 

1.; 4:3!

\ .1... I
r

4
1

 

 

 

 

 

 

,J‘v—aPW. 5 M . ,
n... 1...“...9“... z. 1 .HL 1,.
,‘a 1:1113)‘,‘DOD- [3"

N) y
. 3 ﬁx“ .
1. wauuemwﬂiﬁ, .ﬁnﬁuﬁwﬂ.‘
mm»! :3 $2.. 535$ _ ‘
mum ”tar; 11%...

.3."- . I
, 3......

A a M \b x ,
n5 titawita .gv
. a...» . . . .9 .
“932%... r. 4 . vs :5.
.3 I‘m... ‘ .H.‘ ,
.er f {ﬁlly-Fig.5....
- til.

zkmqul... .1...
.’ 3.. 4. mt u
www.mwﬁuinum

: ﬁt .

‘u’

I .- .QII
it: ‘2...
innai! .

.e‘aiunmapﬁr.
nu.uﬁ.f.uurnhn..mm.

3.30.5401 .

£211
1! 1 v Ll-
. v :wl.l‘v
.1“ 13.6%..
..!..\ \1‘ n...

1‘ 2.2... ¢.
.3 €31.11.

5
.l
I,

.. .8 I
R. 131.15,: a n
. 6413‘ 0.111.. . . .
i 636.64....13 ix! url‘i‘r
1...) (353.1 altar-.130! If um,

 

mm}

2

This is to certify that the
dissertation entitled

CONSTRAINTS TO GREATER BEAN YlELDS IN
HONDURAS

presented by
HUGH EDWARD SMELTEKOP

has been accepted towards fulﬁllment
of the requirements for the

Doctoral degree in Crop and Soil Sciences

 

 

.x/WZJV‘Z: (a i?”

 

Major Professor’s Signature

 

Dec; /é, 3605
Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

LIBRARY i
Michigan State
i Unixersity _

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

SEP 2 929ng ;

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/DateDue.indd-p.1

CONSTRAINTS TO GREATER BEAN YIELDS IN HONDURAS
By

Hugh Edward Smeltekop

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Crop and Soil Sciences

2005

ABSTRACT
CONSTRAINTS TO GREATER BEAN YIELDS IN HONDURAS
by

Hugh Edward Smeltekop

Securing proﬁtable agricultural production for impoverished small-scale farmers is
crucial for economic growth in developing countries. A participatory methodology, like
the Local Agricultural Research Committee model, blends effectively with constraints
ﬁeld trials while supporting other aspects of development, like better access to
information, skills training and access to loans. Research was focused on fertilizer and
common bean variety to identify possible increases in bean production, and weed control
to examine savings from ﬁeld labor to test the effect of traditional versus recommended
inputs for bean production in on-farm trials. In 2003, four full factorial (four
replications) and ﬁve supplemental trials (no replication) were sown in a broadly agro-
ecologically representative area of Honduras called Los Lirnones (Moroceli). Factorial
experiment results indicate that increased fertilizer causes a signiﬁcant increase in yield
in some cases; the supplemental experiments showed with 85% certainty that the
recommended package of improved variety, increased fertilizer and herbicide increased
yield. In 2004, the factorial experiment was repeated without herbicide, with an
additional traditional variety and with two additional fertilizer levels on seven farms
(three farms with three replicates, two farms with one replicate). Due to terminal
drought, one early maturing traditional variety out-yielded the improved variety in one

case, and in an analysis using data from all ﬁve plots. Fertilizer had no effect on yield

under these conditions. On a ﬁeld day in 2005, farmers misidentiﬁed the worst-
producing plot in one case, and misattributed the best production to the local variety
Marciano three times. Misidentiﬁcations of bean variety production were likely due to
variety biases that farmers had based on observations in other contexts. A similar
experiment was performed in another community called Lavanderos (Moroceli) at a
higher altitude. Results from one farm indicate a slight decrease in yield after the
midseason application of urea fertilizer; and a much higher yield ﬁom the traditional
variety compared to the two improved varieties. Surveys revealed that farmers usually
save seed, and one in ten farmers report seed-home anthracnose as important diseases in
2003 and 2004, with similar numbers for mosaic viruses, so farmers likely underestimate
the importance of seed quality. Planting densities varied, and many farmers would
beneﬁt from a higher planting density, up to 60 kg/ha. The Local Agricultural Research
Committee model was an effective way to test new technologies and disseminate new

production information.

 

TABLE OF CONTENTS

LIST OF TABLES ............................................................................ vi
LIST OF FIGURES .......................................................................... ix
LIST OF ABBREVIATIONS .............................................................. xi
ACKNOWLEDGEMENTS .................................................................. xii
PREFACE ...................................................................................... xiv
CHAPTER 1: Introduction .................................................................. l
1.1 Addressing Poverty in Honduras and the World ...................................... 1
CHAPTER 2: Literature Review ........................................................... 6
2.1 History of Common Bean Cultivation ................................................. 6
2.2 Constraints Research: The “Yield Gap” Model ...................................... 8
2.3 Participatory Development .............................................................. 13
2.4 Survey Design and Implementation .................................................... 17
2.5 Development Trends in Honduras ...................................................... 19
CHAPTER 3: Research Justiﬁcation and Objectives .................................... 31
3.1 Justiﬁcation ................................................................................ 31
3.2 Research Objectives ...................................................................... 32
CHAPTER 4: Research Methods ........................................................... 34
4.1 Survey Methods ........................................................................... 34
4.2 Economic Analysis ....................................................................... 37
4.3 Experimental Field Plot Methods ...................................................... 37
4.3.] Experimental Plot Locations ......................................................... 39
4.3.2 Experimental Plot Designs ............................................................ 39
4.3.2.1 Los Limones Plots 2003 ............................................................ 39
4.3.2.2 Los Lirnones Plots 2004 ............................................................ 41
4.3.2.3 Lavanderos Plots 2004 .............................................................. 43
4.3.2.4 Farmer Perceptions of Bean Variety Performance .............................. 45
CHAPTER 5: Results and Discussion ...................................................... 47
5.1 Survey Results ............................................................................ 49
5.1.1 Household and Land Characteristics ................................................ 49
5.1.2 Typical Farming Practices ..................................................... - ....... 52
5.1.2.1 Herbicide Use and Tillage .......................................................... 52
5.1.2.2 Sowing ................................................................................. 53
5.1.2.3 Fertilizer Application and Weeding ............................................... 54
5.1.2.4 Insect and Disease Management ................................................... 55
5.1.2.5 Harvest and Storage .................................................................. 57

iv

5.1.3 Bean Yield Analysis ................................................................... 58

5.1.4 Farmer Financial and Social Factors ................................................ 61
5.1.5 Beneﬁts of LARC Membership ...................................................... 66
5.1.6 Farmer Perception of Production Constraints ....................................... 68
5.2 Field Results .............................................................................. 73
5.2.1 Los Lirnones Plots 2003 ............................................................... 73
5.2.2 Los Lirnones Plots 2004 ............................................................... 77
5.2.3 Lavanderos Plots 2004 ................................................................ 80
5.2.4 Farmer Perceptions of Bean Variety Performance ................................. 81
5.2.5 Economic Analysis of Field Results ................................................. 85
CHAPTER 6: Conclusions and Recommendations ...................................... 89
6.1 Project Summary ......................................................................... 89
6.2 Breeding and Agronomic Recommendations ......................................... 93
6.3 Extension Recommendations ............................................................ 94
6.4 Policy Recommendations ................................................................ 94
6.5 Limitations to the Study .................................................................. 95
6.6 Future Research ........................................................................... 98
APPENDICES ................................................................................. 100
Appendix 1: Summary of Adopted Adjustment Measures .............................. 101
Appendix 2: Principal Constraints for Agricultural Development ..................... 103
Appendix 3: Survey in 2003 for Los Lirnones ............................................ 105
Appendix 4: Survey in 2004 for Los Lirnones ............................................. 121
Appendix 5: Survey in 2004 for communities other than Los Lirnones ............... 132
Appendix 6: Agronomic Results from Factorial Trials in 2003 in Los Lirnones. 150
Appendix 7: Field Day Questions for Farmers in Los Lirnones in 2004 ............... 153
Appendix 8: Common Local (Spanish), Scientiﬁc and Common English Names

for Farmer-Identiﬁed Insects .................................................... 155
Appendix 9: Common Local (Spanish), Scientiﬁc and Common English Names

for Farmer-Identiﬁed Diseases .................................................. 157
BIBLIOGRAPHY ............................................................................. 159

LIST OF TABLES

Table 1. Worldwide production area and gross production for cereals
and pulses from 1961 and 2002 ............................................................ 8

Table 2. Effects of improvements in agricultural technology based on
household initial asset position ............................................................. 9

Table 3. Real producer prices for basic grains, 1989-1996
(Lempiras/hundredweight) .................................................................... 22

Table 4. Surveys implemented in six communities in the department of
El Paraiso, Honduras, 2003 and 2004 ....................................................... 35

Table 5. Locations of communities where agricultural research and surveys
were implemented, El Paraiso, Honduras, 2003 and 2004 .............................. 39

Table 6. Management components for full factorial trials, Los Lirnones,
Moroceli, Honduras, Postrera, 2003 ......................................................... 40

Table 7. Experimental plot soil type, pH, percent organic matter (%OM),
percent nitrogen (%N), Mehlich phosphorus (P) and Mehlich potassium
(K) for nine plots, Los Lirnones, Moroceli, Honduras, 2003 ........................... 40

Table 8. Management components for full factorial trials, Los Lirnones,
Moroceli, Honduras, Postrera, 2004 ......................................................... 43

Table 9. Management components for factorial trials, Lavanderos,
Giiinope, Honduras, Postrera, 2004 .......................................................... 44

Table 9. Experimental plot soil type, pH, percent organic matter (%OM),
percent nitrogen (%N), Mehlich phosphorus (P) and Mehlich potassium
(K) for two plots, Lavanderos, Giiinope, Honduras, 2004 ............................... 44

Table 10. Management components for experimental plots in full factorial
trials, Los Limones, Moroceli, Honduras, Postrera, 2003 ................................ 45

Table 11. Characteristics of Survey Population from Six

Communities, Department of El Paraiso, Honduras, 2004 (Head of

Household unless otherwise indicated) (N=72) ............................................ 50
Table 12. Distribution of bean plot size, number of plots sown,

plot SIOpe and soil quality indicators reported by farmers,

El Paraiso, Honduras, Postrera, 2003 and 2004 (N =72) .................................. 51

Table 13. Distribution of bean plot tenure reported by farmers,

vi

El Paraiso, Honduras, Postrera, 2003 and 2004 (N=72) .................................. 52

Table 14. Distribution of herbicide use and tillage reported by farmers,
El Paraiso, Honduras, Postrera, 2003 and 2004 (N=72) ................................. 52

Table 15. Distribution of number of varieties sown, farmers sowing at
least one improved variety, and rows of beans per row of grain crop
(intercrop) reported by farmers, El Paraiso, Honduras, 2003 and 2004 (N=72)...... 53

Table 16. Distribution of seed source reported by farmers, El Paraiso,
Honduras, Postrera, 2003 and 2004 (N=72) ............................................... 54

Table 17. Distribution of timing of fertilizer application reported
by farmers, El Paraiso, Honduras, Postrera, 2003 and 2004 (N=72) ................... 55

Table 18. Bean yield (kg/ha) for producers L4, factorial
trials, Los Limones, Moroceli, Honduras, Postrera, 2003 ............................... 75

Table 19. Supplemental trial bean yields (kg/ha) for 9 farmers, Los Lirnones,
Moroceli, Honduras, Postrera, 2003 ......................................................... 76

Table 20. Factorial trial bean yields (kg/ha), Producer 1, Los Lirnones,
Moroceli, Honduras, Postrera, 2004 ......................................................... 78

Table 21. Factorial trial bean yields (kg/ha), Producer 2, 2004, Los
Lirnones, Moroceli, Honduras, Postrera, 2004 ........................................... 78

Table 22. Factorial trial bean yields (kg/ha), Producer 3, Los Lirnones,
Moroceli, Honduras, Postrera, 2004 ........................................................ 79

Table 23. Factorial trial bean yields (kg/ha), combination of 5 farms,
Los Lirnones, Moroceli, Honduras, Postrera, 2004 ...................................... 79

Table 24. Factorial tn'al bean yields (kg/ha) of Producer 1,
Lavanderos, Guinope, Honduras, Postrera, 2004 ................. - ........................ 80

Table 25. Factorial trial bean yields (kg/ha) of Producer 2,
Lavanderos, Gilinope, Honduras,Postrera, 2004 ......................................... 81

Table 26. Farmer belief about highest- and lowest-producing plots and

actual production, rainfall and maximum yield, Los Lirnones,
Moroceli, Honduras, Postrera, 2004 ......................................................... 82

Table 27. Partial budget under traditional vs. recommended management,
Los Lirnones, Moroceli, Honduras, Postrera, 2003 (N =9 plots) ........................ 86

vii

Table 28. Partial budget under traditional vs. recommended management,
Los Lirnones, Moroceli, Honduras, Postrera, 2004 (N=5 plots) ......................... 87

Table 29. Partial budget under traditional vs. recommended management
for Producer 1, Lavanderos, Giiinope, Honduras, Postrera, 2004 ...................... 88

viii

LIST OF FIGURES

Figure l. Producers of beans with 5 hectares or less (1993) and

average annual precipitation (1999) in Honduras .......................................

Figure 2. Producers of beans with 5 hectares or less (1993) and

altitude (1999) in Honduras .................................................................

Figure 3. Diagram of the eight steps in the reiterative process

used for organizing a Local Agricultural Research Committee (LARC) ............

Figure 4. Map of Honduras ................................................................

Figure 5. Example of bean plot observed for plot visit by farmers from

Los Lirnones, Moroceli, Honduras at 60 days post-sowing, November 23, 2004

Figure 6. Percentage of farmers (N =72) who reported insect pest

problems, El Paraiso, Honduras, Postrera, 2003 and 2004 ............................

Figure 7. Percentage of farmers (N=72) who reported signiﬁcant infection

by disease, El Paraiso, Honduras, Postrera, 2003 and 2004 ............................

Figure 8. Relationship between seeding density of common beans
and yield where yield was >99 kg/ha (N=121), El Paraiso, Honduras,

Postrera 2003 and 2004 .....................................................................

Figure 9. Relationship between seeding density of common beans
and yield where yield was >850 kg/ha (N=32), El Paraiso, Honduras,

Postrera, 2003 and 2004 .....................................................................

Figure 10. Relationship between proﬁt (Lp) and yield (kg/ha) (N=148),

El Paraiso, Honduras, Postrera, 2003 and 2004 .........................................

Figure 11. Producer-reported level of conﬁdence in available source of
information (N=72), El Paraiso, Honduras, 2004 .............................

Figure 12. Farmer perception of availability of loan for bean production

(N=66), El Paraiso, Honduras, 2004 ......................................................

Figure 13. Self-reported beneﬁts recognized by members of Local Agricultural

Research Committees, El Paraiso, Honduras, 2004 .....................................

Figure 14. Responses to question, “What are the biggest constraints to

increased bean production in postrera?” =66), El Paraiso, Honduras, 2004 .....

ix

..3

.4

..15

.20

46

..56

.57

..59

.59

..62

64

.65

.67

..68

Figure 15. Preference for traditional varieties as reported by farmers (N=22),
Los Lirnones, Moroceli, Honduras, 2003 ................................................... 71

Figure 16. Amadeus-77 plants at 60 days post-sowing in Los Lirnones,
Moroceli, Honduras, November 23, 2004 ................................................. 83

LIST OF ABBREVIATIONS

BGYMV: Bean Golden Yellow Mosaic Virus

CIAL: Comite' de Investigacién Agricola Local (Local Agricultural Research Committee
in English)

CDR: complex, diverse and risk-prone
ha: hectare

kg: kilogram

L: liter

LARC: Local Agricultural Research Committee (Comité de Investigacion Agricola Local
in Spanish)

Lp: Lempiras, monetary unit in the Republic of Honduras
m.a.s.l.: meters above sea level
Mt: metric tons

WUE: Water Use Efﬁciency

xi

ACKNOWLEDGEMENTS

First of all, I would like to thank Dr. Russell Freed who allowed me to develop a program
of study that let me explore what I found most interesting and useful to me as a
professional, who gave me unconditional support in all of my endeavors, and who has

been an incredible mentor and guide for me over the past three years.

I am grateful to Dr. Rick Bemsten, Dr. Jim Kelly, and Dr. Irvin Widders for their help
throughout my studies, especially the time that each one of them gave unselﬁshly, oﬁen

at a moment’s notice.

Dr. Juan Carlos Rosas, the director of the bean research program at Zamorano, was
always available and a supportive collaborator - without his help, this study would have

been impossible.

Byron Reyes, Jorge Venegas, and the whole crew in the bean program at Zamorano were
all hugely important to me, as mentors, teachers and friends during my time in Honduras

and beyond.

The moral support that I received through the members of the bean laboratory at

Michigan State was incalculable, especially Veronica Vallejo, Esteban Falconi, Karolyn

Terpstra, Ann Armenia, Halima Awale, Erica Boerman, and Evan Wright.

xii

None of this would have been possible without the help and support of friends and
family, especially my sister Paula to whom I owe so much in so many ways; my mom
and dad, my step-father John and Bruce in loco parentis, my other siblings Rob, Ellen

and Tyler.

Special thanks to my friends Huy Tran, Suzanne Thiem, Orlando Mejia, Elena Sierra,

Brook Wright, Jeremy Moghtader, Michele Pluck, and many, many others.

The support for the actual research came from the Bean Cowpea/CRSP Grant No. GDG-
G-00-02-00012-00, and support for my studies and travel to Honduras and to professional
meetings came from Dr. Freed himself. I am also grateful for the Lawton Fellowship that
I was awarded to help pay for additional expenses in Honduras, and to the graduate
school and the international studies program for their support for my travels to Bolivia to

present these results.

xiii

PREFACE

I have a compadre named Victoriano Marnani who lives in the small community of
Chovacollo. His is one of forty-odd families with farms carved out of the hillside in a
cloud forest only a few hours from the capital of Bolivia, La Paz. He has three hectares
of coffee, citrus, and coca, interspersed with some food crops like taro, manioc and
vegetables. He owns two cows and a few chickens. He and his wife Cenovia and their
two daughters and two sons work very, very hard, and earn less than ﬁve hundred dollars
a year; most of that income comes ﬁom the sale of coca. And although his coca is

considered legal, coca production is contentious.

What hope is there for Victoriano and the millions of farmers like him, in Bolivia and
throughout Latin American and the world? Economic development may be the answer,
but it is an incomplete answer. Every social system must respect the value of every

human life.

My time at Michigan State, with the generous support of Dr. Freed, has allowed me to
contemplate our fate as human beings one this one, fragile earth, and has renewed hope in
me that I, and every one of us, can make a difference by bringing respect for all human
life — and ultimately all life in earth - to every aspect of our living. Though each of us
has his or her own path, if we strive to bring each moment of our lives into concert with

our values, we can create a world of economic grth and respect for all life.

xiv

CHAPTER 1

INTRODUCTION

A Green Revolution-type strategy for identifying production constraints has helped
address the problems of poor food security and poverty in the past, though hunger
persists. Working with small-scale bean farmers in Honduras can provide insight into

how to relax those constraints and increase food security.

1.1. Addressing Poverty in Honduras and the World

Over one billion people in the world are considered extremely poor, unable to meet their
basic needs for survival, and 800 million are malnourished, with one fourth of these
under the age of ﬁve (United Nations, 2005). Economic development that beneﬁts the
poor must include agricultural development when there are signiﬁcant numbers of poor
living in rural areas, an idea known as structural transformation (Staatz and Eicher,
1998). One model that was successful in staving off large numbers of deaths due to
starvation was the Green Revolution, and current strategies for poverty alleviation beneﬁt

from a look back to those times.

Starting in the 19503, the Green Revolution increased wheat (Triticum aestivum) and rice
(Oryza sativa) production by creating packages of improved varieties and agricultural
input strategies to increase production and feed undemourished populations. When

increases in production are made in the context of the livelihood of the small-scale farmer

- that is, in the context of farmer capabilities and means of living - farmers increase their

production options while conserving the sustainability of their profession.

One potential area for improving the quality of life in the developing world is the case of
small farmers in Honduras and the production of common beans (Phaseon vulgaris).
Common beans are a staple food in the Honduran diet, which, when eaten with maize
(Zea mays), constitute a source of protein nutritionally equivalent to meat. In Honduras,
56% of national production of common beans is produced on farms of ten hectares or

less, which includes 75% of Honduran farms.

In order to better understand the situation of the producers most in need, maps that
represent key factors were generated. First, all farms that grew beans in 1993 (the most
recent census data) (Secretaria de Planiﬁcacién Coordinacion y Presupuesto, 1993) who
had ﬁve hectares or less of land were identiﬁed. Next, this data is compared with average

rainfall (Figure 1) and altitude (Figure 2).

Figure 1. Producers of beans with 5 hectares or less" (1993) and average annual
precipitation (1999) in Honduras.

   

Average annual

 
 
  
  

 

rainfall (mm)
Los Lirnones I" 681 - 1266
, ,3; 1267 - 1588
Lavanderos ‘ Q 1589 - 2005
ti. 2006 - 2533
5 O 2534 - 3283

 

 

 

‘ l blackdot=15farms

Sources: Mapping data from CIAT and Honduras Census (Centro Intcmacional de
Agricultura Tropical, 1999a; Centro Intemacional de Agricultura Tropical, 1999b;
Secretaria de Planiﬁcacion Coordinacion y Presupuesto, 1993)

Figure 2. Producers of beans with 5 hectares or less“ (1993) and altitude (1999) in
Honduras.

    
 
  

_ 0 -400
' 400 - 700

‘ 700 - 900
900 - 1100

, 1100- 1400
' 1400 - 1600
: 1600 _ 1900
:- 1900 - 2100
.2100 _ 2400
: 2400 - 2600

   

Los Lirnones

  
    

Lavanderos

Altitude (meters _
above sea level) ‘

     
 
    
  

 

' 1 black dot=15 farms _
Sources: Mapping data from CIAT and Honduras Census (Centro Intemacional de
Agricultura Tropical, 1999a; Centro Intemacional de Agricultura Tropical, 1999b;
Secretaria de Planiﬁcacion Coordinacion y Presupuesto, 1993)

These maps identify key areas for possible interventions and strategies for improving the
plight of small-scale common bean farmers. From these farms, one-half of production in
eaten, and the other one-half is sold, a fact which highlights the importance of common
beans both as a staple food and as a crop with potential for increasing farm income
(Martel et al., 2000). In addition, since similar conditions exist throughout Central

America and Mexico, this work is consequential for that entire region.

Addressing this issue entails examining agronomic factors (changes in management and

their consequences for the agro-ecology of the farming systems), economic factors (costs

of proposed management changes and availability of funds to make these changes, farmer
attitude about risk and ways to reduce risk), social factors (ability to access information),
and political factors (how government policies support/fail to support agricultural

production).

These issues can be approached with a participatory model that combines farmer
knowledge with researcher knowledge to research production problems relevant to the
farming system in question, and to make recommendations to other farmers and to the
research and extension communities so that this knowledge can be disseminated to others

in similar situations.

CHAPTER 2

LITERATURE REVIEW

Common beans are an important source of food for people worldwide, and have been for
over two centuries. Securing proﬁtable agricultural production for impoverished small-
scale farmers is crucial for economic grth in developing countries, and research to
eliminate production constraints is a useful tool to achieve this. Constraints research has
two parts: surveys to identify constraints, and on-farm ﬁeld trials to test solutions to
production problems. A participatory methodology, like the Local Agricultural Research
Committee model, blends well with constraints ﬁeld trials while supporting other aspects
of development, like better access to information, skills training and access to loans.
Whereas modernization in Honduras has improved the economy in general, more must be
done to speciﬁcally improve the lot of impoverished small-scale farmers. The Honduran
govemment’s plans to eradicate poverty in the rural sector are a step in the right

direction.

2.1 History of Common Bean Cultivation

Evidence of the cultivation of common beans has been uncovered in Mesoamerica as
early as 4000-5000 B.C.E., though the pattern of distribution at archeological sites
suggest that “Phaseolus beans may not have been an important part of the diet until the
Late Pre-Classic and Classic Periods,” around 100-700 C.E. (Kaplan and Kaplan, 1988)

The cited authors continue that

rapid changes in population growth, settlement patterns, land use and other areas of human
activities may have required the production of crops having a short harvest season; in contrast with
the extended season characteristics of vining beans, bush beans may be become more important at

this time. In later Post Classic times, conquest and the exacting of tribute which had become
marked in Classic times no doubt affected agricultural practices by stimulating the production of
short season harvests of storable, transportable crops like maize and bush beans.

Common beans were also discovered at sites in the South American Andes, including
specimens from Guitarrero cave in Peru around 5730 B.C.E. There is evidence of
exchange of crops between Mesoamerica and northern South America, so it is
conceivable that common beans were transported to the Andes and continued to evolve
into a major gene pool; however, species exist in the Andes from which modern cultivars
could have been domesticated, so domestication in the Andes is a possibility. Debouck

has since identiﬁed a third major gene pool in the northern Andes (Debouck, 1999).

Common beans have become the third most important food legume after soybeans
(Glycine max) and peanuts (Arachis hypogaea), and they are the most widely eaten pulse
worldwide (Table 1). They are grown on all continents except Antarctica, and accounts

for 90% of all area sown to Phaseolus species (Singh, 1999).

Common beans continue to be an important source of calories and protein in Honduras.
In 2004, the Food and Agricultural Organization reported that Hondurans produce 69,949
metric tons of common beans per year, equaling a production of about 10 kg/person/year

(Food and Agriculture Organization, 2005b).

Table 1. Worldwide production area and gross production for cereals and pulses from

1961 and 2002.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Area (‘000,000 ha) Production

(‘ 000,000 tons)
Crop / Year 1961 2002 1961 2002
Total cereals 648.1 663.8 876.8 2036.8
Total Julses 63.8 70.9 40.7 57.0
Dry Bean“ 22.8 26.9 11.2 19.3
Pea (Pisum sativum) 7.5 6.1 7.3 9.9
Chickpea (C icer arietinum) l 1.8 10.5 7.7 8.3
Cowpea (Vigna unguiculata) 2.2 10.0 0.8 3.6
Pmon pea (Cajanus cajan) 2.7 4.3 2.2 3.1
Lentil (Lens culinaris) 1.6 3.7 0.9 2.9
Favabean (Viciafaba) 5.4 2.7 4.8 4.2

 

*includes P. vulgaris and other Phaseon and some Vigna species.
Source: Food and Agriculture Organization, 2005a

2.2 Constraints Research: The “Yield Gap” Model

It is widely understood that increasing agricultural production is beneﬁcial to

 

development at all levels. This belief is grounded in the principles of structural
transformation, the idea that agriculture makes important contributions: “it could provide
labor, capital, foreign exchange, and food to a growing industrial sector and could supply
a market for domestically produced industrial goods” (Staatz and Eicher, 1998). The
beneﬁts of growth in the agricultural sector are not evenly distributed, but grth does
unambiguously help net sellers of labor; and all participants in the market beneﬁt as the

costs of production fall (Table 2).

Table 2. Effects of improvements in agricultural technology based on household initial

 

 

 

 

 

 

 

asset position.

Direct and indirect effects of agrEultural technolo improvement‘

Food
Household production Labor
initial asset rises, costs demand Wage Food price
position fall rises rises falls Net effect Comments
Net seller of Shift more land to
food, net + + + - Ambiguous cash crops and sell
seller of more; employment
labor income rises
Net buyer of Food expenditure
food, net + + + + + falls, employment
seller of income rises
labor
Net seller of Shift more land to
food, net + - - - Ambiguous cash crops and sell
buyer of more; wage
labor expenditure rises
Net buyer of Food expenditure
food, net + - - + Ambiguous falls but wage
buyer of expenditure rises
labor
Landless; Food expenditure
buyer of n.a. + + + + falls, employment
food, seller income rises
of labor

 

 

 

 

 

 

 

 

"' “+” indicates a beneﬁcial effect for the actor, “-” indicates a negative effect.

Source: Kerr, 2004

One important model used to investigate the constraints to increased bean production is

the “yield gap” model (Barker, 1979). The “yield gap” is the difference between actual

and potential yields, and it can be deﬁned in different ways. Yield gap can be deﬁned by

the highest yield on a farm or experiment station compared to the national average yield.

This comparison is not necessarily helpful for understanding why farmers are not

producing efﬁciently: the national average is an average of a large diversity of agro-

ecological situations. Another way to conceptualize yield gap as the difference between

an experiment station yield and farmer ﬁelds in a similar agro-ecological zone. This can

lead to more relevant comparisons.

 

Some have attempted to quantify the constraints that cause the yield gap. For example,
Herdt and Wickharn (Herdt and Wickharn, 1975) deﬁned the yield gap as the difference
between the maximum yield at the experimental farm (8 tons rice/ha) compared to the
national average (1.8 tons rice/ha) in the Philippines. They determined that 40% of the
yield gap was due to socio-economic factors (lack of inputs, non-adaptation of
technology, and other economic constraints) and 60% due to environmental factors and

lack of water control.

This kind of partitioning of the yield gap leads to a greater understanding of the
constraints to increasing production. The yield gap can be partitioned into two general
groups that represent different constraints. Yield Gap 1 can be considered non-
transferable technology and environmental differences, factors outside the control of the
farmers. Yield Gap II includes the biological constraints (variety, weeds, disease, pests
and soil fertility) and socio-economic constraints (costs, credit availability, tradition,
knowledge, and input availability). A researcher’s job then becomes measuring the Yield
Gap II, the difference between the actual and potential bean yields in a given area; and
then determining whether changes can be made that would be ecologically sustainable

and proﬁtable for farmers.

The concept of economic slack — the difference between the present product and the
product made when all resources are used optimally — is also useful in a discussion of
yield gaps (Barker, 1979). Economic slack can be created in different ways, and includes

things like advances in technology, programs that increase the supply of inputs, and

10

investment in infrastructure and land reform. This slack falls under the realm of Yield
Gap II and becomes the area of study for determining what changes can be proposed to

farmers.

One problem with examining yield gap without the context of the farm family is that the
goals of experimental farms and farmers are different. Commonly, experiment stations
desire to maximize yield whereas farmers want to maximize proﬁt, as noted by Davidson
and Martin (Davidson and Martin, 1965). This understanding lead to the concept of
economically recoverable yield gap, the amount the gap can be reduced through the
maximization of farm proﬁt. Once the economically recoverable yield gap is eliminated,

the economic slack has been taken up.

In the face of the non-adaptation of new rice varieties where researchers felt that they
were appropriate, and the poor yields on new varieties where they were being used,
researchers at the International Rice Research Institute met and developed the Handbook
on the Methodology for an Integrated Experiment-Survey in Rice Yield Constraints
(Gomez et al., 1979a). The approach used surveys and controlled agronomic experiments
in ﬁelds managed by farmers. The objectives of the study were to compare actual farm
yields with yields achievable through appropriate technologies, and to determine to what
degree these technologies could be proﬁtably implemented, to identify the social and
institutional factors preventing the implementation of these technologies, and to
determine the extent to which these physical, social and institutional factors constraining

yield could be removed.

11

The constraints methodology was utilized throughout Southeast Asia in the late 19705
and the results were mixed and country-speciﬁc. For example, in Bangladesh,
researchers found that new rice varieties were more proﬁtable but that the intensive rice
cultivation did not ﬁt well into the cropping systems (Hoque et al., 1979). Researchers in
the Philippines discovered that institutional arrangements about weeding (pay based on
hand-weeding and harvesting labor) made hand-weeding more desirable, despite an
increase in proﬁt with chemical weed control (Gomez et al., 1979b). Overall, however, it
was found that farmers underestimated the yield effect and economic beneﬁt of nitrogen

fertilizer, especially in the dry season (Herdt, 1979).

Working with other crops, researchers in other areas of the world have also successfully
used this model. Kamanga and Shamudzarira (Kamanga and Shamudzarira, 2001) use a
constraints approach (described as a “mother-baby design” (Snapp, 2002), where the
mother plots are managed by farmers) to test the integration of legumes into maize
cropping systems. They found that farmers had preferences for different legume
intercrops based on many different criteria unique to their situations (e. g. effect of
legume on aphids in maize, suitability as forage). Similar experiments were performed
on-farm to examine constraints to common bean production in Tanzania, comparing
differing levels of fertilizer, insect control and bean variety factors. Researchers found
that soil infertility and poor performing varieties were the two largest constraints to

production (Edie et al., 1991).

12

2.3 Participatory Development

Participatory development is simply the participation of the people who wish to improve
their states of being in that process. The idea of participatory development, as it is
understood today, grew out of the realization in the 1960s and 19705 that packages of
technology were not always helpful to farmers whose situations were complex, diverse
and risk-prone (CDR). CDR farmers usually have livelihood strategies that encompass
many small enterprises. Thus, a failure in some activities is offset by success in others
(Chambers, 1997). Sometimes the technology packages offered by extension agents to
farmers in CDR environments were too expensive, required more time away from other
activities, or were geared toward circumstances that were not applicable to those farmers
(such as non-marginal soils, or non-drought conditions). In areas with farmers who did
not beneﬁt from available technologies, farming systems research and increasing on-farm
experimentation resulted in the idea that farmers would beneﬁt from an integral

participation in the development of useful technologies.

Research into production constraints became more participatory as researchers grew in
awareness of the beneﬁts of direct farmer participation in the selection of on-farm trial
factors like plant varieties. For example, Witcombe (2002) maintains that conventional
breeding approaches were successful in providing rice varieties suitable to low-input
production systems that have been widely distributed. However, “it is certainly true that
the rate of development and adoption of improved cultivars has been lower in rainfed

lowland than in irrigated environments, and that adoption of improved upland varieties

13

has been minimal outside of very favourable environments.” He adds that participatory

variety selection and participatory plant breeding have begun to address these issues.

Participatory variety selection has also been utilized for common beans. Rosas et a1.
collaborated with farmers in Honduras to select segregating plant material, allowing
farmers to use their own selection criteria, and concluded that “a veces es dificil entender
la complejidad de ana'lisis que emplean los agricultores para tomar decisiones. For 10
cual, el incremento de las interacciones con los agricultores es ﬁmdamental para lograr
mayor adopcio’n de variedades mejoradas” [“sometimes it is difﬁcult to understand the
complexity of analysis that farmers use to make decisions. That is why increasing
interactions with farmers is fundamental to achieving greater adoption of improved

varieties.”] (Rosas et al., 2003).

One organizational entity that was born from the participatory development movement
and is compatible with constraints research is the Local Agricultural Research
Committees (LARC, Comité de Investigacion Agricola Local in Spanish). In short, a
LARC is a "farmer-run research service that is answerable to the local community"
(Ashby et al., 2000). LARCs have many advantages to offer constraints research. Since
LARCs are local, they are responsive to local agro-ecological conditions, cultural
preferences, and economic constraints of local people. LARCs reduce the level of
intensity of extension because farmers can test new technologies themselves; and LARCs

can be used to gather information about community needs.

14

LARCs are formed in a community meeting, where community members decide what
problems are most important. Farmers volunteer to work with an extensionist to learn to
design and implement experiments to test new technologies relevant to the problem
identiﬁed by the community. For farmers to execute formalized research requires a
period of training, which includes eight steps. These steps are reiterative: they require a
return to the “Motivation” stage once the process reaches the “Feedback” step. These
steps are Motivation, Election, Diagnosis, Planning, Experimentation, Evaluation,
Analysis, and Feedback (Figure 3).

Figure 3. Diagram of the eight steps in the reiterative process used for organizing a Local
Agricultural Research Committee (LARC).

 

 

 

 

 

  
 

 

 

 

 

 

 

 

 

 

 

Feedback
Analysis
Evaluation
. . Training,
Experimentation Monitoring
Planning and -
Evaluation
Diagnosis
Election
Motivation

 

 

 

 

Source: Humphries et al., 2000
According to Ashby et a1. (2000), the capacity to manage all steps of the process require

at least four cycles of diagnosis, planning, evaluation, analysis and providing feedback to

15

the community; and each cycle includes three phases or experimental trials: a

preliminary/exploratory trial, a veriﬁcation trial, and a production trial.

One criticism of participatory research is that it is not the most efficient way to advance
agricultural research or provide technologies to farmers in a cost-effective way; one
reason for this is that farmer participatory research often serves “certain social and
development goals” (Sumberg and Okali, 1997). ‘ In fact, LARC practitioners on
Colombia have identiﬁed beneﬁts to LARC membership that go beyond technical help.
Ashby et al. (2000) include the following as the possible beneﬁts of LARCs in an

examination of these organizations around Colombia:

Increased local capacity

Improved local planning, management and organizational skills
Higher crop yields

More local experimentation

More experimentation with soil conservation practices

Higher biodiversity in cropping systems

Improved access to credit

Greater availability to improved seed

Improved food security

Establishment of small enterprises

Increased social status of women and other marginalized groups.

Improved access to formal research services and products.
Source: Ashby et al., 2000

Snapp et al. (2003) add that when farmers participate in research, there is greater
accountability and client orientation; and that participation can “address complex training
requirements and organizational linkages.” Similarly, the Centro Intemacional de

Agricultura Tropical evaluated 68 LARCs in 1998 and found that

(a) [LARCs] directly resulted in more rapid technology adoption; (b) the [LARC] process itself
has led to people learning useful skills and forming valuable social linkages; and (c) [LARC]
communities had experienced improvements in welfare. Welfare improvements came about partly
by people starting agro-enterprises based on the results of the experiments and the new skills and
linkages they had developed. For example, some [LARCs] have started to commercially produce

16

the seed of the best crops identiﬁed in their trials. [LARC] communities have been encouraged by
the experimentation to try more new crops, and as a result have more crops and more varieties in
their ﬁelds, than farmers in similar villages without [LARCs]. This diversity helps villages with
[LARCs] to better cope with risk. Moreover, the speed of adoption of technologies was faster in
villages with [LARCs], and the poorest strata of farmers were just as likely to adopt as the richer
strata. Therefore, [LARCs] help communities beneﬁt faster from improved varieties, whether
developed by the formal research sector or farmers themselves. The study also suggests that
[LARCs] may improve food security because farmers in villages with [LARCs] reported fewer
“hungry months” of seasonal food shortage.

Source: Quiros Torres et al., 2004

Humphries et a1. (2000) worked with LARCs in Honduras and found that “participatory
research using formal methods was not found to be sufficiently engaging for very poor
farmers whose primary concerns were more likely to be achieving immediate household
food security.” In light of this, LARCs in Honduras came to emphasize functions in
addition to participatory experimentation, such as building social capital (managing
funds, learning time management, increasing comfort with public speaking, identifying

problems in the community), and providing loans to members and non-members alike.

2.4 Survey Design and Implementation

Surveys use two basic methods to describe human behavior and its results. The ﬁrst type
of method is the quantitative method. Quantitative methods are good for explanation,
description, or evaluation, and especially the development of an understanding that
accurately reﬂects what is happening in the world. This method generally yields
countable things, such as the amount of beans produced in a particular plot. The second
is the qualitative method. Qualitative methods are good for exploration, developing an

“authentic” understanding of a social process or social setting. Qualitative data may

17

describe how a farmer feels about a farming practice, or explain why a farmer chooses to

sow one variety over another‘ (Schutt, 2004).

Sallant and Dillman list four requirements for an accurate survey:

1. The sample is large enough to yield the desired level of precision.

2. Everyone in the population has an equal (or known) chance of being selected
for the sample.

3. Questions are asked in a way that enables people in the sample to respond
willingly and accurately.

4. The characteristics of the people selected in the sampling process but who do

not participate in the survey are similar to the characteristics of those who do.
Source: Sallant and Dillman, 1994

These requirements must be fulﬁlled so that the experience of a few farmers in the survey

can be generalized over a larger population.

Related to accuracy is error, speciﬁcally the minimization of error in the design,
implementation and analysis of survey data. Errors can fall into one of four possible

categories:

1. Coverage error — target population is not represented completely.

2. Sampling error — statistical error, and can be measured. If the sampling error is
too large, the results may be considered inconclusive.

3. Measurement error — difference between respondent’s answer and the “correct”
answer. This happens when the tool used does not measure what we intend it to
measure.

4. Non-response error -— selected people do not answer and their answers would
have been different from others.

Source: Sallant and Dillman, 1994

One might also include causal validity error: the incorrect conclusion that A leads to B,

that one event has caused another (Schutt, 2004). These considerations must be taken

 

' It should be noted, however, that although qualitative data are often descriptive, they can be quantiﬁed,
i. e. coded.

l8

into account for survey data to be precise and free of bias, and to be able to make

generalizations about a population with similar characteristics.

2.5 Development Trends in Honduras

In 1821 , following the Honduran war for independence, productive land was distributed
to military heroes, and much of this land lay idle (Map of Honduras, Figure 4). In
response to the growing number of impoverished farmers who did not have their own
land but desired it to provide basic staples (i. e. maize and beans) for their families, the
Honduran government passed the Agricultural Reform Law in 1962. The principle of
this law was that land was for the people who used it, and idle land was deﬁned as land
that was not currently in use, but also rented land, sharecropped land, and land farmed by
tenant farmers. The Agricultural Reform Law put restrictions to how much land one
family could own and the National Agricultural Institute (Instituto Nacional Agrario) was

formed to enforce it (Thorpe et al., 1995).

19

Figure 4. Map of Honduras.

 

V

B [IA CWbGM Sea
,4:-
GUA
HONDURAS
EL

. ; VADOR

NICARAGUA
Pacgﬂc Ocean %

‘k—l

 

 

 

While agrarian reform moved in ﬁts and starts, most believe that agricultural reform
contributed to political stability. However, little by little the Agricultural Reform Law
lost ground. In 1963, a coup d’e’tat brought in a military regime that repressed farmer
organizations and practically did away with the National Agricultural Institute. By 1975,
a second land reform law was passed, one that supported the formation of farm
cooperatives that required that farmers possess land in common. There was some
redistribution of land, but even these intended reforms were abandoned by 1978 under
new military rule (Franco, 2003). By the 1980s, since governments decided that the
threat of land redistribution was preventing larger landowners from making long-term
investments in their land, it adopted a policy of more intensive land titling of government

and communal land. Thus, most of the basic tenants of land reform were abandoned.

20

In the 19803, Honduras suffered from high ﬁscal deﬁcits that were ﬁnanced externally,
and the balance of payments slowly deteriorated (International Monetary Fund and
International Development Fund, 2001). Finally, in 1990, the government passed the
Agricultural Modernization Law (Ley de Modernizacion Agricola), part of a larger
stabilization program (see a summary of Honduras’ structural adjustment program,
Appendix 1). The government implemented a neo-liberal model that considered previous
agricultural reform as a drag on social and economic development. The new model also
presumed that agricultural reform generated insecurity related to land ownership and

exaggerated the dependence of the farmer on the state (Thorpe et al., 1995).

The program began to correct the internal and external balances. One component was the
liberalization of the exchange rate; related is trade liberalization. Because the exchange
rate appreciation decreased the prices of imported goods, it reduced the rate of inﬂation.
This was paired with the elimination of price supports for basic grains, like beans, in
favor of a price bandz. Large farmers with better access to markets and farmers close to
large towns beneﬁted from the improved prices, whereas farmers ﬁirther from the
markets were not able to capture this guaranteed price because they sold to intermediaries

who paid under this price. Of the basic grains, prices for beans fared the best (Table 3).

 

2 The price band system uses international grain prices and import tariffs to ﬁx ceiling and floor prices.

21

Table 3. Real producer prices for basic grains, 1989-1996 (Lempiras/hundredweight).
Year Avera e
Crop 1989 1990 1991 1992 1993 1994 1995 1996 1991-96
Maize 86.2 114.8 99.0 104.7 104.0 131.0 121.5 122.0 113.7
Beans 68.8 111.5 119.7 67.0 159.8 149.0 128.8 129.7 125.7
Sorghum 79.6 1 18.5 101.9 70.3 74.7 75.3 67.1 65.6 75.8

Rice 87.1 110.2 102.8 101.6 103.3 129.6 118.3 116.9 112.1

Note: 10.09 Lempiras (Lp) = 1.00 US. Dollar (January 1, 1996. 1989 prices are used as an indicator of the
immediate pre-adjustment producer price; the effects of liberalization become apparent in 1990.

Source: Thorpe, 2002

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

While the liberalization of domestic trade, another component of the modernization
program, was expected to help basic grain producers, it did not increase production and
productivity, probably due to lack of access to production inputs and the real appreciation

of the national currency.

All in all, trade liberalization and market deregulation policies “were not associated with
increases in net exports of agriculture in Central America. The main result would appear

to have been import stimulation, with weak effects in exports” (Weeks, 1999).

One key structural factor of Honduran bean markets is the availability of credit to
farmers. Thorpe (Thorpe, 2002) reports that early national banks created to help
agricultural producers favored larger producers. Reforms created to correct this
disequilibrium (like rural credit desks and land banks) have not been effective in reaching
producers due to inadequate ﬁmding and poor management. In the bean sector, credit
disequilibrium has reinforced distinct market channels, with heavily ﬁnanced larger
growers producing more for national and international markets and for value added uses,

and small growers producing for local and regional markets. In addition, national banks

22

have recently pardoned the repayment of loans to farmers, equaling a subsidy to farmers

able to secure loans, i.e. large and medium-sized farmers (n.a., 2003b).

Discussing the absence of a Green Revolution in Sub-Saharan Africa, Lynam notes that
other preconditions are necessary to facilitate such a rapid increase in yields.
These preconditions, such as markets, credit, extension, and information systems,
are developed outside the research sector itself and together with land/labor ratios
and rural population grth deﬁne the stage of crop yield development.
Agricultural research cannot substitute for these larger systems but rather research
must adapt its design strategies to each stage. The important point is that the
relative importance of breeding versus crop and natural resource management
strategies changes at each of the stages, with breeding generating its huge
potential only when crop management systems and larger support systems are
relatively well developed. While this would seem a rational argument, the other
driver in the organization of agricultural research is advances in the biological
sciences. The international agricultural research system, as an example, has
responded to the pull of these two forces in the organization of its research.
Source: Lynam, 2004

To support the development of small-scale common bean farmers in Honduras, the

government intends to support access to loans and inputs for small-scale farmers, even in

basic grain crops; and increase access to reliable information through effective extension

and learning opportunities.

Land tenure also plays a role in ﬁnancing. With a farm title as collateral, lending
institutions are more willing to grant loans. As mentioned above, land reform has been
on the national agenda since the 19603, yet actual land titling has not reached many
farmers. The Global Campaign for Agrarian Reform (F IAN and La Via Campesina,
2000) reports that “according to the ofﬁcial statistics on agriculture more than 126,999
peasant families have neither access to land nor a secure place of employment. This

ﬁgure represents 27% of the rural population. If we add to this group the 80,000 peasant

23

families (at least), who own less than 1 hectare of land, we are confronted with the
disturbing ﬁgure of more than 200,000 families, or 44% of the rural population who have

either no or very limited access to land.”

Thus one way to increase the likelihood of borrowing is to complete land titling. In
circumstances where farmers need collateral for a loan, land titles can serve to monetize
what would otherwise be considered “dead capital” (Soto, 2000). When farmers are
assured that the land they are working will stay in the family, they are also more likely to
invest in that land, including making long-term investments in land architecture (low-cost

terracing, fencing) and more crop fertilizers.

Another possible way to release this constraint is to encourage the development of local
banks. Locally-managed savings and loan cooperatives have been shown to increase the
propensity of borrowing and stimulate local economies (Durm and Arbuckle, 2001; Jain,
1996), and sometimes work to conserve natural resources (Anderson et al., 2002).
Successes with microcredit has been documented in Honduras, including the Caja Rural
de Ahorro y Cre'dito (CRAC, Rural Credit and Savings Bank in English) model
(Economic Commission for Latin America and the Caribbean, 1999). The grth of
microcredit enterprises in rural areas will also likely increase borrowing in rural

communities.

Encouragement of the purchase of agricultural inputs may also be stimulated by the sale

of these inputs at the moment when crops are being sold. To determine whether

24

improved seed and fertilizer use could increase yield economically, an NGO in western
Kenya worked intensively with randomly selected farmers in three communities over six
years to plant small plots of the improved seed with fertilizer or fertilizer only to compare
to a traditionally managed plot (Duﬂo et al., 2004). Increased fertilizer use had positive
farm returns (231% over six seasons with small amounts of top dressing, and less but

positive returns with greater amounts of fertilizer).

The authors note that experimentation with an unproved technology could cause some
time and ﬁnancial losses in the ﬁrst year. They continue, "[i]f a farmer knows that after
this initial investment, he will only be able to use fertilizer on a very small scale for a
long period of time (because he cannot have access to the funds to do it on a large scale),
he might rightly consider it worthless to do the initial investment to master the
technology" (Duﬂo, 2003). In their study, the NGO assumed the ﬁnancial risk for the
farmers, and helped them analyze the results. Therefore, if resistance to experimentation
or incomplete understanding of the results were the main barrier to adoption, the

participants would show an increased use of fertilizer.

In fact, 37% of participants did use fertilizer, whereas only 20% of those farmers who
were not chosen to participate used fertilizer in a given season. The question remains:
why did the other 63% of participants not use fertilizer, even in small quantities on small

part of their ﬁeld?

25

When asked why farmers did not use fertilizer, 98% responded that they had no money.
The authors hypothesized that the farmers found it difficult to save or borrow even the
small amount of money needed to start using fertilizer. To test this hypothesis, the NGO
arranged for farmers to buy fertilizer immediately after the maize harvest, which the
NGO would store for them, or deliver right away. This offer doubled the rate of fertilizer

usage.

What explains this behavior? The authors comment,

We are far from understanding fully why this program was so successful. Did
farmers value the opportunity to commit their money to be used for fertilizer
purchase because they knew that if they kept the money they would be tempted to
use it on other things before the time for purchasing and using the fertilizer occurs
(Laibson, 1991)? Are they particularly keen to exchange cash for fertilizer
because maize is not likely to disappear quite as rapidly? Are they protecting the
cash against themselves (they know now that they need to tie their hands), or are
they worried about their families or their neighbors? Were farmers just lured by
fertilizer as one possible purchase, and would they have been just as happy with a
radio or some alcohol? Were farmers particularly aware of the value of fertilizer,
after we had just weighed and valued the output of the two plots, and would this
have lost its salience if they had waited longer?

The authors turn to three tenants of “behavioral economics” to explain how humans differ
from the unboundedly rational, forward-looking, and internally consistent model that
neo-classical economics proposes. The ﬁrst is the limited ability of humans to analyze,
compute and remember information. The second, a bounded willpower: humans do not

always make choices that are in their best long-tern interest. Third, human decisions are

not purely self-interested.

26

That said, it is further recognized that being poor (and near poor) affects the way people
think and make decisions. The way people think when considering the survival of their
family, social norms and other stresses is possibly radically different depending on their

socio-economic status.

In addition to national tenure changes, the Agricultural Modernization Law repealed the
ban against joint ventures with ﬁrms in other countries. From 1990-1995, joint ventures
generated Lp 30.5 billion (U S$ 6.1 billion in 1995), and eight of 99 joint ventures

included the production of basic grains like beans (Thorpe, 2002).

Although farmers cannot easily get titles to land, opportunity for the use of communal
land and renting land are present. Poor farmers may produce a small plot of basic grains
primarily for consumption but with the intention of selling excess; which represents a
diversiﬁcation strategy that reduces risk in the context of a livelihood. Higher investment
in production technologies (e.g. more fertilizer, herbicide), however, is a barrier for
small-scale bean farmers. Poor soils, lack of appropriate varieties and lack of production
information increase risk, which most farmers are not willing to take; the inability to
secure credit, as mentioned above, is a further limitation to growth into more proﬁtable
bean production. While cooperatives have been in existence since the 19703, they are
generally ineffective due to an inﬂexible structure that does not allow them to adjust to

changing markets (Thorpe, 2002).

27

Recently, the Honduran government acknowledged that to further stimulate the economy
and address rural poverty, it must help small farmers. One attempt was to invigorate the
production of agriculture for small and mid-sized farmers was the Farm Sector Solidarity
Law of 2001 (Ley de Solidaridad con el Productor Agropecuario). This law allowed the
reﬁnancing of debts and access to new loans. However, the law was ignored by farmers
who got better debt readjustment deals and loans under other programs (n.a., 2003a).
Additionally, the most marginalized farmers do not have access to loans due to lack of

access to banks and lack of effective collateral (land titles) (Thorpe, 2002).

Some attempts to stimulate the economy through the agricultural sector encouraged the
Honduran government to adopt policies designed to increase productivity while reducing
exchange and marketing costs. These policies were based on the principles of structural
transformation (Staatz, 1994). World Bank suggestions to the Honduran government
encourage this kind of development, namely investment in agriculture in areas where it is
most proﬁtable and the improvement of access to market information (Appendix 2). The
government of Honduras seems to be adopting this model. This model results in a move
away from basic grains and toward higher value products. While productivity and
proﬁtability for some farmers may increase, its effect on the bean market is not expected

to be signiﬁcant.

28

In 2003, the Vice-minister of Agriculture outlined three steps for eradicating rural
poverty by 2021 through the stimulation of the agricultural sector (Vasquez, 2003). First,
the govermnent will adopt policies that strengthen agricultural development, including 1
popular participation, stronger public sector agricultural services, better links with
institutions, resolution of some land issues and credit availability3 . Farmers from the
bean sector stand to gain from these improvements. In addition, the government will
execute the Central American Common Market (Union Aduanera Centroamericana).
Since the signing of the General Treaty for the Economic Integration of Central America
(T ratado General de Integracion Econo'mica Centroamericana) in 1960, the economic
union of Central America has evolved into a block of countries that wish to establish a
common market and a free trade zone, create a common trade policy with countries
outside of the zone (Superintendencia de Administracion Tributaria, 2003). Since the
stability of the bean market will then depend on the agricultural policies of the Common

Market, farmers have reason to follow its development.

Second, it plans to approve and ratify the Central American Free Trade Agreement
(CAFTA-DR, Tratado de Libre Comercio de Centro Ame’rica in Spanish), institute an
“intelligent market” and complete land titling. In fact, Honduras signed CAFTA-DR in

May 2004, which its congress ratiﬁed CAFTA-DR in March 2005 (n.a., 2005).

 

3 In reality, few effective steps have been taken by the Honduran government to make credit available to
small farmers. For example, the Honduran government has pardoned Lp 2.5 billion in loans to farmers, Lp
60 million of loans from the national agricultural bank BANADESA (Banco Nacional de Desarrollo
Agricola in Spanish), and the rest ﬁ'om private banks (n.a. 2004. Gobiemo Ha Destinado 2,200 millones a1
Agro, pp. 20 El Heraldo, 7911 ed, Tegucigalpa. Honduras). Since banks do not land to farmers without
collateral, and whose programs do not reach rural farmers effectively, the pardoned loans amount to a
subsidy for medium and large farms, putting small farmers at a disadvantage.

29

The third step is the reduction/elimination of tariff barriers, compliance with international
agricultural export standards, and the formation of producer groups called clusters. This
policy is considered unrealistic by farm groups (V asquez, 2003), and again, it represents

another change in policy that will threaten the markets for basic grains including beans.

30

CHAPTER 3

RESEARCH JUSTIFICATION AND OBJECTIVES

Common bean producers can be divided into categories that represent unique agro-
ecological areas, and working with producers in low rainfall areas in the central part of
Honduras can yield results that may be used to generalize for producers in a broader

ecological region.

3.1 Justiﬁcation

Based on the data reviewed in Chapter 1 (Figures 2 and 3), three rainfall and two altitude
regions are identiﬁable. The northwest (Copan) region represents bean producing farms
with intermediate rainfall (1267-1588 mm). Since farmers in this region grow black
beans and production is more similar to Guatemala, the breeding program would not
beneﬁt from research done there. The extreme southwest (Lempira) region with high
precipitation (1589-2533 mm) has the highest concentration of impoverished farmers.
However, it is very isolated and does not have a system of Local Agricultural Research
Committees (participatory research committees that are trained to be collaborators), and it

is currently heavily supported through FAO programs.

Most of the remaining small-scale bean farmers are found in areas with 681-1266 mm of
precipitation; and their farms are distributed over many altitudes. In order to assess
production constraints in this area in 2003 and 2004, research trials were conducted in the

community of Los Lirnones because it represented farms 0-900 meters above sea level

31

(m.a.s.l.). In 2004, trials were conducted in Lavanderos because it represented farms
from 900-1400 m.a.s.1. The trials were focused on bean production in the postrera season
because around 67% of bean production occurs in this season (Secretaria de Planiﬁcacion

Coordinacion y Presupuesto, 1993).

In addition to rainfall and altitude, the presence of Bean Golden Yellow Mosaic Virus
(BGYMV) could be included as an additional criterion. Although farms below 1000
meters above sea level are generally considered effected, the actual presence of BGYMV
must be veriﬁed in the ﬁeld, and interpretations must be sensitive to the presence of

BGYMV.

3.2 Research Objectives
Based on the current limitations to increasing proﬁtability for small-scale producers and
the analysis of the characteristics of those same producers, the following objectives are

proposed:

Objective 1: To quantify the socio-economic limitations to increasing bean yields with

the pr0posed technology.

Objective 2.: To determine which bean management factors can be proﬁtably changed in

small-scale farmer cropping systems in the postrera season.

32

Objective 3: To evaluate the effectiveness of the constraints methodology and the Local

Agricultural Research Committee in order to satisfy objectives one and two.

33

CHAPTER 4

RESEARCH METHODS

Surveys were taken in 2003 and 2004 in 6 communities in rural Honduras with small-
scale producers to identify agronomic, economic and social constraint to production; as
well as the beneﬁts of participation in Local Agricultural Research Committees.
Participatory ﬁeld research in two communities representing two altitudes focused on
fertilizer and common bean variety to examine possible increases in production, and
weed control to examine savings from ﬁeld labor. Farmers were asked to assess bean

production in experimental plots to examine accuracy of farmer observation.

4.1 Survey Methods

Farmers were surveyed in Los Lirnones in 2003 and in Los Lirnones and ﬁve other
communities in 20044 (Table 4). The surveys conducted in 2003 were designed to
describe community households, identify the diversity of farming practices, and identﬁy
farmer perceptions regarding constraints to bean production in the community of Los
Limones (Appendix 3); the 2004 surveys were designed to describe community
households, identify farming practices and costs of production, identﬁy farmer
perceptions regarding constraints to bean production, and discover farmer-identiﬁed
advantages to belonging to a LARC (Appendix 4 and 5). All data collection was

analyzed using SPSS (Statistical Package for the Social Sciences), version 13.0.

 

‘ Communities were chosen based representation of an agro-ecological zone of interest, and on whether the
community had a Local Agricultural Research Committee (LARC) to examine the effect if LARC
membership on farming practices and proﬁtability.

34

Table 4. Surveys implemented in six communities in the department of El Paraiso,
Honduras, 2003 and 2004.

 

 

 

 

 

 

 

 

 

 

2003 2004

LARC* Non-LARC LARC Non-LARC
Community member member member member
Los Lirnones 6 28 6 16
Lavanderos - - 5 4
Tabla Grande - - 4 2
Los Pozos - - 7 2
Mesias - - 3 6
Carrizal - - 5 6
TOTAL 6 28 30 36

 

 

 

 

 

 

‘LARC = Local Agricultural Research Committee

In Los Lirnones, survey participants were chosen by asking key informants to list all
postrera bean farmers, then trying to contact everyone on that list over a period of three
or more days (December 2003 and December 2004/January 2005). Since farmers absent
from the community were not contacted (7 of 41 in 2003, 17 of 39 in 2004), it is possible
that the few farmers not contacted constituted a unique subset of bean farmers (for
example, farmers who found off-farm work to supplement farm income). However, key
informants believed that this was not the case. Survey participants in other communities
were selected due to their participation in the LARC groups, and the direct neighbors of

the LARC members. All but ﬁve participants were male.

According to Sallant and Dillman (1994), there are four requirements for an accurate
survey. The ﬁrst is that the sample is large enough to yield the desired level of precision.
The total sample size was 66, enough for a 95 percent conﬁdence level with i] 0%

sampling error for a 80/20 split on the characteristic in question. A sampling size of 96 is

35

required for a 50/50 split for the same level of conﬁdence and sampling error, which was

not achieved with this survey (Sallant and Dillman, 1994).

The second requirement is that everyone in the population has an equal (or known)
chance of being selected for the sample. An attempt was made to contact all members of
the community of Los Lirnones, which would represent the entire population. Members
of other communities were chosen due to their LARC membership, and all of these
members were chosen; and so were their direct neighbors, to strive for geographical
homogeneity. Randomness of selection was rejected in favor of a reduction in bias due to
geographically determined access to resources, and comparisons within the data set offer
no signiﬁcant differences between LARC members and non-members for the key socio-

economic characteristics determined.

The third requirement was that questions be asked in a way that enables pe0ple in the
sample to respond willingly and accurately. Questions were pre-tested for clarity with
farmers that collaborate with the Pan-American School of Agriculture, and modiﬁed

before administration to the format present in Appendixes 3, 4 and 5.

The fourth requirement is that the characteristics of the people selected in the sampling
process, but who do not participate in the survey, are similar to the characteristics of
those who do. All of the selected participants agreed to participate fully and none refused

to answer a question.

36

4.2 Economic Analysis

Data from the agronomic and survey portions of the study were used for an economic
analysis. First, since the proposed common bean recommendations require only small
management changes, production data was used to construct partial proﬁt budgets
according to Dillon and Hardaker (Dillon and Hardaker, 1993). The necessary data about

prices was taken from the survey data.

Second, the survey data was used to identify relationships between high common bean
yields and speciﬁc management practices; and to examine the relationship between yield
and proﬁtability. All regression statistics were calculated using SAS 9.1 (SAS Institute

Inc., 2003).

4.3 Experimental Field Plot Methods

The agronomic experiment was designed to identify what physical management factors

contribute to the yield difference between farmer yields and experiment farm yields (the
phenomenon called “yield gap”) based on the data collected in the survey. Farmer-level
and recommended-level inputs were used as test factors. A typical experimental design

for a yield gap experiment is a factorial design with three levels:

1) The complete factorial experiment permits analysis of the interactions between

all test factors at farmer and recommended levels. This identiﬁes the interactions

of all combinations of factors.

37

2) The mini-factorial experiment permits analysis of the interaction of practices
where all but one practice is at recommended levels. This allows the
identiﬁcation of the contribution of test factors when we expect the absence of
one single factor to be signiﬁcant.

3) The supplemental experiment permits a comparison of farmer and
recommended practices and simply allows the measurement of the yield gap.

(Barker, 1979)

In Honduras, variety selection, fertility management, and weed management are three
very signiﬁcant farm practices that impact bean yields. Disease management was not
included as a factor because farmers typically use fungicide ineffectively (farmers apply
fungicide after the disease is well established in the ﬁeld, and they often use fungicides
that are not appropriate for the disease present). Insect management was not included
because the producers seemed content with their control methods (despite that they did
not follow university recommendationS), and generally do not have a great impact on

bean yields when managed by farmers.

Weed management was not included as a production constraint factor in 2004 for two
reasons. First, the recommended weeding practices did not contribute to yield differences
in trial plots in Los Lirnones in 2003; Secondly, farmers typically applied herbicide
before plowing, which adequately controlled weeds at least until the bean seedlings
emerged. Since this was the time when herbicide application was recommended, it was

deemed unnecessary, and was excluded from trials in 2004.

38

4.3.1 Experimental Plot Locations
The experiments were executed in the community of Los Lirnones in 2003, and in Los
Lirnones and in Lavanderos in 2004 (Table 5).

Table 5. Locations of communities where agricultural research and surveys were
im lemented, Department of El Paraiso, Honduras, 2003 and 2004.

 

 

 

 

 

 

 

Community Municipality Altitude (meters Latitude Longitude
above sea level)
Los Limones Moroceli 700 14° 5’ 54” -86° 50’ 10”
Lavanderos Gilinope 1,380 13° 56’ 20” -86° 56’ 38”
San Antonio

Tabla Grande de Oriente 1,240 13° 59’ 39” -86° 55’ 38”
Los Pozos Moroceli 860 14° 8’ 36” -86° 49’ 29”
Mesias Moroceli 1,100 14° 7’ 9” -86° 47’ 22”
Carrizal Moroceli 900 14° 12’ 10” -86° 50’ 26”

 

 

 

 

 

 

 

4.3.2 Experimental Plot Design

4.3.2.1 Los Lirnones Plots 2003

On four farms, the eight treatments were installed with four replications in a randomized
complete block design (Table 6). On ﬁve other farms, the supplemental experiment plots
were installed with only one replication. The recommended management package
corresponded to treatment number one and the farmer practices corresponding to
treatment number eight. Plot size was three meters by ﬁve meters, with 50 cm between
rows, and 10 cm between plants in the row. Soil pH was calculated with a 1:1 soil to
water ratio; percent soil organic matter (%OM) with Walkley and Black (Walkley and

Black, 1934); percent soil nitrogen (%N) as 5% of %OM, soil phosphorus (P) and

39

potassium (K) with Mehlich-3 extraction (Mehlich, 1984). These results are listed in

Table 7.

Table 6. Management components for full factorial trials, Los Limones, Moroceli,

Honduras, Postrera, 2003.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Treat- Common Fertilizer Fertilizer Weed management
ment bean variety type quantity (kg/ha) strategy
1 Amadeus-77 18-46-0 130 One chem. application
urea 32.5 and by hand
2 Amadeus-77 18-46-0 130 By hand only
urea 32.5
3 Amadeus-77 18-46-0 65 One chem. application
and by hand
4 Amadeus-77 18-46-0 65 By hand only
5 Seda“ 18-46-0 130 One chem. application
urea 32.5 and by hand
6 Seda 18-46-0 130 By hand only
urea 32.5
7 Seda 18-46-0 65 One chem. application
and by hand
8 Seda 18-46-0 65 By hand only

 

*Also known as “Rojo de Seda”

Table 7. Experimental plot soil type, pH, percent organic matter (%OM), percent
nitrogen (%N), Mehlich phosphorus (P) and Mehlich potassium (K) for nine plots, Los
Lirnones, Moroceli, Honduras, 2003.

 

 

 

 

 

 

 

 

 

 

 

Farmer Soil type pH* %OM %N PM K***
(ppm) (ppm)
1 Sandy loam 6.3 1.6 0.08 7 190
2 Clay loam 5.9 2.2 0.11 51 180
3 Sandyloam 6.0 1.3 0.07 l 16 178
4 Sandy loam 6.4 3.2 0.16 4 216
5 Loam 6.3 3.8 0.19 29 534
6 Loam 5.8 2.7 0.13 25 264
7 Sandy loam 5.7 1.9 0.10 25 172
8 Loam 6.4 1.8 0.09 4 278
9 Sandy loam 5.8 2.3 0.11 11 250

 

 

 

 

 

 

 

 

"‘ Normal range 6.0 to 6.8

" Normal range 13 to 20 ppm.
"‘ Normal range 98 to 195 ppm.

40

 

Researchers and farmers chose the research sites together, taking care that the plots were
as homogenous and free from border effects as possible. Planting dates were chosen by
the farmers, and farmers and researchers staked and planted the seed and applied the
NPK fertilizer together. Researchers applied 0.5 liters per hectare of paraquat herbicide
to herbicide treatment plots after sowing. Farmers undertook the daily management of
the plots, including the subsequent weeding (usually at 20 days post germination), urea
(46-0-0) application (concurrent with weeding, spread next to plants and mounded up
over bean row), disease and insect management. Plots were visited at peak ﬂowering to
determine plant population (two replicates of plants per 1 meter of row), disease and
insect damage (0-9 scale), and weed biomass with representative cuts (two replicates of
0.5 m2 per plot). In the complete factorial plots, rain gauges were installed and rainfall
data were collected every week for 60 days starting at planting. At harvest, two rows of
beans were harvested (two meters each), beans hand-threshed, analyzed for moisture
content and weighed, and the yield was calculated at 13.5% moisture. Farmers received

bean seed to replace the beans that were harvested from the ﬁeld.

4.3.2.2 Los Limones Plots 2004

On three farms, the 12 treatments were installed with 3 replications in a randomized
complete block design, representing the full factorial design (Table 8). In four other
farms, the 12 treatments were installed with only one replication, allowing for a spatial
5

comparison of all treatments as well as a traditional versus recommended comparison .

Plot size was 3 meters by 5 meters, with 50 cm between rows, and 10 cm between plants

 

5 Of the four unreplicated plots, two were lost due to insufﬁcient rainfall.

41

in the row. As in 2003, researchers and farmers chose the research sites together, taking
care that the plots were as homogenous and free from border effects as possible. Planting
dates were chosen by farmers, and farmers and researchers staked and planted the seed
and applied the NPK fertilizer together. Farmers undertook the daily management of the
plots, including the subsequent weeding (usually at 20 days post germination), urea (46-
0-0) application (concurrent with weeding, spread next to plants and mounded up over
bean row), disease and insect management. Replicated plots were visited at peak
ﬂowering to determine plant population (two replicates of plants per 1 meter of row) and
disease incidence (0-9 scale). Rain gauges were installed and rainfall data were collected
every week for sixty days starting at planting. At harvest, two rows of beans were
harvested (two meters each), beans hand-threshed, analyzed for moisture content and
weighed, and the yield was calculated at 13.5% moisture. Farmers received bean seed to
replace the beans that were harvested from the ﬁeld. Plot soil characteristics were not

reevaluated, and plots in 2004 correspond to the numbers listed in Table 6.

42

Table 8. Management components for full factorial trials, Los Limones, Moroceli,
Honduras, 2004.

 

 

 

 

 

 

 

 

 

 

 

 

 

No. Common bean Fertilizer type Fertilizer quantity
variety (kg/ha)
l Seda 18-46-0 65
2 Seda 18-46-0 65
urea 32.5
3 Seda 18-46-0 130
4 Seda 18-46-0 130
urea 65
5 Marciano 1 8-46-0 65
6 Marciano 1 8-46-0 65
urea 32.5
7 Marciano 1 8-46—0 l 30
8 Marciano l 8-46-0 1 30
urea 65
9 Amadeus-77 1 8-46-0 65
10 Amadeus-77 18-46-0 65
urea 32.5
1 1 Amadeus-77 1 8-46-0 130
12 Amadeus—77 1 8-46-0 1 30
urea 65

 

 

 

 

 

 

4.3.2.3 Lavanderos Plots 2004

A full factorial experiment was performed in Lavanderos in 2004 to permit comparisons
with the plots in Los Limones (Table 9). On one farm, the two experiments were
established with three replications in a randomized complete block design; on another
farm, the experiments were installed with two replicationsé. Planting dates were chosen
by farmers, and farmers and researchers staked and planted the seed and applied all
fertilizer except urea together. F arrners undertook the daily management of the plots,
including the subsequent weeding (usually at 20 days post germination), urea (46-0-0)

application (concurrent with weeding, spread next to plants and mounded up over bean

 

6 Experiments were sown with a total of six farmers from Lavanderos, but three were lost due to excessive
rainfall which caused seedling death from Rhizoctonia root rot, and one was planted very late in the season
with irrigation thus inappropriate for comparison. Three replicates were planned, but the second farmer did
not have enough land for a third replicate.

43

row), disease and insect management. Urea was applied by farmers and researchers
together at the ﬁrst and only weeding, at about 20 days after germination. Plots were
visited at peak ﬂowering to determine plant population (two replicates of plants per one
meter of row) and disease incidence (0-9 scale). Rain gauges were installed and rainfall
data were collected every week for 60 days starting at planting. At harvest, two rows of
beans were harvested (two meters each), beans hand-threshed, analyzed for moisture
content and weighed, and the yield was calculated at 13.5% moisture. Farmers received
bean seed to replace the beans that were harvested from the ﬁeld. Plot size was three
meters by ﬁve meters, with 50 cm between rows, and 10 cm between plants in the row.
Soil pH calculated with 1:1 soil to water ratio; percent soil organic matter (%OM) with
Walkley and Black (Walkley and Black, 1934); percent soil nitrogen (%N) as 5% of
%OM, soil phosphorus (P) and potassium (K) with Mehlich-3 extraction (Mehlich,
1984). Results listed in Table 10.

Table 9. Management components for factorial trials, Lavanderos, Giiinope, Honduras,
Postrera, 2004.

 

 

 

 

 

 

 

 

Fertilizer quantity Common bean
No. Fertilizer type (kg/ha) variety

4 18-46-0 65 Tio Canela

5 18-46-0 65 T10 Canela
urea 32.5

6 1 8-46-0 65 Amadeus-77

7 1 8-46-0 65 Amadeus-77
urea 32.5

8 1 8-46-0 65 Paraicito

9 1 8-46-0 65 Paraicito
urea 32.5

 

 

 

 

 

44

Table 10. Experimental plot soil type, pH, percent organic matter (%OM), percent
nitrogen (%N), MehliCh phosphorus (P) and Mehlich potassium (K) for two plots,
Lavanderos, Giiinope, Honduras, 2004.

 

 

 

 

 

 

 

 

 

 

Farmer Soil type pH* %OM %N P** K***
(ppm) (ppm)

1 Sandy loam 5.4 2.6 0.13 3 130

2 Sandy loam 5.9 4.1 0.20 65 164

 

 

" Normal range 6.0 to 6.8
" Normal range 13 to 20 ppm.
"‘" Normal range 98 to 195 ppm.

4.3.2.4 Farmer Perceptions of Bean Variety Performance

Toward the end of the 2004 growing season in Los Lirnones, the farmers with
experimental plots visited all of the research plots as a group to give their impression of
the study and a visual appraisal of the plots. Farmers were asked to judge the quality of
the soil, report their highest yield from each parcel in the past, identify experimental plots
that they expected to have best and worst yields, and assess which variety would produce
the highest bean yield (Appendix 7). Plants in all plots had at least 50% of pods with
fully developed seeds, with traditional varieties Seda and Marciano developmentally

ahead of the improved variety Amadeus-77 by about 10 days (example in Figure 5).

45

Figure 5. Example of bean plot observed for plot visit by farmers ﬁom Los Limones,
Moroceli, Honduras at 60 days post-sowing, November 23, 2004.

 

Images in this dissertation are presented in color.

46

CHAPTER 5

RESULTS AND DISCUSSION

In surveys taken in 2003 and 2004 in 6 communities in rural Honduras with small-scale
producers, farmers identiﬁed drought as the largest constraint to production, followed by
the lesser constraints of insects, disease, weeds and lack of fertilizer. Since farmers said
they can adequately control insects and disease, research was focused on fertilizer and
common bean variety to examine possible increases in production, and weed control to

examine savings from ﬁeld labor.

Experiments were designed to test the effect of traditional versus recommended inputs for
bean production in on-farm trials. In 2003, four full factorial (four replications) and ﬁve
supplemental trials (no replication) were sown to examine the effect of variety, fertilizer
and herbicide use in a broadly agro-ecologically representative area of Honduras called
Los Limones in the province of Moroceli. Factorial experiment results indicate that
increased fertilizer causes a signiﬁcant increase in yield in some cases; the supplemental
experiments showed with 85% certainty that the recommended package of improved
variety, increased fertilizer and herbicide increased yield. The results of an economic
analysis indicated that the marginal beneﬁt/cost ratio was 7.0 using the improved package

of inputs.

In 2004, the factorial experiment was repeated without herbicide, with an additional

traditional variety and with two additional fertilizer levels on seven farms (three farms

47

with three replicates, two farms with one replicate). The additional fertilizer treatments
were to try to identify the level of fertilizer that yielded the highest possible farmer
income. Due to terminal drought, one early maturing traditional variety (60 day) out-
yielded the improved, longer-season (70 day) variety on one farm and over all ﬁve farms.

Fertilizer had no effect on yield under these drought conditions.

A similar experiment was performed in another community called Lavanderos at a higher
altitude, representative of another agro-ecological area. Results from one farm indicate a
slight decrease in yield after the midseason application of urea fertilizer; and a much

higher yield from the traditional variety compared to the improved varieties.

On a ﬁeld day in 2005, farmers in Los Lirnones were asked to assess bean production in
each experimental plot. Farmers misidentiﬁed the worst-producing plot in one case,
believing that the improved variety Amadeus-77 produced the least instead of the local
variety Marciano; and attributed the apparent poor production to blight when the plants
were in fact only drying down early from drought stress. Farmers also misattributed the
best production to the local variety Marciano three times, when in fact the other local
variety Seda had produced the most. Misidentiﬁcations of bean variety production were

likely due to variety biases that farmers had based on observations in other contexts.
Surveys revealed some information relevant to agronomic and economic

recommendations. Farmers usually save seed, and one in ten farmers reported seed-home

anthracnose as important diseases in 2003 and 2004, with similar numbers for mosaic

48

viruses, so farmers likely underestimate the importance of seed quality. Planting
densities varied, and many farmers would beneﬁt from a higher planting density, up to 60
kg/ha. About half of the farmers fertilized at planting, which may result in an economic
loss in drought conditions; farmers that split applications or apply fertilizer later in the

season can decide not to apply if the crop is suffering a debilitating drought.

Members of local agricultural research committees (LARCs) reported access to technical
information as an important beneﬁt, along with access to loans, group cohesion, help for
their communities, and other educational opportunities. These beneﬁts, however, were

not correlated to other production or economic beneﬁts.

Misperceptions about bean variety performance on the part of the farmers emphasize the
importance of collaborative strategies for development of new bean varieties. Not only
do farmers lack some technical skills that would aid them in making better management
decisions (like the ability to identify disease and understand effective treatments), but
also prejudices based on experiences that are generalized from similar but not identical
experiences (like assuming that improved varieties always do poorly in drought

conditions).

5.1 Survey Results
5.1.1 Household and Land Characteristics
Survey participants were generally in their mid-forties with incomplete primary schooling

in a household with ﬁve members (Table 11).

49

Table 11. Characteristics of Survey Population from Six Communities, Department of El
Paraiso, Honduras, Postrera, 2004 (Head of household unless otherwise indicated)

(N=727).

 

 

 

 

 

 

 

 

 

 

 

Mean Median Standard Deviation
Age 44.3 42 14.8
Years of school 3.8 4 2.5
Family members 5.0 5 2.1
Total land cultivated 1.75 1.4 2.1
Total land owned 2.01 0.7 4.8
Percent without land title 42 %

 

 

There were notable differences among the farms where surveys were conducted. About
one-half the farmer families sowed less than one hectare of beans, while around three
quarters of families sowed less than two hectares (Table 12). Around half of the farm
families sowed only one plot, whereas the other half sowed two to four total plots.
Around 70% of farmers described their plots as ﬂat, and very few (6-7%) described their
plots as steep. About one-half of farmers described their soil as good, and most of the

rest described as fair, with few farmers considering their soil poor.

 

7 In some cases a survey question was improperly recorded, so whereas a total of 72 individuals were
surveyed, the number of responses for a given question may be as low as 50 respondents in this and all
subsequent tables unless otherwise noted.

50

Table 12. Distribution of bean plot size, number of plots sown, plot slope and soil quality
indicators reported by farmers, El Paraiso, Honduras, Postrera, 2003 and 2004 (N=72).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

YEAR

2003 2004
Bean Plot Size % %
0.1 — 1.0 ha 45.2 45.0
>1.0—2.0 ha 30.1 35.0
>2.0 — 5 ha 24.7 20.0
Number of Bean
Plots Sown
1 58.9 45.0
2 27.4 41.7
3 11.0 11.7
4 2.7 1.6
Plot Slope
Flat 71.7 68.0
Gradual slope 21 .7 26.0
Steep Slope 6.6 6.0
Soil Quality
Good 56.7 54.0
Average 35.0 42.0
Poor 8.3 4.0

 

 

 

 

 

Although 42% of farmers do not have titles to the land that they farm to beans, becaseu
they had access to other family land — considered borrowed land — most farmers did not
have to rent land (Table 13). In fact, 81 to 95% of farmer plots were owned by the
farmer or borrowed, and some of these same farmers (7%) also rent land for beans. Some
farmers do not have access to “free” land to sow to beans in postrera, so they rent land;

and a small number (4%) illegally occupy land without the land owner’s consent.

51

Table 13. Distribution of bean plot tenure reported by farmers, El Paraiso, Honduras,
Postrera, 2003 and 2004 (N=72).

 

 

 

 

 

 

 

YEAR
2003 2004
Bean Plot Tenure % %
Own 50.7 47. 1
Borrowed 30.4 48.0
Rented 21 .7 7.9
Occupied 4.3 3.9

 

 

 

 

 

Note: Percentages do not add to 100% because some producers farm both their own and rented land.

5.1.2 Typical farming practices
There were some marked trends in how farmers managed their bean crops across

communities and landscapes.

5.1.2.1 Herbicide Use and Tillage

While 39 to 50% of farmers applied paraquat herbicide to their bean parcel before any
tillage operations, 49-57% did not use herbicide before tillage (Table 14). Farmers then
tilled their land with oxen (58-76%) or by hand (22-36%). Planting density was very
variable: from 20 to 60 kg seed per hectare (average 35 kg/ha).

Table 14. Distribution of herbicide use and tillage reported by farmers, El Paraiso,
Honduras, Postrera, 2003 and 2004 (N=72).

 

 

 

 

 

 

 

 

 

 

 

YEAR

2003 2004
Herbicide Use °/o %
None 48.6 56.9
Paraquat 5 1 .4 43.1
Tillage
BLhand 37.0 22.2
Oxen 58.9 75.9
No preparation 4.1 1.9

 

 

 

52

5.1.2.2. Sowing

Fifty-nine to 76% of farmers sow only one bean variety in signiﬁcant quantities, while
22-40% sow two varieties and only 1-2% sow three (Table 15). Of the varieties sown, 17
to 26% of farmers sow improved varieties. Sixty-ﬁve to 78% of farmers plant beans as a
monocrop, while the rest intercrop beans and maize or sorghum, mostly with three or four
rows of beans for each row of the cereal crop.

Table 15. Distribution of number of varieties sown, farmers sowing at least one improved

variety, and rows of beans per row of grain crop (intercrop) reported by farmers, El
Paraiso, Honduras, Postrera 2003 and 2004 Q\I=72).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

YEAR

2003 2004
Number of varieties sown % %
1 76.7 58.3
2 21.9 40.0
3 1.4 1.7
Farmers sowing at least one
improved variety 25.6 16.7
Rows of beans per row of
maize/srﬂhum
No intercropping 77.6 65.0
1 1.5 1.8
2 3.0 3.3
3 9.0 13 .3
4 7.5 13.3
5 1.5 3.3

 

 

 

 

 

Of the beans sown, about 90% are from saved seed from the past harvest and/or from a
neighbor (Table 16). Forty percent of farmers select their seed prior to sowing based on
visual characteristics (uniformity, size, color, lack of damage and disease). The
remaining farmers obtained seed from a local NGO, Zamorano, or a local supply store.
Only two percent had even done a germination test to know how well their seed would

germinate.

53

Table 16. Distribution of seed source as reported by farmers, El Paraiso, Honduras,
Postrera, 2003 and 2004 (N=72).

 

 

 

 

 

 

 

 

 

 

 

YEAR

2003 2004
Seed source % %
Past harvest 60.3 80.0
Neighbor 35.7 30.0
Local Ag. Research Committee 4.1 5.0
NGO 2.8 1.7
Store 1.4 3.4
University 2.7 0

 

 

 

Note: Percentages exceed 100% because some producers acquire seeds from more than one source.

5.1.2.3 Fertilizer Application and Weeding

In 2003 when there was less concern about rainfall, 46% of farmers who applied fertilizer
only applied it at planting; another 18% split fertilizer application between planting and
at ﬁrst weeding (Table 17). Thirty-one percent waited until they weeded, presumably to
ensure that there was enough moisture to make a fertilizer application worthwhile. In
contrast, in 2004 when there was more concern about drought because of low rainfall in
the ﬁrst primera season, only 21% of farmers fertilized at planting, with 29% splitting the
fertilizer application between planting and ﬁrst weeding. Forty-eight percent waited until
the ﬁrst weeding to apply fertilizer. In both years, about 19% of farmers used a foliar
fertilizer spray, and a few of the foliar sprays were concocted by the farmers themselves
with local materials. The average fertilizer application was 15.9 kg nitrogen/ha and the
median was 11.7 kg nitrogen/ha, with a range of zero to over 100. Most plots were
fertilized far below the 26.8 kg nitrogen/ha as recommended by the university at
Zamorano (Rosas, 2003). No farmers applied Rhizobium inoculant to encourage

nitrogen-ﬁxing noduation on bean plant roots, and although no systematic observations

54

were made about nodulation, occasional plants that were observed in the field during this
study did not have nodules.

Table 17 Distribution of timing of fertilizer application as reported by farmers, El
Paraiso, Honduras, Postrera, 2003 and 2004 (N=72).

 

 

 

 

 

 

 

 

 

 

 

YEAR

2003 2004
Timing of fertilizer application % %
None 8.2 13.3
At planting 42.5 20.0
At ﬁrst weeding (about 21 days 28.8 41.7
after sowing)
Split between planting and ﬁrst 16.4 25.0
weeding
Foliar spray 16.4 16.7

 

 

Note: Percentages exceed 100% because some producers use both a granular fertilizer and a foliar spray.

Ninety percent of farmers weeded only once throughout the bean growing season,

generally at about three weeks after germination.

5.1.2.4 Insect and Disease Management

Over 25% of farmers surveyed noted that bean pod weevil (Apion godmani), ﬂea beetle
(Aphis 3p.) and whiteﬂy (Bemesia tabaci) were their principal insect pest problems in
postrera (Figure 6). Other insect pests were serious pests for a smaller number of
farmers. These pests included chrysomelids (Diabrotica spp, Cerotoma spp), leaﬂroppers
(Empoasca spp), wireworrns (Family: Elateridae), and defoliating insect larvae (Order:
Lepidoptera) (see Appendix 8 for local names). Twenty-one to twenty-two percent of
farmers chose not to use chemical control for insects; a few farmers used a self-made
preparation with local plants extracts to control insects. When farmers opted'for insect

control, 72-79% used parathion, and another 6-9% used parathion mixed with another

55

chemical. Other insecticides used in small quantities included metharnidophos,

chlorphoxim, malatlrion, and lambda-cyhalothrin.

Figure 6. Percentage of farmers (N=72) who reported insect pest problems, El Paraiso,
Honduras, Postrera, 2003 and 2004.

50
45
40
35
30
25
20
15
10

Percent farmers with pest problem

 

Insect pest

Note: insect names identiﬁed by farmers

Most farmers did not report signiﬁcant amounts of disease in their bean crops in postrera
in 2003 and 2004 (Figure 7). Web blight (Thanatephorus cucumeris) was the most
commonly reported disease (see Appendix 9 for local names). Chemical control of
diseases was very uncommon, including only 10-18% of farmers. When chemical
control was used, farmers used a large variety of products, including mancozeb, copper
sulfate, othiocarbarnate, calixin, copper oxychloride, copper hydroxide chlorotalonil, and

benomyl.

56

Figure 7. Percentage of farmers (N=72) who reported signiﬁcant infection by disease,
postrera, 2003 and 2004, El Paraiso, Honduras, Postrera, 2003 and 2004.

70

    
  
  
  
  
  

0'10)
00

NO)
00

Percent of farmers affected
8 3

Note: diseases identiﬁed by farmers
* indicates a seed-home disease

Slugs were reported as problems in 12-17% of farmers’ bean ﬁelds. By the time slugs
were a pest in a bean ﬁeld, farmers opted for chemical control with a commercial

molluscicide.

5.1.2.5 Harvest and Storage

Once beans reached harvest maturity, the bean plants were generally pulled up by hand

and dried in bunches upside down in the ﬁeld, and then threshed with long poles on the

57

ground once they were completely dried. Bean seeds were stored in 100 pound woven

polypropylene sacks and stored for sale or future use.

5.1.3 Bean Yield Analysis

Bean yields were highly variable, ranging from zero to 2,080 kg/ha. According to
farmers, low yields were caused mainly by drought conditions, and infrequently by slug
and vvireworm infestations. There were no statistically signiﬁcant relationships between
yields and fertilizer use or other input uses; and no relationship between planting density
and yield. However, when low-yielding plots resulting from drought, ﬂooding or severe
slug infestations were removed from the analysis (yields < 99 kg/ha), a moderate linear
relationship between planting density and yield (r = 0.403, P=0.0001) become apparent
(Figure 8). When only high-yielding plots are considered (yields > 850 hg./ha), the linear
relationship improves (r = 0.464, P=0.0086) (Figure 9). This data supports the Pan-
Arnerican School of Agriculture recommendations of planting densities of 170,000 -
215,000 plants per hectare, or about 50-60 kg viable seed per hectare for common beans
with a weight of 30 grams per 100 seeds, typical for the beans in this study (Rosas, 2003;
Tshering, 2002). Some of the other variation unexplained by the regression analysis was

probably caused by differences in germination.

58

Figure 8. Relationship between seeding density of common beans and yield where yield
was >99 kg/ha (N=121), El Paraiso, Honduras, Postrera, 2003 and 2004.

 

 

 

 

 

 

2500
2000 ° °
y = 13.409x + 160.09
A R2 = 0.1625
g 1500 °
E
.2 1000
>-
500 -
o .

 

 

 

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Seeding density (kg seed/ha)

Figure 9. Relationship between seeding density of common beans and yield where yield
was >850 kg/ha (N=32), 2003 and 2004, El Paraiso, Honduras.

 

 

 

 

 

2500.0
2000.0 ° °
y = 12.14x + 661.64
2 _
’0? R - 0.215 .
3 1500.0 “
O O
c: e e e
V O
E e . e
.2 1000.0 ° ’ ‘ ’
>1 ./.°: : '
500.0
0.0 T I —T —I I

 

 

 

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Seeding density (kg seed/ha)

59

There is no apparent relationship between N use and yield, neither looking at the plots
without very low yielders, nor looking only at high-yielding plots. Since Rhizobium
inoculant was not used in farmer-managed systems and fertilizer nitrogen was present,
there was no expected biological nitrogen ﬁxation; as previously mentioned, although no
systematic observations were made about nodulation, occasional plants that were

observed in the ﬁeld during this study did not have nodules.

Some survey questions were chosen in order to give insights into why some farmers were
better producers than others. These questions were speciﬁcally about intrinsic farm
conditions like land slope and soil quality, and extrinsic factors such as soil fertility
management, pest management, cropping systems, variety choice, and cropping systems.
These factors occur often enough to predict management choices that result in higher

plant yields, especially in uniform environments.

Despite the general similarities in rainfall and altitude, microclimate and unnoted
management differences likely played a role in confounding the ability to make
generalization about successful management strategies. These could include precipitation

patterns, residual fertilizer, rotation effects, and micronutrient deﬁciencies.
Generalizations about successful management strategies may be confounded .if farmers

have different risk strategies for economic gains. For example, if a farmer wants to

spread out farm income over many different income-generating projects, he or she may

60

spend less on inputs and accept a mediocre harvest in any given season (the “minimax”
risk strategy mentioned in section 5.1.4). Another farmer may invest more resources in
fewer projects, and accept losses in a season unfavorable to production while beneﬁting
more in a favorable season, a strategy with higher tolerance for risk. If some farmers in
this survey used a minimax strategy and other a more risky strategy, the correlation

between management and production may not be apparent.

5.1.4 Farmer Financial and Social Factors
If farmers did not have animal traction, they hired oxen to plow the bean ﬁeld, but
otherwise farmers generally did not hire others to help them except for weeding and

harvest.

Once beans were harvested, farmers sold a portion of the harvest to pay loans and to
supply cash to the household; otherwise, beans were stored and sold when the household
needed more cash. Beans were generally sold to intermediaries (74%) or to buyers in the
market in the capital city (23%). Only 24.2% of farmers did not sell any of the beans they
harvested in postrera; the remaining farmers sold from 17 to 98% of their harvest.
Storing beans was preferred to selling immediately because the price generally went up
as harvest grew distant and beans grew scarce. Farmers desired to be self-sufficient in
beans, but many households (42%) reported selling beans for cash and then purchasing

them at a later date.

61

One way to determine the proﬁtability of producing common beans is to calculate the
cost of production (including family labor) and the value of the crop, and comparing the
net return to the returns from the next best investment. Data for all direct bean
production costs were collected with the survey; and family labor was estimated using
data reported by Tshering (T shering, 2002) and valued at the average price paid to ﬁeld
workers in the survey (Lp 45.50/day). Returns from the sale of beans were ﬁxed at Lp
13.20/kg, the median price at harvest for 20048. It was discovered, as one might expect,
that as yield increases, so do proﬁts (R2=0.890, r=0.943, P<0.0001) (Figure 10). Most of
the costs of bean production (65%) were accounted for by labor costs when family labor
was included in the calculation.

Figure 10. Relationship between proﬁt (Lp) and yield (kg/ha) (N =148), El Paraiso,
Honduras, Postrera, 2003 and 2004.

30000

 

 

25000 y = 10.951x - 1868.9 '

2 _
20000 R —0.8901 . /

15000 /

1 0000

 

 

 

Proﬁt (Lp)

 

5000

 

 

 

’ g- 500 1000 1 500 2000 2500

 

 

 

 

Yield (kg/ha)

 

1' Prices at harvest time are the lowest, so proﬁtability will be underestimated if farmer is able to sell some
of the crop when the price rises as beans become scarcer.

62

One way to determine whether farmers are making more money on a plot of beans
relative to other activities is to estimate the returns ﬁ'om that plot per day of family labor,
and comparing that to what a farmer might earn as a day laborer. To calculate this
hypothetical wage, costs of production (paid labor and input costs) are subtracted from
gross proﬁts to calculate the net returns before farm labor. When net returns are divided
by farm labor, the result can be compared to day laborer wages to show the proﬁtability

of growing beans.

When this type of analysis is performed, 68.9% of farmer ﬁelds yielded enough beans to
make farming more proﬁtable than working as a day laborer. Theses results are not
surprising considering that 12.2% of the ﬁelds in this analysis had no yield at all, and

another 6.1% harvested under 99 kg beans/ha.

Farmers differed in where they got their information, and how much they trusted the
information that they got. Many farmers (44.2%) reported getting their bean production
information from their neighbors, and trusted that information (Figure 11). Almost forty-
one percent of farmers had access to technicians from Zamorano, and conﬁdence in that
information was also high. The remaining farmers, 14.7%, got information from supply
stores, and conﬁdence in this information was not as high (43.6% high and 38.2%
moderate) because “los agropecuarios quieren vender sus productos nomds” (“store
owners just want to sell their products”) and “los vendedores no son productores de ﬁijol

y a veces no saben” (“the salesmen are not bean producers and sometimes they do not

63

know [about bean production]”). Interestingly, 16.4% of farmers claimed that they do

not actively seek information about bean production.

Figure 11. Producer-reported level of conﬁdence in available source of information
(N=72), 2004, El Paraiso, Honduras.

100

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”l IHigh
5° : E Moderate
,0 . El Low

 

 

 

 

 

 

Percent Conﬁdence (%)

 

 

 

 

 

 

 

 

 

 

 

 

 

Neighbor Zamorano

Source of Information

When farmers were asked about where they could get a loan for bean production, 40% of
farmers responded that they could get loans from family members, with interest rates
from 0 to 10 percent over the period of the loan (four months, the length of the bean
season) (Figure 12). Thirty-two percent of farmers revealed that they could get loans
from the community’s LARC or another community based organization, and the same
percentage of farmers mentioned community banks as an option; both of these options
charge 2 or 3% to members and up to 7 percent to non-members. Seventeen percent of
farmers mentioned intermediaries, who normally charge 20%, or buy beans at a lower-

than-market price to cover the amount of the loan. Six percent spoke of the Honduran

64

national bank BANADESA, who charged 1%. Twenty-two percent of farmers said that
they had no options for loans for bean production, and many of those who listed options
said that they were unwilling to take loans even if they were available. Why farmers were
shy about taking loans was not asked, although most farmers who expressed this

sentiment said that they were afraid of debt.

Figure 12. Farmer perception of availability of loan for bean production (N=66), El
Paraiso, Honduras, 2004.

45
4o
35

Percent (%)
-‘ r— N N 1»
Ln O Ur 0 U1 O

 

O

~51 9° 4‘" .40
Q69 0 e

025‘ 55"» 9919
0&6 ‘0'? §Y9 %

Q
.9” ‘ 1’

Potential Source of Loan

Note: CBC = Community Based Organization, including Local Agricultural Research Committees

(LARCs); BANADESA = Banco Nacional de Desarrollo Agricola, the Honduran National Agricultural
Development Bank.

Farmers generally did not rely wholly on bean and other crop production for income.
When farmers were asked what other activities farm families engaged in to earn money in
the 2004 survey, 52% of the 66 respondents claimed that someone in the household

worked as agricultural day laborers. Twenty percent responded that someone sold eggs,

and 20% of households grew and sold coffee. Nine percent of families had members who

65

rented out their oxen and worked in construction; seven percent baked and sold bread or
ran a store ﬁ'om their home. Six percent had a household member that worked in the city
or made clay roof tiles. Fewer than ﬁve percent of households had members who cut
sugar cane, worked as a carpenter or mason, repaired fences, worked for a patron in
another household, and raised and sold animals. Only one of the 66 households surveyed
reported that a family member washed clothes for pay, collected and sold wild fruit, sold
soft drinks from their home, worked in a maquila (a clothing factory in the city), ran a
pool hall, worked as an elementary school teacher, prepared and sold food, worked at a
local brick factory, slaughtered animals, sewed clothes, managed dairy cows for others,
sold fresh milk, ran a bar, sold beans for seed, used his vehicle as a taxi service in his
community, ran a small saw mill, cut fence posts, grew and sold bananas and sold
vegetables. Only four of the 66 households in the 2004 survey claimed that they had no

other source of income apart from bean and corn production.

5.1.5 Beneﬁts of LARC Membership

It was difﬁcult to determine what made the members of LARCs different from other
farmers in the communities where the research was conducted. One survey question
asked what beneﬁts LARC members perceived by being a member of their LARC. Of
the 33 LARC members surveyed, practically all members (82%) mentioned access to
technical information (Figure 13). Thirty-nine percent mentioned access to loans as a
beneﬁt of LARC membership. About a quarter of members noted that LARCs created a
sense of togetherness (“compai‘ierismo”) among members, and a quarter of the members

also mentioned their skill as experimenters. About 20% noted that membership permitted

66

access to technical help from experts, 15% mentioned the leadership and organizational
opportunities, 9% stated that participation in LARCs helped their communities solve
production problems, 9% stated that LARCs permitted them to visit and learn from other

communities with similar problems, and 6% noted that a beneﬁt was getting free seeds.

Figure 13. Self-reported beneﬁts recognized by members of Local Agricultural Research
Committees, El Paraiso, Honduras, 2004.

30
254
20- —-
15~
101
5.
o-

 

 

 

 

 

 

       

Number of respondents

 

 

Free seed

Access to loans

5
.2
a
a:
O
O
G.
5
O
h-
9

Aids community

Access to technical
Information
Ability to experiment
Access to experts
regions

and organization
Visits to farmers in other

Opportunities for leadership

Beneﬁts of LARC membership

The results of some of these beneﬁts one might expect to see reﬂected in other aspects of
the survey, notably increased bean production, greater adoption of improved varieties,
more mention of Zamorano and NGOs as sources of information, and greater access to

loans. These beneﬁts, however, were not reﬂected in the survey.

67

One concern is that LARCs favor farmers that are more educated, thus exclude farmers
with more need for education (Humphries et al., 2000). In this study, however, there was

no relationship between LARC members and education.

5.1.6 Farmer Perception of Production Constraints

When asked about constraints to production, most farmers answered that their biggest
constraint was drought (Figure 14). There are two major kinds of drought in Honduras.
Intermittent drought is a break in rainfall during the development of the plant, and
requires that the plant limit growth-related activities until rainfall begins again. Terminal
drought is when rainfall ends and does not resume over the life of the plant. In the region
under study, and in a signiﬁcant part of Honduras in postrera, terminal drought is the
most common form of drought. The agronomic study supports the idea that drought is a
constraint to better beans production, and bean varieties bred to produce well in areas that
suffer terminal drought is one solution to this constraint. Irrigation, another solution to
drought in general, is generally not considered a viable solution due to the low price of
beans compared to other, higher value crops. However, irrigation is required for bean

seed production.

68

Figure 14. Responses to question, “What are the biggest constraints to increased bean
production in postrera?” (N=66), El Paraiso, Honduras, 2004.

Number of Respondents

 

70

60‘

501

 

I Not important
I Somewhat important
I Very Important

40-

30-

 

 

 

20-

 

 

 

 

 

 

 

 

«c1 -c\

\\ \‘

$0 (0 0° “.1 00% go
(‘0 Q\ ,

“9°\ 9,00

s. e s s
of? 0‘? «if? “’6 590 8‘80 \°°"

Farmer Response

The other constraint that was most evident was the lack of fertilizer. In years of

sufﬁcient rainfall, fertilizer has the potential to increase yield and proﬁt, as this research

in Los Lirnones in 2003 demonstrates. Castro-Zr'rniga, in a similar study in central-

eastem Honduras, showed that increased fertilizer signiﬁcantly increased yield in 45% of

the cases under study (Castro-Zuniga, 1998). However, only 14% of farmers consider the

lack of fertilizer in their beans in postrera very important (or extremely important), and

only 36% consider it not important at all. It is possible that farmers are risk-adverse, so

do not consider increasing fertilizer a priority when its success is not guaranteed.

Farmer’s dislike of credit supports this assertion.

69

Insects, disease and weeds are identiﬁed as greater constraints than a lack of fertilizer in
2004. Farmers noted that since these constraints can generally be controlled through
chemical application and weeding, they are much more under the control of the farmer
than precipitation. Still, in a drought year where fertilizer has a lesser effect on
production than in a normal rainfall year, farmers perceive fertilizer as less important

even though the question was about postrera in general.

Twelve and eleven percent of farmers reported that ﬂooding and slugs, respectively, were
very important. Since ﬂooding is speciﬁc only to certain physical areas, and considered
undesirable for crop production, it is not surprising that ﬂooding is not a very important
constraint to bean production. However, since 12% of farmers did respond that ﬂooding
was very important, it is probable that some farmers are sowing beans in poor soils due to
a lack of available good quality soils; and farmers may be more willing to sow in soils
that are prone to ﬂooding in the postrera season since the chance that they will not ﬂood

when rainfall is lower.

Similarly, slugs in beans in postrera are generally problematic where they were present in
the previous maize crop but not eradicated at that time (Rosas, 2003). Most farmers seem

to control slugs effectively and they are not considered problematic.
Seed quality was considered very important for only 5% of respondents, and not

important at all by 83%. Since many diseases are seed-bome, seed quality is an

important factor in controlling important diseases like common mosaic virus, common

70

bacterial blight, and anthracnose. In fact, about ten percent of farmers report mosaic
viruses and anthracnose as important diseases in postrera (Figure 7), and anthracnose is
probably more common in the wetter primera season. These results indicates that
farmers are not aware that seeds can be major sources of disease, and this lack of

understanding may limit success to encourage farmers to buy clean seed.

Lastly, soil quality was considered very important by only 3% of respondents. Since
inherent soil qualities are not under the control of the farmer, farmers give it very little

importance when thinking about how to improve production.

In 2003, farmers in Los Lirnones were asked why they chose the variety they planted that
year. Over half (59%) reported that they chose a traditional variety because of its
marketability; half responded that the traditional variety produced well. About one
quarter (27%) claimed that the traditional variety is what they prefer to eat, and 22%

claim that they just used what seed was available to them at the time (Figure 15).

71

Figure 15. Preference for traditional varieties as reported by farmers (N=22), Los
Lirnones, Moroceli, Honduras, 2003.

 

 

 

 

 

Number of Respondents

 

 

 

Marketable Produces well Good to eat Used what was
available

Criteria

Note: Multiple answers acceptable.

Survey results indicate a 5% price discount for darker-colored red beans like Amadeus-77

and T10 Canela, which plays a part in the reluctance of farmers to sow these varieties.

The question of whether traditional varieties produce well is relative to the farmer’s risk
strategy. Farmers who are very risk-adverse may choose to produce a low-yielding
variety that yields consistently than choose to produce a high-yielding variety that may
fail in an exceptional year, the aforementioned “minimax” strategy. Nevertheless, a
farmer may choose to plant some ﬁelds to traditional varieties and others to improved
varieties. This way the farmers can take advantage of the high potential yields in ﬁelds

sown to improved varieties in normal rainfall seasons while guaranteeing yield security

72

with the harvest from the ﬁelds sown to the traditional varieties in low rainfall seasons.
This strategy can maximize farmers’ returns in the long run without sacriﬁcing yield

security.

Male farmers in Los Limones claim that the taste of dark red Amadeus-77 and traditional
light red varieties like Seda are identical, and that the color of the beans themselves are
indistinguishable when cooked. These observations seem to indicate that differences in
common bean seed coat color create a perception that there is a difference in taste where

there is none (though these conclusions have not been studied objectively).

Ninety percent of farmers in this study saved seed from a past harvest to sow in the
future, and farmers view purchasing seed as an unnecessary expense. As mentioned
above, reluctance to buy high-quality seed is due in part to a poor understanding of
disease transmission by infected seed; and one in ten farmers report seed-bome
anthracnose as important diseases in 2003 and 2004, with similar numbers for mosaic
viruses. As farmers understand the connections between disease and seed quality, the

desire to buy disease-free seed may increase.

5.2 Field Results

5.2.1 Los Lirnones Plots 2003

There were similarities between the bean yields among the four complete factorial plots
(Table 18) (more detailed tables are given in Appendix 6). When considered alone, the

yields of the improved variety Amadeus-77 and the traditional variety Seda (also known

73

as Rojo de Seda), were similar in all four factorial plots. Data from preliminary trials
with Amadeus-77 in farmer ﬁelds in 2002 indicated that Amadeus-77 has a yield
advantage under local, non-drought conditions (Rosas and Escoto, 2002), so the lack of
difference was unexpected, and not correlated with weed density, disease pressure, or
insect damage; although overall, variability in the ﬁeld was high and a lack of differences
may have become apparent with more replications. The addition of urea played a
signiﬁcant role in increasing bean yield in two of the four plots, indicating that under
some circumstances, the plants were able to increase production at the levels of fertilizer
used in this experiment. There was no apparent effect from the use of paraquat herbicide;
in retrospect, due probably to the fact that it was used at a time when weed seedlings
were not actively germinating. Weed densities at mid-ﬂower in the experimental plots
were measured at 0 to 658 kg/ha dry matter, though no correlation was discovered

between weed biomass and plot yield. There were no signiﬁcant interaction effects.

74

Table 18 Bean yield (kg/ha) for producers 1-4, factorial trials, Los Lirnones, Moroceli,

Honduras, Postrera, 2003.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CONSTRAINT FACTOR* PRODUCER
Trt Variety Fert. Weed 1 2 3 4
mgt. mgt.
1 "' "‘ R R R 1,740 1,240 1,660 1,070
2 R R T 1,760 1,830 2,440 730
3 R T R 1,490 1,430 2,040 560
4 T R R 1,640 1,570 2,500 870
5 R T T 1,410 1,490 2,110 640
6 T R T 1,700 1,790 2,230 940
7 T T R 1,340 1,170 2,170 460
8*" T T T 1,290 1,490 2,160 710
PRECIPITATION IN 60 DAYS 140 148 134 120
(mm)
COMPARISONS OF INDIVIDUAL FACTORS
Variety
R 1,600 1,500 2,070 760
T 1,490 1,500 2,260 760
P-value 0.35 0.81 0.33 0.87
Fertilizer maggement
R 1,700 1,610 2,210 900
T 1,390 1,400 2,1 10 600
P-value 0.01 0.36 0.67 0.04
Weed management
R 1,540 1,360 2,100 740
T 1,540 1,660 2,230 760
P-value 0.93 0.18 0.50 0.86

 

* R=Recommended, T=Traditional.
" Equivalent to the recommended management package.

 

""' Equivalent to the traditional farmer package.

The results of the complete factorial experiments give insight into how the constraint
factors interact; while the results of the supplemental experiments compare the relative
performance of the recommended versus the traditional management strategies. To
analyze the supplemental plots, the results from the complete factorial trial that represent

the traditional plots (Seda, less fertilizer and no herbicide) and the recommended plots

75

(Amadeus-77, more fertilizer and herbicide) were added to the supplemental experiment
results and analyzed as a single data set to maximize the amount of data (Table 19).

Table 19. Supplemental trial bean yields (kg/ha) for 9 farmers, Los Lirnones, Moroceli,
Honduras, Postrera, 2003*.

 

 

 

 

 

 

 

 

 

 

 

 

RECOMMENDED TRADITIONAL

PACKAGE MANAGEMENT

Improved variety Traditional var.

+ extra fertilizer - typical fertilizer
Producer + paraquat - paraquat Difference
1 1,740 1,290 + 450
2 1,240 1,490 - 250
3 1,660 2,160 - 500
4 1,070 710 + 360
5 910 1,170 - 260
6 3,340 2,130 + 1,210
7 3,410 770 + 2,640
8 1,630 930 + 700
9 1,290 1,030 +260
AVERAGE 1810 1300 +510

 

 

 

 

 

" Differences determined with Tukey’s mean separation, signiﬁcant at a=0. 15

In general, the plots using the recommended practices (Amadeus-77, more fertilizer and
herbicide) produced more beans than plots using the traditional farmer practices.
However, in three cases, the traditionally managed plots yielded more beans (Table 19,
Farmers 2, 3 and 5). Farmers gave no explanation for this, and weed, disease and drought
notes also do not explain this difference. In addition, one recommended practice plot
produced much better than the traditional farmer management plot (Table 19, farmer 7).
The farmer claimed that this difference was due to a severe attack of web blight that
affected the traditional variety but not the improved variety. When analyzed as a whole,
we can say with only 85% certainty that there were true differences between the farmer
and recommended practices. The economic implications of these results are discussed in

Section 5.1.5.

76

5.2.2 Los Lirnones Plots 2004

In four of ﬁve cases, rainfall in the Los Lirnones experimental plots was lower in 2003
than in 2004 (that is, under 120 mm precipitation over 60 days). Although rainfall data
alone does not adequately reﬂect the quantity and timing of water available to plants, it
gives insight into the water constraint that deﬁnes the area in which these producers grow

their crops, and serves as a proxy for drought conditions.

Other researchers have calculated water use ratios for beans in different cropping systems
and environments. Researchers in Kansas (Hattendorf et al., 1988) calculated the daily
water use rate (WUR) for pinto beans and calculated it to be between 3.95 and 5.2
mm/week (mean=4.76 mm/week) over the 80 to 90 days the crop was growing; seed
yields were 1,458 to 2,909 kg/ha (mean=2,077 kg/ha). Similarly, Yonts (Yonts, 1996)
reports that common beans use 5.6 mm/week at peak growing (rapid vegetative grth
and ﬂowering and pod development), and less during stand development and pod ﬁll and

maturation.

On the farm where rainfall was 124 mm from September to December, there were no
signiﬁcant differences in bean yield due to variety or fertilizer, and yield was below what
the farmer reported as the typical yields for that ﬁeld (900-1100 kg/ha) (Producer 1,
Table 20). When rainfall was around 100 mm, there were signiﬁcant differences between
all three varieties (Producer 2, Table 21), with the traditional variety Seda yielding better
than either the other traditional variety Marciano, and the improved variety Amadeus-77.

In Producer 3’s ﬁeld, where rainfall was only 72 mm, yields were extremely low, and

77

there were small differences between Amadeus-77 and Seda, but not with Marciano
(T able 22). In no case were there differences due to fertilizer. A combination of the
replicated complete factorial ﬁelds and the un-replicated plots shows differences between

Seda and the remaining two varieties; and no differences in fertilizer (Table 23).

Table 20. Factorial trial bean yields (kg/ha), Producer 1, 2004, Los Lirnones, Moroceli,

Honduras, Postrera, 2004*.

 

 

 

 

 

 

 

 

 

 

 

 

Variety Seda Marciano Amadeus-77
Fertilizer (trad) (trad.) (improved) AVERAGE
65 kg/ha 18-46-0 840 535 822 733 a
65 kg/ha 18-46-0 638 772 679 696 a
32.5 kg/ha urea
130 kg/ha 18-46-0 793 673 956 807 a
130 kg/ha 18-46-0 910 787 1122 939 a
65 kg/ha urea
AVERAGE 795 a 691 a 895 a
Interaction fertilizer x variety: P=0. 0.8260

 

 

Note: Rainfall during growing season was 124 mm.
" The different letter in a row or column indicates a signiﬁcant difference in yield (Tukey’s mean ,
separation, a=0.05).

Table 21. Factorial trial bean yields (kg/ha), Producer 2, 2004, Los Lirnones, Moroceli,
Honduras“.

 

 

 

 

 

 

 

 

 

 

 

 

 

Variety Seda Marciano Amadeus-77
Fertilizer (trad) (trad) (improved) AVERAGE
65 kg/ha 18-46-0 420 309 173 301 a
65 kg/ha 18-46-0 420 355 204 326 a
32.5 kg/ha urea
130 kg/ha 18-46-0 505 354 194 351 a
130 kg/ha 18-46-0 474 351 312 379 a
65 kg/ha urea
AVERAGE 455 a 342 b 221 c
Interaction fertilizer x variety: P=0.9210

 

Note: Rainfall during growing season was 103 mm.
" The different letter in a row or column indicates a signiﬁcant difference in yield (Tukey’s mean
separation, a=0.05).

78

Table 22. Factorial trial bean yields (kg/ha), Producer 3, Los Lirnones, Moroceli,

 

 

 

 

 

 

Honduras, Postrera, 2004*.
Variety Seda Marciano Amadeus-77
Fertilizer (trad) (had) (improved) AVERAGE
65 kg/ha 18-46-0 105 73 41 73 a
65 kg/ha 1846-0 94 67 60 74 a
32.5 kg/ha urea
130 kg/ha 18-46-0 62 66 43 57 a
130 kg/ha 18-46-0 117 57 44 73 a
65 kg/ha urea
AVERAGE 95 a** 66 a 47 a

 

 

 

 

 

 

Interaction fertilizer x variety: P=0.907 3

 

Note: Rainfall during growing season was 72 mm.
* The different letter in a row or column indicates a signiﬁcant difference in yield (Tukey’s mean

separation, a=0.05).

** Seda and Amadeus-77 were signiﬁcantly different at the a=0.06 level

Table 23. Factorial trial bean yields (kg/ha), combination of 5 farms, Los Lirnones,
Moroceli, Honduras, Postrera, 2004*.

 

 

 

 

 

 

 

 

 

 

Variety Seda Marciano Amadeus-77
Fertilizer (trad) (trad) (improved) AVERAGE
65 kg/ha 18-46-0 410 265 289 321 a
65 kg/ha 18-46-0 302 342 209 284 a
32.5 kg/ha urea
130 kg/ha 18-46-0 403 272 317 331 a
130 kg/ha 18-46-0 400 328 317 348 a
65 kg/ha urea
AVERAGE 379 a 302 b 283 b

 

 

Interaction fertilizer x variety: P=0.4320

 

* The different letter in a row or column indicates a signiﬁcant difference in yield (Tukey’s mean

separation, (F005).

79

 

 

5.2.3 Lavanderos Plots 2004

The factorial experiment showed signiﬁcant differences for variety and fertilizer
treatments in the plot of Producer 1 and no differences in the plot of Producer 2 (Tables
24-25). Farmer one experienced much higher yields with the traditional variety Paraicito
compared to the improved varieties Tio Canela and Amadeus-77 (Table 24). These
differences cannot be explained by variety alone, though observations of weed pressures,
disease and insect damage did not reveal the cause of these extreme differences in yield.
Perhaps the reduction in yield due to the absence of disease pressure on the traditional
variety accounts for some of the difference. The addition of urea fertilizer also had a net
negative effect when examined across all varieties. The urea fertilizer was applied at the
same time that the plots were weeded so the soil around all plants was disturbed by the
hoeing; however, the additional action of incorporating the urea fertilizer, plus perhaps
some toxicity from root contact with the fertilizer, may have caused the reduction in
yield. Another explanation could be that the urea stimulated weed growth which
debilitated the bean plants at a critical stage in growth, causing the lower yield. The
application of urea may also have further acidiﬁed the soil (Table 9), causing some
toxicity from low pH, like aluminum toxicity which is common at pH under 5.0.

Table 24. Factorial trial bean yields (kg/ha) of Producer 1, 2004, Lavanderos, Giiinope,
Honduras, Postrera, 2004*.

 

 

 

 

Variety Tio Canela Amadeus-77 Paraicito
Fertilizer (improved) (improved) (trad) AVERAGE
65 kg/ha 18-46-0 500 360 1370 740 a
65 kg/ha 18-46-0 290 420 890 530 b
32.5 kg/ha urea
AVERAGE 390 a 390 a 1130 b

 

 

 

 

 

 

 

Note: Rainfall during growing season was 235 mm.
* The different letter in a row or column indicates a signiﬁcant difference in yield (Tukey’s mean
separation, a=0.05).

80

Table 25. Factorial trial bean yields (kg/ha) of Producer 2, 2004, Lavanderos, Giiinope,

 

 

 

 

Honduras".
Variety Tio Canela Amadeus-77 Paraicito
Fertilizer (improved) (improved) (trad) AVERAGE
65 kg/ha 18-46-0 680 370 440 500 a
65 kg/ha 18-46-0 690 620 590 630 a
32.5 kg/ha urea
AVERAGE 680 a 500 a 520 a

 

 

 

 

 

 

 

Note: Rainfall during growing season was not recorded, estimated at 235 mm.
* Statistical analysis with Tukey’s mean separation, a=0.05

5.2.4 Farmer Perceptions of Bean Variety Performance

On the ﬁeld day in Los Limones in December, farmers responded to questions about their
perceptions of soil quality, bean performance and personal preferences for cultivation.
When asked about the qualities of the soil where the beans were planted, the farmers
identiﬁed soil properties that correspond to better production, albeit in their own
vocabulary. Speciﬁcally, to explain that the soil has a better water holding capacity, one
farmer said, “Ese suelo tiene mds barro, y agarra mejor el agua” (“This soil has more
clay, and holds onto the water better”). And to explain that the soil was richer in
nutrients, different farmers at different times said that the soil had more “vitaminas”

(“vitamins”).

After asking how much the parcel had produced in the past, farmers were asked why they
thought that piece of land produced that much. In every case, the farmers answered that

it depended on the quantity of rain that fell that year. In the farmer survey in Los

81

Lirnones, farmers identiﬁed drought more often than any other factor as a very important

constraint to increased common bean production in postrera.

At each experimental plot, the farmers were asked to examine all of the plots in one block
(one replicate) and choose the best- and worst-producing plots (Table 26). Farmers
always chose an Amadeus-77 plot as their example of the plot that would produce the
least. In most cases they were correct that Amadeus-77 had the poorest production, and
chose a plot that, if not the least production, was close to the worst. However, in the plot
with the highest rainfall (Plot 1, Table 26), farmers chose an Amadeus-77 plot when the
plot with the lowest production was actually of the local variety Marciano.

Table 26. Farmer belief about highest- and lowest-producing plots and actual production,
rainfall and maximum yield, Los Lirnones, Moroceli, Honduras, Postrera, 2004.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Highest yielding Lowest yielding .
variety variety 60-day Max.
Farmer Actual Farmer Actual Rainfall yield"
Plot belief production belief production (mm) (kg/ha)
1 Marciano Seda Amadeus-77 Marciano 124 974
2 Marciano Seda Amadeus-77 Amadeus-77 103 742
3 Seda Seda Amadeus-77 Amadeus-77 72 909
4 Seda Seda Amadeus-77 Amadeus-77 70 935
5 Marciano Seda Amadeus-77 Amadeus-77 107 742

 

*Maximum yield is the highest yield that the farmer could recall for that ﬁeld since he had been farming it.

The reason that the farmers gave for the poor production of Amadeus-77 was that it

produced more slowly then the traditional varieties by about 10 days, and thus performed

poorly in drought conditions. Farmers also stated that Amadeus-77 was suffering from

“hielo” (“blight,” usually anthracnose), and that was another reason that it produced less.

In truth, Amadeus-77 was not suffering any signiﬁcant disease (except a little bacterial

82

 

blight) (Figure 16), but rather dying back due to lack of water before reaching

physiological maturity.

Figure 16. Amadeus-77 plants at 60 days post-sowing in Los Lirnones, Moroceli,
Honduras, November 23, 2004.

 

It is probable that since farmers expected Amadeus-77 to produce poorly in all plots due
to less than average rainfall, their judgment about what plants would produce poorly was
clouded by this supposition. Their belief that the Amadeus-77 plants were suffering
blight probably also contributed to their prejudice against the production abilities of

Amadeus-77.

Farmers also stated that Seda was more resistant to drought than Marciano, and their
decisions about what plants would produce the best reﬂected that: Farmers stated that
Marciano would produce better in moderate drought conditions, and that Seda would
produce better in more extreme drought conditions. This belief was incorrect: In every
block where farmers looked at plants, there was at least one plot of Seda that produced

better than every other plot.

83

Use of local variety Marciano became more popular in the community over the course of
this work. In 2003, LARC members in Los Limones wanted to use local variety Seda as
the variety of comparison with the improved variety Amadeus-77; in 2004, the LARC
members wanted to use both Seda and Marciano for comparison, stating that Marciano
was as good a producer as Seda, and in some cases better. This trend appears to be true
for the entire community: according to 2003 survey data, 79% of farmers planted Seda,
and only 26% had sown Marciano; one year later, only 55% of farmers were planting
Seda, and 64% were sowing Marciano. Since Marciano was gaining in popularity yet was
not as extensively cultivated in the community, farmers did not have much experience
producing Marciano yet, and it could be that farmers had an exaggerated idea about the

production potential of Marciano.

When asked which bean variety farmers would use to maximize their bean production
based on their experiences in the past, all farmers stated that they would plant Seda and
Marciano, insisting that these two varieties performed differently based on soil and
rainfall, but that both produced well in their environments. They pointed to the plots
sown with Amadeus-7 7 to explain why they do not have conﬁdence in this variety,
stating that they did not want to risk having a low yield in any year. Farmers in complex,
diverse, risk-prone situations often try to minimize the maximum economic losses for all
income-generating activities (a “minimax” strategy), so choosing a variety that produces

some beans in both non-drought and drought years (i.e. a variety with a good yield

84

stability over the range of local precipitation) instead of a variety that produces well in a

non-drought year and poorly in a drought year is likely a rational management decision.

Researchers often idealize how much farmers understand about their farming systems and
the potential and performance of their crOps, even about basic skills like making yield
comparisons by examining crops in the ﬁeld. After all, farming these crops is a full-time
profession for these men and women, and their skill as farmers in large part determines
their survival. Yet these data indicate that the generalizations that farmers make about
variety performance based on direct observation can be incorrect, and can bias farmers to

the facts of variety performance.

Part of the bias may be explained by the way that farmers understand experimentation.
Stolzenbach explains that farmers understand experimentation as “a continuous and
innovative element of the craﬁ of farming,” and that farmers are not accustomed to
looking at individual elements in an experiment, but rather perceive improvement in a
farm activity as a continuous series of experiments, as “reﬂection in action”
(Stolzenbach, 1994). It seems likely that farmers were responding to other cues in the
ﬁeld, like the perception of disease in the improved variety, which skewed their

perception of plot yield.

5.2.5 Economic Analysis of Field Results

We can compare the strategy that most resembles the area’s typical cropping system with

the strategy that is recommended to the area’s producers to see if the recommendation

85

makes sense for the producers (Table 27). In the case of Los Lirnones in 2003, one
would compare the traditional variety Seda, and 65 kg/ha 18—46—0 fertilizer, and no
herbicide with the improved variety Amadeus-77, 130 kg/ha 18-46-0 fertilizer, and one
pre-germination herbicide application. Savings from reduced weedings, initially
expected from the use of the herbicide, was not included because the herbicide
application was considered ineffective at the time of application (see section 4.1).

Table 27. Partial budget under traditional vs. recommended management, Los Limones,
Moroceli, Honduras, Postrera, 2003 (N=9 plots).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Item Seda Amadeus-77
Variable costs

18-46-0 fertilizer @ Lp 5.83 Lp 379 Lp 758

/kg 65 kg/ha fertilizer 130 kg/ha fertilizer
Urea fertilizer @ Lp 4.8 Lp 156

/kg 32.5 kg/ha urea
Paraquat application @ Lp Lp 136

95 /L 1.43 L herbicide
Total variable costs Lp 379 Lp 1,050

Additional investment Lp 671

Revenue

Weeding (no savings expected)
Light red seed 1,300 kg x Lp ll/kg

@ Lp 11 /kg = Lp 14,300

Dark red seed 1,810 kg x Lp 10.5 /kg
@ Lp 10.5 /kg =Lp 19,005

Total revenue Lp 14,300 Lp 19,005
Additional returns Lp 4,705

 

 

 

 

Note: 18.16 Lempiras (Lp) = 1.00 US. Dollar (January 1, 2005)

In this case the recommended variety Amadeus-77 plus the greater amount of fertilizer
and herbicide was economically worthwhile because with the additional costs of Lp 671
per ha (cost of additional fertilizer and herbicide), the farmers receive an additional Lp
4,705. That is to say, farmers can expect, on average, a Lp 4,705 increase in income

from the additional Lp 671 spent on extra inputs. The marginal beneﬁt to cost ratio is 7.0

86

(Lp 4,705 / Lp 671), a seven-fold increase in returns with the additional investment of

inputs.

In Los Lirnones in 2004, the traditional variety Seda with 65 kg/ha 18-46-0 fertilizer is
compared against the improved variety Amadeus-77 with 130 kg/ha of 18-46-0 fertilizer
without the addition of herbicide (Table 28).

Table 28. Partial budget under traditional vs. recommended management, Los Lirnones,
Moroceli, Honduras, Postrera, 2004 (N=5 plots).

 

 

 

 

 

 

 

 

 

 

 

 

 

Item Seda Amadeus-77
Variable costs
18-46-0 fertilizer at Lp 5.83 /kg Lp 379 Lp 758

65 kg/ha fertilizer 130 kg/ha fertilizer
Urea fertilizer at Lp 4.8 /kg Lp 272

32.5 lg/ha urea

Total variable costs Lp 379 Lp 1,030
Additional investment Lp 651
Revenue
Light red seed 410 kg x Lp 11
@Lpll /kg =Lp4,510
Dark red seed 317 kg x Lp 10.5
@ Lp 10.5 /qq = Lp 3,328.5
Total revenue Lp 4,510 Lp 3,328.5
Additional returns (Lp 1,181.5)

 

 

 

 

Note: 18.16 Lempiras (Lp) = 1.00 U.S. Dollar (January 1, 2005.

In this case, where drought resulted in low yields for both varieties, the traditional variety
Seda out-yielded the improved variety Amadeus-77, and Seda proved more proﬁtable.
With the additional fertilizer costs of Lp 651 per ha, the farmers also lose Lp 1,181.5.
The marginal beneﬁt to cost ratio is -1.8 (Lp -1,181.5 / Lp 651), a decrease in returns

with the improved variety and the additional fertilizer investment.

87

A similar comparison can be made with the data from Lavanderos. Looking at the data
from Farmer 1 (the trial that showed signiﬁcant differences due to variety and fertilizer
use), the traditional variety Paraicito with 65 kg/ha 18-46-0 fertilizer is compared with
the recommended variety Amadeus-77 and the same fertilizer plus the addition of 32.5
kg/ha urea fertilizer (Table 29).

Table 29. Partial budget under traditional vs. recommended management for Producer 1,
2004, Lavanderos, Giiinope, Honduras.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Item Paraicito Amadeus-77
Variable costs

18-46-0 fertilizer @ Lp 5.83 Lp 379 Lp 379

/kg 65 kg/ha fertilizer 65 kg/ha fertilizer
Urea fertilizer @ Lp 4.8 /kg Lp 156

32.5 kg/ha urea

Additional cost of urea Lp 50
application @ l day/ha

Total variable costs Lp 379 Lp 585
Additional investment Lp 206

Revenue

Light red seed 1370 kg x Lp 11
@Lill /kg = Lp 15,070

Dark red seed 420 kg x Lp 10.5
@ Lp10.5 /kg =4410 Lp

Total revenue IA) 15,070 Lp 4,410
Additional returns (Lp 10,660)

 

 

 

 

Note: 18.16 Lempiras (Lp) = 1.00 U.S. Dollar (January 1, 2005)

The difference in yield between Paraicito and Amadeus-77 for Farmer 1 in Lavanderos
was large, and lead to a large difference in beneﬁts for the traditional system over the
recommended system. With the additional fertilizer and application costs of Lp 206 per
ha, the farmers also lose Lp 10,660. The marginal beneﬁt to cost ratio is -51.7 (Lp -
10,660 / Lp 206), a large decrease in returns with the additional fertilizer investment and

improved variety.

88

CHAPTER 6

CONCLUSIONS AND RECOMMENDATIONS

This study concludes with a summary followed by recommendations for breeders and
agronomists, extensionists, and policy makers. Limitations to the study and

recommendations for future studies are also included.

6.1 Project Summary

Common beans are an important source of food for people worldwide, and have been for
over two centuries. Securing proﬁtable agricultural production for impoverished small-
scale farmers is crucial for economic growth in developing countries, and research to
eliminate production constraints is a useful tool to achieve this. Constraints research has
two parts: surveys to identify constraints, and on-farm ﬁeld trials to test solutions to 1
production problems. A participatory methodology, like the Local Agricultural Research
Committee model, blends well with constraints ﬁeld trials while supporting other aspects
of development, like better access to information, skills training and access to loans.
Whereas modernization in Honduras has improved the economy in general, more must be
done to speciﬁcally improve the lot of impoverished small-scale farmers. The Honduran
govemment’s plans to eradicate poverty in the rural sector are a step in the right

direction.

In surveys taken in 2003 and 2004 in 6 communities in rural Honduras with small-scale

producers, farmers identiﬁed drought as the largest constraint to production, followed by

89

the lesser constraints of insects, disease, weeds and lack of fertilizer. Flooding and slugs
were important constraints to a few farmers, and soil and seed quality were considered
important to even fewer. Since farmers said they can adequately control insects and
disease, research was focused on fertilizer and common bean variety to examine possible

increases in production, and weed control to examine savings from ﬁeld labor.

Experiments were designed to test the effect of traditional versus recommended inputs for
bean production in on-farm trials. In 2003, four full factorial (four replications) and ﬁve
supplemental trials (no replication) were sown to examine the effect of variety, fertilizer
and herbicide use in a broadly agro-ecologically representative area of Honduras called
Los Lirnones in the province of Moroceli. Factorial experiment results indicate that
increased fertilizer causes a signiﬁcant increase in yield in some cases; the supplemental
experiments showed with 85% certainty that the recommended package of improved
variety, increased fertilizer and herbicide increased yield. The results of an economic
analysis indicated that the marginal beneﬁt/cost ratio was 7.0 using the improved package

of inputs.

In 2004, the factorial experiment was repeated without herbicide, with an additional
traditional variety and with two additional fertilizer levels on seven farms (three farms
with three replicates, two farms with one replicate). The additional fertilizer treatments
were to try to identify the level of fertilizer that yielded the highest possible farmer

income. Due to terminal drought, one early maturing traditional variety (60 day) out-

90

yielded the improved, longer-season (70 day) variety on one farm and over all ﬁve farms.

Fertilizer had no effect on yield under these drought conditions.

In 2004, a similar experiment was performed in another community called Lavanderos at
a higher altitude, representative of another agro-ecological area. This experiment
involved two farmers with two levels of fertilizer and three varieties (two improved and
one traditional). Results from one farm indicate a slight decrease in yield after the
midseason application of urea fertilizer; and a much higher yield from the traditional

variety compared to the improved varieties.

On a ﬁeld day in 2005, farmers in Los Lirnones were asked to assess bean production in
each experimental plot. Farmers misidentiﬁed the worst-producing plot in one case,
believing that the improved variety Amadeus-77 produced the least instead of the local
variety Marciano; and attributed the apparent poor production to blight when the plants
were in fact only drying down early from drought stress. Farmers also misattributed the
best production to the local variety Marciano three times, when in fact the other local
variety Seda had produced the most. Misidentiﬁcations of bean variety production were
likely due to variety biases that farmers had based on observations in other contexts.
Misperceptions about bean variety performance on the part of the farmers emphasize the
importance of collaborative strategies for development of new bean varieties. Not only
do farmers lack some technical skills that would aid them in making better management
decisions (like the ability to identify disease and understand effective treatments), but

also prejudices based on experiences that are generalized from similar but not identical

91

experiences (like assuming that improved varieties always do poorly in drought

conditions).

Surveys revealed some information relevant to agronomic and economic
recommendations. Farmers usually save seed, and one in ten farmers reported seed-home
anthracnose as important diseases in 2003 and 2004, with similar numbers for mosaic
viruses, so farmers likely underestimate the importance of seed quality. Planting
densities varied, and many farmers would beneﬁt from a higher planting density, up to 60
kg/ha. About half of the farmers fertilized at planting, which may result in an economic
loss in drought conditions; farmers that split applications or apply fertilizer later in the
season can decide not to apply if the crop is suffering a debilitating drought. No farmers
used a Rhizobium inoculant, a practice that can greatly increase production in the absence

of sufﬁcient fertilizer (Giller, 2001).

About half of the farmers do not own the land they use for beans, precluding the use of
land as capital in these cases. Farmers most commonly relied on local sources for loans,

and far fewer took loans from intermediaries. Farmers were generally wary of credit.

Members of local agricultural research committees (LARCs) reported access to technical
information as an important beneﬁt, along with access to loans, group cohesion, help for
their communities, and other educational opportunities. These beneﬁts, however, were

not correlated to other production or economic beneﬁts.

92

Half or more of farmers surveyed identiﬁed marketability and high production as the

important characteristics of their preference for traditional varieties.

6.2 Breeding and Agronomic Recommendations

This study emphasizes the need for common bean varieties that produce well in
environments that are typical and in environments that are prone to terminal drought,
such as early maturity (drought avoidance) or some form of drought tolerance. Farmers
also desire improved varieties that have market characteristics similar to their traditional

varieties, especially color and size.

These results also indicate that farmers would beneﬁt from raising their planting densities
to 50-60 kg of viable seed per hectare to maximize economic yield, and wait to apply
some fertilizer until the ﬁrst weeding in order to assure that there is enough rainfall so
that plants can take up the nutrients provided my fertilizer. Three of the ﬁve replicated
factorial trials showed a signiﬁcant increase in yield due to fertilizer (two in 2003 and
one in 2004), so increased fertilizer use for very low-input farmers to up to 26.8 kg
nitrogen/ha can be recommended with the expectation of a yield increase in more
situations than not. Results from 2003 indicate that packages of improved technologies
most likely beneﬁt farmers on average, and that improved plant varieties have potential to

improve farm proﬁtability.

Some farmers in the survey used foliar sprays made ﬁ'om local materials; a few farmers

also used green manures, despite the loss of production this rotation entailed. Crop

93

rotation, conservation tillage and practices that increase soil organic matter can also have

highly beneﬁcial results for farmers, and should be recommended when appropriate.

6.3 Extension Recommendations

Farmers have a unique and detailed understanding of their own environments, but lack
skills in some areas that would make them better farmers. Learning formalized
experimental methods, and understanding the theory behind crop system ecology (like
pest and disease cycles) and newer technologies (like how pesticides are different and
when to use them most effectively) would help farmers make better decisions about their

own local situations.

Support of farmer experimentation with the use of low-cost technologies like cover crops,
soil conservation and locally-made foliar fertilizers and insecticides may also help

farmers increase their proﬁtability and take advantage of low-input technologies.

6.4 Policy Recommendations

Many farmers in this study (22%) perceive themselves without loan options for the
purchase of farm inputs for common bean production; and other farmers are resistant to
the idea of taking a loan even when an option is available. Understanding farmer
reticence to debt accumulation would be useful in order to relax that constraint once
farmers decide how to increase production. Creating the conditions for loan taking, such
as changing community norms, ensuring collateral, providing insurance, increasing

opportunities to take loans, and other factors can help the development of policy to

94

increase agricultural production through the purchase of effective technologies that are
currently unaffordable. Some kind of crop insurance, if it can be proﬁtably conceived,
would possibly help farmers in risky production areas. Local banks may also be part of
the solution since farmers often increase borrowing when local banks are available (Dunn

and Arbuckle, 2001; Jain, 1996).

6.5 Limitations to the Study

On-farm research typically has a very large amount of variability between farms, thus
requires many replications in order to observe a trend in the experimental variables. The
applicability of the constraints methodology, an on-farm methodology, depends in large
part on the number of replicates. With only nine replicates in 2003, and ﬁve in 2004,
being able to identify signiﬁcant differences in yield due to management practices was
difﬁcult. Future constraints studies would beneﬁt from increasing the number of farms,
especially the more cost-effective supplemental experiments that do not require much
follow-up. Because of the year-to-year difference in rainfall, studies that are conducted
over many years also help to reﬁne understanding of responses to inputs under different

moisture regimes.

The large differences in responses to inputs between farms beg the question of whether
the constraints methodology used for making blanket recommendations to farmers is
useful in this situation. In areas where soil and water resources are more homogeneous,
being able to make generalizations about production is useful for the farmers that live in

that agro-ecosystem. This was the case with Southeast Asian rice producers with

95

irrigation where the constraints methodology was developed. In areas where conditions
are quite variable and out of the practical control of the farmers, one may be able to make
generalizations about production on an “average farm,” but when applied to individual
cases, individual farmers may not be better off with the recommended technology. This
is perhaps the case in this study: even on farms with four replicates (2003), fertilizer had
a signiﬁcant effect in only two of four cases, and no differences due to variety were
evident. However, when analyzed together, nine farms in the same area showed an effect
due to variety and fertilizer. This result makes it tempting to recormnend that all farmers
use more fertilizer and an improved variety. If farmers in a given community pooled all
of their bean harvests and divided their proﬁts, such recommendations make perfect
sense; when considered as separate entities, farmers beneﬁt more with a strategy tailored
to his or her farm, with some farms using the package of technology that is

recommended, and other using parts or none of it.

How can farmers come to know their speciﬁc agro-ecological situations better, and
decide whether to use the technologies that are available to them? When the constraints
methodology is used as a way of involving farmers in the research process so that they
can develop an understanding of how technologies perform on their speciﬁc farms,
farmers learn to make decisions based on their own agro-ecosystems. The addition of the
LARC technique to the constraints methodology created a community-wide dynamic
where technologies could be vetted by interested “early adaptor” and “innovator” farmers

and discussed in a community forum for the diffusion of the appropriate technologies,

96

and advice based on the experiences of individual farmers as well as the LARC group as

a whole.

Whereas no differences between LARC members and non-LARC members were
identiﬁed in this study, more survey questions might have elucidated differences. For
example, research with LARCs in Honduras have noted higher participation in leadership
positions (Humphries et al., 2000), and stronger group leadership, knowledge of research
methodologies, use of improved varieties, and implementation of sustainable farming

practices (Zurita Ramirez, 2003).

This study lacked some important information to truly understand the dynamics proposed
in the objectives. For example, did water availability explain all of the variability in
yield? If not, what other factors played a role, and how does that effect whether a farmer
should consider the use of a given technology? Rainfall did not explain yield very well,
but total water availability might, so the inclusion of data that predicts water use would
be helpful (especially soil water content at the beginning of the season, but also
estimations of solar radiation, and minimum and maximum air temperatures, all of which

can be recorded by a farmer or school).

The data ﬁom this study show that very few farmers in this group had adopted improved
varieties. Although farmers were asked why they preferred traditional varieties, more
could be known about why these farmers have not adopted newer varieties. Poor

performance of improved varieties in Los Lirnones in 2004 gives a hint about why

97

farmers in low-rainfall areas are reluctant to plant them, but questions to get at this were
not included in the survey per se. Also, most major gains seen by improved varieties are
in areas that are highly affected by Bean Golden Yellow Mosaic Virus (BGYMV)
(Mather et al., 2003), and this study area did not register a large BGYMV intensity. The
conclusions about variety performance in this study, thus, should be interpreted as

relevant for areas where BGYMV is not considered important.

6.6 Future Research

Current improved varieties are not as tolerant to drought as local, traditional
varieties. Breeding for beans that perform well in drought situations is worthwhile,
especially considering the world-wide shift of beans to drier and less fertile
environments. This would be an interesting task for participatory breeding since drier

regions are often also more diverse, complex and risk-prone.

No single variety will perform well in every situation. The agronomic trials in 2003
indicate that improved variety and greater fertilizer application may result in higher
yields in years with adequate moisture. Unfortunately, it is impossible to know if a
drought will occur in a given postrera season. However, it is possible to use historical
rainfall data to predict over a long period of time how frequent droughts will occur.
Given this data, it is possible to determine whether there it is possible to plant some area
to a traditional, drought-tolerant variety and the remaining area to an improved (though
less drought-tolerant) variety to maximize the production and economic returns in a

farming system over a longer period of time.

98

Farmers do not use a nitrogen-ﬁxing Rhizobium inoculant. Although present in other
areas, Rhizobium inoculants are not used by the small-scale common bean farmers in
south-central Honduras, nor do they appear to be in the soil. Rhizobium inoculants have
high potential to increase bean yields, especially in the absence of mineral fertilizers.

Research into the use of these inoculants should be done.

Small-scale farmers are not using all of the low-input technologies available to them.
Low-cost techniques that take advantage of natural processes and changes to the
physical growing environment should be tested and promoted to give farmers a

”toolbox ” of techniques to use to increase production and proﬁts. These include soil
conservation techniques, organic fertilizers, cover crops, integrated pest management and

other techniques used in low-input and organic agriculture.

99

 

APPENDICES

100

APPENDIX 1

Summary of Adopted Adjustment Measures

101

Exchange Rate Liberalization. The ﬁxed exchange rate was abandoned, and a ﬂexible
mechanism to determine the rate of exchange was adopted. At the present time this rate
is adjusted on the basis of differences between expected domestic inﬂation in Honduras

and that its major trading partners, together with an auction system.

Liberalization of Internal Trade. Various price controls were eliminated for
agricultural and non-agricultural products.

Interest Rate Liberalization. This was a gradual proves that ended in November 1992.
This liberalization sought to achieve positive real interest rates that would help mobilize
savings and provide a larger supply of loaned funds.

External Openness. The country achieve full membership in the General Agreement I
Tariffs and Trade (GATT) and full entry into the World Trade Organization (WTO); non-
tariff trade barriers were eliminated; tariffs were reduced from a maximum of 90% to a
range of 5-20% and the process of regional tariff uniﬁcation has advanced.

Fiscal Sector. Tax-revenue administration and restructuring and simpliﬁcation of the tax
system was initiated simultaneously with a search for better incentives to investment.
Recent major changes (April 1998) are: 1) a reduction of the marginal tax rate from 42%
to 25% and increased tax exemptions for individuals; 2) an increase in the sales tax from
7% to 12%; and 3) a signiﬁcant reduction in export taxes. In 1997, a tax code was
approved that establishes an integrated legal framework for the tax system, including
general principles, administrative norms and sanctions that include prison.

[from (International Monetary Fund and International Development Fund, 2001).]

102

APPENDIX 2

Principal Constraints for Agricultural Development

103

Taking advantage of agricultural export markets can be one component in a healthy
agricultural sector. In 2003, the World Bank identiﬁed constraints to economic
development via the agricultural sector in Honduras, and stated that these must be
overcome for the country to become competitive in the high value agricultural sector.
These constraints were:
0 Low productivity and diversiﬁcation
Low adaptation of technology
Lack of skilled labor
High costs of distribution
Lack of integration between small farmers and distribution ﬁrms
Trade restrictions in Central America
High barriers to entry into the USA
Lack of systems to certify food safety
Lack of credit for small producers
Lack of land tenure and titling
0 Lack of general support for these changes

(World Bank, 2003)

104

APPENDIX 3

Survey in 2003 for Los Lirnones

105

wee—.8 a" $132.» ~93

—. OEOHEmH—O>m car #5226
$1326 Ea «8.31» Edna—5...? 82.3 ..a name 55 an .8 .88 «ca—.338 no: ..3...— a: .3 5.55:. was.

9.5— one"

20 a1 $39.28

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. a. Ede...
o .3 5.32— 929.919“. 392.8 2.39.2.8. 3.68. Beans—Ea...
o _e 35:38»
a _e 983..."? an a: new» ABE-:8 e 5.9383qu
4 .3 8:9:— g. men—e...
o a. 832.3» .3— .8»...
a new. 8 o. Ede 8.». an .8 33.38 as» 2.55: game—8 w. 8.8...‘
Fe? >3» Car—2. n.a.—22:0 Ema—3am: Awe—Eu.— a». male n.a.—game
r08 Egan 8928 o E:
maman 58 B8 II 8558 0 II E:
1&5...
H888 .08 B8 II 8358 0 II KB
Esau
ﬁeﬁwe»...

$35.38.. N "@536. NHaShSone. .035. «33.33%
95.33 8% «3&9. N "3839 ~u3ew~333€§ km E 33%.33. w "how‘s
3uen§ame\u..3 838333.338

588323.. 33.358 3 E3 mm 5 «.8: 3E 333883.

H3833... N "@333. Nnhwwaame. w "335.3%.

106

 

 

 

=. .Zmﬂgcm < ”Hg—3.5.4.6 UH rOm POﬁHm UH n.a.—On.

P. >22.» 20 «88...» 2.0828...» 82.0» .8 .» 82...» ..8 8.. 32.2.8 ..122.» N98.

20 .8. 3.02.22.

 

 

 

 

 

 

 

 

 

 

o 08 5.12.2. .8 3...... 823...... ..2....8...
o .10 82...... 2. 8.8.0 0 88.2... 2... 8..» 2.52...
o 9......» 2.0» 82...... 8.. 88 2.1.2.2....
o 0:... 8 .» 2...»; ..o 22.2.»...
o .8 2.28 2.2.... 88 82...»...
o 2.82.228 ».m.:.8 2.2.88.8 ..8 ..»8.. .2» .222.» .8 «2.2.28.2. u. 2.8 ..8 ..0 .0 .28... O8 ..»8 8.2....
0 8.08.2... .8 82...»... a: ..8 2022......
o 0:... 823...... .8 82...» 82...... 2. 88 2.8..
o 02.. 20 0. 82.2.2.8 .8 88 .08»
o 9.... 8 .» 22...... .8. 2...» 8*. 22.8 2.2.358. 288. 22.228...
0 0.8 98...... 8 2...» » .» 882.»...

.108 <»...o..».. 88.8.2. >8» c... 5.2.8 .8 028...»: .82.. c... 58......

8E... 82.2.2... 2... 82...» .8 82...» 2.2.2 9».—
..08 28 @220
82.2.0 .08 28
1mg..."
H0820 .08 28 @228

 

 

 

 

 

 

 

 

 

 

 

 

 

632.8...

.0838 8 8.2.5... . "c.8838 23.8.0.8. NH8.3....0 8 03.0 2.2320... m "8.3....0 8 0.2.30 OZQ\....0.<88. «08.32.22 2. 33a

0808.38.82... u 023 «88133.

107

20 ..0. 32.8.0.

NM 2.2.» 2.. «8......» 6822.2...“ 82.2. .8 2. 222.0 .8 32.2.82... .8. 28.0 .. 81:28.2. 2. 2.... .08... 2. 5.22... 3......

0020 .22.»... a. 28.0...
GEE. 81:22.88...

02...... 81.52.88 ......n..... H038 8.8 v. ..0 088...
9.2.8 .8 a...» 81.528 ......N......
02.2.0 2:3 .8 81.52.88... .32.... 2. 2.8 2022.8 .. ..0 0.3...

 

 

o 2928 ....0....22.8 8.2. .2 2.2.8.0 ..8 8 ..»2» 202......8 u. 0...8 ..0. .88 2.2.8.0 3;...» » .8 2.8... ..o .8

...»...8 ..a 9...?— » 2.82. 220... GEE... 2. 202......8 ..o 3.0. ......»..8 ... 8.82.22...

 

 

 

 

 

 

 

 

..08 $802.82.... GEN... .500 ..8 0.2....2. ..2..0..0 ..8 C80 .8
..8 28.0 ..2.....N»..8 81:22.8 Arm. 2:88.02 202......8

92.8»... $2.890
r08 28 ..o w. ..o m.
@828
womgao .08 ..o m. :0 m.
28 magma
..288 .08 ..0 m. so 2.
28 mama..."

 

 

 

 

 

 

 

 

 

6.0.2.2....

58288.0... .8. 28.0.. .n..8....320.= 32.2.... mniﬁﬂd... 2.5.... .2328. ..n.8 32.8... MUSE «828.33%
$2.28 .8 2.58.2... .nuxm...m§...§ Nun .2 2.2.1.3. mummuﬁﬁu ..8 W832: ..na .mEQEnES. ..n23 «8.88%.8...

108

20 mi 50952

Nu >53 Ba «53....» Eon—58...." .526» 9a a: «£38»? an 8.5.:— ao Sigma 2. $152... 33.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o 0533 <88 mam—53o; a 535...
o n.a.—8m amen mom—Emu an E «SE—=6 nan—5:89..»
o >3an ran—.333 5.3—8 .5..»
0 O5: ..3—.33» :23...» 5.3.5 88 w .8 33...
o 052:? ”6:3 a. ran—.339...
r3» <6on Eva non—Emu an .» man—53:3: :8 no 223:6 $12.: an 3:283:
@859.”qu can mamas—a3 .5.—.33» no
a BEE 9.335 32.3%..»
r08 8% macaw no mm
mamﬁao 58 so a
aghagan
H283 ~08 8% so a
«85%
grams.

$24.28 an atknanmmau N unwmummSwB. Nun 3 2.8339 w “taunt? um mmxsmza: «n23 «$313»

109

n... 2.2... .2. «=23... gang—.31? .826» no 2...; 3.2.2538 2.0 :8 2. $12....» ~25.

o :93»: 2mm»... 2. a. .98 no 3&2... 05:2. 2.2....

2.. .5— 2.39.22

 

 

 

 

 

 

 

 

 

 

0 >23... n.a.... 2.522.... a 8.5.. 2. 88m 258...
o 0.5—am 2.5222... 2.2.... $2.25 2.8 2.2—=2: w. E. 22......
o lain: £95.... 292...... a." nunowﬁawaam... 0.5—m... 2.2....
o 0:28 252.22 553... 5......» «2.... 2.2—=28 E. 23......
o 9.2:... 2.:3 2...... 2.2—=22...
Fa... 2.53 Cu: an 2252... an M33533... :3 no 22:2... $2.2... 2.
2.522..» E 2625... .E 95:82....
92:3... Human...» 9.28:. 2.2—=2:
he... 8% ..o m. ..o m.
manna ..o 8. ..o m.
momcsao so mm ..o E.
58 an... so 2. :o m.
@250
H288 so 2. ..o m.
88 8mm :0 2. so m.
@359.

 

 

 

 

 

 

 

 

 

Emlyic an 23:82....“ Nuﬁxmhxsvg. mum 5 2.21.5. w Magnum. mm m3§xmx ..na 520320.... muc?c «mmhmngnat

110

.... U>H0m mO0.0.wOOZO.S.OOm
H... >........ B» «8.2.... ....»m..:......» 826.. ..6 .8 2.62.2.8 ..6 .9 55...... ....» 3.2.26: 2.: .8 .2..6:8 2. 2...... 6: 5.52...

was.

20 a». 32.8.2

o 05.38 52.258 36:» .u 3.25:... O... 6...... 26:»: .8 :38... €68.58 ....» 36:»: B8 ....» .m .58 ..» 6...... a»

»..8...»....: 2.2.8..

U» 8.8 .3858. .....»: 8.2... 6: .: ...6..2...»....: ..6 8:62....
026: 8...»... 6: .a 8.62....» ..6— 81.62....

02...: 8.2... 6: 6. ..8........... ..6. 8:62....

02.: 8...... »: .9 6866..» .6. 8:62....

 

>......8

>......8

2.58

2....8

 

H2... 6: .n
.93...»

 

...6..2...»...: .6
2.»...

 

92:2...

 

..8...»....»

 

 

0866....

 

 

 

 

 

 

111

Zo ..». 309.28
up >......: B» «8.2.: 262.32..» :»»..»: ..» .: :.::.. ..» 2...: .2» 2.3.3: »: 5.22.: 2.8.
o 6:3..3: 2:2. ..» 2...: »: 5.22.: N68... _
o ..2.: .2: ....»..:»...:8 ..». 2.......»».:: 2.3.2: .:::.. ..» 2...:...

 

 

 

 

 

 

 

 

 

F...» :2. ..» .:::.. ..» O..2.:».:: ..» 2.....:»».:: A.:...»:...
2...: 83.32.:
9:...»2.
he.» 28 WEE.» :o m. ..m m U O ..
m»m::..o .2» B8 :0 m. ..m m U 0 ...
mama.»
.888 .2» 28 :o m. ..m m U O ..
Im~::..»
E

9.262638 ..» 2.2.822... ..hhﬁa»..2.:n.:: ..» 2.2:. Mum.»5....:. Unb»...=.»..:». Quﬁou»».§ N." 28......»

7.. F.Z.H>0.0zwm > ..> w.»OUCOO—OZ
.w: 2... .»....»:8. »..:.8 2.: .8 322.8 38.22.1838 .2» «24.3.2: 232.»... ..»:...:..»38 ..» ..»...2 »: 2.22:... 9.82.2.
.: ..2: 2.: .8 .....»..»38 222.8 w. ..2. we ..»...28 :. 2.2...22. :2... 8.82.8 ....:..» ..3. 28 2.8.8..

 

112

 

 

m»..:.: H:..»..:.»..:..8 $2.255.»
.5288 6232. ..» m»:....: .:..:..:».::
.88.... n:....:.. ..» 9.»...

 

 

 

 

 

 

 

 

 

..»—.8 N. ..8...2.» ~58
.. OES>EW=G>m UMP HHEZO

.582... 88 «8......» Ban—8.2.8 82.8» .8 3..» 5:. .8 .8 888 88...»..8 2... a... . 2. .» $8.12.» ~26. 0:... 2.».

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o o. 9.8»...
o .» 8...». .8. 9.8» 8.“. 8888 2.23.2.8. .288. 82888....
a .» ..2....2.8...
o .» n.a.»..a.» .8 .2. a8» .8858 9 5388.33...
0 .» 8...». .8. 28......
o a. 8.8.8.» ..». .98...
o 0:». a... a. 9.8» 8.». .8 .8 88.2.8 ..8 2...?! 3.9.8 w. :88...
~88 >88» 68...». ..2...8..8 5.8.28.» 08...». .8. £2.28.»
m8...
be... 88 E8 II 8888 0 ll
.8
woman... .08 88 I. 8.8.8 0 ll
1%.»..8 8
.288 .08 88 .II 8.8.8 0 II:
@888 *8
SEE»...

338.838.. .nb.§o. NHQSESQRP w USS. 83....»an
n.a....»a 8. 28.9. . "98:9 Nn=o~3».\=.h..8 «.8 .» 8%.».3 w "@898

\. Hogwamcx...» 888:...5838

6888...»... 3.8.8 o .3. mm .a n8» 8. 86829.
52.2.2.9. . HESS». Nnnxmuaic. w "c.2238.

113

29 .8. 5.9.8.9..
... .ngcm a. ”HZ...3.HZHO U... ..Om Poawm UH Ewan...
N. >..2.» 88 5.22..» 6389.28 283» .8 .» 88...» ..8 8.. .528 ..182.» N98.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o O... 8.8.». .o 3.9. 882...... ..938...
o ..9 88...... 2. 3.8.9 9 8988.9 89.. 9...» a.....<9...
o 0......» .3» 88...... 99.. 8.» «.28.»...
o O»... 8 .» 2...». .o 88...».
o .8 ......a 9.....<9 8.» 88...».
o H..a9.....»88 ».n.=.8 2.9.8.28 ..8 ..»89 ..»» .22....» .8 «2.8.98.2. v. 9...8 ..8 ..9 .9 ..»8... O... ..»8 8.8....
o 8.829.... .».. 88...»... m... a... 8982.8...
o O... 88...». .o 88...» 88...... 2. 88 ..3»...
o 0.... ..8 2 39.8.2.3 .3 88 .98..
o 0....— 8 .» 2...». .3. ..3» 8*. 82.8 28.—2.8. .»38. 8282.8... M
o O... .8....9 8 .2... » .» 8982.»... 1
O
..98 <2..o.». >3» c... 5.2.8 .48..» .8 02...». 52.38 C... ..8....9
.8 .......9. 88.2.2.9 .». .3 “2.8.88.2. .8 88...» 88.9 .».
8......» $8.82..
r98 8...... mania ..o w.
..o 8.
898.9 .98 88 ..9 8.
2.2.8 ..o m.
H.883 .98 88 9.8.0 ..o m.
..o m.
6......n8.

3.2.8 .8. 88...... . "c.9828... 2.82.88. N828...» .8 2.... 2.9.8.9... .w "88...... .8 in»... QZQ\....9.829. «”8836... 2. .5.
8898882.... unis 382......

20 .8. 30.88.

PM >82.» 88 «8.2..» ..3m.....»...8 883» .8 2. 88.8... .8 ..3..»..»8.... .8. 2.8... V. 81.88.... 8.. 2.8 .».8 2. .5882.» use”...
0.8.. ..3..»... 8. 88......

c...». ..8......N»...8..
0......8 ..8......N»...8 ......s... ..»...»8 8...... w. E. 8.82....
05...... .8 8».» ..2.....N»..8 5...»...
0...»... »....8. .8 ..8......N»...8.. 5.2.... 8.. 8.8 8882.... u. 8.. 8......

 

 

0 >525... .........8.2.8 .82. ..» .........8... ....8 8 ..»8» .»..8=.»...8 v. 8...... 8... H88 .........8... 3....» » .»8 ..».88 .8 .»8

..»...»8 .8 3.... » 2.888.. 88...... 0....8... .... 888...»...8 .8 .......... .5.»...8 .» 8.88.8».

 

 

 

 

 

 

 

 

......8 ....8..»..»8.... 5.8.. ........ .8 08...». ..8........ .8 C2. .8
.8 2.8... ..8......»»...8 ..8......N»...8 Arm. »....8»8.... .»..8...»...8

..»..8»... ..»..8»...
r08 88 ..o w. ..o m.
«..»»...
M8858 .08 ..o m. ..o w.
8.8 88.8
88383 .08 ..o m. ..o m.
8.8 m8»...

 

 

 

 

 

 

 

 

 

E

3.2.88...» .8. 2.8.9. . "888.82.... 5.85.. .3382... 8.5.... .1888: «use 88...... .. n23 «8.488.338.
$82.28 .8 8.88.8.8. .nhx8883... mu. .2 8.23.8. ..u88»... .8 «2.8.8: A". 3.262.... .. no.3 «838.53...

115

20 no. 5098::

Nu >32.» Ba «53:» Edna—.310 wan—.2. no a: 83.38%» .3 8.56— % Sign... a: $25.2.» N98.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o n.a.—Sm <38 9859—35 a 325...
o 0.338 a»... mam—Ema an .9 £235.» 38.5:33
o 358 32.53%... 5.3.8 .5...
o GS: ..»—.335» 3:3» moi—em 88 w. .3 33...
o n.a.—ac “i=8 a. 32.3393
he? <82 Emu nan—Emu no 3 «2...:—=53: :3 no Zea—:6 :12? an 3:233:
985...»; can mama—£35 32.303: .3
a 325 c5225 ..»—.52..»
r08 8% mum—an so a
montage .08 no mm
9% magma
H388 .08 8% no &
[magma
63%.?»

33.2? An aimnanmmau NuuanESwB. Num 5 h8g3. .w "956%.... km mmziznw. «v.23 32.3.3»

116

NA >33.» 08 n83...» Emu—522.... .823» .3 02.8 $305538 £8 83 2. $82.2.» N38.

20 .3. 309.28

 

 

 

 

 

 

 

 

 

 

o :12»: 0.5»... 2. a. .08 .3 2.30.... 05:8 2.2....
.- >23... in...» 0.533...» » 3....» .3 8.»... 0.58...
0 02.3.8 0.53338 2.2.... $03.8 88 0300020 w. ..0 02.8...
o =»...».. Ens—.8 32.08 .3 222.52.»..8... 9.3.8 2.2....
o 9.3.8 0.0.2.88 02.33... 003.8 8.8 3.0.2.88 ..0 02.8...
0 9.3.30 »...3... 8.8 0.2.8.8...
r08 23»... C80 .3 2053.... .3 .» .w..~8...08..»38 :80 .3 2033.... 02.2.0 .3
03.9.3... 03.8.3...» .26....20 .3. 3:38....»
» 2.2.3»... 3.09.20
@003»...
..08 B8 :0 m. :0 m.
@323 :0 m. 00 m.
momgmo ..0 m. ..0 m.
.08 B8 :0 m. :0 m.
magma
.3880 :0 m. ..0 2.
.08 B8 :0 m. :0 2.
8:8

 

 

 

 

 

 

 

 

 

502.000 an. 01302.0..." . “.38....330. NH0 .0 2.2330. wummwhmmw «.8 833.30... «H0 53032.0... ....H0.x0 88.08.28...

117

u... U>H0m mO0.0.wOOZO.S.OOm
u... >......» ..8 «83...» 039.82.... 826» .3 .8 5.23.2.8 .3 .» ..»B...» ..»o »w.....»...... 2... .8 83.2.8 .3 2...... 2. ..822.»

N3“...

20 .8. 509.20.

a 05.38 .88....8 22.8 .» 35...»... 0.8 2.2. 22.2. .8 0....8... 3.882.»... ....o 32.2. 0.8 ..»a 5 »n8 .3 8...... .8
2.8.02.2. 2.5.8..

 

 

o .8 8.8 02.8038. ..».2. 8.8.... 2. .» 32.2.8...» .3 82.2.0...
o 0.22. 8.8.... 2. .» 83......» ..n. 8.2.2.0...
o 0...... 8.8.... 2. a. ..8...2..8 .3. 82.2.0...
o 0...... 8.8.... 2. .» 388..» .3. 83.2.0...
>......8 >......8 2...:... 258
H8». 2. .»
..»E...»

 

$808.83.. ..a
8.8...

 

m.2......»

 

58...»...8

 

 

088..»

 

 

 

 

 

 

118

20 an; $89.08..

PM >32.» 3» «gag...» Eda—531a 326» an _» B»: a» 3:.» £8 8:2.»8 a: $25.2.» 38.
o 02.2%» S»: no 2:.» a: wean—.2.» 25w...» .
0 $3.» new 52.33:... an. 33:83» 8.5.»; B»: an e_=.»..V

 

 

 

 

 

 

 

 

 

r30 Cu» no 325 an 012.33: 9» $39.83:
:5.» 832:2.» gamma»;
9:235
r08 Eggaa no mm mum m U 0
mamEEO .08 8% so mm Hum m U 0
magma
H.383 58 5% no & mm m U 0
[mean
665%..

Q§-anm¢=§ a» Eggnnmmau wmuwamhagam: an .329 mummnSuE. bnbaimlvm. Gnmeumnra

Wu >5...» Ba «58:» Eon-.53...» .83.» um Emu—ac.

a Guam .8 83.32.» .5 539.22. a» :25.» 856—3»... Ami—ma»; a»
o dune..— .EE 3..» ”..»—.».? 5.2.» a» _» H55» 9...»:8 $826.»... 9:333 .5 a.
a 0:558» Ewen Emma» a: _» 82.3%v

2.3m.

‘
m—

‘

119

20 an. 3.02.22.

7.. 53:50—02Hm .9 Cr wwOUcnn—OZ
H» .28 32.2.8. 2:28 82. .8 ..»228 28 .2624»...8 a...“ «2.232.. 232.8 32:22.88 .3 2.:..— 2. c8322... A3882.
—» r2.» 2... .8 332.228 5.22.8 V. 32. N: «1.28 »_ 2.2—=22. ..».» 8.8228 :2...» a»! 28 2.228..

 

 

 

mam...» 2.3.22.2»... mans—m2...»
3288 02...».— ..o m..2.=» 22.222.
2.828 02...».— ..a we»... w»_.8»u

 

 

 

 

 

 

 

 

 

c.22— 32... _» 2.2.22. ..o 82. _8 2828 2.228 2... 2 2...?» ..» an: _ 2. $122.» ~23 .2» 8.. 2. $122.» N98...
A295»... 2. 2
mm _» ..8—28.» 8 2..

0 ..3—22.» 2..» 2.25»... A293»... 2. 2 m. .8222. 2. 982.2».
552..» 2.... ..»—4:223... 92:0»... 2. 2 mm .822...» 2. 382.3».
..3—mu»... 82.28.» ..» 22.2.. ..a 2288... A2325 2. 2 m. ..8_......_o 2. 982%.».
I»! 2.... ..»22. ..»a w... 2. 22.22.»... :25»... 2. 2 mm ..8—.2...a 2. ..82.=.».

 

 

 

 

GEO—>m n.a.—w m: Nisan—5902.2

120

APPENDIX 4

Survey in 2004 for Los Limones

121

20 ..». $39.20..

$82.2.» 3....

.. 0>.w>0...>.w.m.$.0>m UHF Hmwgzc

$222.0 2» «8.2..» 2.8.2.2..» »»2.»» ..» 3..» .20 ..» .8 .28 82.2.2.8 ..2. ..»...0. 8.» 20. $2. ..»<0.. .222» .2. .........0
..» .8 .28 ....» 832...... a2. 2....0. 2. $82.2.» 8.» 2.0. 02.. 2.». .

2 2...»... .22»... .» .......».. ..». 26». ».n. .»..»8. 22.228.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o .» $2.222»...
o .» 2.22.2» ..» 2. 8...» .2228 0 5.0223»...
c .» o»....».. ..». 2.20...
a .» .»..2.2» ..». .2»...
o 02.. 8 ». 2d» .2». ..» .8 .2223... ..»» 22...». 3.0.8 v. 02.8... 52.2.0 ..» 8.8 .22.»..8. 2.220 8 2. .2220 .»....»..0...
o m» ..»...» ..» .2» ..2»22. ..» 2......8. 8 ..»».... .2» 2....» ..» 2......8 82.2.2.8 ..»...8 w. ..» .2» 22.2.» 832...». C22.
22.» ..2»20..8 .....»8... m. :22. ..»» ..2»22.8 ..» 2......8. 3.23.82» 820 80...
F0.» >..»» 02...»: $2.22..» 5.22.2» 0»....».. ..». m..».0 ...»..2.2»
.0.» 2E2...» 2... 0 .2.
m»m.2..0 .0.» 28 was...“ 2... 0 ..2
H2020 .0.» 28 m2...» 2... 0 .2.
>..»» .2». ..»» 2......» .0000. 00000000000. 0000000A V0000.
>8» 20% .0000. 00000000000. 0000000n 00000.
0.22%....
$2.22.». .n$.»..0. Nu»...$....2.0. .33.... 2.2.22.0 6.8.2.2.... 3.328 0 .9. a.» .2 08» ..». 2.02.22.
02.2... ..». 2.2.... .2283... NH..0.....».\...$..00 ..» .2 82...... .w "$0.28 22.2.2... . H.202... NHE..8.»».0. .. ”2.02220.

A "00.22202... 00.82.23.220

 

 

$082028 v. 8028.028 ..» 2......8.

 

 

122

20 .8. $30220.

I. .ZmGZOw < ”02....SEZHO UH ..Om .2.me 5.. 2.2.0..
P. >22.» .8 «222...» $1222.28 82.0» .8 .» 8.2...» ..8 22.. v. a. 82.20 .8 .» 0082.» .8 88 $822.».
0 02... 222.2. 8 2....0. 8.2.2.0...
F0 82...... 2. 8.8.0 0 8022.0 002 02.» 2222...
0222.0 .8 2.2. 82.2.... 88.. 00.. 8.» 222.2... 220...... .» 22.22. 8. 2...».
.8 .828 02.5.0 8.» 82...»...
02... 2.2222. 8 82....» 82.2... 2. 88 2...»... e222... .» 22.22. .8 ..80.
0.2.2.0 .8 2....0. 2.2.50 88 .08... 220...... .» 22.22. .8 ..80.
08022228 2228 2.0.2.228 ..8 ..»8» :2» .222.» .8 «2.2.22.2. v. 02.8 ..8 20 .0 ..»82. 0.8 ..»8 28.2....

 

o >.m228 .2.0...8.0..8 8.8202»: .». 220.8 82...»... ..»..» .» 22828 222—2.». 0. 02.8 282.22. .» 82...» 00.20 8.
0.8 ..»8 8.2.... m. 8.8202» 82...». 82 ..28. 20282.0 .0 ..»8...

 

o H2 .08.. 22.2.0 8 .......0. 0080.... 88.. 88 2.0... 2222... 22...»..8 .8 ..80.
o .8 .28 o. 2....0. ..8 88.. 0080.... 88 2.0. .88.. 0022228... ».m0... m. ..8 8.. 2.22.0 8 0022.22.20... 2222...

 

 

 

22...»..8 .8 ..80. > ..8 .2820...
o > ..».2. 8 82...... .8 2.2.0.8... 0.2.2.8 82.22.88 ..».. n2 2. 00222.22...
0 C22. 82...... .......0.8. w. ..8m0 002.80...
.08 <»...2.2. >800. >8» C2. $22.8- 02.222. 02... .8222 c... 002.2.
.8 8.8. 8.2.82.0 ..».. 8.2...» .8 8.2...» 2. 82.0 ..».. 2.3

 

..08 28 9.2.8

 

 

mnmgao .08 28
92.8

 

 

 

.8880 .08 2.8 22.8

 

 

 

 

 

 

 

 

 

 

 

 

 

002.822.0028 .8 82.20.. N ”0080...... 0:82.08... NH8§E0 8 080 00.2.0.0: . ”.05....0 0.8 0.230 02000820. «H8382...
02 .20 020005822. MH0§0 «0.801..0.-002$~0&02 . ”28280.30. Nu..0=00 02 .0 220.00.. wu080 A80.0.2.0. 0 ”020 «0803.05

123

 

20 ..n. 58988.

Nb >52... Bo «5.35....» 550M532? 522.2. ..o a: 5.25.... ..o 552.2525: ..a. 2...... u. 31.53.55 «5 2.8 .38 «5 8.8 $8.52....
095.. ..3—.».... o. 28......

62:5... ..oZ...nw5n8...

655.8 nous—.5558... ..»..an

0555... ..a 2...» 31.3553 52.5.... 05.5... 88......

0555.... 51.2.. .8 ..»..»...55588...

>.m..5..u 5.6....838 5855 ..5 ..........2.. ....o 81.5.55 .585558 w. ..»....u 5... an... 8 ..5 .57... 55“.... ....a 3.5..» 5 .5...
55.58 ..o .58 5.55.8. ..a 3.... 5 2.82. 5.8.5... CBS. ..5 .585558 ..o 3...... ..5358 .5 8.25555...

o 0555.... 5.32. ..5.2.o 85...... .5.... 3...... .8 3.4.5558... 5.8—55.5.8. .35.? 558...... 58.2. 8... .5.......5..n.~.5...

 

 

 

 

 

 

 

 

r58 wan—.5552? C55... 5.... ..a 0553...... ..nmn... ..2........ ..o :85 ..» 5.55...
..o 8.8... ..»..»...5558 3.4.5558 Arm. 55:28....5 .559 ..o

$55.55... c5535... ..5......5..

r08 .58 50 m. so m. .

mama...

mom55ao .08 50 8. 50 m.

.58 95......

H0880 .08 5o 8.. 50 w.

.58 mag...“

 

 

 

 

 

 

 

 

 

 

 

65.....n3.

$355323. km. 2.5.5.. ~n~§€a§2§ 355...... N33525: 55.5.... .1523. «use 53555.. M no.3 «858.38%
$25.5E5 an 51.5523... .uuxmmESwg. mu: .5 2.2553. w "$8558 5.5 w2§=ﬁa Ann .5\.535...5. u “2.... «8.58.553.
5.52:5 5.5 ..=\S§52.5.=.. TEEN... Nuamxohmnanxﬁ ..HDNGS... «um... w. m no.3 Q8885

124

20 ..5. 58055.0.

Pu >52.» .55 «58.8.» 555N538? 825» ..5 8.. 8...».5m.» ..5 505...... ..5 35.558 55 8.5 58.52.».

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 0555.8 «.558 585.355.. » .5»5.....
5 0555.8 52.8058 .5.—..»..»..05 ..»..» 58.55.55: 3.... 5. .08...
5 0555.8 5.8 .5»..».»....5 ..»..» 825.55.. 5. 58.51558...
5 05.5.8 ...8 58.5.8 ..5 .» 8.55....» 55835.55...»
5 25.5... 52.5.5.5»...
o 05... 52.5.5...» »....5.....
o 0555.... »....5... 5. 52.5.5.5»...
o 0.55.... 8.5.. ..5.2.5 ..»—.2. .58 8......5 .8 52.35.58 5.8—.05....8. .55.... 555....5 8.5.. 8.» .5.......5»5....5...
..08 <558 ..58058- 6.8 58.5.8 :80 ..5 20.5....5 $510.... ..5 5.55 5.55.5
58.55.55... ...8 ..5 .» 52.5.5...» ..5 55:555....5 .0 ..5
w 5550 «555.555....5 c535»... .52....5...» .5.......5.
..55 58.5».55...
..08 .58 «B505 50 m.
momgao .08 50 m.
.58 @5505
..5858 .08 5.8 50 m.
«3505
05.5.?“

555.555 55 55.8555...” . "5.8.5.535. NH5 .5 5.5535. 5 ”5.8555... 55 ”535.55... «”55... E85155.
3.558 5.5 ..5.\5.§55..5:.. . "55$... NH5®S5555519 wu~u~059 «MEAN u "5.3 «5851......

125

N... >52.» .55 “8.5.5.5 5..5N55.»...5 552.55 55 52.8 5n..5..5.5..58 ..55 85 55 555.525 N58.

25 55. 585558..

 

 

 

 

 

 

 

a ..55.55 5.5n8 55 5. .5.5 55 3.5.... 055.8 2.55...
o >555... 5.9.5 5.55.5.5» 5 558» 55 55.8 5.5n8...
o 055.8 5.55.5.5». 5..»5... 0555.5 58.5.55... 0555.5 55 .5 5.35.5.5» 8... 55 5. 55...<5...
5 055555 8.55 ..5.5..5 5552. .555 555.5 .8 5.5m5.5.58 58555.58. 55555 58..55 8.55 8.5 .5..5...555....5...
F5... 5.5n8 C55 55 25.5555 55 .5 5.55... 55.. 0555555 55 5555.5 55
5.5m5.5.5» 5.5m5.5.55 55.555 53555.5 .5..5..5.»5....5
$55.55... 855..
50.5 .55... 5o m.
@558 50 m.
m5m555o 50 m.
.95 5.8 50 m.
«.5555
55.553 50 m.
.95 .58 55 m.
@5555

 

 

 

 

 

 

 

 

 

5.5535 .55 ..5\.5..355..55.. . “SEW... Nu5m55555=555 m ”b.0555 5 “5.5.... M ”55.5 E55515...»

126

 

 

 

 

 

 

 

 

 

 

 

 

o =55.55 55558 5.»..58 55 5535555558.. 055.5... 2.55...
5 055.8 53555.8 5.5.555
. 055555 55555 8.8 5358.8...
o 05555.. 8.55 5.5..5 5552. 5.8 .55... .8 ..55 .5.58 5... 55.5.8 55555 58555 8.55 8.» 3.5.5.5255...
F53 H555355555u 6.5 55 25.5555 52.555 55 5.55.5 55.. 0555555 55 5555.5 55
53555... 55. 555555.55 55.55.. 53555... 355355.55
555.5»... 53555.5 8555
55.5 .58 55 5.
E5555 5.. m.
m5m555o 55 m.
.5... .58 55 m.
5555
55.553 55 m.
.58 .58 5.. m.
@5555

 

 

 

 

 

 

552.555 55 555555.55. .n.5..55..53555. NM5 .5 5.5.55.5. .wu555555... 55 5555.55... 5H5 355555.55. u "5.3 «8555.555...

3555.5 55 3.3555255. 5 “5....W5. NH5®35555519 .wub5055. AME... ..n5...5 «55553.5...»

127

 

 

25 55. 53555.8

:5 U>HOm m00.0—w0020.¢=00m
9. >553. 5.5 «8.55.5 555555.515 555.55 55 .8 3.55.5.8 55 .5 5.5.5.5 ..55 5555535 555 .8 .5....558 55 5....5. 55 58.55....
N55“...
5 0555.55 5555555... ..555 .5 3.5.55... 055 5555 ..5555 .8 5.58... Q5835... ..55 ..5555 .555 ..55 .m 558 55 5555 55
5555.55.55 55558..
U5 55.55 558558. 55.55 55.5..5 55 .5 5..555..55.55 55 .5....555...
05.55 55.5..5 55 .5 5.5.55.5 55. .5....555...
05.55 55.8..5 55 5. 585.5555 55. .5....555...
05.55 55.5..5 55 .5 585555 55. .5....555...

 

>55..8 >55..55 2.58 2.58

 

5.5.5. 55 .5
..55....5

 

5355.55.55 55
555.5

 

95.55....

 

5555.255

 

 

 

 

 

 

 

085555

 

128

 

 

 

wk ..»—.2... So «53...... 53“....31a 98.6.. ..a r. 3...... an .5.... a..." 8.5.3... 2. 55.3.... N93.
0 02.5.3 5...... an 9...... 2. $8.3... N98...
0 5...» a: 3233...". ..»— encmgaaau 2......»3 5...... no 2...»... > £5. 38......
o 9.9.3... 332...... .3me3: n.a.... 2:... 327.32.... 9.9.8.. n.a.. c.3332. 2. a...» 29.32....

 

 

 

 

 

 

 

 

 

 

r98 :8 an 5...... an 0338...... an 33:83.. $522.... 5.8...
3...» 833:...»
9.32.5
he... 8% magma so u. ..m m U 0
$9858-...me
momgao .0... 8% ..o mm mm m U 0
Imagaa 33885.8“
H288 .08 8% ..o m. ..m m U 0
mama wonmoawmdmmm”
g

Qﬁnsan€=mu ma bgmznnmmau wmnwxmhawanmo... am .358. mummmSuS. bubmarmmwi. Quﬁoumnra

129

20 00. ....00:0.0.

Pu m. 880 :22... 2.8.2. 0.02.0 ..»..» m: a:...<0 08 .......0.. 2.0.8 82. 8:... 0.8.2.8... 0:... 8 .» 02.0.8.2. ..»..0 0:8 880
2.8.2.» 8. 0.02.0...

 

2.2.2.» 0.. 2.8.2. 02.0.8.2. 08. 2.8.200

 

$8.». 00 :0 2:50

 

5.8.». 0.... 58280.30

 

38.0.. 08 E.» 2...:» .02..

 

.48.». 0o 2. ..»..00

 

 

0.8 380005

 

 

 

E
00:08.0: 08. 2.8839. . n2... 00:08.0... Nuns: Bu: 8 .388... «20.08 E08288. 015.0 «080....2i

u... >02.» ..8 «8.2..» 2.82.8.8 80...... 8:... 0...8 »n..<.0»08 ..:o .8 .8.......».. .0280... 02:8 82. .8 »a..<.0»08 0:8

..8:..».. 2. .» «2.2.8.» 08 0.02.0. .2. 08 880 820 08 .8 0...8 .8800»... ..2. 82.5.8: 8: ..»2...»...
.. N.

 

w. A.

 

m. 0.

 

um >02.» ..8 «8.2..» ....om:...2._a 80....8 a. 2.2.5.
o a: a: 2.0 ..»»... 880...
o 0:2..8 .30... 8.8.... 2. .» 82...»... 2.08.

130

 

20 00. 08000.0.

 

.<. r..(.....>0.02wm > F> 0WOCC00.0Z
>..2.» 020.2.» ....0«::.2..0 00.2.0 .8 ..»0.2.8 ..:0 ..00:02. :0 ..:0: ..0:0.....0:.0 00 .......0. 0: ..8...2.». .w8..0:00 0 .8
0.0m::.8 00: 2.5 .300.~H>2.:u. 0000 .300ﬂﬁ>2...w. 0 2O .300.~H>2....w.

 

 

 

 

 

 

 

 

 

 

 

2...:.» 0000200008 ..»..» 00 00......0008
30.008 2.0.» 02.000 00 m0.:...» .:::000.0:

.8028 2.0.0 02.000 00 m:0.0 ..»—.88

E

.u .5.... 0:00.220. .0." 0000 2.00.230. ..u 20 9.00.230.

<. 0OZ....>2N> > P>m ..CHZHﬁm 6.. .200§>0.02 mOwg 0.~O..C00.0Z
>02.» 050.2.» 0.0m::.2..0 0:0:.» 00:00:00 ..0:0 8.00 0.. .8 0080.8 00 .0 ....00:00.0: 00 .400. A8 .0 00.. .8 0.m:.0:.8
.8...:0.2.8 0 0080:»... ”800:00 0 .8 000m::.8 00: .Sd0m.> 0OZE>2N>. F00 U... 0020.>ZN>. 0 0O0>

0OZ...>2N>.

 

 

 

 

 

 

 

 

 

<00...0 N02002:.
>m..0..00:»...» g0.
5.0.5 .030.
00....»0.

.u .58.... 0003.08... Nu 0W0 00 000.008... mu 000.. 0000.08».

5. 00203025 6.... .L>m 002..me ..H .2<...m....0>0.02 >0a0OF> PO0>P
$0.0 ..»..» 2.02.2.8 00. 0.>... 0:...8 02. .8 02.2.0.8 0:0 8.00 02.0...0 00 00.. 2.02....» 00 0: 0.>.....

DEO.>m 00w 0.. ..>.~....0.0>0.02

131

APPENDIX 5

Survey in 2004 for communities other than Los Limones

132

 

 

Zo n.a. $39.23

$82.2.» 2.8

.. 0>ﬁ>0ﬁ>am§0>m UHF HMEZO

$1.52... So «53...» Ban—.52.? 2.2.0» ..a 3..» E... ..» .8. .39.. «2...:»..8 2... ..»..... 2. _» $92.2.» 2...... $2. ..»<.... rung.»
E. 2......» ..a .8 .3»... A...» 22......» 2... 5...». 2. an.» ..»..» 2. $82.2.» ..». 2..» N93. 0:». 2.».

a. 9...»... can»... .» 5.3.... ..». ..3». 9m. .»..a»... B»..n»..»u.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o .» ..».....2.8...

o .» $228.» ..a 2. 8m» 9......8» » 55.52.62...

. .» 8.2.2. ..». 2.2»...

o .» 852.2» ..o. .98...

o 0....— ou ». m3» 8.». ..» .8 82.2.8 ..»» 2:9». 3.28. v. 33.4.... 0:... on a. ..3» .3». ..a a: ..3—..»..»....
..».» >3» can»; $2.32.? 5.22.2» 0»....».. ..». mg... €2.25.»
.68 B...» «8....» B... o .3.

Mama...» .08 3% «.»....» a... o .8.

H988 .08 8% mg...» 8... o .3.

>8» 8.». be» 2...?» .0000. EXXXVOOOA VCCCCOCn E
>8»B~o% VCCOCA VOOOOOCCOCCA VCOCOCOA é
Crag».

$2.5.»32 N "$8.26. Nu.:n~..=»».o. quS. Slims.» 5.25.2.9 3.328.. c .3. mm .2 8.8 n.a. $38.29.
0332. a». 2.29. . H.328. NuzexSQSNEwQ a» .a 3%.»... w who...» $2832.». . "$3.26. Muhamuaio. .w "n.a....ako.

A nenzhamox...» no...§:.§..§8

133

2.. ..0. $8288..

... .chgcm < gzu—EHZHO UH 20m ..Q—ﬁw an ...—2......
N. >......» 50 «8.2..» $..0n:...»...0 808» ..0 .» 8......» ....0 ....» w. 0. 28...... ..0 .» 2.82.» ..»..»...0 $82.2.» N28.

0 0..» 820...... ..0 2......— 85...».

o F» 85....» 8:22... 2... :2.» 2:28... 0:2..8 8:88 ..0 2.... ... 2.2.28 ..0. 02... 0:52....

0 0:2.8 ..0 2.0» 85....» :22. 2... 8.» 822...»... 252:... .» ..»...2. ..0. 2.0».

0 U0 ..»....0 .....:<.. 8.» 85...». w. 2.5$..» .» 85E». » ..:0. $82.. .» 2.5$..»...

o 0:» 02.2....» ..0 85...» 85....» 0.. 8.0 2.0»... 252:... ... ..».....» ..0 $8...

. 0:». ..8 2 8.2.5.2.... ..0 8.0 ....0... 28...... .» 5......» ..0 $8...

0 >55...» $328.28 8.00295»: .».. 50......8 8......8 $»..» ... 28.2.8 205...». w. .....8 205...»: .» 85...» 2.5.. 8.
0..» ..»00 8.2.... m. 8.002....» 85...». 2. 2:» 5.52.... ... ..»00...
H: 2...... 0:2.2. ..0 2...... 2.82.» 8.0.. 8.0 3... $250..» w. $822.»... 252:... :5...»._.8 ..0 $8...
0 =0 .».... 2 2...... ..:0 8.0.. 0802.» 8.0 2.... 8.0.. 2.52.2»..N» ».m..... m. 2.0 8.. .52.... 80 2.52.222»... 28......

..»...»28 ..0 $8... 0:» $82.. 82...»...
o > .....2. 8 8:...» .8 ......28 2...2.50...»...... $:.$2..». ......0m»...
0 $8» ».n:_.» <0» ....0 8:...» ......28 w. .8»: .0...» ..8 2.5$..»...8...

 

 

 

 

.....0 52.2.2. >80. >8» C... $8.20- $82.. 0»......».. C... .w0.....5 c...
..0 2...... 85.2.2... ..».. 85...» 85...» ..0 8......» ..».. .2.... ..»..

 

..o.0 58 $»..»...

 

 

m0m::..o ....0 58
82.8

 

 

 

...0803 .20 ......» @820

 

 

 

 

 

 

 

 

 

 

 

 

 

022"»...
3.0...» ..0 83...». . "0800.5... 8.02.80... ..."..03...» ..0 ......» $328.8. .w H85...» ..0 2%....» 9202.382... ..n»»......»..» 0.. ....»
»w..»0»...0..0.».. hue...» 2.08....»».

134

20 ..0. $89.08.

N.» >......» 50 «8......» $8m:...».._0 82.0» ..0 «: 52...... ..0 $8$»..»2».. ..0. «:0... v. 2 :8 ..0 ..0......u»...0« 0.. «8 ....0... 0..

$82.2.» 2.8.
o 0»3.. $..0$»..» 2 «8......

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o m. ..0:$» .308... 2. 0. 2.2... u. 2 «:....:0.. ..0. ....0. 05...... 08.» $.... ...». w. 0:»...»« ...8 02$»...
0 m. ..0:$» 388 0.. .» $8$»..»2».. ..0. ....0. 0:»...8 388 ..0:$» ..:..»...0 0:2..»« ...»«... > ..:0 $82.. $»«» .8 388...
o ......u» m..»3..u....0 : ...... ..0.....2..» 0.. ... $..0$»..»2».. ..0. 80...... 05...... 8...»... w. » ..:0 $82.. $.... ::...».....
o 98$» 388 0.. .» »$..0»2».. ..0. ..0.....2..»... 0:2..»« .28... > ..:0 $82.. $»m» .8 388...
.....0 $8$»..»2».. ...... ..:0w.0« ..«c ..0 3..»..« ..0.....2..» 0.32....» $82... :2. 5.88
..0 «:0... ...8 $82.. ..0. ...»« $82.. 25...»... :.....».. .28 82..
.68 38
[magma
M09520 ....0
58 @250
...0803 ....0
58 man»...
02.58..

3.2.5.8....» 8. «.8.». .n$8$»..»2.»= .88.... Nn§»8.»» 92.3.... m H.308... .. ".8 $8.88». u no...» «880.28»...

135

 

 

 

 

 

 

 

 

o 0:28 nag—Ewaow 33.58 553...
o DES—3 an 3..» 31:55? 33:...» ..3—5....» > ea 3.83 8:51. _» 2...:—Emu» 0.5—ac 2E3 3m noun—waving
o Cam—mug Emma arc—5 3:2...» OE»...v GEE—8 an 2.28 85: 553... > ea 3.33 8.3.3... as ea 3952.8 3 ”5:23
o 02.: Ban: on _w 3:883: no stone 3:2...» GEM—:3 $5..» > A...» 3.8? tuna .8 538...
~58 Ciao ﬂea an 02:52: 5693 3123 an :3 an 538
..0—4:553 33.5.3» .532. 5.3»; 3:583:
Eamon-G E3 383
r08 Bum no mm
Manama
momcaao 58 so mm
5% magma
H0803 38 so mm
5% main

 

 

 

 

 

 

 

 

 

wax—.2? an aﬁananaﬁ NHhxmthwS. Num 5 b.9332. .w ”hone kahuamm km “@3333. «HE 383:0». m n23 «mmﬁmnSnaQ

136

N... 2.2... 3» «=22...- ....a«2.......o 22.2. .... 2. 8.38m... .... 23...... ..o 32.3 2. 5.2.2... 3......

2.. ..a. $39.22.

 

 

 

 

 

 

o 9.2.3.. <88 ..3—.... ... ....o a 32......
o 9.2:... v2.2.3... .3.............. ......» ...—.22. 8.... o. .29..
o 9.2.2... ...... 2.2.2.2.... ......a 23.3... a 23.22...
o .... 8.8 22.2.36 ....a 23.222. 2 .2... 2.228 2.2. 3.38...
o 9.2.8.. ...... 22.2.2 ..a ......2. ...—2.... 2 2...... 23.22...
0 >25. «...—3.32.0 = .5... .2.—22.... 28.2.... ..a ......2. 3.2.... 2 2...... ......» 22.2.22. 2 32.8...
o 9.... .2.—22.... 9.2.2...
o a... ....m 3.32.... 9.2.2. ... ..2....2....... 9.2.... ......N... w. a ...... 2.02.. ...... 222......
0 O2..... 3.3.. 2. ... 3.2.2.2.... .... ..»..—22...... 9.2.36 .22.... > ....o 2.02.. .25.. ...» 3.52....
.....o <82 ....a3.... ..2... .....u 2.3.2.9 ..2........ .... G329... ....8... ...... ....
..3—.... .— ..3....... ....u. .... 22.222. 3.222.. .2229... 33....
32... 22... .....a 338 M23. .8222...— .....u 2.8.
.....a 3% 2.2.20
mamgao .2...
3...... magma
.682... ...8 3....
6.2:...

 

 

 

 

 

 

 

 

 

 

 

 

 

137

20 cl $09.22

Pm >33.» 5... «.581» Eon—53.1» moo—.2. an 92.8 33:583....“ ...—a :8 a: $83.2.» uccm.

 

 

 

 

 

 

 

 

o :93»: exam: 2. a. .30 no mac—w him—om 2.9:...
o 2:39 n.a.”: 299:0me » even» no 826 Erna...
o 0:». Eva—.33» 2.»...
o 02.2. Some 2. .9 “5:233: an Ema-.333... Gum—:3 ﬁg... > £5 Eda? can... .8 Sega...
H58 Emma :8 no 235—=6 no .u 02:5».— wwoac due no 538
Eva—:39. Eamioav €532: c552:
Ami—maul .55 @3an
r08 8% mango so mm
so mm
mnmEEo 58 3% so 2.
magma so mm
A1288 58 3% magma so a
so mm

 

 

 

 

 

 

 

 

 

138

 

 

 

 

 

 

 

 

o Earn»: mam—58 was—Sm no gnu—.5338... Dam—am 2.2....
o 63:!» n.a....— EémgoS E...» 3963—. o 2.2:. E:— 2.3—.3992.»
o 0:33: :3an 88 539.33
a 0269 Eoue on ... 3.533: an 88 Eda-.23 9.9.3.4. E3... > ago 3.85 Rim. :5 53ch
~58 Mano—.5838 due no 223:6 02:32. 523 C8 ma 338
c3958 %— A==Exav A5138
:2:an E6988 9.3 5.83
r08 8% magma so mm
so mm
momcamo 88 8% so mm
manna so mm
H.288 58 5% magma so a
so ,9.

 

 

 

 

 

 

 

 

 

139

20 ..0. $39.28
$82.2.» Na...
.... 0EOH>Emd0>m UHF .8520
$132.: ...a «:82.» 33:88.? 826» ..a ..»..» :8. .... .8 .88 85.8.2.8 2.: ......o. a: .» $82.2.» N2... $8. ..»<8. ..»m»...a
:: ...-.5.» .... .8 .88 ..8 22......» 2.: ......c. 2. 8..» ......» a: $8.2...» ..a. N23. 0:... 2.».
a. 96»... 9.2:... .» ::...».. ..a. ....a». PM. 3.68. 338:8.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. .» $25.38...

o .» 9.885.» ..a ...: 08» .»....88 9 555838....

c .» 8...». ..a. «:2»...

o .» 8:25.» ..o. .88.

o 0:... 8 a. ......» .8». ..a .8 336:8 ..8 8.9.». ......38 w. 82...... 0:... 8 a. ....o» .8». ..a 8: 8.3.2.2....
F8» >2.» 0:3»: $322.8 0.883.» 08...»: ..». mg... H.332»
r80 88 main 3... o .8.
momgao .80 58 m3»... 3... o .3.

...883 .08 3.8 «88.0 B... o ...:

>8» .8». ...... 2......» .0000. 00000000000. 0000000. 00000.

>3» $8...» .0000. 00000000000. 0000000. .0000.
0».8.n8..

$88.88.. . "$92.9 mutating. .w n35. .8....»9.» 5.88.2.9. 3.388 c .9. &m N» «.8» a... $38.28.
0838. m». 28:... 71.98:». NH.8.§».\=.$.8 an F 8%.»... . ”$8.8 $8.82.»... .uhxth. Nunxmuaio. .w "n.a....nme.

. "82.5.8.8... 8823.38...»

140

Z» :0. $8808..

 

.<. .ZmGZOm < NHZU—ZEZHG Us .85 POHHm ..0 n.a.—..0...

.... >5...» 50 «8.2..» ....0n..8»...0 82.0» :0.» 85...» ..:0 8.. v. 0. :8..:» :0.» 0800..» ::..»80 $82.2.» 2......
0:0 <»...0:»: :0 2...... 85.2.»...

n... 85...... 800.»:0 0»: 8..» 0:...<.... 05.88 05.08 :0 2...... w. 0::88 :0. 8.... 0:...<.....

0:8... :0 ....0» 85...... 8.0: 0»: 8.» <»...0:»:.. ..:0.:... .» ::.:»: :0. 2.0».

..0 ::::0 :85... 8.» 85...»... m. 005...... .» 85...». » ..8 ....00... .» 005......

0:0 0»:..:»: :0 85...» 85...... 0: 8.0 2.0»... .505... .» ::.:»: :0 ..8...

0:... ..8 0. 8:85.08» :0 8.0.80... 2:05... .» ::.:»: :0 ..8...

>.m::8 ..3::0.....8 8.000.055 .8 50.8.8 85...». ..»..» .» 0.8.080 0.058.». w. .....8 205...»: .» 85....» 0:...» 8.
O8 ..»00 8.0:... m. 8.00025» 85...». 0: ..8 50508.. ... ..»00...

 

H: .8».. 05.8.. :0 2...... 0800.... 8.0: 8.0 8... $0.50..» w. ..8...0..»... ..:0.:... ::.:»:8 :0 $8».
o U0 8:» 0. 2...... ..:0 8.0: 0800..» 8.0 8.... :80: 005000.»..0: in»... m. ..:0 8.. 05.8.. 8 0..50..0.»..N..... ..:0.:...

::.:»:8 :0 ..8... 0:. ....00... ..00....:...
o > :82. 8 <0::.: .8 2...:—8 28250:.»1... .5.—.01». 20:08....

 

141

 

 

 

 

 

 

F80 <»...0:»: >80. >8» 0:. $880. $.00... 0»:..:»: 0:. ”0:85 0:.
:0 2...... 85...»:» :»: 85...» 85...» :0 85...» :»: .08.. ..»:

r80 58 88:0

M0853 .80 58

@260

...082.» .80 58 «38.0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0»... 8..
3.080 :0 808...... . "0800...... 03.0188. N803...» ..0 83 $380.»... .w "803...... ..0 8%....» 020008.003. A n»».......»..» 0.. ....»
0.0308382». u H83 2.80.8».

 

20 mi $09.99

..N >5...» Ba «53:.» Ban—58...» :83» no m: 332.: .3 32.9238: %— man... u. a. :8 an. mans—mnuaam a: 2.... 58... a:
$82.2.» N23.
a 09—5 ..3—.26 a «..0—ow

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

o mm .5.—3 3:38 a: n- 2.2:. a a. 2.:—=3 %— 58. 2.9.8 253 as. an. a 2.955 23 85:3
o mm 92.: Song a: :— uqouwgnau mg .98. 2538 538. :26... 9538 ecu—.8..“ $3... > ...—a 3.8? Rio .8 53:3
o c953 «3583.3 : 33 32.53..» a: _u 1.6—332% Ew— an...» n.a.-=8 ..3—ms? % a ...—a Eda? E... .552...»
g 02:3 388 a: .m 3:988: %— ..»..—:39...» n.a.—3a. .55... > ...—a 2.8? can... Eu Banana
r98 wanes—.33: :3 5:85.. due an 538 £2.53? Gasman.— wnonmi Cu: 533
an «an... E»... 5.83 .5. 28 bag? :53an .532— amau 32:
r08 83
magma
manna—o .08
8% magma
H0803 58
8% @253
ﬁamﬁaﬂ

waaﬁasanmma ha 288.. Nuhwmhaxanwm: 593E. NM§§Q§ 95.3%. .w "3883 « Hg Banal» M "23 «muhmnﬂnni

142

 

 

 

 

 

 

 

 

o 0:98 no2:_:us..8u 3:558 55%...
o 059.8 .3 3&— 33533 33.5» 553... > ...—m Eda... 85:3 F 2...:—...»... 9.9:... 9::8 Em nag—E9383
o GEE: ”5...: ”:98 3:9...‘ 05:... 9.9.8 an ”:95 8:5. 553... > cam :39: 85:3... H: 9:." 3952.8 3 95:23
0 02%... Base 2. _» ”6:399. an 395 3:3... 0:95; :2... > ...—o 6.69.. can... :5 338..»
be? can: duo no 0939.2: 3.2:: :19... no :8 no 538
gag—Ewan 33:55:" .55»: 5:9:— 2:899.
@5933 a: 3.83
r08 33 so m:
magma
mmmzsao 88 so w:
3% Magma
#10820 .08 :o m:
Bmw @330

 

 

 

 

 

 

 

 

 

$84.28 ta gananmmau Nuhxmtawsgﬁ Nun 5 2.2339. wnhomc kmwﬁzmw um meSEQ: «Hi £8329 MUSE «mung—33¢

143

 

 

 

 

20 An. 38:29.
A; >32..— 50 mean—...: anon-581a 32.8 no 2. 8.38m? no 8.56— :0 32.8 on $89.2... ~23.
o 05:56 <88 =33: a. .98 a 528...
o Gen—=3 $2.258 .Bauhang 2:... :53: .2.: a. 58...
o new—:8 E8 393.93.. 2:.» 833...... a 5:32....
o Ga 8.: 33:55 ago =53»qu a. .93. n.a.—Sn 2.2— 338...
o 0.338 as: non—Ema no ..»—.2. .5an «:1?— .533
o 2&3 «...—Bongo = 3.6 .5.—.35» 986:? no 552. ESE: o. 3....— ?? 833:... a. 52.8...
o new. .5.—.33» .5523
o a: ...—m 32:38 96:3 3 32.33%... 0:553 533. w. a a5 3.8? .5.. .552...»
o 62.3 Base on a. 5:233: no won—.595 9.5.58 22...» > ...—o 3.8? ...-«9 :3 53%..»
has <88 53:... 2:.» E: 22:56 :12... an 0259:— 560? Cu: a»
.358 a :33: can. no 5:258: €532: €532: Sauce.
525 no? .5: Bowen wan—u. ran—mamas ER ..3.
P08 Sum «35%
mom—Ego 58
8mm magma

 

 

Hanna—.0 58 8%
«35%

 

 

 

 

 

 

 

 

 

 

 

 

 

-.ﬁﬁu

144

20 an. 38:29

am >32.» Ba «.5819 Eon—.53.? 32.2. ..» 3.6m swan—::..Bm ...—a :2. a: $82.2.» ~23.

 

 

 

 

 

 

 

 

o ..»..—E: Emma... a: a. .3.“ ..a name...» 05:8 2.:—...‘
. >35. 9.9.:— u.wm=.a...w » 2:5» ..0 88m ...»num...
o 05.. ..Fmio...» 2.»...
. 02:... Sen: 2. .n ”...—.93....— no Erwin...qu 05.53%. ...»m... > ...—o Eda... ..»na .8 Samoa...
.38 Sun: :8 no 23......» ..n .9 03:.an .48.: :8 an 538
awn—.3...» ...»mia...» €55»... Asia»...
92:89 ...»u 3.8.:
r08 33 magma so m.
:0 m.
mnmgao .08 8% :o m.
magma so w.
Hanna—.0 .08 53 magma so w.
:o m.

 

 

 

 

 

 

 

 

 

145

 

 

 

 

 

 

 

 

o ...»...05 0.350... 3......8 .... 8:82.80»..an 05.8... 2.2.... .
o .2.—.3 9.3.... 3.0....20 3:... 3.9.2.... 0 2...: ...... 2.8—.52.»...
o 05.2... ......3 88 3.2.220...
o 02...... 33.. 8.. ... 2:88.00 .8 88 3.2.8.0... 03.58... ......”V > ...8 3.8.0 3.“... .8 2.38...
..08 Hanonanawmnm C... .... 22......» 0003...... ....oo... d... .8 Ben...
3.0....20 .8. Asia... 30...»...
9.0.2... 3.0....20 ...... E8...
r08 8.... mama..." ..0 m.
:0 m.
momcsao .08 ......w :0 m.
magma :0 m.
H0820 .08 ......m macaw ..0 w.
:0 m.

 

 

 

 

 

 

 

 

 

146

20 ..n. $2.22....

<. ...»...Om mOO—OMOOZOZ—uﬁcm

m. >......» ...» «8.2..» 1.62.22? 326» ..o .8 2.2.2.2.... .... .» ..»E...» ....a 3.2.2.2. 2... .8 822.8 ..» 3...... 2. $822.»
~23.

. 9.2:»... 522...»... ......» .» ..»E...»... 0.5 2.2. 32.2. .8 ......8... 2.2.82.8 ....a ..2.2. .....u ....» .m 28 ..a 2.2. an
2.8.322. 2.2.8..

.... 8.»... 32.82»... .....2. 8.2... 2. .» 22.2.3.2. ..a 822......

0...... 8.2.... 2. .» 22......» ..a. 822......

0......» 8.2.... 2. a. ..8...2...a ..o. 822......

0...? 8.2... 2. .» 2.26..» ..n. 322......

 

>......8 >......»m 258 2...»...

 

H2». 2. .»
..»E...»

 

52.2.3.2. ..»
.2.»...

 

92......»

 

58:2...»

 

 

 

 

 

 

 

O82...»

 

mb >2...» .2. «8.2..» 262.22.? 8......» u. 2.22..
o a... ....m 2.... ..»n... 8.2....
o 02.28 2.8 8.2.... 2. .» 2.2.5.»... 2.8.

147

 

 

m... >......» ...» 28.2..» 262.22.... 2...:. 88 .....8 8.7.52.8 ...... ... 2.22.2. 2.2.8.8. 0:28 8.... .8 8.32.2.8 ....a
..8....».. 2. .» «2.2.2» .... .....2.... .2. ..o 8.2. 2...... .... .8 .....8 .88....8 28 2.22.2... 8.. 22...»...

.. N.

 

 

w. ...

 

 

m. a.

 

 

<.. ....S....>0.0zmm > r> ...wOUc00.0Z
a.— >......» .....22.» wanna—..»..... 8.....6 .8 222.8 ....a .2252. .... ....2. .2........2.... .... ......c. 2. 2822.». ”82.2... » .8
2625.8 2... 2.5 92.0.3522... ..O0O .3..O.~...>Z....w. .. 20 .ZwOwH>2...w.

 

 

 

 

 

 

 

 

 

 

 

man...» 2.22.2.2... ..»..» .... ”2.222.?
32885.9...» 2.»... 0»....».. ..o moB...» .....2.»o.2.

.8828 2.2» 0»....».. ..a man... ..»..88

E

.n .5.... 23.8.22». Nu ..»..... ......»22..m. mu 2.. 222.229

. ab m. 8.2. .....26 2.8.2. .....2... ..»..» 8.. 2......» ..o 2...... 2.28 8.... 28 2.22.8... 0:». 8 .» 2.2.8.2. ..»..: ....a 8.2.
2.8.2.» o. .....2......

 

2.2.2.» ..» 2.8.2. 02....22. ..a. 2.8.2....

 

....88 .... 2. ......mo

 

....82. ..a. ....on...2..»...o

 

2.8.». .... ....» ..».» .2.».

 

38.2 ..o 2. ..»..oo

 

 

 

 

0.8 382......

 

022.28..
08.2.2.2. 8:22.839. . "...... 2.32.2.2. NUS: .8» 8 .228... «...»...an ......8..§.Q .w n22. «882......»

148

 

 

Zo n.a. 309.20.

 

<... 0OZ...>2N> > F>m ......UZHHm UM .2...O.w.<.>0.02 mOwaw 2.0.2.0023;
... 0:2...» 8.6.. ..25 .... 2622..» 2. 8.. ......2»... .....2. .58» ..»..» ..62—.... .... 258.... ..»..» 8.5.52. 6. 2......25»...

 

...u >......» ..»...2.» ...62....2..a 2.2..» 25...»..N» ..2... 8.2. 2. .8 2586...... ..6 .» ..........8...... ..6 9...... ....6 .6 ..».. .8 8.2—.228
.8....8.....8 .. ..2258. ”82.2.6 » .8 2625.8 2... 3:0...» 0OZ....>ZN>. #06 ....U 002.9925? .. ..O0>
0OZ....>ZN>.

 

0.......»..2.8

E625...»

N».......»....
9......

0.2.2..

..I. .58..» n..:\.»:~». N" 2.2. 8 8.52.8. an .68 83.3.88.

 

 

 

 

 

 

 

149

<.... ....UZH...0.0m .... ..>m 097....Hm .... .2<Hm....0>0.02 >0EOOF> ..O0>..
Am...» ..»..» 2.25.2.8 ..6. 0...»... 0.....8 8.5 .8 v2.2.6.8 ....6 8.2. .22.... ..6 82. 2.25....» ..6 2. 0.>.....

0:0..9m ..0” m: ..>.w....0...>0.02

APPENDIX 6

Agronomic results from factorial trials in 2003 in Los Limones

150

Farmer 1 bean yield in kilograms per hectare (Tukey’s mean separation, beans at 13.5%

 

 

 

 

moisture).
Amadeus Amadeus Amadeus Amadeus
+ urea + urea - urea - urea
+ paraquat - paraquat + paraquat - paraquat
1740 kg/ha 1760 kg/ha 1490 kg/ha 1410 lgha
Seda Seda Seda Seda
+ urea + urea - urea - urea
+ paraquat - paraquat + paraquat - paraquat
1640 kg/ha 1700 kg/ha 1340£yha 1290 kglha

 

 

 

Amadeus total: 1600 kg/ha (P =0.35)
Seda total: 1490 kg/ha

With urea total: 1700 kg/ha (P =0.01)
Without urea total: 1390 kg/ha

With paraquat: 1540 kg/ha (P =0.93)
Without paraquat: 1540 kg/ha

 

Farmer 2 bean yield in kilograms per hectare (T ukey’s mean separation, beans at 13.5%

 

 

moisture).
Amadeus Amadeus Amadeus Amadeus
+ urea + urea - urea - urea
+ paraquat - paraquat + paraquat - paraquat
1240 kyha 1830_kg/ha 1430 kg/ha 1490 kg/ha
Seda Seda Seda Seda
+ urea + urea - urea - urea
+ paraquat - paraquat + paraquat - paraquat
1570 kg/ha 1790 kg/ha 1170 kg/ha 1490 kg/ha

 

 

 

 

 

Amadeus total: 1500 kg/ha (P =0.81)
Seda total: 1500 kg/ha

With urea total: 1610 kg/ha (P =0.36)
Without urea total: 1400 kg/ha

With paraquat: 1360 kg/ha (P =0.18)
Without paraquat: 1660 kg/ha

 

 

151

 

Farmer 3 bean yield in kilograms per hectare (Tukey’s mean separation, beans at 13.5%

 

 

moisture).
Amadeus Amadeus Amadeus Amadeus
+ urea + urea - urea - urea
+ paraquat - paraquat + paraquat - paraquat
1660 kg/ha 2440 kg/ha 2040 531m 2110 lg/ha
Seda Seda Seda Seda
+ urea + urea - urea - urea
+ paraquat - paraquat + paraquat - paraquat
2500_kg/ha 2230 kg/ha 2170 kg/ha 2160 kg/_ha

 

 

 

 

Seda total: 2260 kg/ha

 

Amadeus total: 2070 kg/ha (P=0.33)
With urea total: 2210 kg/ha (P =0.67)
Without urea total: 21 10 kg/ha

With paraquat: 2100 kg/ha (P =0.50)
Without paraquat: 2230 kg/ha

 

Farmer 4 bean yield in kilograms per hectare (Tukey’s mean separation, beans at 13.5%

 

 

moisture).

Amadeus Amadeus Amadeus Amadeus
+ urea + urea - urea - urea

+ paraquat - paraquat + paraquat - paraquat
1070 kg/ha 730 kg/ha 560 kg/ha 640Ma
Seda Seda Seda Seda

+ urea + urea - urea - urea

+ paraquat - paraquat + paraquat - paraquat
870 kg/ha 940 kg/ha 460 kg/ha 710 kg/ha

 

 

 

 

Seda total: 760 kg/ha

 

Without paraquat: 760

Amadeus total: 760 kg/ha (P =0.87)

With urea total: 900 kg/ha (P =0.04)
Without urea total: 600 kg/ha

With paraquat: 740 kg/ha (P =0.86)

its/ha

 

 

152

 

APPENDIX 7

Field Day Questions for Farmers in Los Lirnones in 2004

153

Field analysis

1. What are the qualities of this soil?

2. What is the highest yield from this ﬁeld in the past?

3. What do you think are the reasons that this piece of land produced so much?
Treatment comparison

4. Which treatment will produce the best this year? Why?

5. Which treatment will do the worst this year? Why?

6. Based on your experiences in the past, which bean variety would you use to maximize
your bean production?

154

APPENDIX 8

Common Local (Spanish), Scientific and Common English Names for Farmer-

Identified Insects

155

 

 

 

 

 

 

 

Local name Latin name English name
Picudo de la Vaina Ayiongodmani Bean Pod Weevil
Pulgén Systena blanda and others Flea beetle
Mosca Blanca Bemesia tabaci Whitefly
Tortuguilla Diabrotica spp., Cerotoma gap. Chrysomelids
Crisomelido Diabrotica spp., Cerotoma spp. Chrysomelids
Lorito Verde Empoasca spp. Leafhoppers

 

Gusano Desfoliador

Order: Lepidoptera

Defoliatinginsect larvae

 

Gusano Alambrique

Family: Elateridae

Wireworms

 

 

Gallina Ciga

 

Phyllophaga spp.

 

White grubs

 

156

 

 

APPENDIX 9

Common Local (Spanish), Scientific and Common English Names for Farmer-

Identified Diseases

 

 

157

 

 

 

 

 

 

 

 

 

 

 

 

 

Local name Latin name English name

Mustia Hilachosa Thanatephorus cucumeris Web blight

Hielo Negro Colletotrichum lindemuthianum usually Anthracnose

Xanthomonas campestris pv. Common Bacterial

Bacteriosis phaseoli Blight

Mancha Angular Phaeoisarigsis griseola Angular leaf slot

Antracnose Colletotrichum lindemuthianum Anthracnose

Roya Uromyces appendiculatus Rust

Sclerotinia Sclerotinia sclerotium Sclerotinia

Mal de Talluelo Rhizoctonia spp. or Pythium spp. Damping Off

Rhizoctonia Rhizoctonia solani Rhizoctonia Root Rot
Bean Common Mosaic
Virus or Bean Golden

Mosaico Yellow Mosaic Virus
usually Bean Golden

Hielo Amarillo Yellow Mosaic Virus

 

 

 

158

 

 

BIBLIOGRAPHY

159

 

BIBLIOGRAPHY

Anderson, C.L., L. Locker, and R. Nugent. 2002. Microcredit, Social Capital and
Common Pool Resources. World Development 30:95-105 (11).

Ashby, J .A., A.R. Braun, T. Gracia, M.P. Guerrero, L.A. Hernandez, C.A. Quiros, and
J .I. Roa. 2000. Investing in Farmers as Researchers: Experience with Local
Agricultural Committees in Latin America Centro Intemacional de Agricultural
Tropical, Cali, Colombia.

Barker, R. 1979. Adoption and production impact of new rice technology: the yield
constraints problem, p. 1-26 Farm-Level Constraints to High Rice Yields in Asia:
1974-1977. International Rice Research Institute, Los Baﬂos.

Castro-luniga, IA. 1998. Study of the Productivity and Acceptance of Improved Bean
Varieties in Central-Eastern Honduras [transl.], University of Puerto Rico,
Mayagiiez.

Central Intelligence Agency, U.S. 2005. The World Factbook: Honduras [Online].
Available by Central Intelligence Agency

http://wwwcia.gov/cia/publications/factbook_/geos/ho.html (posted 1 November
2005; verified 15 December 2005).

Centro Intemacional de Agricultura Tropical. 1999a. Curvas de Nivel - Honduras
[Online]. Available by CIAT http:ﬂgiswebciat.cgiar.org/Mitch/index.htm.

Centro Intemacional de Agricultura Tropical. 1999b. Lluvia Anual, Temperatura
Minima, Temperatua Maxima, Elevacion (5 Km) de Honduras [Online].
Available by CIAT http://gisweb.ciat.cgiar.org/Mitch/index.htm.

Chambers, R. 1997. Whose reality counts? : putting the ﬁrst last Intermediate
Technology, London.

Davidson, BR, and B.R. Martin. 1965. The relationship between yields on farms and in
experiments. Australian Journal of Agricultural Economics 9:129-140.

Debouck, D.G. 1999. Diversity in Phaseolus Species in Relation to the Common Bean, In
S. P. Singh, ed. Common Bean Improvement in the Twenty-ﬁrst Century. Kluwer
Academic Publishers, Dordrecht, the Netherlands.

Dillon, L.D., and J .B. Hardaker. 1993. Farm Management Research for Small Farm
Development Food and Agriculture Organization of the United Nations, Rome.

160

Duﬂo, E. 2003. Poor but Rational. Massachusetts Institute of Technology.
(ed.) 2004. Behavioral Economics, Public Economics and Development Economics,
Ithica, New York. May 28-29, 2004.

Dunn, E., and J .G. Arbuckle. 2001. Microcredit and Microenterprise Performance:
Impact Evidence from Peru. Small Enterprise Development 12:22-33 (12).

Economic Commission for Latin America and the Caribbean. 1999. Centroamérica :
Cambio Institucional y Desarrollo Organizativo de las Pequeﬂas Unidades de
Produccion Rural United Nations: La Sede Subregional de la Comision
Econémica para América Latina y el Caribe, Mexico, D.F.

Edie, O.T., J .M. Maimu, O.S. Mbuya, P. Ndakidemi, C.S. Mushi, and DJ. Allen. 1991.
Progress in on-farm research with beans in Lushoto district, Tanga Region,
Tanzania. Bean Improvement Cooperative.

FIAN, and La Via Campesina. 2000. Problems with the Implementation of Agricultural
Reform in Honduras. Food, Land and Freedom Annual Campaign for Agrarian
Reform, Heidelberg/Tegucigalpa.

Food and Agriculture Organization. 2005a. FAOSTAT Agricultural Data - Agricultural
Production - Crops Primary [Online]. Available by FAO
http:Hfaostatfaoogjfaostat/ (posted July 14, 2005; veriﬁed October 17, 2005).

Food and Agriculture Organization. 2005b. FAOSTAT Agricultural Data - Food Supply -
Crops Primary Equivalent [Online]. Available by FAO
http://faostat.fao.mjfaostat/ (posted August 27, 2004; veriﬁed October 17, 2005).

Franco, J. 2003. Tenencia de la Tierra a Otto Ensayo La Prensa, Tegucigalpa. Honduras.
Giller, KR. 2001. Nitrogen Fixation in Tropical Cropping Systems. 2nd. ed. CAB
International, Wallingford, Oxon.

Gomez, K.A., R.W. Herdt, R. Barker, and SK. DeDatta. 1979a. A methodology for
identifying constraints to high rice yields on farmers' ﬁelds, p. 27-47 F arm-Level
Constraints to High Rice Yields in Asia: 1974-1977. International Rice Research
Institute, Los Baﬁos.

Gomez, K.A., L. Lopez, M.L. Novenario, R.W. Herdt, and VP. Marciano. 1979b.
Constraints to High Rice Yields, Laguna, Philippines, p. 173-190 F arm-Level
Constraints to High Rice Yields in Asia: 1974-1977. International Rice Research
Institute, Los Baﬂos.

Hattendorf, M.J., M.S. Redelfs, B. Amos, L.R. Stone, and RE. Gwin, Jr. 1988.
Comparative water use characteristics of six row crops. Madison, Wis, Agron-J.

161

Herdt, R.W. 1979. An overview of the constraints project results, p. 395-411 Farm-Level
Constraints to High Rice Yields in Asia: 1974-1977. International Rice Research
Institute, Los Baﬁos.

Herdt, R.W., and T. Wickharn. 1975. Exploring the gap between potential and actual rice
yields in the Philippines. Food Research Institute Studies in Agricultural
Economics, Trade, and Development 14:163-181.

Hoque, M.Z., E. Ahsan, and M. Hossain. 1979. Constraints to High Rice Yields,
Bangkadesh, 1975 to 1977, p. 49-83 Farm-Level Constraints to High Rice Yields
in Asia: 1974-1977. International Rice Research Institute, Los Baﬁos.

Humphries, S., J. Gonzales, J. J ., and F. Sierra. 2000. Searching for Sustainable Land Use
Practices in Honduras: Lessons from a Programme of Participatory Research with
Hillside Farmers 104. Overseas Development Institute, London.

International Monetary Fund, and International Development Fund. 2001. Poverty
Reduction Strategy Paper:

A peoples' commitment towards a better Honduras 22661-HO. International Monetary
Fund, International Development Fund, Tegucigalpa.

Jain, RS. 1996. Managing Credit for the Rural Poor: Lessons from the Grameen Bank.
World Development 24:79-89 (11).

(ed.) 2001. Seventh Eastern and Southern Africa Regional Maize Conference. February
11-15, 2001.

Kaplan, L., and L. Kaplan. 1988. Phaseolus in Archeology, p. 613, In P. Gepts, ed.
Genetic Resources of Phaseolus Beans: Their Maintenance, Domestication,
Evolution, and Utilization. Kluwer Academic Publishers, Dordrecht.

Kerr, J. 2004. Effects of agricultural technology improvement based on household initial
asset position, In RD826, (ed.). Michigan State University, East Lansing,
Michigan.

Laibson, D. 1991. Golden Eggs and Hyperbolic Discounting. Quarterly Journal of
Economics 112:443-477.

(ed.) 2004. Proceedings of the 4th International Crop Science Congress, Brisbane,
Australia. September 26 - October 1, 2004. The Regional Institute, Ltd.

Martel, P.V., R.H. Bemsten, and M.T. Weber. 2000. Food Markets, Policy and

Technology: The Case of Honduran Dry Beans 78. Michigan State University,
East Lansing, Michigan.

162

Mather, D.L., R. Bemsten, J .C. Rosas, A. Viana Ruano, and D. Escoto. 2003. The
Economic Impact of Disease Resistance Research in Honduras. Agricultural
Economics 29:343-352.

Mehlich, A. 1984. Mehlich 3 soil test extractant: A modiﬁcation of Mehlich 2 extractant.
Comm. Soil Sci. Plant An. 15: 1409-1416.

n.a. 2003a. Produceién de granos basicos perﬁla inicio de reactivacion agricola La
Prensa, Tegucigalpa.

n.a. 2003b. Perdon de deudas y mas créditos para productores La Prensa, Tegucigalpa.

n.a. 2004. Gobiemo Ha Destinado 2,200 millones a1 Agro, pp. 20 El Heraldo, 7911 ed,
Tegucigalpa. Honduras.

n.a. 2005. Honduran Congress Votes to Ratify the DR-CAF TA [Online]. Available by
Business Coalition for U.S.-Central American Trade
http://www.uscaﬁaorg/articles/view.a_sp?ID=41 (posted March 3, 2005; veriﬁed
October 17, 2005).

Quiros Torres, C.A., B. Douthwaite, J .1. Roa Velasco, and J. Ashby, (eds.) 2004.
Colombia, Latin America and the Spread of Local Agricultural Research
Committees (CIALS): Extension Through Farmer Research, Vol. Volume 3.
Demand-Driven Approaches to Agriculture Extension. The International Bank for
Reconstruction and Development / The World Bank, Washington, DC.

Rosas, J .C. 2003. Recomendaciones para el Manejo Agronomico del Cultivo del Frijol
Programa de Investigaciones en Frijol, Escuela Agricola Panamericana,
Tegucigalpa, Honduras.

Rosas, J .C., and D. Escoto. 2002. Amadeus-77. Escuela Panamericana Agricola,
Zamorano, Tegucigalpa.

Rosas, J .C., O. Gallardo, and J. Jimenez. 2003. Mejoramiento genetico del frijol comun
mediante enfoques participativos en Honduras. Agronomia Mesoamericana 14: 1-

9.

Sallant, P., and DA. Dillman. 1994. How To Conduct Your Own Survey John Wiley and
Sons, Inc.

SAS Institite, Inc. 2003. The SAS system for windows. Release 9.1. SAS Institite,1nc.,
Cary, NC.

Schutt, R.K. 2004. Investigating the social world : the process and practice of research.
4th ed. Pine Forge Press, Thousand Oaks, Calif.

163

 

 

Secretaria de Planiﬁcacion Coordinacion y Presupuesto. 1993. IV Censo Nacional
Agropecuario 1993. Direccion General de Estadistica y Censos, Comayaguela.
Honduras.

Singh, SP. 1999. Production and Utilization, In S. P. Singh, ed. Common Bean
Improvement in the Twenty-ﬁrst Century. Kluwer Academic Publishing,
Dordrecht, the Netherlands.

Snapp, S. 2002. Quantifying Farmer Evaluation of Technologies: The Mother and Baby
Trial Design, p. 9-17, In M. R. Bellon and J. Reeves, eds. Quantitative Analysis of
Data from Participatory Methods in Plant Breeding. CIMMYT, Mexico, D.F.

Snapp, S.S., M.J. Blackie, and C. Donovan. 2003. Realigning Research and Extension
Services: Experiences from Southern Africa. Food Policy 28:349.

Soto, H.d. 2000. The Mystery of Capital : Why Captitalism Triumphs in the West and
Fails Everywhere Else Basic Books, New York.

Staatz, J .M. 1994. The strategic role of food and agricultural systems in developing
countries: experiences from Latin America 94-3 9. Michigan State University,
East Lansing, Michigan.

Staatz, J .M., and C. Eicher. 1998. Agricultural development ideas in historical
perspective, In C. Eicher and J. Staatz, eds. Agricultural Development in the
Third World. Johns Hopkins Press, Baltimore.

Stolzenbach, A. 1994. Learning by improvisation: farmer experimentation in Mali, p.
155-159, In I. Scones and J. Thompson, eds. Beyond Farmer First: Rural People's
Knowledge, Agricultural Research and Extension Practice. Intermediate
Technology Publications Ltd., London.

Sumberg, J .E., and C. Okali. 1997. Farmers' experiments : creating local knowledge
Lynne Rienner Publishers, Boulder, Colo.

Superintendencia de Administracion Tributaria. 2003. Avances de la Union Aduanera
Centroamericana [Online]. Available by Government of Guatemala
http://www.sat.gob.gt/pls/publico/portal.documentos?codigo=1077 (veriﬁed 8
April 2004).

Thorpe, A. 2002. Agrarian Modernisation in Honduras E. Mellen Press, Lewiston, N.Y.

Thorpe, A., H.N. Pino, P. Jimenez, A.L. Restrepo, D. Suazo, and R. Salgado. 1995.
Impacto del Ajuste en el Agro Hondurefio. 1. ed. POSCAE/UNAH, Tegucigalpa,
Honduras.

Tshering, C. 2002. Proﬁtability Analysis of Bean Production in Honduras 616. Michigan
State University, East Lansing.

164

 

United Nations. 2005. The Millennium Development Goals Report 2005. United Nations,
New York.

Vasquez, I. 2003. Mesas agricolas, puro papel La Prensa, Tegucigalpa, Honduras.

Walkley, A., and LA. Black. 1934. An examination of the Degtjareff method for
determining soil organic matter and a proposed modiﬁcation of the chromic acid

titration method. Soil Science 37:29-38.

Weeks, J. 1999. Trade Liberalisation, Market Deregulation and Agricultural Performance
in Central America. Journal of Development Studies 35:48-75.

J. R. Witcombe, et al. (ed.) 2002. Breeding Rainfed Rice for Drought-prone
Environments: Integrating Conventional and Participatory Plant Breeding in
South and Southeast Asia, Los Baﬁos, Philippines. IRRI and DIFD.

World Bank. 2003. HONDURAS - Trade Facilitation and Productivity Enhancement
Project AB41. World Bank Group, Infoshop.

Yonts, D. 1996. Irrigation Requirements, p. 24-31, In H. F. Schwartz, eta1., eds. Dry
Bean Production and Pest Management, Vol. 562A. Colorado State University,
Fort Collins, CO.

Zurita Ramirez, CE. 2003. Evaluacién de irnpacto de la metodologia de investigacion
participativa en tres CIAL de la region del Yegilare, Honduras. Licenciatura,
Escuela Agricola Panamericana, Zamorano.

165