provide more effective extension services than are available in Grand Gedeh.
There is no ADP in Grand Gedeh. The County Agricultural Office, under the supervision of the Regional Agricultural officer, is responsible to provide extension services to farmers in the area. Without an ADP, there are very few resources
available for extension-related activities.
0.2 Constraints Imposed on Farmers by Project Activities:
The ADPs have been very useful in providing extension, credit, and other advisory services to farmers in the study areas. However, it seems that projects and the extension personnel are introducing too many interventions to individual farmers at one time without considering the ability of the farmers to manage properly his farming operations and the demands on his time.
Most farmers in the area surveyed (particularly in Nimba and Bong Counties) are very diversified. They are involved in many different farming activities. Besides the usual upland rice
fields, most farmers have a cocoa and/or coffee field, swamp rice field, and cassava or a sugar cane field. Additionally, they are required to donate 2 to 3 days per week to FDA and/or community farm activities. As a result, farmers who cannot afford to hire labor neglect some of their farm enterprises at the expense of others. For instance, when a farmer with tree crops and an upland rice field is faced with labor constraints, he may choose to spend much of his time on the upland field at the expense of underbrushing the tree crops. Even where cash credit is provided to hire labor, it seems that a number of important farm operations are not completed on farms. Therefore, projects should

exercise caution regarding the number of project activities in which farmers are encouraged to participate in a given year.

A. Access to Land:
Usufruct arrangements predominate in all the villages
surveyed. Land is not being purchased nor is land actually being rented. A considerable proportion of the land is still in high bush due to low population densities. Rights to access of land are gained by being the first individual to clear high bush from an area. Permission from the individual or his heir must be obtained by other individuals wishing to make a farm on this land.
In quest of high bush, farms are often located considerable distances away from the village. Due to such distances, satellite villages are sometimes created.
B. Upland Rice
Upland rice farming in Grand Gedeh is different from that in Bong and Nimba primarily because of the availability of high bush
and use of the dibble method of planting. Fields are often made on land in fallow up to 20 years or more. This long fallow
strongly affects rice cultivation by reducing the presence of weeds, groundhogs, and birds. Felling trees requires more labor and is much more of a constraint than where trees are immature.
Site s ele c t ion is usually based on the presence of plants which are known to indicate good soil or on traditional soil tests. Such indicators are important criteria for selecting sites
for farms in high bush areas. The landscape seems flatter than
in Nimba or Bong. Soil degradation and erosion losses appear
low, perhaps due to the flatter landscape, the minimum tillage dibble planting technique, and the effect of long fallow on soil structure.
Farming operations take place somewhat earlier in Grand Gedeh than in Bong and Nimba because the rains begin 4 to 6 weeks earlier. Brushing typically begins between December and March, but may begin as early as October so the farmer can concentrate on felling trees during the principal land preparation period of January to March. Felling trees is an important constraint in high bush where large trees must be removed. Chainsaws are available for rent in some villages at $25 a day, but few farmers are using them. Farms in Grand Gedeh appear to be -somewhat
smaller than in Nimba or Bong, although land appears to be
relatively more plentiful. The additional constraint of felling trees probably restricts the amount of land prepared. In
addition, high bush fields should be somewhat more fertile and less affected by weeds and pests so that higher yields might be expected. If in fact high bush fields do produce higher yields, then less area is required to feed a family of a given size.
Burning takes place from March through May. A farmer may burn
as soon as he finishes felling trees and cutting and piling the branches. If he finishes early, he may burn immediately to assure

a good burn or wait until the first rains begin to minimize weed growth prior to planting. If a reasonably good burn is achieved, farmers do not attempt to clear the field of remaining trunksland debris. This is practical because the dibble method of planting used is a minimum tillage technique and is not preceded or accompanied by any soil cultivation. Planting may begin as early as March and can be completed within a few days if all the planting is done at one time. Some women wait up to a week between plantings to help spread out the harvest. Other women rely on varieties of different maturities to space out *the harvest.
Most families grow several varieties of different maturities ranging from 3 to 5 months. Short maturing varieties are planted first on a hungry farm or a portion of the major farm to help meet hungry season food requirements. Hungry farms tend to be found only in families with more than one wife. Varieties cited by farmers include Desemah, Menonkor, Queekor, Doodoe, Koryea, and Vlayonedu. Desemah is found in most of the villages
surveyed and farmers claim that it is very productive. Improved upland rice varieties are not widely distributed in the area. Rice is typically intercropped with cassava, corn, yam, eddo, sweet potato, banana, plantain, and numerous other vegetables including bitterball, okra, squash, pumpkin, pepper, eggplant, and plato.
Weeding begins as early as March. Secondary bush may require a second weeding in June or July. Farmers hope that weeding will not be required on high bush fields but they may weed in July if it becomes necessary. Bird watching is not practiced in a the majority of the villages surveyed in Grand Gedeh, and groundhog fences 4re usually not built to protect the rice. Farmers say that it is easier to make a larger field so some rice can be shared with the groundhogs, than to fence it.
C. Cas sava
Cassava is normally intercropped with rice. In Grand Gedeh, cassava is planted 4 to 6 weeks after the rice on only a portion of the rice field. The spaces left by the random nature of
dibble planting make this delayed planting practical. The delay, coupled with intercropping cassava in a portion of the rice field where a fast maturing rice variety is planted, allows the rice to mature before the cassava grows tall enough to shade the rice. In this manner, the competitive effect of cassava on rice is
minimized, allowing the cassava to be planted at a higher density than in other systems observed. Bananas and plantain are not planted .on the same portion of the field as cassava, since these
---------------------------20 Varieties are listed in the appendix by county, village, and
approximate length of maturity.

crops will also be competitive as they mature. To plant the
cassava, shallow trenches are made between rice pockets and 3 or
4 cuttings are laid parallel in each.
Cassava is also planted in the rice field after rice harvest or on a separate field. Planting in a separate field often takes place in January or February with the intent of producing cassava for the hungry season. Such fields are seldom intercropped except with a few vegetables. Farmers feel that the rice fields are too
large to fence, but will often make a groundhog fence for the smaller cassava field. Other pests include porcupine and ground squirrel.
Most farmers plant traditional local varieties of cassava such as Mornfo, Boutoh, Coco, and Banweh. In addition, Say-ton-pon, an introduced variety, is grown. These varieties are not resistant to cassava mosaic, since it is observed in every field.
D. Swamp Rice:
Family members tend not to have individual swamp rice farms in Grand Gedeh. Swampy areas are often farmed as part of the traditional upland rice field but are not treated as being distinct from the rest of the field. Improved swamp farms are not common and appear to be concentrated around Zwedru and other large towns. Improved swamp farms tend to be privately operated by families or individuals but often are in areas where some defunct development project originally built the bands and
canals. The general lack of swamp farming and its pattern of concentration appear to be related to the availability of land for upland rice farming. The prevalence of high bush in Grand Gedeh reinforces this thesis concerning land availability.
E. Other Field Crops:
Sugar cane and groundnuts are not commonly grown in the village surveyed in Grand Gedeh.
F. Tree Crops:
Cocoa and coffee are both important cash crops in Grand Gedeh.
Cocoa is a little more common, and farmers probably have been growing it a little longer than coffee. A few farmers began
growing both crops over 20 years ago after obtaining seed from the Ivory Coast. Originally, seeds were planted directly but
more recently farmers usually acquire seedlings or make nurseries of their own. Site selection is based on traditional soil tests or plants which indicate good fertility, but is strongly influenced by folk wisdom concerning the best type of site. Folk
wisdom concerning cocoa suggests sites with good soil and where roots will have access to moisture. Folk wisdom on coffee is

mixed, indicating: 1) that it requires sites similar to cocoa with good soil and also good drainage; and 2) that coffee can grow anywhere, even on gravelly soils. Cocoa is most often planted in a forested area which provides shade. Coffee requires. less shade and may be planted in either a forested area or on the upland rice field along with various other intercrops. The
spacings employed for both tree crops are always quite dense. This assures that a shade canopy will develop when the trees mature, which will minimize the growth of underbrush and the need for underbrushing. Cocoa and coffee seedlings are underbrushed twice a year in December-February and June-August. Mature cocoa is underbrushed only once if at all, prior to harvest. Most farmers try to underbrush coffee twice since they recognize that it produces less shade than cocoa, but labor and financial constraints often limit them to only one underbrushing prior to harvest.
Many farmers begin harvesting both cocoa and coffee as early as July and continue periodically through January. Cocoa tends to be fermented 3 to 7 days and then dried 3 days to 2 weeks in direct sunlight. Coffee is dried about a week and sold as cherry coffee without any milling.
As in the other counties, blackpod is a serious cocoa disease problem and stem borer is a serious problem for both crops. A variety of wild animals plague cocoa producers, given the
prevalence of wildlife in the area.
Bananas and plantain are important secondary crops in Grand Gedeh. They are intercropped with a variety of field and tree crops as well as planted around the house and/or rice kitchen. Farmers try to avoid intercropping bananas and plantain with cassava. Citrus production seldom consists of more than 3 or 4 trees grown for family consumption. Oil palm is an important crop for commercial concessions in the area but is seldom cultivated by small farmers. Villagers harvest wild oil palm and produce a number of oil palm products for both consumption and sale. Rubber is not an important crop in the survey area of Grand Gedeh.
G. Livestock:
Few animals are observed in villages in Grand Gedeh. Goats are the most prevalent type of livestock, except for poultry. More domestic animals are observed in Grand Gedeh than in the other two counties, but the numbers are still very small. Important health problems observed are "Zeh" (perhaps mange or scabies), particularly in sheep, and blindness (maybe infectious conjunctivitis), particularly in goats. "Zeh" is so serious in some area's that farmers have stopped trying to raise sheep. Goats seem to the most commonly used livestock for important social obligations and feeding kuus.

H. Marketing in Grand Gedeh County:
Markets and marketing channels are less developed in Grand Gedeh than in the other two counties. This can be attributed to the following factors:
1) The area is less densely populated and villages seem more
widely scattered;
2) The road conditions are bad and some are inaccessible,
particularly during the rainy season.
LPMC does not have buying stations in the county and the few marketing cooperatives in the study area are non-functional. Farmers have few alternatives with regard to marketing channels. In the absence of both LPMC and efficient marketing cooperatives, private traders, (mostly Mandingo) are the major buyers of cocoa, coffee, and rice in the study area.
Good roads are critical for marketing and market development, and farmers need alternative marketing channels to enable them to negotiate a good price.
I. Other Sources of Income:
Grand Gedeh's isolation in terms of marketing opportunities and distance from population centers limit other sources of
income for farmers more than in Nimba and Bong. Local off-farm employment is less likely and the sale of palm wine is less
important. The sale of wild meat and animal products stand out as an important source of non-farm income. Farm labor is l e ss common, but more expensive than the other two counties ($2.50 per day). Fishing, crafts, and money sent from relatives also provide income for some families.
3. Credit in Grand Gedeh County:
Credit seems to be a critical production constraint for farmers in Grand Gedeh. It appears more constraining than in
Nimba or Bong counties because:
1). Active farmers cooperatives do not exist;
2). Grand Gedeh has no agricultural development projects.
The Agricultural Cooperative Development Bank (ACDB) was established in 1978 to provide short, medium, and long-term credit to both individuals and farmers' organizations and to mobilize rural savings. Because of internal operational constraints and the need to make its credit program more manageable, the ACDB channels its credit to farmers through
farmers' cooperatives at relatively low interest rates of about 10-12 percent. Although there is an ACDB branch in Zwedru, small
traditional farmers in Grand Gedeh seem not to benefit from the bank because the cooperatives in the area are not functional. Also, it is difficult f or individual small farmers to obtain
loans directly from the bank because of collateral requirements

and other credit conditions which they must meet.
The fact that there is no agricultural development project in the area compounds the credit problems. These projects, either directly or through cooperatives, provide credit and other services for farmers. LPMC and its subsidiary corporation, LPPC, are present in Grand Gedeh. Their activities are, however, limited to oil palm development, an activity in which very few farmers are engaged.
Therefore, farmers in the area rely on small village credit clubs and associations, friends and relatives, and Mandingo traders as their major credit sources. It appears that credit clubs themselves are limited in their ability to meet the credit needs of farmers since their only source of loanable cash is from members' contributions. As mentioned in the general discussion, the interest rates are high. The repayment required for a three month loan period ranges from 125 percent to 150 percent of the loan amount. The credit provided by Mandingo traders is given primarily for underbrushing cocoa and coffee. It may not be available at the time when farmers need credit to hire labor for other farming operations, even though these operations may be more critical than underbrushing tree crops.
K. Food Consumption:
Because of low population densities and the prevalence of high bush, farmers in Grand Gedeh have more access to wild game than farmers in the other two counties. Wild meat is used both for home consumption and as a source of income. Such wild meat is also used to feed kuu labor groups. The tendency for farmers to sell wild meat rather than consuming it themselves may lower family protein intake levels considerably. Fresh water fish is also consumed when it is in season.
L. Government Interventions in Grand Gedeh County:
Unlike Bong and Nimba Counties, the Ministry of Agriculture, through the local County Agricultural Office, is responsible for providing extension services to farmers in Grand Gedeh. As
previously mentioned, there are no agricultural development projects. Although LPPC and LPMC operate in the area, they are mainly involved with the development of oil palm estates. Since few farmers grow oil palm in the study area, the corporations only serve as sources of off-farm employment for local farmers. The Liberian Cocoa and Coffee Corporation (LCCC) does provide some extension services for some farmers.
To be effective, extension organizations and/or personnel must have something to deliver and the means of extending it to the farmers. It appears that the regional office and the extension program are constrained by two major factors:
1) The lack of technical packages to deliver;

The lack o f logistical support (transportation i n
The lack of technical packages and the means to deliver them make it difficult for the extension program in the area to be effective. It is not logical to teach farmers about the application of chemical fertilizers or the advantages of improved rice varieties without having these inputs to demonstrate to
farmers their superior qualities. Even where technical packages are available, there must be a means of getting them to the farmers, i.e, extension personnel must also be mobile.

A. Access to Land:
The tenure situation is more complex in Nimba than in Grand Gedeh because of the higher population density characterizing t h is area. Land pressure seems considerabl 'e. In addition to usufruct rights, cases of land purchases and land rentals are f ound Land purchases inc lude individual purchases as well as vilIlage purchases Two forms of rental arrangements are found. One is the traditional quasi-rental pattern of giving a portion of rice at harvest for the right to farm an area. This amount could be as much as two bags of paddy r ic e. The second form consists of cash payments of ten to twenty dollars for one year's re nt .
Very little high bush is available. Thus, most of the land accessible to farmers is secondary bush. In addition, land pressure continues to shorten fallow periods. Even secondary bush
may only be available at considerable distance from the village. Both the prevalence of secondary bush and the shorter fallow
periods are having a significant impact on the farming system found in the area.
B. Upland Rice:
Uplan 1d rice farming in Nimba is characterized by the cultivation of secondary bush and by the integration of soil tillage and planting into one activity. scratch-planting. Problems associated with farming secondary bush, i.e. groundhogs, weeds, and rice birds, require the addition of several operations not common in Grand Gedeh.
Site selection is more restricted by land availability than in Grand Gedeh. Farmers or their fathers typically have had
experience cultivating the various tracts of land available to them. They use their knowledge of the tracts and the length of time each has been in fallow to determine which site should be used in a given year. Soil fertility may be limited by the short fallow period and erosion may also be a problem due to the
cultivation of fairly st e ep slIop e s. These problems are. exacerbated by the use of the scratch-planting technique which loosens soil over the entire field.
In the absence of large trees, brushing is the major f ield preparation activity and is usually done between January and
April. Felling trees is a secondary activity and is often done at the .same time as brushing. Kuus of both men and women are
often used, in which women do much of the brushing and men fell trees and help brush. The field may be burned as early as March, but late May is more common. Clearing the field of debris to facilitate hoeing and planting may be done as a separate activity
from April through June. However, in 1984, the rains came early and were heavier than normal. Many farmers are still in the process of clearing, hoeing, and planting in late July.

Scratch-planting is usually delayed until 4 to 6 weeks after burning, i.e., beginning in April or May through July. This is done to make sure that weed seed in the soil has germinated before hoeing. Hoeing the entire field is much more time consuming and requires more labor than the dibble method, but it reduces the labor required for weeding later in the season. Broadcasting provides more uniform planting and perhaps a higher plant population than the dibble method. This may help compensate for the lower fertility expected after a short fallow period. Broadcast planting also eliminates the option of planting intercrops after the rice, since some of the rice would be destroyed. Intercrops include cassava, corn, yam, eddo, sweet potato, and a variety of other vegetables. Bananas and plantain are less common than in the other two counties because cassava is intercropped over the entire rice field.
As in Grand Gedeh, most families plant several rice varieties of different maturity ranging from 3 to 5 months. However, in Nimba this is done to spread out the planting season rather than the harvest. Because of the serious bird problem, farmers do not want any portion of their rice to mature ahead of the majority of the rice acreage. Therefore, they plant their longer maturing varieties first and plant shorter maturing varieties later in the season. This need to delay the time at which rice begins to mature reinforces the advantage of waiting a number of weeks after burning before beginning to hoe and plant. The bird
problem also eliminates the possibility of using an early maturing variety to help meet hungry season food requirements. Instead, cassava is relied upon much more heavily as a hungry season food among the Mano, and as the main staple all year among the Gio. A number of farmers are growing improved upland rice varieties, including both red and white Lac 23. Local varieties grown include: Nakatua,2fankanoeh, Leebay, Meleken, Plantee, Conko, Gwesiah, and Lesah.
Birdwatching must be done for a week after planting and for about a month while the rice is heading. Farmers often build fences to protect rice fields from groundhogs. This is done from June through August, and requires a month or more depending on the amount of labor available. A second weeding is often done at this time as well.
C. Cassava:
Cassava is more important in Nimba than in Grand Gedeh or Bong. It plays a more important role in the diet and is planted over larger areas than in the other two counties
Because the rice is broadcast planted, it is impractical to
21 Rice varieties are listed in the appendix by county, village, and approximate length of maturity.

plant cassava after the rice. And since cassava is planted over the entire field, except for swampy areas, it is not only planted with f a st maturing varieties of rice. Therefore, competition
between the cassava and rice is unavoidable and cassava may shade the rice before the rice reaches maturity. To minimize this competition, farmers plant cassava in low densities and strip the lower branches and leaves to limit the area shaded. Since bananas and plantain are not considered compatible with cassava, they are eliminated as a rice intercrop, except perhaps in swampy areas. As in the other counties, the method followed for planting cassava consists of laying 3 or 4 cuttings parallel in shallow
Cassava is also planted as a second crop after rice or in separate fields. Separate cassava f ielId s may be an adaptive strategy in cases where the farmer had trouble with rice-cassava intercropping. The fear that cassava will attract groundhogs to the rice is also cited as a reason to plant cassava as a second crop or separate from the r ic e. Cassava planted in either of these two ways is not commonly intercropped, except with a few vegetables. Planting of such fields tends to be in November
through December and/or March through April to have cassava for the hungry season. Separate cassava fields may also be planted during the rainy season.
Most farmers in Nimba have adopted an introduced cassava
variety called Matadi which does not contain hydrocyanic acid. The popularity of Matadi has practically stopped the cultivation of traditional varieties in the villages surveyed. Matadi is
often attacked by cassava mosaic. Porcupine and ground squirrel
are other cassava pests found in addition to the groundhogs mentioned earlier.
D. Swamp Rice:
Traditional swamp rice farms are more prevalent in Nimba than in Grand Gedeh. These swamp rice f arms are often ,the responsibility of women who may be assisted by a man. Women have limited access to upland fields, and swamp rice is one of few domains open to them as a source of personal income. In contrast
to Grand Gedeh, these swamp farms may not be in close proximity to the upland rice field.
Improved swamp rice cultivation is also more prevalent in
Nimba than in Grand Gedeh. Swamp rice farms are operated
communally by project motivated groups as well as privately by families or individuals. Group swamp farms have been initiated by the NCRDP. Village work groups called FDA's have been established in many villages and are strongly oriented towards swamp rice cultivation as a group activity. In some cases, members spend one to two days a week during peak labor seasons working on the FDA swamp farm. The limited input approach of this
project contrasts with that of the BCADP in Bong County. Private participation in improved swamp rice farming again tends to be

more common around major towns like Saclepea.
E. Other Field Crops:
Sugar cane and groundnuts are both grown by some farmers in Nimba, increasing the diversity of the cropping system relative to Grand Gedeh. Groundnuts may be cultivated in pure stands on a portion of a previous year's rice field or intercropped with cassava. Groundnuts are grown as a secondary crop for both consumption and cash income. Sugar cane is more likely to have a major impact on a farmers' activities since it can provide a major source of income. It may be grown in lieu of tree crops and may even influence the size and location of the upland rice field.
F. Tree Crops:
Cocoa and coffee are probably of equal importance as cash crops in Nimba County. As in Grand Gedeh, both crops have been cultivated by a few farmers for over 20 years. Although some early farmers obtained their planting material from the Ivory Coast, the Cocopau plantation stands out as a major source of seed and seedlings. Most cultivation operations for these tree crops follow similar patterns as those found in Grand Gedeh, except that harvest appears to be somewhat later, especially for coffee. Coffee harvest does not typically begin until October or November. Farmers in Nimba place somewhat more emphasis on
processing tree crop products than farmers do in Grand Gedeh. Cocoa is often fermented for 6 or 7 days, and is stirred on the third and sixth day. Some farmers also mill their coffee after drying it, and sell it for a higher price as clean coffee. It is not clear whether the presence of LPMC buying stations or project activities have increased farmer involvement in cocoa and coffee processing.
Other Tree Crops
Unlike the other counties, banana and plantain are rarely intercropped with rice because cassava is generally planted over the entire rice field. Bananas and plantain tend to be planted in pure stands often near the rice kitchen, or intercropped with cocoa, coffee, or sugar cane. Citrus appears to be a little more common in Nimba than in Grand Gedeh but it is still not
significant as a source of income. A number of small farmers are planting oil palms provided by the NCRDP. The project policy is to provide enough seedlings to plant one acre a year for up to ten years. This is a new project activity and the oil palms planted have not yet begun to produce. Rubber is an important commercial tree crop for plantations and concessions in Nimba

county but it is only occasionally grown by small farmers.
G. Livestock:
Husbandry patterns in Nimba are similar to those found in Grand Gedeh. Compared to Grand Gedeh villages, villages in Nimba seem to have a few more sheep, about the same number of goats and chickens, a few less cattle, and many more pigs. Pigs are cited as an important source of meat for feeding kuus.
H. Other Sources of Income:
In Nimba, the opportunities for off-farm employment seem
greater than for Grand Gedeh, but perhaps somewhat less than in Bong. Palm wine is sold, but perhaps less than along the highway in Bong. The sale of wild meat remains important, but less so than in Grand Gedeh. Migration, crafts, fishing, farm labor and money sent from relatives are additional sources of income. Farm laborers receive about $2 a day on average.
I. Marketing in Nimba County:
Marketing Cooperatives are more prevalent in Nimba than in Grand Gedeh. Although they are relatively ineffective in providing marketing services to farmers, the coops do provide some limited services. Compared to Grand Gedeh, it seems that farmers in Nimba have better marketing access. The road
conditions are better. There are more major market centers, probably due to a higher population density and the prevalence of larger towns and villages. The LPMC sub-buying center in Ganta provides marketing services to farmers and farmers' cooperatives. Some of the FDA's sponsored by NCRDP are also beginning to take on some of the responsibility for marketing farmers cash crops.
In spite of all of the above, Mandingo traders appear to have a monopsony on produce buying in Nimba. The Mandingo and their sub-agents seem to be more organized and have established some informal but relatively efficient produce buying networks in the Nimba area of study.
J. Credit in Nimba County:
In Nimba, village credit associations (banks) are commonly found and serve as important credit sources. Compared to Grand Gedeh, it appears that there are more and better organized credit clubs, perhaps because susus are more traditional in Nimba. As in Grand Gedeh, credit club interest rates are high (i.e., usually the repayment is 125 percent to 150 percent of the loan amount for a three month period).
The presence of the Nimba County Rural Development Project

(NCRDP) is an advantage for farmers in the area. The project has adopted a low cost and low-key approach which emphasizes "selfhelp" rather than financial aid. It arranges credit for small tools, improved rice seeds, fingerlings, tree crop seedlings (cocoa, coffee, and oil palm) rice mills and rototiller services for farmers, especially those participating in FDA's and village working groups in the project area.
In a limited way, Mandingo traders also provide credit for some cocoa and coffee farmers in the area. In addition, farmers who do not belong to any credit association obtain credit from friends and relatives.
K. Food Consumption:
The Gio farmers in Nimba County indicate that they prefer cassava prepared as gigbah (a dumboy like substance) more than rice. They claim they eat cassava at least once a day throughout the year. The Mano, on the other hand, do not consume as much cassava as the Gio, although they do eat it during the hungry season.
As for meat consumption, less wild meat is available in Nimba than in Grand Gedeh because of the secondary bush and high population densities. Farmers rely more on freshwater fishing, market-purchased meat (especially marine fish) and domestic animals for their sources of meat. As stated earlier, domestic pigs are commonly used to feed kuu labor groups because of the amount of meat obtained from this animal.
L. Government Interventions in Nimba County:
As pointed out before, the Nimba County Rural Development Project has had some influence on the farming systems in the project area. The project's interventions include the development of FDA's and village working groups, improved swamp rice farms, fish ponds, rice mills, tree crop (cocoa, coffee, oil palm) seedlings and the use of rototillers for swamp development. The scope and intensity of these activities has caused certain
farmers to become over extended and to neglect or reduce some of their personal farm enterprises.
NCRDP provides logistical and other support services for the MOA extension personnel assigned to the project. The assistance (motor bikes, gas, etc.) facilitates the work of the extension staff in the county.

A. Access to Land:
The tenure s it ua t ion i s al1s o complicated in Bong, but fo r different reasons than in Nimba. Land pressure is due to the prevalence of concessions (mostly rubber) and private estate ownership. Usufruct tenure arrangements a re s t ill common, but land purchases and land rentals are also found. The only rental arrangement identified is the traditional one of giving a
quantity of rice at harvest to obtain land access. However, the amount of rice given is much more than the traditional token given in the past (now as much as two clean bags of rice is given) in one case, a farmer claimed that he had not been able to make a rice farm the previous year because he had been unable to negotiate land rental with any of the landowners. Land pressure caused by the prevalence of rubber plantations, private estates and government farms has had a considerable impact on the farming system. Similar to Nimba, very little high bush is still accessible to farmers. Farmers are making farms on secondary bush.
with short fallow periods (5-6 years) and are often obliged to farm at considerable distances from the village.
B. Upland Rice:
Upland rice farming in Bong, like Nimba, is characterized by the use of the scratch-planting technique and other strategies which counter the problems associated with farming secondary bu s h. The timing and practices employed in field operations in Bong are even more variable than in Nimba because of the d iv er s ity of farm enterprises and other family a ct iv it ie s typically found in Bong.
SSite selection for upland rice fields tends to be constrained by the availability of land. Given the short fallow periods,
selection is usually based on previous experience with the piece of land being considered. Some of the villages surveyed are
located in very hilly terrain, and the slope of some fields may approach 45 degrees.
Brushing may start as early as October or as late as April but
is commonly done between January and April. Villages differ on whether women participate in brushing or not. The timing fo r
felling trees varies but is centered in February and March. It may be associated with brushing as part of a kuu activity in which both men and women participate, or done separately. Burning
may take place from early March to early June and often falls in April. Clearing takes place from March through June, but may be done all at once by kuu labor or be done progressively in association with hoeing and planting. Hoeing and planting follow burning by two weeks to two months, starting as early as April and continuing through August. Women weed the field again anytime between April and August, while the men spend 4 to 6

weeks building a groundhog fence. Once again, farms appear
slightly larger than in Grand Gedch and practically no "hungry farms" are found. Farmers seem to reserve their cassava for consumption during the hungry season rather than eating it as often during the rest of the year as in Nimba. Buying more rice and eating fewer meals per day seem to be other common strategies for dealing with the lack of food during the hungry season.
C. Cassava:
As in the other two counties, cassava is often intercropped with rice. The pattern in Bong is to plant the cassava 2 weeks to 2 months before the rice. Farmers must then hoe around the cassava, but it does not otherwise interfere with broadcasting the rice. Shading of rice is a problem, but stripping cassava leaves and branches is less commonly practiced. In many fields, low density alone is the only control used to compensate for this competition. The area of the rice field planted to cassava also varies from a portion to the entire field. Bananas and plantain are still usually found separate from the rice. Similar to the other counties, cassava planting consists of laying 3 or 4
cuttings parallel in a shallow trench.
Cassava may also be planted as a second crop after rice or in separate fields. The reason often cited for this practice is the fear that cassava will attract groundhogs to the rice. The time of planting of separate cassava fields is quite variable. There also is a strong tendency to intercrop the cassava with a crop other than rice, usually groundnuts. Intercropping with coffee is also observed.
Many farmers are using Matadi, the introduced variety. But most farmers also continue to grow traditional varieties as well. These include Tusan, Behuna, Gbarkpalin, Kpelemana, Two cents, Awakana, and Gorbu. As in the other counties, groundhogs, porcupines, and ground squirrels are cited as the major pests of cassava.
D. Swamp Rice:
Traditional swamp farming is basically the same as that in Nimba. Improved swamp farming, on the other hand, though project-motivated, is essentially individual oriented rather than group oriented as in Nimba. The BCADP encourages swamp rice
production by providing farmers with loans to buy tools and to hire labor to build the canals and bunds and maintain them for the first two years. The BCADP agents also furnish the technical expertise. This approach, based on providing credit for private farms, contrasts with the low input and communal approach of the NCRDP.

E. Other Field Crops:
Groundnuts and sugar cane are additional sources of crop diversity in Bong County. In some villages surveyed, groundnuts appear to be an important secondary crop and are commonly present
either as an intercrop with cassava or in separate pure stands. In other villages, groundnuts are cultivated by very few farmers or are not present at all. Sugar cane is frequently grown by a few farmers in each village. Cane juice may be produced primarily for *a local market or oriented towards the Monrovia market depending on the location and size of each operation.
F. Tree Crops:
Cocoa i~s an important cash crop in Bong County In one case a farmer claimed his family had been growing it for over 30 years. Coffee, on the other hand, appears to be considerably less common than in the other counties, and of fairly recent origins. many of the coffee growers have only planted coffee since the BCADP began
providing seedlings about 5 years ago. The BCADP provides loans to develop coffee and cocoa fields and to hire labor for underbrus hing during the first two years. Project agents advise farmers on site selection, but farmers complain that in the past these agents have not always been successful in choosing the right sites for cocoa. Project supported cooperatives also have an important impact on marketing alternatives.
Banana and plantain are important secondary tree crops which are both intercropped with rice and grown in pure stands. Citrus is somewhat more important in Bong than in the other two counties. This is probably due to the access to city markets afforded by the Monrovia-Ganta highway. Oranges are the most
common type of citrus grown followed by grapefruit and tangerines. Grapefruit are less favored than oranges because the tr e es a re reputed to be short-lived and because grapefruit generally will not sell while oranges are available. Since grapefruit can remain on the tree f or a cons iderab le period without spoiling, they are often held until the orange season is over.
As in Grand Gedeh, oil palm is not cultivated by small farmers
but wild oil palm is exploited for both consumption and cash income. Rubber is a very important cash crop in Bong county
because of the concentration of concessions and private estates which grow i t. FolIlowing this lead, more small farmers have adopted rubber as a cash crop than in the other two counties. The number and distribution of concessions and estates which are
willing to buy latex or cuplumps makes marketing of rubber more practical than in other areas surveyed. Many farmers do and/or have worked on rubber estates in the past. This experience provides them with a source of knowledge about rubber cultivation practices and access to seed.

G. Livestock:
As in the other two counties, the total amount of livestock observed is quite small. The amount of livestock per village seems to be similar to Nimba except that there are fewer pigs.
H. Marketing in Bong County:
Markets and marketing channels for both the major and minor cash crops are comparatively more developed in Bong than in Nimba and Grand Gedeh. It appears that there are more and relatively good farm-to-market roads in the Bong area of study.
Because of Bong's proximity to Monrovia, the Bong farmers also have access to a larger market. City market women and traders attend the local markets to buy produce for the Monrovia market.
In Bong county, LPMC has an active buying station at Gbarnga. In addition, the Tungban cooperative is active throughout the county and is a major produce buying organization. It is
supported by BCADP which also provides other commercial services, particularly credit to Bong county farmers. There are also
private traders and other local merchants who are involved in the cocoa, coffee, and rice trade. Despite the existence of Tungban as a county-wide marketing organization, it appears that Mandingo still dominate the produce trade much as they do in the other two counties.
I. Other Sources of Income:
The presence of concessions, estates and/or rubber farms make off-farm employment more prevalent than in the other counties. Rapid access to Monrovia via the highway makes the sale of palm wine an important income generating activity. Wild meat sales seem somewhat less common in Bong than in Nimba. Fishing,
crafts, farm labor, and money sent from relatives are additional sources of income for farm families. On the average, Bong farm laborers receive $2 per day.
J. Credit in Bong County:
Village credit associations (susus) are as common in Bong as in Nimba. Most of the farmers interviewed during the survey belong to susus. Additionally, unlike the NCRDP in Nimba, the BCADP provides development and seasonal loans, and chemical
inputs to farmers in the project area. Due to BCADP programs, credit facilities seem to be better in Bong than in Nimba and Grand Gedeh.

K. Food Consumption:
Fish is the most common meat source for most farm families in Bong. In addition to fresh water fish, dried marine f is h from Monrovia is of ten purchased in local markets. Fish also is a common meat used to feed kuus. Wild meat, market purchased meat
(other than fish), and domestic animal meat are less frequently consume d.
L. Government Interventions in Bong County:
The BCADP does have some influence on farming systems in the project area. Its interventions include the introduction and
distribution of improved tree crop (cof fee and cocoa) seedlin-gs, chemical inputs, development and seasonal loans to farmers, improved rice and cassava varieties, swamp rice development, wells and other extension services.
Phase II of the project which started in 1983 is expected to cover a large part of the county and provide additional services to more farmers.

Two basic farming systems are found in the three counties surveyed. The key factors which distinguish the systems are the length of fallow, the rice planting method, and the degree of diversity. The rice-cassava intercrop pattern is also important in distinguishing sub-system differences in Bong and Nimba.
I. Grand Gedeh:
Grand Gedeh farmers typically make their upland fields in high bush which has been in fallow for up to 20 years or more. The
presence of numerous large trees makes tree felling the most difficult and constraining field preparation activity. Groundhogs, weeds, and birds are less of a problem than where the fallow period is normally shorter. Rice is planted using the dibble method and no soil tillage is performed. If a reasonable burn is achieved, farmers do not attempt to clear the field of remaining tree trunks and debris.
Cassava is planted on a portion of the rice field about one month after the rice is planted. This delay prevents cassava from shading the rice and minimizes competition. Bananas and plantain are intercropped with the rice on a different portion of field. A number of tubers and other vegetables are intercropped with the rice throughout the entire field. These include: cocoyams, yams, sweet potatoes, corn, okra, peppers, bitterball, pumpkin, squash, plato, watergreens, tomatoes, and sesame.
Both early- and late-maturing rice varieties are usually
planted. The early variety is planted first so that rice harvesting can begin as early as July or August.
Farmers tend not to build fences to protect their rice from groundhogs. Instead, they try to increase the size of their field and/or locate their field far from areas of secondary bush where groundhogs are likely to be numerous.
In high bush, the dense shade of the mature forest limits the presence of weeds. With little weed seed in the soil to germinate, rice gets a head start and dominates weed competition. Farmers generally do not expect to weed their high bush fields.
Few farmers have either traditional or improved swamp rice farms. A portion of the upland rice field often descends into a swampy area. However, it is considered part of the upland rice field and is cultivated in the same manner.
Women often plant a short maturing rice and intercrops in a small "hungry farm" separate from the main field. It is usually placed on a portion of the previous year's rice field or some other area of secondary bush. This field assures the family's rice needs until the main field is ready for harvest. The remaining produce of the "hungry farm" is used to meet the personal needs of the woman.
The reciprocal kuu system is not generally practiced in Grand Gedeh villages and farmers rely more on family labor. Although some labor is hired to help with many farming activities, hired

labor is more expensive and less frequently employed than in the other counties surveyed.
Cocoa and coffee are the dominant cash crops in Grand Gedeh. In addition, rice, cassava and minor crops are sometimes sold. The sale of wild meat and produce collected from the forest (e.g. wild palm oil fruit) are other important sources of income. Farmers own and produce few animals even though they are important in meeting social obligations and as a source of ready cash for emergencies.
II. Nimba and Bong:
Many of the characteristics of the farming system in Nimba and Bong counties are similar. Both counties are experiencing some degree of land pressure, which has shortened fallow periods, and compels farmers to make their farms on secondary bush. In Nimba, this pressure comes from a high population density, while in Bong, it comes from the presence of concessions and from the private ownership of large estates. Problems closely associated with farming secondary bush such as groundhogs, weeds, and rice birds, have an important impact on the characteristics of this farming system.
Farms appear to be slightly larger in Nimba and Bong than in Grand Gedeh, perhaps due to some combination of the following factors: I) felling trees is less of a constraint; 2) when large kuus do the brushing, more land may be cleared than would be otherwise by family labor; 3) poor fertility and pests associated with secondary bush may oblige families to farm more extensively to meet 'food requirements; and 4) the absence of "hungry farms" may both permit and oblige the cultivation of a larger main upland rice field.
Few large trees are present, so the major land preparation activity is brushing rather than felling trees. More effort is also devoted to clearing small tree trunks and debris after burning so that it will be easier to hoe the field. Hoeing (or scratching) the field does not begin until 2 to 8 weeks after burning in order to give weed seed in the soil a chance to germinate. Broadcast planting is done at the same time as the scratching in order that this single hoeing might both eliminate the weeds and cover the seeds. After hoeing, the weeds are removed from the soil and piled so they will not have a chance to grow back. A second weeding often seems to be necessary, at least in years of high rainfall. This may also be related to the severely reduced fallow period.
In Nimba, cassava is planted at low density at the same time as the rice. It is typically intercropped on the entire rice field, eliminating bananas and plantain as an important rice intercrop. If the cassava gets too tall relative to the rice, the lower branches and leaves are stripped to reduce shading. In Bong', the cassava density is perhaps even lower than in Nimba and it is planted 2 to 4 weeks before the rice. It is often planted on only a portion of the rice field and stripping the cassava

leaves to prevent shading is less frequently practiced. Cassava, in Nimba, seems to have a more important role in both production and consumption than in the other two counties.
In Bong, some of the corn is planted with cassava before the rice is pl1a n ted. Some of the vegetable intercrops tend to be planted after the rice in both counties.
Early- and late-maturing rice varieties are usually planted, but the pattern is not consistent. Where birds are a serious problem, the late-maturing varieties tend to be planted first. The fear is that an early-maturing field will be decimated by the birds, but the attack will be spread across more fields later in the season. This strategy may lengthen the "hungry season" by delaying the new harvest by 2 to 3 months.
Farmers regularly build fences in Bong and Nimba to protect rice from groundhogs. This will occupy the men for .a month or more, while women finish the planting and weeding. Bird watching is also common for a week after planting and for about a month while the rice is heading.
A few farmers have swamp rice fields, but men tend to be more involved in the improved rice fields sponsored by projects. "Hungry farms' are not common, but in some c a ses swamp r ic e f ielIds will1 take their place as well as provide a source of income for personal needs.
The diversity of farm enterprises is one of the bi ages t differences between the farming system in these two counties and that of Grand Gedeh. Farmers are often involved in both cocoa and
coffee and perhaps sugar cane or groundnuts, as well as in their upland f ie ld and occasionally swamp r ic e. Rubber, c it r us or cultivated oil1 palm may alIs o be present. These numerous activities in any one family tend to strain family labor resources and management capabilities. Farmers are al1s o more likely to be involved in some form of off-farm employment. For these reasons, hiring labor and hiring kuus, as well as using reciprocal kuus are very prevalent in these two counties. Farmers
in Nimba often raise pigs with the intention of using them to feed kuus. The use of hired labor is further encouraged by lower day wages than those prevalent in Grand Gedeh.

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EX943HHQN_F55O3E INGEST_TIME 2018-12-11T16:44:21Z PACKAGE UF00054823_00001