• TABLE OF CONTENTS
HIDE
 Title Page
 Dedication
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Abstract
 Introduction
 Literature review and methodol...
 Cotton production functions
 Linear programming
 Conclusions and recomendations
 Appendix A. Canete video
 Appendix B. Survey questionnai...
 Appendix C. Survey data
 Appendix D. Production function...
 Appendix E. Linear programming...
 References
 Biographical sketch






Title: Farm problems, solutions, and extension programs for small farmers in Canete, Lima, Peru
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00091272/00001
 Material Information
Title: Farm problems, solutions, and extension programs for small farmers in Canete, Lima, Peru
Physical Description: Book
Language: English
Creator: Cabrera, Victor E.
Publisher: Victor E. Cabrera
Place of Publication: Gainesville, Fla.
Publication Date: 1999
 Subjects
Subject: Farming   ( lcsh )
South America   ( lcsh )
Agriculture   ( lcsh )
Farm life   ( lcsh )
Farming   ( lcsh )
University of Florida.   ( lcsh )
Spatial Coverage: South America -- Peru
North America -- United States of America -- Florida
 Notes
General Note: A thesis presented to the Graduate School of the University of Florida in partial fulfillment of the requirements for the degree of Master of Science
 Record Information
Bibliographic ID: UF00091272
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved, Board of Trustees of the University of Florida

Table of Contents
    Title Page
        Page i
        Page ii
    Dedication
        Page iii
    Acknowledgement
        Page iv
        Page v
    Table of Contents
        Page vi
        Page vii
        Page viii
    List of Tables
        Page ix
        Page x
    List of Figures
        Page xi
        Page xii
    Abstract
        Page xiii
        Page xiv
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
    Literature review and methodology
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
    Cotton production functions
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
    Linear programming
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
    Conclusions and recomendations
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
    Appendix A. Canete video
        Page 85
    Appendix B. Survey questionnaire
        Page 86
    Appendix C. Survey data
        Page 87
    Appendix D. Production function data
        Page 88
    Appendix E. Linear programming models
        Page 89
    References
        Page 90
        Page 91
    Biographical sketch
        Page 92
Full Text










FARM PROBLEMS, SOLUTIONS, AND EXTENSION PROGRAMS FOR SMALL
FARMERS IN CANETE, LIMA, PERU.














By

VICTOR E. CABRERA


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


1999





























Copyright 1999

by

Victor E. Cabrera























To my father














ACKNOWLEDGMENTS

I gratefully acknowledge the support of the Kellogg Foundation for the awarded

fellowship that allowed my enrollment in the Graduate School at the University of

Florida. I want to thank to Dr. Heliodoro Diaz, Dr. Robert A. DeVries, and Dr. Rebecca

Hernandez for making possible the grant. I also want to thank to Dr. Marcos Kisil and Dr.

Marc D. Carnersi for their excellent guidance during my fellowship and to Denise

Alvarado for her permanent help.

I want to thank to the Valle Grande Rural Institute for agreeing with my graduate

studies and for helping me in the field phase of this investigation. I am very thankful to

David Baumann and Oscar Sebastiani for conferring me confidence. Appreciation is

extended to the sondeo team members: Juan M. Carrillo, Jorge Bernaola, Sabino Julian,

Polo Echacaya, and Willy Pinedo for their disinterested labor in the first phase of this

study.

I would like to express my profound gratitude to the Cafiete community in general

and especially to the small farmers for their willingness to participate in this research.

I wish to thank to the University of Florida and the Agricultural Education and

Communication department for allowing me to complete my master's degree and this

thesis. I am especially grateful to the members of my supervisory committee for their

guidance and assistance during the elaboration of this research: Dr. Mathew T. Baker, Dr.

Peter E. Hildebrand, and Dr. Edward Osborne. Recognition is expressed to Dr. Matthew

T. Baker, my adviser, for his patience, advice, and contributions in all the aspects of this








study. My special thanks go to Dr. Peter E. Hildebrand who empowered my sensibility

toward the more deprived small farmers, awakened in me abilities for Farming Systems

analysis, provided substantial contributions, and helped in all phases of this research.

I wish to express my immense gratitude to my wife, Milagritos, for her permanent

support, encouragement, dedication, and help.

I would like to express my eternal gratitude to my mother, Rosa, who is, since

always, one of my greatest supporters. My sincere thanks go to my brothers and sisters

who, although far away, are my permanent encouragement: Pedro & Chuza, Abelardo &

Ana, Maria Silvia & Tofio, Juan Carlos & Frin6, and Rosita.

Thanks are extended to all my friends who have given me support, confidence,

and encouragement. They deserve special recognition for their help. These include: Abib

Araujo, Karla Rocha, Paul Litow, Amy Sullivan, Patricio Moya, Olanda Bata, Marcos

Freire, Noelle Connor, Carl Pomeroy, Andrea Snyder, Abdullah Al-Shankiti, Paulanco

Thangata, Raphael Pierre, Guillermo Zanelli, Manuel Bedoya, Pablo Puertas, Justino

Llanque, Jorge Hurtado, and Elena Acurio.

Gracias a los amigos que me acompafian, de cerca o de lejos, en todos los

moments importantes de mi vida. Seria impossible mencionarlos a todos.















TABLE OF CONTENTS
page


A C K N O W LED G M E N T S .............................................................................................. iv

LIST O F TA BLE S ................................................................................................ ix

L IST O F FIG U R E S................................... .......................................... ................... xi

A B S T R A C T .............................................................................. ............................x iii

CHAPTERS

1 IN TRO D U C TIO N ................................................... ........ ........ ............... 1

Cafiete: Agricultural Valley................. ......... ..................... 1
Study Justification .................................................................... 1
P purpose and O bjectives........................................................................................3
The V alle G rande R ural Institute ................................................ ......... ..........4
Significance ..................... ............ ... ........ ..... ......... ........ ........................ 5
Background Information of the Cafiete Community .................................. 5

2 LITERATURE REVIEW AND METHODOLOGY ..................................... 12

L iteratu re R ev iew .................................................................................................... 12
Introduction ......................................... ..... ........................ 12
Farming Systems Research & Extension................................ 12
Stages in Farming Systems Research & Extension................................... 13
Methodology ................................................... 18
Population and Sam ple................. ..................... ...................... 18
D ata Collection and D esign .......................................................... .................. 20
Production function .....................................................................................20
Linear program ing ...... ............ .............................. ................. 21
D ata Analysis ........................................................ ................ ... ...... 22

3 COTTON PRODUCTION FUNCTIONS.................................... 23

Introduction ........................................................ .. ... ................... 23
A analysis ..................23...........................................................
C erro A legre ............................................. ........................... ................... 27
L a Q uebrada .............................................................................. ................. 28
Palo Isla ..................................... ................. ............... ....... .......... 29

vi

('








Santa B arbara .................................... ..... .......................................... .. 29
S an B enito .......................................................... ....................................... .. 30
San F rancisco ....................................... ............... ............................................. 3 1
Quilmana ...................... ............. ................................................... 31
Cotton Yield Analysis Based on Production Functions..................................... 32

4 LINE dAR PROGRAM M ING............................................................................ .. 39

Introduction ..... ........................ .. ............... ................................... .. 39
Assumptions for Linear Programming.......................................................... 40
Linear Programming M odel Validation...................................... .................. 49
Prices, Alternative Crops, and Discount Analyses Based on Validated
M odel .. ............................................................... .. ................ ... 52
O ne Y ear M odel ....................................... .................. ................ ... 52
Average Price Scenario ......................................................... .............. ..... 52
Below Average Price Scenario ................ .......................................... 54
Above Average Price Scenario.......................... .................. 55
Six Year M odel........................... ................... .... ......... ....................... 59
Average Price Scenario ......................... ...... .................. ..................... 59
Grape and asparagus activities analysis ......................................... 60
Discount rate analysis ...................................... 61
Below Average Price Scenario ....... ....................................................65
Above Average Price Scenario............ ........................................ 66

5 CONCLUSIONS AND RECOMMENDATIONS .................... ...................... 67

Production Functions ..................... ........................... ...... ................... 67
Linear Programming ....... ................. .. .................... ........ ..... ............68
Extension Programming........................................... 70
Introduction ...................... ........................ 70
Identifying Target Publics.............................. ... .. ........................ 70
Collaborative Identification, Assessment, and Analysis of Needs
Specific to Target Publics....................................................................70
Design and Implementation ...................... ... 71
Implementing the Planned Program ................. ........... ... ..................... 72
D developing Plans of Action...................................................................... 72
Cafiete first major program: improvement of traditional crop
management (cotton, sweet potato, and maize).............................72
Cafete second major program: introduce new crops
(horticultural, fruits, and annual crops) .......................................74
Cafiete third major program: credit administration ............... ......... 76
Cafete fourth major program: marketing for Cafiete growers ..............77
Cafiete fifth major program: apply farm management techniques......... 78
Cafiete sixth major program: agricultural legal issues ...................... 79
Cafiete seventh major program: organize farmers in strong
associations ..................................................................... 80









Cafiete eighth major program: learn healthy and affordable
nutrition practices ................... .... ............ .................... ... 82
Cafiete ninth major program: decrease pesticide use and increase
environmental awareness................... ................................. ...... 83

APPENDICES

A CANETE VIDEO .......... ..... ........................... ................ ..... .. ..................... 85

B SURVEY QUESTIONNAIRE.................... ................................................... 86

C SURVEY DATA......... ............................................ 87

D PRODUCTION FUNCTION DATA..................................... ....................... 88

E LINEAR PROGRAMMING MODELS................................................................... 89

LIST OF REFERENCES.................................................. ......................... 90

BIOGRAPHICAL SKETCH ................. ... .. .................................................... ... 92















LIST OF TABLES


Table page

1-1 Educational institutions in C aiete ........................................ .......................6

1-2 Formal education stratification in Cafiete ................................................ 7

1-3 Land and income in Cafiete ... ...................... .. ............... ................... 10

3-1 Range of fertilization factors by geographic zone, kg/ha ............................ 24

3-2 Data used in production functions ................................. ..... ..............26

3-3 Cotton production function in Cerro Alegre ...............................................28

3-4 Cotton production function in La Quebrada ..............................................28

3-5 Cotton production function in Palo Isla ..................................................... 29

3-6 Cotton production function in Santa Barbara.................. ................. 30

3-7 Cotton production function in San Benito................. ................. 31

3-8 Cotton production function in San Francisco ..................................... 31

3-9 Cotton production function in Quilman. ................................................ 32

4-1 Resources and constraints for linear programming, average household......... 44

4-2 Crop activities and resource use for linear programming, per ha.................. 45

4-3 Resource use for house and livestock activities, average household .............45

4-4 Income from traditional activities, per ha .............................. ................ 46

4-5 Asparagus resource needs in six-year model, per ha ................................... 46

4-6 Grape resource needs in six-year model, per ha.........................................47

4-7 Asparagus & grape income, per year per ha................................................47

4-8 Household composition and land for each farm.........................................48









4-9 Total land used (t-test assuming equal variances) ................................... 50

4-10 Maize land used (t-test assuming unequal variances)....................................51

4-11 Sweet potato land used (t-test assuming equal variances) .............................51

4-12 Cash at the end of the year in different scenarios, economic variability ........ 58

4-13 Analysis of the asparagus activity regarding household characteristics .........60

4-14 Analysis of discount and crops.......................................... ................ 63














LIST OF FIGURES


Figure page

1-1 Caiete location..................................................................................... 2

1-2 Cafiete Valley................................................ 3

1-3 The Valle Grande Rural Institute ...........................................................5

1-4 Small farmers planting potatoes ........................................ ................. 11

2-1 Hierarchy in programming. ........................................ ................. 15

2-2 Survey interview locations. .......................................... .. ................ .20

3-1 Range of fertilization factors by geographic zone, kg/ha ......................... 24

3-2 Annual environmental index for cotton yield in Cafiete ............................. 27

3-3 Estimated cotton yield in response to phosphorus and
potassium fertilization rates, Cerro Alegre ......................................33

3-4 Estimated cotton yield in response to phosphorus
fertilization rates, La Quebrada........................... ................... 34

3-5 Estimated cotton yield in response to nitrogen and
potassium fertilization rates, Palo Isla............................................ 35

3-6 Estimated cotton yield in response to phosphorus and
potassium fertilization rates, Santa Barbara.................................... 36

3-7 Estimated cotton yield in response to nitrogen and
potassium fertilization rates, San Benito ...................................... 37

3-8 Estimated cotton yield in response to nitrogen, phosphorus,
and potassium fertilization rates, San Francisco.............................. 38

3-9 Estimated cotton yield in response to potassium
fertilization rates, Quilman ................... ........ ............. ........ 38

4-1 Total area planted in cotton for different price scenarios.............................. 54

4-2 Households with water constraints in different price scenarios ............... 55









4-3 Range of cash at the end of the year in different price scenarios ................ 56

4-4 Hired labor in different price scenarios.................................................... 56

4-5 Land rented out in different price scenarios.............................................. 57

4-6 Different economic variability and land to person index.......................... 58

4-7 Capability of planting asparagus and number of children in household...... 61

4-8 Capability of planting asparagus and number of
adult m em bers in household .................................. ............... ... 61

4-9 Capability of planting asparagus and farm size......................................... 62

4-10 Capability of planting asparagus and household education ........................ 62

4-11 Area of crops planted in different stress levels....................................... 64














Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

FARM PROBLEMS, SOLUTIONS, AND EXTENSION PROGRAMS FOR SMALL
FARMERS IN CANETE, LIMA, PERU.


By

Victor E. Cabrera

August of 1999

Chairman: Mathew T. Baker
Major Department: Agricultural Education and Communication

This was a study about ways of improving the Cafiete small farmer community

(4,800 households, 18,080 ha) through agricultural extension. The study framework

focused on Farming Systems Research and Extension, which was complemented with

Targeting Outcomes of Programs, Participatory Rural Appraisal, and extension

programming approaches.

Several procedures were used to gather data. A sondeo (qualitative data collection

technique or rapid appraisal survey) was conducted to obtain a general understanding of

the community and to support the research procedures, especially in constructing a

formal questionnaire. A survey of 60 random household interviews was also conducted to

obtain information of the community to be used in further analyses.

Ex post facto or secondary data were also gathered. These data came from records

maintained by the Valle Grande Rural Institute (local extension agency with which the








investigation was coordinated), from records maintained by the city government, and

from records of Peru's Ministry of Agriculture.

The analyses included production functions, linear programming, and extension

programming. Production functions for seven geographical zones were generated based

upon multiple regression of cotton yield as a function of fertilization and environmental

factors.

Linear programming was used to simulate and better understand the current

situation of individual households. Following statistical validation, a projection of future

production, income, and consumption was undertaken at the household level. These

simulation models are "interactive working models."

Based upon the sondeo, survey, production functions, linear programming, and

secondary data a list of nine extension programs was proposed. These programs were

based upon priority needs as identified by small farmers.














CHAPTER 1
INTRODUCTION


Cafiete: Agricultural Valley


Cafiete is a Valley in the central coast of Peru. It is located between 12"54' and

13030' south latitude and between 7600' and 76024' west longitude (Figure 1-1), it

covers about 22,600 ha of agricultural land and its elevation varies from 0 to 700 meters

(Figure 1-2). The Valley is irrigated by Caniete River water. The river flows east to west,

and it runs continuously all year long. The weather is desert-like; there is little rain. The

temperature varies from 120C in the winter to 32C in the summer, with an average of

180C. There are 152,379 people living in Cafiete; 41,000 of them constitute the rural

population, who are directly engaged in agricultural production. Agricultural production

constitutes the main source of income of the whole province.

Study Justification


Cafiete has 4,800 small farmers with 12 ha or less (80% of the 22,600 ha). In

general, the environmental conditions in Cafiete including soils, climate, and water

resources are not agricultural constraints. Compared to other regions of the country, these

agro-climatic conditions are perceived as being very favorable. The Cafiete Valley is an

influential agricultural production area in Peru.
















Peru


N








CA NE I


L ~',----- A --,



-" *






L 'J '1r-, -- "
Truli0 Pt-al.




lb i .Plw.i Ic"
'I* | c" ", .. 1

Usr'1" _1j,,vqg woL-'
Lg d CUy lt em


Figure 1-1: Cafiete location



Despite the favorable agro-climatic advantages, on average, households (7

members) have an annual income of US$ 1,420, and many Cafiete residents live below

international standards for nutrition, health, housing, services, and other basic needs

(Valle Grande Rural Institute, 1997). In this context, it is highly relevant to find ways to

alleviate the present conditions. The researcher believes that the Farming Systems

Research and Extension approach is a very appropriate theoretical framework to use in

developing interventions to improve the current situation. These research findings will

provide information for extension programs to address the needs of Caiiete's small

farmers.












.*-- *.

%M- I ._ ..





















Figure 1-2: Cafiete Valley



This research is specially needed in the long-term programming efforts of "Valle Grande

Rural Institute."





Purpose and Objectives


The overall purpose of this study was to identify problems and needs and to

design extension programs to improve the livelihood of limited resource farmers in the
Caete Valley. The specific objectives of the study were to
The ovel p--ral pro ofw ti sdy ws tp


































SA non-government organization which works with Cafiete's small farmers and with
which this thesis was coordinated.
which this thesis was coordinated.








(1) Develop production functions that explain the current cotton
production enterprise and use the production functions to
predict future yields of small farmers in the Cafiete Valley.
(2) Use linear programming to simulate individual household's
livelihood systems and to explore production alternatives in
different scenarios of the small farmers in the Cafiete Valley.
(3) Propose a priority list of future extension programs to meet
the needs of small farmers in the Cafiete Valley.

The Valle Grande Rural Institute


Valle Grande is a development institution that has been in existence for more than

30 years in the Cafiete Valley promoting rural improvement through extension and

education programs designed for low income farmers. Valle Grande reaches more than

1,000 small farmers in different programs annually. The headquarters is located in Lima,

the capital, and it is a non-profit foundation, which has the responsibility to raise the

needed funding for programming. Its budget comes from different sources;

approximately one-third of it is from local resources (services have a minimum charge),

another one-third is raised through local and national donations, and the additional funds

are provided from international institutions.

Valle Grande currently runs a coastal extension office, a mountain extension

office, an entrepreneurial development office, a soil laboratory, and an agricultural

college. With the exception of the mountain extension office, the offices, laboratory, and

college collaborate in serving the needs of the Cafiete Valley farmers. Undoubtedly Valle

Grande (Figure 1-3) has an excellent reputation as an extension agency in Cafiete Valley

and perhaps in the whole middle coastal region of Peru.

Among other development agencies in the Valley is CECOACAM (Cooperative

Headquarters of Cafiete and Mala), which may reach as many as one-half of the total








small farmers in the Valley (2,400) but it has a poor reputation due to a number of

questionable experiences. Most of the farmers work with CECOACAM seeking credit and

low-cost machinery. CECOACAM collaborates to a small extent with the government,

however it is not an official agency.


Figure 1-3: The Valle Grande Rural Institute



Significance


The significance of the present study is two-fold. First, there are 34,000 direct
beneficiaries (farmers), and about 7,000 indirect beneficiaries (laborers) who might
obtain some benefit from extension programming. Second, the Cafiete Valley is a highly
prestigious agricultural zone in Peru and any methodological success could be diffused
quickly throughout the country.


Background Information of the Cafiete Community


In an analysis, including the elements of Cafiete's social system, one could

observe that in general, the majority ofCafiete's community members practice the Roman

Catholic Religion; they strongly believe in God. Many family-related decisions (i.e.








number of children) are based upon church doctrine. Community members are also

fervent participants in religious festivals and attend mass regularly.

Cafiete's farmers are aware of the importance of education, especially formal

education. Currently the percent of young people enrolled in primary or secondary school

is about 95%. The educational conditions in the community are summarized in Tables 1-1

and 1-2. Table 1-1 indicates the educational institutions in the Cafiete community. The

state schools are funded directly by the government. Although these institutions offer free

services, they are perceived to be of low quality, with a few exceptions. The private

schools charge a fee for their services and generally provide a good quality education.

The private and state schools as well as the church schools receive part of their

funding from the government. These are perceived as the best schools in the zone

because they offer an excellent quality education at low cost, although admittance is very

difficult. More than 90% of the Valle Grande targeted audience with some formal

education attended state schools.

Table 1-2 reveals the conditions of the current Cafiete student community

regarding formal education (people currently enrolled), stratified by age, gender, and

level. Most of the population is young people (40% between 5 and 10 years, and 78 %

between 5 and 14 years) and they are mainly kindergarten, elementary or high school

students. Some students younger than 14, who attend high school, also enroll in college

(technical degree of three years duration), while others are not enrolled in school at all.










Table 1-1: Educational institutions in Cafiete

Education Level Total State Private Private & Church
School School State school School

1. Kindergarten 116 53 52 10 1
2. Elementary 184 110 55 17 2
3. Adult elementary* 8 5 8 -- -
4. High school 68 38 23 4 3
5. Adult high school* 15 9 6 -- -
6. College 23 6 17 -- --
7. Special school** 5 5 -- -- --

Total 419 226 156 31 6

Source: Peru's Education Ministry, 1998.
Note: Ages of the students, (1) four to six years, (2) six to twelve, (4) twelve to seventeen, (6) eighteen to
twenty-one.
* Adult students. ** Schools for special populations.


Less than seven percent (2,961) continue for post secondary studies (college,

technical degree or university, BS degree). The number of people who complete higher

education programs is considerably fewer, only 0.4% obtain a technical degree, and only

0.3% obtain a BS degree. Little difference exists in education levels based upon gender.

On the other hand, the perception of the adult towards informal education or

continuing education is not so good. Although the attendance in informal programs is

good, in many cases, the participants' expectations are to obtain easy solutions to their

often-complex problems. Such high expectations create conflicts between the local

development agencies and the clientele. It is difficult to communicate to clients that the

problem-solution process is often slow and complex.










Table 1-2: Formal education stratification in Cafiete

Age

5to 10to 15 to 20 to 30 to 40 to >65
Total 9 14 19 29 39 64

Total 46,792 18,745 17,720 9,101 2,561 685 368 112

No education 332 304 10 11 7 -- --
Kindergarten 3,769 3,651 37 39 19 5 18 --
Elementary 25,138 12,302 11,538 738 187 168 137 6,666
High school 13,326 -- 5,191 7,067 727 212 104 2,020
Some College 1,770 -- 747 400 890 110 28 88
College complete 165 -- 15 102 41 7 --
Some University 888 -- 267 513 73 28 7
University complete 138 -- -- 66 58 14 --
Not specified 1,266 488 444 224 50 18 32 11

Males 23,496 8,353 8,668 4,600 1,302 344 176 80

Females 23,296 8,392 8,552 4,501 1,259 341 192 50

Source: Peru's Education Ministry, 1998.


A recent survey ofValle Grande (1997) showed that young people are more

interested in careers such as military, police, and the like. Many young people do not

want to study agriculture or related careers because these careers are not as prestigious as

others. Currently, the recruitment of young people for agrarian careers presents a major

challenge to educators.

Almost all small farmers are "born farmers," most of them have always lived in

Cafiete, and others are recent settlers who had practiced agriculture in other regions of the

country before relocating. In addition, the Valley has an agricultural economic base.

Consequently, the farmers have a great deal of practical and empirical knowledge of









farming. It is not a surprise that in many cases, the farmers possess agricultural

knowledge beyond the knowledge of the change agent.

In terms of environmental conditions, the soils where these farmers work are

classified as fair or good. The weather conditions normally do not present barriers for

agricultural activities. Despite dry weather conditions, the whole valley is crossed by an

important river (Cafiete River) that supplies water for irrigation all year long. Four paved

main roads and many secondary dirt roads traverse the Valley; almost all the farmers

have easy access. For agricultural practices, farmers have access to inputs, such as

pesticides, fertilizers, and herbicides. In addition to that, they use tractors and trucks for

production and transportation. Most of the farmers' houses have services such as water

and electricity. Telephones are uncommon. The Valley has a geographic advantage due to

its location near Lima, which is Peru's largest market (eight million people).

Farmers, business people, and authorities define the social context of the Cafiete

Valley. Farmers are in charge of agricultural production, which is the principal activity of

the community. Although the business people provide services for the main activity, they

do not give a great deal back to the community. Business people serve as middlemen who

usually benefit most from the production-based economy.

The major authorities recognized in each district include the priest, the mayor, and

city commissioners. The priest exercises a powerful position. The people seek his advice

frequently, but the role of the priest does not include political affairs, just spiritual

concerns.

The community's income stratification is shown in the Table 1-3. Farmers are

stratified into four categories. The medium and large farmers have more than 12 ha of









land, an average annual income of US$ 6,090, and control 20% of the land in the Valley.

These farmers are not the target of Valle Grande but they could be collaborative leader

farmers to help in programming.

The small farmers constitute the very small, small and parcel categories, and they

are the target of the agency. They have less than 12 ha of land, an average income of US$

1,220, and control 80% of the lands of the Cafiete Valley. However, some small farmers

with less than 12 ha that have higher incomes (more than US$ 5,000) and are not clients

of Valle Grande Rural Institute.

Table 1-3 includes a column about "optimum income" for each farmer

classification. This information was generated in Valle Grande based upon the leader

farmers in each category and economic projections.


Table 1-3: Land and income in Cafiete

Farmer Area Number Total Average Total Actual "Optimum
(ha) farmers (ha) (ha) land income income"
(%) (US$/farmer) (US$/farmer)

Medium
& Large >12 260 4,530 17.42 20.00 6,090 35,235
Parcel* 3.5-12 2,002 10,362 5.18 45.70 1,820 10,530
Small <12 2,482 7,273 2.93 32.10 1,015 5,866
Very small < 1 715 491 0.69 2.20 245 418

Total 5,495 22,656 6.55 100.00

Source: Valle Grande Rural Institute, 1997.
* Small farmers resulting from Peru's Agrarian Reform (1969-1985)


Medium and large farmers are mostly white Hispanic. Small farmers have an

ethnic background called "mestizo" (Figure 1-4, Appendix A) that represents a racial mix









between Hispanic and natives. In the last twenty years, natives have migrated from the

highlands to Cafiete Valley. These immigrants are mainly small marginal farmers.


Figure 1-4: Small farmers planting potatoes


There is a small group of black people, few of whom own land. Some segments of

medium farmers are "Nikkei" (Japanese ascendance). They are usually hard and efficient

workers and maintain strong economic networks.














CHAPTER 2
LITERATURE REVIEW AND METHODOLOGY


Literature Review


Introduction


The present chapter schematizes the theoretical framework for this study. Farming

Systems Research & Extension (FSR&E), the spinal column, was complemented by other

approaches: Targeting Outcomes Programs (TOP), Participatory Rural Appraisal (PRA),

and Extension Programming (EP). These other frameworks were used not to contrast

with FSR&E, but to enrich the research process.

Farming Systems Research & Extension


FSR&E represents a unique approach to agricultural research and extension; it

was formulated in response to the complex and diverse production methods encountered

on small often-mixed farms in the developing world. (Zandstra, 1983).

According to Hildebrand and Waugh, 1982, FSR&E,
Deals mostly with conditions inside the farm gate. It is concerned with
technology generation, evaluation, and delivery. It emphasizes on-farm
biological research, and it is applied, farmer-oriented, agro-biological
research, supported by the social sciences in a team effort, which includes
extension responsibilities. The principal product is the technology and the
primary clients are the farmers (p. 13).

According to D.W. Norman (1982):

A Farming System is the result of a complex interaction of a number of
interdependent components. At the center of this interaction is the farmer,

12








who is the central figure. Moreover, both farm production and household
decisions of small farmers are intimately linked and should be analyzed in
Farming Systems research. A specific farming system arises from the
decisions made by small farmers or farming families with respect to
allocating different quantities and qualities of land, labor, capital and
management to crop, livestock, and off-farm enterprises in a manner
which, given the knowledge the household possesses, will maximize the
attainment of the family goals (p. 8-9).

The farming systems perspective is especially appropriate for such a process-

oriented approach to gender analysis, which includes a focus on small farm households

and on the participation of farmers in the research and extension process (Poats,

Schmink, & Spring, 1988). The same authors believe that a farming systems emphasis on

reaching low-income groups can illuminate women's roles in agricultural development,

and there is a high and growing proportion of female-headed households emphasizing the

economic importance of decisions made by women in poor populations.

Stages in Farming Systems Research & Extension


Norman (1982, p. 10) proposed four stages in Farming Systems Research &

Extension. Norman stated the following regarding the first stage:

The descriptive or diagnostic, in which the actual farming system is
examined in the context of the total environment to identify constraints
farmers face and to ascertain the potential flexibility in the farming system
in terms of timing, slack resources, etc. An effort is also made to
understand goals and motivation of farmers that may affect their efforts to
improve the farming system.

In this first stage, the researcher looked for needs and/or problems that Cafiete
small farmers' currently face.
Need can be defined as a deficiency, imbalance, lack of adjustment, or gap

between the present situation and a set of societal norms believed to be more desirable

(Boone, 1985). The first sub-process of Boone's Extension Programming, planning,









includes: linking the organizations to its publics, in which he discusses the identification

of the publics to be served with effective mapping, as well as identification and

interfacing with the leaders. He concludes that the most important task is the

collaborative effort to identify needs of the publics.

Participatory Rural Appraisal (PRA) describes a growing family of approaches

and methods to enable local people to share, enhance and analyze their knowledge of life

and conditions, to plan and to act. (Chambers, 1992). PRA is shared and owned by local

people; the behavior and attitudes of outside facilitators are crucial, including relaxing

not rushing, showing respect, "handing over the stick," and being self-critically aware.

The same author posits that the PRA has been used in natural resource

management, agriculture, programs for the poor, and health and food security. Evidence

to date shows high validity and reliability in information shared by rural people through

PRA.

Targeting Outcomes of Programs (TOP) developed by Bennett and Rockwell

(1995) proposes that program planning first targets SEEC -social, economic, and

environmental conditions- then the practices necessary to achieve the targeted conditions,

and the KASA -knowledge, attitudes, skills, and aspirations- needed to realize adoption of

the practices.

The second FSR&E stage is the design stage. According to Norman:

The design stage, in which a range of strategies is identified, is thought to
be relevant in dealing with the constraints delineated in the descriptive or
diagnostic stage. Heavy reliance at this stage is placed on obtaining
information from the "body of knowledge."









In this second stage, the researcher analyzed different alternatives in different

scenarios through linear programming simulation of individual households. Production

functions were also used to analyze the cotton production enterprise in the Cafiete Valley.

The second sub-process of Boone's Extension Programming, design &

implementation, includes: Plans of action, where he devotes great attention to develop

specific educational strategies and learning experiences based in the planned program and

action strategies, which are devoted to develop the plans of action through marketing

techniques and using leader resources.

Targeting Outcomes of Programs (TOP) proposes that the program targets the

reactions needed to ensure sufficient participation in program activities that enable

learning the intended KASA. Finally, the resources necessary to support the activities are

acquired. (Figure 2-1).



SEEC
Practices
KASA
Program planning Reactions Program performance
Participation
Activities
Resources
Figure 2-1: Hierarchy in programming.



The third FSR&E stage is the testing stage. Norman revealed the following

regarding this stage:

The testing stage, in which a few promising strategies arising from the
design stage are examined and evaluated under farm conditions to
ascertain their sustainability for producing desirable and acceptable
changes in the existing farming system.









The FSR&E approach lies strongly in that the limited resource farm households

live and work on farms characterized by a high degree of both biophysical and

socioeconomic diversity. Technologies developed solely under the conditions common

on experiment stations are rarely transferable directly to small scale limited resource

farmers (Hildebrand & Russell, 1996). Adaptability of experimental treatments to

particular environmental conditions must be identified and incorporated into a specific

experimental design ex ante, that is, before the trial is put in the field.

The fourth stage in the FSR&E model is the extension stage. According to

Norman:

The extension stage is the last stage in which the strategies that were
identified and screened during the design and testing stages are
implemented.

Boone (1985) defines Extension Programming as:
A comprehensive, systematic, and proactive process encompassing the
total planned, collaborative efforts of the adult education organization, the
adult educator in the roles of change agent and programmer,
representatives of the learners, and the learners themselves in a purposive
manner and designed to facilitate desirable changes in the behavior of
learners and the environment in which they live (p. 41).

In objective three of this study, the researcher proposes a priority list of future

extension programs for the Valle Grande development agency for the low resource

farmers in Cafiete Valley, Peru.

The TOP model suggests that program performance expends targeted resources to

conduct the program activities intended and obtain targeted participation with positive

reactions.








Affirmative reactions help program participants acquire targeted KASA leading to

the adoption of targeted practices. Use of such practices helps achieve the targeted SEEC

changes, according to Bennett and Rockwell (1995).

De los Santos and Norland (1990) used Bennett's theoretical framework to study

extension programming in the Dominican Republic. They found that the farmers gained

mostly knowledge and skills toward the program and that older farmers tended to hold

negative attitudes toward information provided by extensionists.

The third and final sub-process ofBoone's extension programming is evaluation

and accountability. Evaluation and accountability include determining and measuring the

program outputs, assessment of program inputs, and, using evaluation findings for

accountability purposes.

After extension programs are initiated, there is a need to monitor the programs.

According to Rossi and Freeman (1985) there are several reasons monitoring of programs

is required. First, monitoring provides judgment information. Second, monitoring is

required for accountability purposes. Third, it is adjunct to impact assessment. Fourth,

monitoring evaluations often are instrumental in decisions to continue, expand, or

terminate ongoing programs. The authors mention that the monitoring of programs is

directed at two key questions: (1) whether or not the program is reaching the appropriate

target population, and (2) whether or not the delivery of service is consistent with

program design specifications.

The social sciences provide reliable, useful information to biological scientists.

Social science research is useful in making suggestions concerning the consequences of

technology change (DeWalt, 1985).








Sociological and economic variables play complementary roles in the innovation

decision process; the sociological variables had more impact in the adoption stage, while

the economical factors were more predictive in the implementation-confirmation stages

(Sapp & Jensen, 1997).

It was found that farmers preferred neighbors as sources of agricultural

information. The rate of early adoption was influenced by the farmers' level of education

for attaining knowledge and land holding. Organizational assistance and communication

showed a higher correlation with adoption of farming innovation (Sandhu & Allen, 1974;

Pfeffer & Gilbert, 1989).

Methodology


Population and Sample


The population and sample differed based upon the multiple data collection

methods used by the researcher. For example, in order to develop the cotton production

functions, the researcher used a population of small farmers who borrowed money

through the Valle Grande Rural Institute during the period 1992 through 1998 (N=

1,860). A purposeful sample (n= 622) consisting of farmers with complete records was

used to develop the production functions.

In terms of the linear programming model, the researcher used data from

numerous sources including a sondeo, survey, and selected secondary data. First, the

researcher conducted a sondeo consisting of a sample of 22 farmers in the area. A sondeo

is an open-ended, non-structured interview technique (Hildebrand, 1976).








Second, the researcher conducted a survey consisting of structured questions

developed based upon personal knowledge of the Cafiete Valley, and the sondeo results.

A questionnaire consisting of 70 items was developed by the researcher (Appendix B).

The instrument contained three sections. The first section had three subsections: (1)

household information, (2) agricultural factors; and (3) economic information. The

second section consisted of seven open-ended needs assessment questions. The final

section included 13 open-ended questions regarding farm problems and concerns.

The population for the survey consisted of limited resource farmers in the Cafiete

Valley (N=4,800). A random sample of 60 farmers was selected for participation in the

survey; the raw data can be found in Appendix C. In an effort to collect information that

was reflective of the population, the researcher took a map of the Cafiete Valley and

divided the area into 60 zones. One zone was then randomly selected at a time by a

computer. The researcher subsequently randomly selected a limited resource household

to interview in each zone. Any small farm household had the same chance to be selected.

Sometimes the interviews were accomplished in the farms and other times in the houses.

Survey data were collected from a broad cross section of valley residents and the study

area was completely covered, Figure 2-2.

For both the sondeo and survey, households had to meet the following criteria (1)

farm less than 12 ha of land, (2) have a net annual income less than US$ 5,000, and (3)

generate the majority of the household's income from agricultural production.









12" 54,- -
I-- --- '"


12" ,. J =- ,. _


S- --- .- ... -- / *
,,._. -.s --
*A"* \ *-. -. L-. 1.
A
.- - -, *,-
.-- ', .. .., -. ... .
\.-S '-"A ^llr










HOUSEHOLD INTERVIEWED
Data Collection and Desig











Production function" '
.(resources) into outputs (commodities). is a rule for assigning




























to each value in one set of variables (the domain of the function) a single value in another
; .. '., .* ... ., -






























set of variables (the range of the function). A general way of writing a production

function is y= f (x), where y is an output and x an input (Debertin., 1986).
-"- ,",.:-'-.4. "-..T ,, -

.-a -* *- s. "


1. ---"f .--- .
3 ....A. n --- -a -s -


Figure 2-2: Survey interview locations.



Data Collection and Design


Production function

A production function describes the technical relationship that transforms inputs

(resources) into outputs (commodities). Mathematically, a function is a rule for assigning

to each value in one set of variables (the domain of the function) a single value in another

set of variables (the range of the function). A general way of writing a production

function is y= f(x), where is an output and x an input (Debertin, 1986).








Regression analysis is the way to transform data representing an existing

phenomenon (physical, socioeconomic or biological) into a mathematical expression to

help understand the relationships among the dependent and independent variables

(Hildebrand, 1997). The production function data were gathered from records maintained

by the Valle Grande Rural Institute.

Linear programming

Linear Programming (LP) is a method for simulating and analyzing family farm

livelihood systems by determining a combination of farm and non-farm activities that is

feasible, given a set of fixed farm constraints. It maximizes (or minimizes) a particular

objective or family goal. The model, according to Hildebrand and Araijo (1997), requires

the following in each farm family situation:

(1) The farm and non-farm activities and options with their respective
resource requirements and any constraints on their production, (2) the
fixed resources and other maximum or minimum constraints that limit
farm and family production, (3) cash costs and returns of each activity;
and (4) a defined objective or objectives. (p. 3).

In a first approximation to a linear programming methodology in Cafiete, a family

livelihood system with real data was selected. Using data that are accurate on various

types of households will reveal ways to improve the complex farming systems in Cafiete.

(Cabrera, 1997).

In terms of the linear programming data, numerous sources of data were utilized.

A sondeo was coordinated by the researcher between May 11 and May 15, 1998. A

sondeo is a qualitative data collection technique, also referred to as a reconnaissance

survey, informal survey, and exploratory survey. The sondeo is an important needs

assessment tool used in FSR&E (Hildebrand, 1976). A properly conducted informal








survey can provide accurate and comprehensive information on the ecology of farming

and related practices (Rhoades & Bidegaray, 1987). According to Franzel (1984), the

sondeo has the following four distinguishing characteristics:

(1) Farmer interviews are conducted by researchers themselves, (2)
interviews are essentially unstructured and semidirected, with emphasis on
dialogue and probing for information (questionnaires are never used), (3)
informal random and purposive sampling procedures are used, and (4) the
data collection process is dynamic. (p. 1).

The researcher used an interdisciplinary team consisting of six professionals with

expertise in extension, economics, and technical agriculture. The sondeo or rapid rural

appraisal team spent one day driving through the valley making general observations.

Next, the six researchers split up into two teams and interviewed selected household

representatives. Finally, the six pooled their findings and wrote a final report.

The survey information was collected through personal interviews conducted by

the researcher between May 18 and July 17, 1998. Each interview lasted between one and

two hours. At least one adult household member was interviewed. In addition to the

interviews, the researcher made and recorded personal observations regarding each

household.

The researcher also collected information for the linear programming model from

Valle Grande Rural Institute records, records maintained by the city government located

in Cafiete, and Peru's Ministry of Agriculture.

Data Analysis


The data were analyzed using Microsoft Access 97 SR-1, Microsoft Excel 97

SR-1 Microsoft Visual Basic, and Microsoft Visual Basic.














CHAPTER 3
COTTON PRODUCTION FUNCTIONS


Introduction


The purpose of this analysis was to generate production functions based upon

multiple regression of cotton yield based upon fertilization and environmental factors.

Table 3-2 shows the summary of the data used in the production functions. All the

raw data can be found in Appendix D.

The dependent variable in all cases was the cotton per ha yield in quintals (100

lb). The independent variables tested were: nitrogen (N) in kilograms, phosphorus (P) in

kilograms, potassium (K) in kilograms, annual environmental index (average production

per ha for the specific year) (EI) in quintals, and the interaction between the fertilizers

and the annual environmental indexes, EIxN, EIxP, EIxK.

Analysis


The fertilization rates used by the farmers in the cotton crop were the amounts

recommended by Valle Grande Rural Institute. These amounts had little variation within

zones (Table 3-1, Figure 3-1), and consequently the subsequent analyses are only on a

small data range. Therefore, it is recommended to complement these results with the

results of on-farm trials, in which the independent fertilization variables will have greater

ranges. It would also be desirable to include more zones in other analyses to generate

more complete production curves.









Table 3-1: Range of fertilization factors by geographic zone, kg/ha
Cerro Alegre La Quebrada San Benito San Francisco Santa Barbara


Palo Isla


Quilmana


N 170-240 180-230 190-245 200-250 110-229 200-240 200-240
P 46-110 46-120 46-120 30-103 46-100 80-100 80-100
K 40-100 50-100 50-100 40-90 25-95 90-100 90-100


1l mItI


Nitrogen


Phosphorus




IIIII1"


Potassium



IInIII"


do aC < < (g C C j (
2 M LL a q L L
0 c 0(0 IBo 0 ( -B
1 ( ) 0 M I0
0 j 0 0j C Co
Zone

Figure 3-1: Range of fertilization factors by geographic zone, kg/ha



The following is an analysis of seven multiple linear regressions (one for each

association or geographic zone) of the following form:

Y =a + biX + b2X2 +... bkXk


Where:


is the estimated value of cotton yield in one zone in
quintals (100 lb),


X1, X2, ...Xk are the independent variables,

a is the intercept, and


are the partial regression coefficients.


300
250
200 -
S150
100 -
50 -
nA


bi, b2, ...bk










The hypotheses tested in each case were:

F test for the whole regression equation:

Ho: R = 0 in the population.

H1: R > 0 in the population, and

t-test for the independent variables:

Ho: bk = 0 in the population

H1: bk < 0 in the population

The coefficients that were statistically significant at 95% confidence level (a =

0.05) are the only independent variables reported in each specific equation because they

can be used for prediction purposes. Although curvilinear variables (x2) were included in

the regression analyses, none were significant at this level of confidence.

The annual environmental index (EI) is the result of calculating the average of all

available production data for each year. As seen in Figure 3-2, the annual environmental

conditions are responsible for drastic changes in the yield variable of the cotton crop. For

analyses and recommendation purposes the production years are divided into good (more

than 60 qq/ha), fair (more than 45 but less or equal to 60 qq/ha), and poor (less or equal

to 45 qq/ha).

The interactions of the environmental index variable (EI) and the macronutrient

variables (N, P, K) were the result of multiplying both values creating the interaction

variables (EIxN, EIxP, EIxK).










Table 3-2: Data used in production functions
Production Season Association Number of Farmers

92/93 37
Cerro Alegre 14
San Benito 23

93/94 110
Cerro Alegre 19
La Quebrada 3
Santa Birbara 28
San Benito 51
San Francisco 9

94/95 72
Cerro Alegre 16
Santa Brbara 2
San Benito 46
San Francisco 8

95/96 138
Cerro Alegre 13
La Quebrada 18
Palo Isla 19
Quilmana 22
Santa Birbara 24
San Benito 41
San Francisco 1

96/97 124
Cerro Alegre 8
La Quebrada 22
Quilmana 19
Santa Birbara 18
San Benito 44
San Francisco 13

97/98 141
Cerro Alegre 25
La Quebrada 15
Palo Isla 4
Quilmana 18
Santa Birbara 23
San Benito 38
San Francisco 18


Seven associations


Total













75.00 -

70.00

65.00 -

60.00

*- 55.00
0*
2 50.00

45.00 -

40.00

35.00 -

30.00
92/93


Good Year


64.30


Bad Year


93/94


94/95 95/96
Production Year


Figure 3-2: Annual environmental index for cotton yield in Caniete




Cerro Alegre


The cotton production function for Cerro Alegre is presented in Table 3-3. Fifty-

one percent of the variance in cotton production was accounted for by four factors:

phosphorus (P), potassium (K), annual environmental index (EI), and the interaction of

phosphorus by the annual environmental index (EIxP). A statistically significant

proportion of the variance in final cotton production in Cerro Alegre was explained by

those factors (F = 23.89, p < 0.001), with a very high multiple correlation (R = 0.72).


2 Interpreting strength correlations according to Davis (197 1)


96/97


97/98









Table 3-3: Cotton production function in Cerro Alegre


Variables Standard error b t p

Phosphorus (P) 0.89 -4.01 -4.50 <0.001
Potassium (K) 0.14 -0.34 -2.45 0.016
Environmental Index (El) 1.51 -6.16 -4.09 <0.001
Interaction (EIxP) 0.02 0.071 4.62 <0.001
Intercept 88.79 430.43 4.85 <0.001

R = 0.51 R = 0.72 F = 23.89
Adjusted R2 = .49 p< 0.001
Standard error = 12.99 Observations = 96

Y' in Cerro Alegre = 430.43 4.01P 0.34K 6.16 El + 0.071EIxP


La Quebrada


Table 3-4 summarizes the cotton production function for the La Quebrada zone.

Fifty one percent of the variance of the cotton production was explained by two factors:

phosphorus (P) and the interaction of phosphorus and the annual environmental index

(EIxP). A statistically significant proportion of the variance in final cotton production in

La Quebrada was explained by those factors (F = 28.87, p < 0.001) with a very high

multiple correlation (R = 0.72).


Table 3-4: Cotton production function in La Quebrada


Variables Standard error b t p

Phosphorus (P) 0.21 -0.91 -4.44 <0.001
Interaction (EIxP) 0.00 0.014 7.60 <0.001
Intercept 15.08 77.69 5.15 <0.001

R = 0.51 R = 0.72 F = 28.87
Adjusted R2 = .49 p < 0.001
Standard error =18.39 Observations = 58
Y' in La Quebrada = 77.69 0.91P + 0.014EIxP









Palo Isla


Table 3-5 summarizes the cotton production function in Palo Isla. As indicated,

eighty-four percent of the variance of the cotton production was explained by three

factors, potassium (K), nitrogen (N), and the interaction of nitrogen by the annual

environmental index (EIxN). A statistically significant proportion of the variance in final

cotton production in Palo Isla was explained by those factors (F = 34.42, p < 0.001) with

a very high multiple correlation (R = 0.92).


Table 3-5: Cotton production function in Palo Isla


Variables Standard error b t p

Potassium (K) 0.82 3.58 4.35 <0.001
Nitrogen (N) 0.32 -1.72 -5.44 <0.001
Interaction (EIxN) 0.00 0.012 9.19 0.010
Intercept 72.57 -81.50 -1.12 0.280

R2 = 0.84 R = 0.92 F = 34.42
Adjusted R2 = .82 p < 0.001
Standard error = 11.91 Observations = 23

Y' in Palo Isla = -81.50 + 3.58K 1.72N + 0.012EIxN


Santa Barbara


Table 3-6 summarizes the cotton production function for the Santa Barbara zone.

Thirty percent of the variance of the cotton production was explained by three factors,

potassium (K), the interaction of potassium by the environmental index (EIxK), and the

interaction of phosphorus by the annual environmental index (EIxP). A statistically

significant proportion of the variance in final cotton production in Cerro Alegre was









explained by those factors (F = 13.02, p < 0.001) with a substantial multiple correlation

(R = 0.55).


Table 3-6: Cotton production function in Santa Barbara


Variables Standard error b t P

Potassium (K) 0.29 -1.66 -5.72 <0.001
Interaction (EIxK) 0.00 0.02 4.32 <0.001
Interaction (EIxP) 0.00 -0.006 -2.19 0.003
Intercept 17.12 119.45 6.98 <0.001

R2 = 0.30 R = 0.55 F = 13.02
Adjusted R2 = .28 p < 0.001
Standard error = 22.37 Observations = 95
Y' in Santa Barbara = 119.45 1.66K + 0.02EIxK 0.006EIxP


San Benito


Table 3-7 shows the cotton production function for San Benito. Thirty six percent

of the variance of the cotton production was explained by four factors, potassium (K), the

interaction of potassium by the annual environmental index (EIxK), nitrogen (N), and the

interaction of nitrogen by the annual environmental index (EIxN). A statistically

significant proportion of the variance in final cotton production in San Benito was

explained by those factors (F = 33.46, p < 0.001) with a substantial multiple correlation

(R = 0.60).








Table 3-7: Cotton production function in San Benito


Variables Standard error b t p

Potassium (K) 0.42 1.58 3.74 0.000
Interaction (EIxK) 0.00 -0.025 -3.37 0.000
Nitrogen (N) 0.22 -0.87 -4.03 0.000
Interaction (EIxN) 0.00 0.016 4.92 0.000
Intercept 27.08 44.57 1.65 0.101


RP = 0.36 R = 0.60 F = 33.46
Adjusted R2= .35 p <.000
Standard error = 15.90 Observations = 243

Y' San Benito = 44.57 + 1.58K 0.025EIxK 0.87N + 0.016EIxN


San Francisco


Table 3-8 shows the cotton production function for the San Francisco zone.

Seventy seven percent of the variance of the cotton production was explained by five

factors, nitrogen (N), phosphorus (P), potassium (K), the interaction of phosphorus by the

annual environmental index (EIxP), and the interaction of potassium by the annual

environmental index (EIxK). A statistically significant proportion of the variance in final

cotton production in San Francisco was explained by those factors (F = 29.54, p < 0.001)

with a very high multiple correlation (R = 0.88).

Quilmana

Table 3-9 reports the cotton production function for Quilmand. Fifty-four percent

of the variance of the cotton production in Quilmana was explained by two factors,

potassium (K) and the interaction of potassium by the annual environmental index

(EIxK). A statistically significant proportion of the variance in final cotton production in









Quilmana was explained by those factors (F = 32.28, p < 0.001) with a very high multiple

correlation (R = 0.73).


Table 3-8: Cotton production function in San Francisco


Variables Standard error b t p

Nitrogen (N) 0.15 0.46 3.09 0.003
Phosphorus (P) 1.38 4.90 3.55 <0.001
Potassium (K) 1.36 -5.57 -4.08 <0.001
Interaction (EIxP) 0.02 -0.088 -3.51 <0.001
Interaction (EIxK) 0.02 0.103 4.17 <0.001
Intercept 34.62 -63.01 -1.82 0.075

R2 = 0.77 R = 0.88 F = 29.54
Adjusted R2 = .74 p< 0.001
Standard error = 11.18 Observations = 49

Y' in San Francisco = -63.01 + 0.46N + 4.90P 5.57K 0.088EIxP + 0.103EIxK




Table 3-9: Cotton production function in Quilmand


Variables Standard error b t p

Potassium (K) 0.31 -0.84 -2.75 0.008
Interaction (EIxK) 0.00 0.01 6.89 <0.001
Intercept 29.63 52.06 1.76 0.084

R = 0.54 R= 0.73 F = 32.28
Adjusted R = .52 < 0.001
Standard error = 18.27 Observations = 59
Y' in Quilmana = 52.06 0.84K + 0.01EIxK


Cotton Yield Analysis Based on Production Functions


Based on the function equations, the fertilization ranges, and the environmental

indexes, an analysis of the estimated cotton yields is presented for all and each one of the









different zones. Three levels of each fertilizer factor were used, lowest, medium, and

highest, according to the range of the data of each zone (Table 3-1). Three levels of

environmental conditions were also included as proposed in Figure 3-2: good year (65

qq/ha), fair year (52.5 qq/ha), and poor year (37.5 qq/ha) in each particular analysis. Sixty

quintals of cotton per ha is considered the acceptable, good, or objective yield in each

case. The fertilizer factors reported were those with statistical significance and therefore

those that can be used in predictions with confidence.

Figure 3-3 shows the estimated cotton yield in Cerro Alegre in function of

phosphorus and potassium fertilization rates in different environmental conditions.


140
120 ---- Poor Year --- Fair Year -a Good Year
S100
S80
V 60
40
20
0 C----,---------------------C--C
o 0 0 0 0 0 0 0
-r, i- r- r-- r- o o o
Co cO cO CT) 0
0- 00 -
Phosphorus Potassium (P-K) fertilization in kg/ha


Figure 3-3: Estimated cotton yield in response to phosphorus and potassium fertilization
rates, Cerro Alegre



A poor year as previously defined might become a good year for this zone

because the best yields are found over that curve with lowest rates of both phosphorus

and potassium. Drastic changes in the yield curves would indicate phenomena of

synergism and antagonism among the fertilizers and the soil, which enhance or deplete

the productivity. In a good year, Cerro Alegre could obtain reasonable yields with highest









phosphorus and lowest potassium. In a fair year, lowest rates of fertilizers would be

appropriate to obtain the best yields.

Figure 3-4 shows the curves of La Quebrada production functions. The explaining

factors are the phosphorus fertilizer and the interaction of phosphorus with the

environmental conditions.




100 ---Poor Year -- Fair Year -+-Good year
S80-
c- 60
S40
>- 20
0
46 83 120
Phosphorus Fertilization (kg/ha)

Figure 3-4: Estimated cotton yield in response to phosphorus fertilization rates, La
Quebrada



In all cases, poor, fair, and good year, there is no reason to apply more than 46

kg/ha of phosphorus because it would not increment the yield of cotton. Indeed, applying

more phosphorus, in poor or fair years, would decrease the yields. This negative effect of

the phosphorus fertilizer could be explained in this zone for two reasons. First, these

small portions of curves are part of major curves of production, what are conceived as

curvilinears, in which they might be located in the top or decreasing segments. Second,

La Quebrada is a zone with pest problems. These problems would become more severe in

presence of more phosphorus fertilizer. It is probably better not to recommend cotton

production in poor years because the yields would not reach the acceptable limit.









Figure 3-5 shows the Palo Isla production functions explained by nitrogen and

potassium fertilizers and the interaction of nitrogen with the environmental index.




100 Poor Year ---Fair Year -- Good Year
80
C" 60 --
-o 40---
>- 20


0) 0) 0 0) 0) 0 0) 0) 0
C 0 T- 6 6 6 T -
o o o0 o0
o o 0 K C( '0 V 0
(N 04 O CN C0 C%( C(N CN (
CN (N (N

Nitrogen Potassium (N-K) Fertilization in kg/ha

Figure 3-5: Estimated cotton yield in response to nitrogen and potassium fertilization
rates, Palo Isla



Palo Isla is not a good zone for cotton production. In poor years, it is difficult to

obtain any cotton yield. In fair years, there is also a great chance of failure unless there

would be minimum nitrogen (200) and maximum potassium (100). The same rates of

fertilizers would be recommended in a good year to obtain the best yields.

Figure 3-6 shows the curves of the cotton production functions in Santa Barbara.

The curves indicate higher yields with lower fertilizer application, especially with lower

potassium. The Santa Barbara soils are saline and this problem could be aggravated with

potassium applications. On the other hand, cotton is a crop adapted to saline soils and it

could result in good yields even in poor years if fertilization is balanced. In all years, the

lowest rates of phosphorus and potassium would be recommended.











100 -.- Poor Year --- Fair Year
90 Vzz
80T\'K '.


0i


-A- Good Year


70
60
50

30
20
10

O) o o IO o o t 0 0
(C (D 1 (C D 0 C0 <0 0)
(DPhos s (P-) Fn in

Phosphorus Potassium (P-K) Fertilization in kglha


Figure 3-6: Estimated cotton yield in response to phosphorus and potassium fertilization
rates, Santa Barbara



Figure 3-7 presents the curves of cotton production for San Benito. As seen in the

curves, San Benito is a good cotton zone in good years. In a good year, good cotton

yields are expected with almost all fertilization N-K combinations, although best results

could be obtained with highest rates of nitrogen and lowest rates of potassium. In fair

years, best yields are predicted with maximum amount of potassium (100 kg/ha) and

minimum amount of nitrogen (190 kg/ha). In poor years, it would be better not to

produce cotton in San Benito.

Figure 3-8 shows the curves of the cotton production functions for San Francisco

zone based on nitrogen, phosphorus, potassium, and environmental factors.

In good years, the yield is enhanced with the highest rates of nitrogen (250 kg/ha)

and potassium (90 kg/ha) and lowest rate of phosphorus (30 kg/ha). The amount of

potassium has more effect than nitrogen and phosphorus: the peaks in the curve are

determined by this factor. In the fair years, the curve increases with more nitrogen and









phosphorus, but decreases slightly with more potassium. The best yields in fair years

could be obtained with highest rates of nitrogen (250 kg/ha) and phosphorus (103 kg/ha)

and lowest rate of potassium (40 kg/ha). In San Francisco, as it was in Cerro Alegre, the

poor years might become good years. The best yields are found over this curve. In poor

years, the best yields could be reached with the same fertilizer rates as fair years. In fair

and poor years, the peaks in the curves are determined with the lowest amounts of

potassium.


90


60 _- --- ------.-
70 ____- ___


50
S40
30
20
10 --- Poor Year -U- Fair Year -+-Good Year
0
o .o o U -- 0 C
0 a- M -



Nitrogen Potassium (N-K) Fertilization in kg/ha

'Figure 3-7: Estimated cotton yield in response to nitrogen and potassium fertilization
rates, San Benito



Poor years and, in to a lesser extent, fair years are characterized by the presence

of major pest problems that could be the reason that potassium becomes a deleterious

factor because it increases the chance of pest infestation.

Figure 3-9 shows the production functions in Quilmana based on potassium

fertilization and environmental factors. As seen in the curves, Quilmana is not a good

cotton zone, the yields are low. The potassium fertilizer negatively affects yield for the









three types of years. The least affected is the curve of the good year. Probably, it is better

not to raise cotton in Quilmana.


-0- Poor Year


- Fair Year


-- Good Year


CF
c


40 Poor Year -- Fair Year Good Year
35- A ,





15 ,
10 -
30
25


15
101



90 95 100
Potassium (K) Fertilizer in Kg/ha

Figure 3-9: Estimated cotton yield in response to potassium fertilization rates, Quilmana


160
Atf I


I40 1
120


804



20






Nitrogen-Phosphorus-Potassium (N-P-K) Fertilizers in Kg/ha

Figure 3-8: Estimated cotton yield in response to nitrogen, phosphorus, and potassium
fertilization rates, San Francisco


t














CHAPTER 4
LINEAR PROGRAMMING


Introduction


The purpose of this analysis was to study the various farming systems in the

Cafiete Valley community in order to evaluate different scenarios. Linear programming

was used first to simulate the households' current situation and after statistical validation,

predict different farmer's responses to various scenarios.

The simulation of the Cafiete community was done using the 60-household survey

sample. Independent linear programming models -one for each household- were

developed. That fact was critical for analyzing the overall community while maintaining

the diversity of the systems. In all cases, there was a one-year model and a six-year

model.

The simulation results were statistically compared with the real data to validate

the models. The validation process resulted in verifying that the models represented the

same population as the sample, and consequently would allow an extensionist to use the

models with confidence in the Cafiete community. The simulation results were

aggregated, grouped, and analyzed at the community level in different scenarios.

The simulation models are "interactive-working models," which, even though

satisfactorily simulating the community, should be updated and reviewed frequently. The

prices and production conditions could change through time; different technologies can

be available; new activities could be added, and ultimately, the models can be used with









each specific household, after updating with appropriate data. Use of Visual Basic,

embedded programming language of Microsoft Excel 97 SR-1, allowed the researcher

to make the interactive process quick and efficient. Appendix E contains two interactive-

working Excel files, the one-year and the six-year linear programming models of the

Cafiete community.

Assumptions for Linear Programming


Based on the data gathered in the survey, the situation was simulated by a linear

programming model that maximizes discretionary cash at the end of the year after

satisfying all basic family needs (Table 4-1). The linear models include the following

activities and constraints:

1. There are two well-known production seasons in Cafiete. The matrix was

divided into these: First season, August 15t to April 14th, and second

season, April 15th to August 14th.

2. Land is a limited resource in the Cafiete Valley. Its use is intensive.

3. Renting land from both the owners' perspective and the renters' perspective

is a common practice in the community that was included in the models.

The farmers have the opportunity to rent out their lands or take or rent in

other farmers' lands for a reasonable price. The cash received when renting

land is less than the cash needed to take or rent in others' land because of

some fixed costs (i.e. irrigation fee, municipality tax). The commercial

transaction requires in both cases five units of household management

(explained in number 7 below).









4. Labor is a limited resource. It is determined by the number, age and gender

of the household members (Table 4-5). Each child younger than five years

requires 0.75 day-labor per day, each child between 5 to 14 years

contributes 0.5 day-labor per day as well as the males older than 65 years

and the females older than 75 years. The males between 14 and 65 and the

females between 14 and 75 years contribute 1.00 day-labor per day to the

household. The female labor is more limited than the male because they

attend the children, the house, and most of the livestock.

5. The household has the opportunity to hire people in labor intensive-seasons

(labor is available in the community). It is also common that the household

members work for others (off-farm labor) to supplement household income.

The cost to hire someone is S/. 10.00 per day, however the cash contribution

to the household income when working for another is only S/. 8.00 per day

because the member keeps 20% for his/her own. At least 50% of the total

household labor is provided by its members. The house and livestock

activities do not use hired people.

6. Water is not a limited resource in the first semester, but it is, in some cases,

in the second semester. The availability of water is determined by the

frequency, time, and flow of water received in each household. The house

and the livestock consume the same irrigation water.

7. Management is an aggregate index computed by summing the total years of

education of all members in each particular household (the following values

were utilized in the index: 1= illiterate, 3= primary, 7= secondary, 10=









superior). Management is not a limited resource in the basic one-year model.

It becomes a serious limiting factor in the six-year model because the new

activities (asparagus and grapes) require more management skills than

traditional crops and because in this simulation many households would be

able to rent in or rent out land, activities that require management resources.

8. Credit is an available resource for cotton and maize in the first semester and

for maize in the second semester (development agencies, industry). This is a

limited resource, mostly in the second season. The interest is 10% in the first

semester and 8% in the second. The farmers may get cash credit (retailer,

intermediaries, pesticides shop, etc.), and may be assessed as much as 100%

interest for just one season of credit. Credit is also available for activities

such as asparagus and grapes in the six-year model.

9. Each household has some cash at the beginning of each season. This money

is used for household expenses, livestock, and production activities. The

quantity was computed based upon family activities, and production records

as well as previous debts. The household cash is a limited resource in both

semesters.

10. The basic activities of the household in the first semester are general

household expenses (food, services, etc.), livestock (mainly poultry, but also

rabbits, Guinea pigs, or sheep), cotton, maize, and sweet potato. In the

second semester, there is no cotton production. The production and prices

vary in maize and sweet potato according to semesters (Tables 4-2 & 4-4).









11. The family and livestock consume maize and sweet potato produced on the

farm. The family requires a certain amount of livestock produced by the

household (Table 4-3). These quantities are computed according to the

number, age, and gender of the household members.

12. The cash, if not used in the first semester, can be transferred to the second

semester. If the cash in the second semester is not used, it is transferred to

the end of the year cash. The needed cash is transferred to the first semester

of the next year, in the six-year model.

13. The cash at the end of the year could be negative in the one-year and in the

six-year models. Negative cash at the end of the year indicates a non-

sustainable system.

14. In the six-year model, the perennial crops -asparagus and grapes- have

different resource needs and production rates through the years, Tables 4-5,

4-6, & 4-7. The land used by these crops is not available for others crops

during the six years.

15. The traditional crops (cotton, maize, and sweet potato) require one unit of

management per enterprise. Similarly, the house and livestock activities

require one unit of management each per semester. The new activities, such

as asparagus and grapes, require more management resources estimated at

five units per semester per crop.






44



Table 4-1: Resources and constraints for linear programming, average household


Resources and Constraints Sign Unit Amount

Land I <= ha 4.72
Land II <= ha 4.72
Male labor I <= days 449.00
Male labor II <= days 224.00
Female labor I <= days 466.00
Female labor II <= days 233.00
Male hired I <= days 449.00
Male hired II <= days 224.00
Female hired I <= days 464.00
Female hired II <= days 230.00
Water I <= m3 64,076.88
Water II <= m3 32,038.44
Management I <= unit 31.48
Management II <= unit 31.48
Credit for cotton and/or maize I <= Soles 9,700.58
Credit for maize II <= Soles 3152.69
Household cash I <= Soles 4,999.99
Household cash II <= Soles 3300.00
Livestock consumption I = unit 8.00
Livestock consumption II = unit 4.00
Maize consumption (house) I = Kg 543.11
Maize consumption (house) II = Kg 273.46
Maize consumption (livestock) I = Kg 960.89
Maize consumption (livestock) II = Kg 478.55
Sweet potato consumption (house) I = Kg 497.07
Sweet potato consumption (house) II = Kg 900.13
Sweet potato consumption (livestock) I = Kg 900.13
Sweet potato consumption (livestock) II = Kg 600.08

Note: I is first semester: August 15t to April 14th
II is second semester: April 15th to August 14th
Average of the 60 sample households. Specific information for each household can be
found in Appendix D.

The one-year model maximizes the cash available for discretionary spending at

the end of the year for each household after meeting all households needs. The six-year

model maximizes the sum of the end of the year cash for all six years. In both cases, the


maximization is after meeting the family needs.










Table 4-2: Crop activities and resource use for linear programming, per ha


Cotton Maize I Sweet
Potato I


Maize II Sweet
Potato I


Male labor I (days)
Male labor II (days)
Female labor I (days)
Female labor II (days)
Water I (m3)
Water II (m3)
Management I (unit)
Management II (unit)
Credit I (Soles)
Credit II (Soles)
Household cash I (Soles)
Household cash II (Soles)


8,000 9,500 6,5(


3,128 1,986


1,8L


-- 55
00 --
7,500
1 --
-- I1

1,500
18 --
-- 386


There are two non-elective activities in the models: 1) household activities -cash

for food, health, education, transportation, etc., and farm products -and 2) livestock (the

livestock activity is mainly for consumption).


Table 4-3: Resource use for house and livestock activities, average household


Resource


Male labor I (days)
Male labor II (days)
Female labor I (days)
Female labor II (days)
Water I (m3)
Water II (m3)
Management I (unit)
Management II (unit)
Household cash I (Soles)
Household cash II (Soles)


House


25
12.5
126
42
1,519.20
1,000.14
1
1
5,367.84
2,638.92


Livestock


40
20
64
32
1,000.00
500.00
1
1
800.00
400.00


60

50

5,000

1



1,534


Note: Average of the 60 sample households. Specific information for each household can
be found in Appendix D.


Resource









Table 4-4: Income from traditional activities, per ha


Activity


Unit


Yield


Below
Average
Price


Average
Price


Above
Average
Price


Cotton Quintal 60.83 68.00 103.50 121.50
Maize I Kg 5607.28 0.43 0.48 0.58
Sweet potato I Kg 19845 0.15 0.25 0.35
Maize II Kg 7300.14 0.45 0.85 0.90
Sweet Potato II Kg 24630.4 0.08 0.20 0.45
Rate of price change (average) -36.62 % +41.80 %


Asparagus and grapes are two introduced crops in the Cafiete Valley. They are

perceived as complex but profitable. Currently, the development agencies are

recommending these crops for the small farmers as alternatives to improve their

livelihood. Indeed, development agencies are financing these crops. The six-year model

was used to test the viability of these alternatives from the small farmers' perspectives. In

the case of the asparagus, the development agency requires that the small farmer be able

to plant at least one hectare due to harvesting and marketing concerns.


Table 4-5: Asparagus resource needs in six-year model, per ha

Years

Resource 1 2 3 4 5 6


Male labor I (days)
Male labor II (days)
Female labor I (days)
Female labor II (days)
Water I (m3)
Water II (m3)
Management I (unit)
Management II (unit)
Credit I (Soles)
Credit II (Soles)
Household cash I (Soles)
Household cash II (Soles)


15 15
15 15
2 4
2 4
3000 3000
1500 1500
5 5
5 5
2500 2500
1500 1500
500
500


45
45
15
15
3000
1500
5
5
2500
1500
500
500


30
30
12
12
3000
1500
5
5
2000
1500
500
500


30
30
12
12
3000
1500
5
5
500
500
500
500


30
30
12
12
3000
1500
5
5
2500
2000
500
500












Table 4-6: Grape resource needs in six-year model, per ha
Years

Resource 1 2 3 4 5 6

Male labor I (days) 20 15 30 30 30 30
Male labor II (days) 20 15 30 30 30 30
Female labor I (days) 5 5 30 30 40 40
Female labor II (days) 5 5 30 30 40 40
Water I (m3) 2500 2500 2500 2500 2500 2500
Water II (m3) 1500 1000 1000 1500 2500 1000
Management I (unit) 5 5 5 5 5 5
Management II (unit) 5 5 5 5 5 5
Credit I (Soles) 3000 3000 3000 2500 1000 2500
Credit II (Soles) 1500 2000 2000 2000 500 2500
Household cash I (Soles) -- 1000 1000 1000 500 500
Household cash II (Soles) -- 500 500 500 1000 1000


Table 4-7: Asparagus & grape income, per year per ha
Years

Perennial crop 1 2 3 4 5 6

Asparagus -- 12000 20000 26000 35000 48220
Grape -- 2000 10000 22000 28000 35000



Analyses were conducted from different perspectives. The researcher attempted to

explain the overall household system dynamics. After the aggregation of the sixty model

solutions, they were compared with the original data to understand "why" the simulation

selected some activities over others. Without losing the system diversity, there were some

naturally occurring household groupings. As suspected, family composition: number of

members, ages and gender were critical characteristics as well as the land resource. A

summary of those characteristics for the 60 households is shown in Table 4-8.







48


Table 4-8: Household composition and land for each farm

Members, gender, and age


N <5 5-14 15-65male 14-75female


>65male >75female


01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45


(Continued in next page)


Land


3.00
1.00
4.50
2.84
4.00
4.50
0.70
7.50
6.00
3.00
6.00
5.07
3.45
1.50
5.70
7.00
4.00
4.25
4.70
4.68
3.50
4.37
6.00
5.75
8.33
6.77
4.50
4.40
6.70
6.13
5.10
3.00
3.00
4.70
4.20
3.00
5.40
2.25
3.20
3.00
3.25
1.00
7.00
3.00
5.10











Table 4-8: Continued
Members, gender, and age Land

No <5 5-14 15-65male 14-75female >65male >75female (ha)

46 0 0 1 1 1 0 5.00
47 0 1 1 3 0 0 6.84
48 0 0 5 3 0 0 11.00
49 0 0 1 1 1 1 8.70
50 0 0 0 1 1 0 2.00
51 0 0 1 1 0 0 10.00
52 0 0 3 3 1 0 6.00
53 0 3 2 2 0 0 2.75
54 0 0 0 3 0 0 11.75
55 0 1 3 4 0 0 5.00
56 1 0 1 2 0 0 3.50
57 0 0 4 6 0 0 3.00
58 0 0 2 3 0 0 2.00
59 0 3 2 2 0 0 1.00
60 0 0 2 2 0 0 9.00




Linear Programming Model Validation


Before continuing with further analyses, the researcher examined the 60 one-year

models to determine the degree to which they represent the same population as the

sample of 60 households. The results of the 60 individual simulation models were

statistically compared with the data from the household survey. Three items were used

for this analysis: total land used in the second semester, land planted with maize and land

planted with sweet potato in the second semester because they were the most accurate

data from the survey.

Both an F-test and a t-test were used to test the models. The F-test compared the

data variances in order to determine if both sets of data had equal variances. The t-test

compared the means (with equal or unequal variances) in order to determine if significant









differences existed between the two sets of data: simulated and survey. Both tests were

based on probability level of 0.05.

The null hypothesis in each case was that variances and means were not equal:

F-test:
Var (solution data) = Var (real data)
Var (solution data) o Var (real data)
t-test:
Mean (solution data) = Mean (real data)
Mean (solution data) <> Mean (real data)

Table 4-9 shows the analysis for the total land used in each household in the

second semester. The null hypothesis for variance is rejected (F = 1.50, p = 0.062)

indicating that we can accept that the variances are not different. The t-test, assuming

equal variances, leads us to reject the null hypothesis and accept that the sample and the

results of the simulated models represent the same population. (t = 0.135, p = 0.893).


Table 4-9: Total land used (t-test assuming equal variances)

Land used n Mean Variance t p

Model simulation 60 3.68 3.71 0.135 0.893
Real data 60 3.73 5.55

(F = 1.500, p = 0.062)


Table 4-10 analyzes the maize land planted in the second semester. The F-test

indicated unequal variances (F = 2.967, p = 0.004). The t-test for unequal variances

indicated that both sets of data -simulation & real- have approximately equal means and

therefore the sample and the simulated models represent the same population. (t = 1.684,

p = 0.122).









Table 4-10: Maize land used (t-test assuming unequal variances)

Land used n Mean Variance t p

Model simulation 26 2.00 0.75 1.684 0.122
Real data 26 2.54 2.22

(F = 2.967, p = 0.004)


Table 4-11 shows results for sweet potato land planted in the second semester.

The F-test for the simulation and the real data indicated equal variances (F = 0.674, p

0.176). The t-test, assuming equal variances, revealed no significant differences in means

indicating that the sample and the simulated models represent the same population. (t = -

1.408, p = 0.166).





Table 4-11: Sweet potato land used (t-test assuming equal variances)

Land used n Mean Variance t p

Model simulation 24 2.65 2.46 -1.408 0.166
Real data 24 2.06 1.66

(F = 0.674, p = 0.176)


Based on this validation analysis, we conclude that the linear programming

models adequately simulate the population sampled: Cafiete's small farmers. These

models can be used with confidence to project production, income, and consumption in

different scenarios for any Cafiete's small farmer household.








Prices, Alternative Crops, and Discount Analyses Based on Validated Model


The following section presents analyses with the results of the solutions of the

different models (60 solutions -one for each household- in each scenario). The specific

numerical results can be reviewed in Appendix D.

One Year Model


A one-year model was run for three different scenarios in all the 60 households:

average, below average, and above average product prices. The average scenario model

represents the most current information, the normal or average prices in all crops and it is

the same as the simulation used in the validation. The below average and above average

scenarios estimate responses to external extreme price changes (Table 4-4).

Average Price Scenario

In this scenario, with the most common prices, cotton is produced by almost all

the households, in variable quantities. Seven households that do not produce any cotton

have land, labor, and cash constraints. In all households, more sweet potatoes than maize

are produced in the first semester, and more maize than sweet potato is produced in the

second semester. The maize in the first semester and the sweet potato in the second are

produced primarily to cover the family and livestock consumption needs.

Female labor is more limited in the second semester than in the first. In both

semesters, female labor is more binding than male labor. There is more male off-farm

labor than female and more in the first semester. All of the farmers in the model, with

exception of three households, rent out a portion of their land in both semesters. The

three households that chose not to rent out land (NO 42, 50, and 59) have very little land








(two ha or less) and plenty of labor available. One of them (No 50) has very low

management resources as well. Little land is rented in by the farmers in either semester.

Water is a constraint only in four households in the first semester. It is a constraint

in about 50% of households in the second semester. The management resource is a

constraint in twelve households.

Fifty households (83.33%) have credit constraints, mostly in the second semester.

The ten households without a credit constraint have no characteristics in common. Six of

them have plenty of labor resources; the other four have low labor resources but cash

available. Five of these households have small farms (two ha or less), the others have

medium or large farms (3 to 8.7 ha). Four households have management resources below

25 units; the others have plenty of this resource.

Cash at the end of the year varies greatly. There is one negative (S/. -2,823,

household No 7). This household is characterized by having very little land (0.7 hectares)

and large number of family members in the household (including many small children).

All other households have positive year end cash, with a maximum of S/. +57,013

(household No 60). The cash at the end of the year is highly related to factors such as

farm size and household demographics (size, age, and sex composition). All of the

households require a certain amount of credit for agricultural production. This is

generally in the form of credit for seed, fertilizer, or pesticides. However, nine

households borrow cash in addition, from retailers, intermediaries, and/or pesticides

shops.









Below Average Price Scenario

In difficult economic times (Table 4-4), farmers would tend to plant more

hectares of cotton (Figure 4-1), and subsequently fewer hectares of sweet potatoes and

maize. The sweet potato and maize production would be primarily for household

consumption.

In the below average scenario, farmers would be financially unable to hire labor,

and would more likely seek off-farm employment to subsidize cash income. In addition,

more land would be rented out, and less land would be rented in by these small farmers.

What little land that is rented, would be rented in the first semester only.





120.00


80.0co




20.00
20.o
Below Average Average Above Average

Figure 4-1: Total area planted in cotton for different price scenarios



In terms of water, it would be a constraint only in six of the households for the

first semester. It would not be a constraint in any of the households during the second

semester (Figure 4-2). Eighteen households would experience management constraints

during both semesters.









There would be three households with negative cash holdings at the end of the

year (Households No 7, 25, and 26). The maximum cash holding at the end of the year

would be S/. +39,971 (Figure 4-3). End of the year cash is related to farm size and

household demographics.

40
35
*30

20
15


o] ----. .-








Only four households use expensive cash credits. On average, the prices included

in this model are 36.62% less than the prices included in the previous model. The end of

the year cash in this model is 27.40% less than the end of the year cash in the average

model.

Above Average Price Scenario


The most optimistic prices were utilized in this model. Only about one-half of the

average hectares would be planted of cotton. Maize production in both semesters would

be only for household consumption. However, sweet potato would become the major

cash crop.










120000.00

100000.00

?. 80000.00

o 60000.00

>. 40000.00 -

S20000.00 -, it

0.00 i-m il
Below Average Average Above Average
-20000.00


Figure 4-3: Range of cash at the end of the year in different price scenarios



In this model, there would be more cash available to employ labor (Figure 4-4).

Consequently, there would be less income brought into the household from off-farm

labor. In an effort to maximize hectares, more land would be rented in (50% of the

households) than in the average model (Figure 4-5). The land expansion (by renting)

would be mostly in the second semester.



16000.00
14000.00
0
12000.00
10000.00-
.: 8000.00- 114
S6000.00 i
S4000.00 -:
< 2000.00- ,t:"TT.
4 ~i l
0.00 T-
Below Average Average Above Average

Figure 4-4: Hired labor in different price scenarios









Water constraints in this model would be very similar to the water constraints in

the average model. Management constraints were only slightly higher in this model. The

household with a negative cash income in the average model still would experience a

negative cash income in this model.

The maximum end of year cash would be S/. +110,869 (household NO 60). In this

scenario, the product prices are, on average, 41.80% higher than in the average model and

the end of year cash in all households is 52.46% higher than in the average model.


350.00
300.00
250.00
S200.00 -
150.00
100.00 Til.


0.00
Below Average Average Above Average

Figure 4-5: Land rented out in different price scenarios



An inspection of Table 4-12 reveals that households were differently affected by

prices changes. Some households were highly affected, while others were slightly

affected. According to Table 4-12, 26.67% of the households experience a high degree of

economic variability and 20.00% of the households experience a low degree of economic

variability. The other households (53.33%) experience a moderate economic variability in

the three price scenarios. In an attempt to assess potential risk or potential windfall for a

general price decrease or increase, the researcher calculated a land to person (number of

people in the household) index.










Table 4-12: Cash at the end of the year in different scenarios, economic variability

Economic n Below Average Above Range Land/Person Index*
Variability Average Price Average
Price Price

Low 12 13,382 16,237 25,856 12,474 0.57 ha/person
Moderate 32 17,673 24,355 46,546 28,873 1.08 ha/person
High 16 25,480 36,857 74,153 48,313 1.36 ha/person

*Land/Person index -land in hectares and number of people in household- is proposed to
explain the household economic variability characteristics in different price scenarios.


Households with more land per person experience more economic variability than

households with less land per person. (Figure 4-6).


60000
S50000
0 40000
0-I_^


1.5

1 X
-- r l m o


20000 0.5 C
C
0 10000
0 0
Low Moderate High
I End Year Range of Cash -l Land/Member Index

Figure 4-6: Different economic variability and land to person index.


Households with high land per person index (large area or few family members)

experience more potential risk or potential windfall in any enterprise than any of the other

groups. They would obtain very high incomes in "good" price years, but they would

obtain very low incomes in "bad" price years. While the incomes obtained in "good"

price years would be almost double than family needs, the incomes in "bad" years would

be not enough to cover household needs.


C








Households with a low land per person index (little area or many family

members) experience less potential risk or potential windfall. This last group of

households would obtain acceptable incomes in "good" price years and in "bad" price

years. These incomes would be enough for the household needs.

Six Year Model


The six-year model was run in the same three different scenarios: Average, below

average and above average cotton, maize, and sweet potato product prices. In the average

or normal scenario the two new alternative crops were included (grapes and asparagus) as

well as an analysis using discount rates.

Average Price Scenario

Seventy-five percent of the households do not produce cotton in this scenario. Of

the 15 households that do produce cotton, they added the crop in the sixth year, but

produce no cotton during the first five years. All of the households produce maize for

household consumption. However, during the final two years they also produce maize for

the market. Sweet potato production increases each year. Sweet potato is grown for both

household consumption and as a cash crop.

Every year in this six-year model, less land is rented out. Those that rent in land

for expansion tend to have more land to begin with and a higher degree of management

capability. Households with greater labor resources rent out less land. In terms of land

expansion, more land is rented in for production each year.

Eight households do not hire any labor during the six-year period. They have less

land to care for and more family members to care for their land. The end of the year cash

increases every year with one exception (household No 7).









Grape and asparagus activities analysis


In this average price scenario, no household is financially capable of investing in

grape production. However, 46 households would be able to raise some asparagus,

twenty-five of which could produce over one hectare (Table 4-13).


Table 4-13: Analysis of the asparagus activity regarding household characteristics

Grouping Members, Age, and Gender Land Zone* Management
1 2 3 4 ha or Education
No Asparagus 43% high
14 out of 60 0.50 0.79 1.71 1.64 4.35 43% medium 20.69
(13.33%) 14% low

< 1 ha Asparagus 48% high
21 out of60 0.19 0.67 2.24 2.14 4.11 38% medium 31.90
(35.00%) 14% low

>=1 ha Asparagus 20% high
25 out of 60 0.08 0.56 2.56 2.60 5.45 48% medium 38.19
(41.67%) 32% low

Solution for
"Average" 0.21 0.65 2.25 2.22 0.18 -------- 31.48
Household
0.84 ha Asparagus

Note: 1, males and females of less than five years.
2, males and females between five and fourteen years of age.
3, males between fourteen and sixty-five years of age.
4, females between fourteen and seventy-five years of age.
*The Cafiete Valley can be divided in three well-known zones: low, medium, and high, according to
elevation differences.


Those households that could devote a greater amount of land to asparagus

production are characterized as having fewer children living at home (Figure 4-7), and

having more available adult labor (Figure 4-8).










m 30 0.6
025 0.5
S20- 0.4 >
o 0
z 15 0.3 v


No Asparagus < ha Asparagus ha Asparagus
-10 0.2


z 0 1 1 0
No Asparagus < 1ha Asparagus >=1 ha Asparagus

SNumber of Households C -Children < 5 years old

Figure 4-7: Capability of planting asparagus and number of children in household



30 6

0 25-+ 5
E
S20 4 L
o d
I: 15- 3 3



0 0
O 0 0
No Asparagus < 1 ha Asparagus >=1 ha Asparagus
Number of Households -*.*Adult Males & Females

Figure 4-8: Capability of planting asparagus and number of adult members in household



In addition, the households that could devote a greater amount of land to

asparagus have larger farms to begin with (Figure 4-9). They tend to have more land in

the lower to middle valley range (probably the most productive lands in the valley). They

are also more highly educated (Figure 4-10).

Discount rate analysis


Three levels of discount rate were analyzed: low (50%), medium (100%), and

high (150%). These discount rates reflect the degree of stress (necessity to produce food









in the short run) that any household could perceive at any time. The higher the stress, the

higher the discount rate.







S40 6
S35
G 30
(A-
o 25 4
o0
I 20 3
4-'
o 15
--2
10
E 5 1
z 0 0
No Asparagus & < 1ha >=1 ha Asparagus
Asparagus
Number of Households *Number of Hectares per Household


Figure 4-9: Capability of planting asparagus and farm size




30 -45
-40 o
S 25
S35 3
20 30
W
025 2
15 0 0
20
10 15

5 5
0 0
No Asparagus < 1ha Asparagus >=1 ha Asparagus
Number of Households -$-Education

Figure 4-10: Capability of planting asparagus and household education



The total or general findings pointed out that no grapes would be produced at any

discount rate. At a relatively low 50% discount rate, asparagus would increase from 48.83









ha (no discount) to 60.39 ha. With a medium discount rate of 100%, asparagus area

planted would decrease back to 48.85 ha and at a high discount rate of 150%, asparagus

would decrease to only 14.41 ha planted. (Table 4-14, Figure 4-11).

Cotton and maize would increase in area planted at all discount rates. Cotton

would increase from 30.27 ha (with no discount) to 31.25 ha with 50% discount, while it

would increase to 32.45 ha at 100% and to 45.36 ha at a 150% rate. Maize would increase

from 46.20 ha to 47.30 ha, to 51.94 ha, and to 63.63 ha at the highest discount rate.

Sweet potato would vary less than the other crops. It would decrease from 507.37

ha to 491.82 ha, to 505.42 ha, and to 503.93 ha at the highest discount rate.

With a zero discount rate, 25 households would produce asparagus. With the

highest discount rate reflecting a strong preference for short term gain, only five

households would produce asparagus.


Table 4-14: Analysis of discount and crops
Rate of Discount

Crop (0%) (50%) (100%) (150%)

Asparagus
ha Planted 48.83 60.39 48.45 14.41
Households (%) 25 (42.67) 30 (50.00) 23 (38.33) 5 (8.33)
Cotton
ha Panted 30.27 31.25 34.45 45.36
Households (%) 15 (25) 17 (28.33) 16 (26.67) 21(35.00)
Maize
ha Planted 46.20 47.30 51.94 63.63
Household (%) 60 (100) 60 (100) 60 (100) 60 (100)
Sweet Potato
ha Planted 507.37 491.82 505.42 503.93
Household (%) 60 (100) 60 (100) 60 (100) 60 (100)











I--Asparagus -Cotton --Maize Sweet Potato

60.00%

4 40.00%

p o 00.00% --'--


.E
S8 -20.00%



S o. -40.00%
>o RL -40.00% ----------------
o
S-60.00%

-80.00% ,
50% 100% 150%
Discount Rate

Figure 4-11: Area of crops planted in different stress levels



The discount rates affect to all crops beginning the second year. One difference is

that the asparagus starts generating income in the second year (and income increases until

the sixth year). The others generate income from the first year, but the amount is the same

through the years. Therefore, the other crops (cotton, maize, and sweet potato) would

have a priority of resource use in the first year, when these are able to generate income.

However, the high profitability of the asparagus makes this crop competitive even at low

or medium discount rates.

Another difference is that the asparagus requires a set of fixed resources during

the six years, which would not be available to the other activities, and the other crops

only need resources in the season of production. This gives the household more flexibility

in resource use. For example, the cotton, although not very profitable, requires very small








amounts of resources. This explains the fact that this crop would increase in area planted

when the asparagus would decrease at medium or high rates of discount.

Finally, other differences among these crops are that the asparagus and the cotton

are cash crops and the maize and the sweet potato are mostly food crops, required for the

family and livestock consumption. This explains the fact that the priority of resource use

at medium or high discount rates would go to the sweet potato and maize. However, this

situation varies in each household according to composition, available resources, and

activities.

Below Average Price Scenario

In this scenario, cotton would be produced in the sixth year by 30 of the

households. This is compared with only 15 of the households producing cotton in the

average scenario. Maize that would be produced is only for household consumption. The

quantity of sweet potato produced would be drastically decreased in this model. Sweet

potato would be only produced for household consumption. Asparagus production would

increase from 48.83 ha in the average model to 86.92 in this model. Only two households

would not produce asparagus in this model (households N 7 & 5). Both households are

large in terms of family members, have limited hectares, and little available cash. Sixty-

five percent of the households would produce more than one ha of asparagus. No one in

this model would produce grapes.

In this scenario, the farmers would tend to rent more land out to maximize cash

coming into the household. Only one household would expand production by renting in

land (household No 42). This household has limited land, and more available labor,

management resources, and cash. In this model, only nine households would have the









resources to hire labor. Two households would have negative year end cash for each of

the six years model (households No 7 & 5).

Above Average Price Scenario

In this scenario, only 14 households would produce cotton in the sixth year,

compared to 15 in the average model. Maize would be produced only for household

consumption. More land would be used for sweet potato production following a similar

trend observed in the average scenario. Less asparagus would be produced in this

scenario. A total of 43.71 hectares would be devoted to asparagus production by the 60

households. Twenty-three households would produce more than one hectare of asparagus.

As in the other scenarios, grapes would not be produced.

More land would be rented in this scenario and less land would be rented out,

especially after the third year. Similarly, little difference would be observed in hired labor

and the increasing trend of greater amounts of end of year cash. The amount of end of

year cash would be greater in this model, as one would have anticipated.














CHAPTER 5
CONCLUSIONS AND RECOMMENDATIONS


Production Functions


1. The analysis of cotton production functions demonstrated enormous variability

among geographic zones in relation to yield and its response to fertilizers and

environmental factors. This fact points out the need to make fertilization

recommendations on an individual basis (geographic zone) using the production

functions, according to the anticipated environmental factors. It is also

recommended to complement the production function results with on-farm trials.

2. The production functions demonstrated that, contrary to common belief, higher

yields are not necessarily reached with higher amounts of fertilizers. Actual

recommended fertilizer amounts are mostly too high; they are probably based

upon trails conducted on the very best soils in good years. It is suggested to

recommend balanced amounts of fertilizers based on annual environmental

conditions (good, fair or poor year) for each specific zone using the appropriate

production functions. Lower rates could probably be recommended but should be

based on on-farm research. It is also recommended to generate more production

functions for other zones and other crops.

3. The production functions may become predictor tools. For example, a "poor year"

anticipated, would be a "good year" for Cerro Alegre and San Francisco if









fertilized adequately. This opportunity would be much better with an anticipated

better cotton price because of less Cafiete total production.

4. These equations may also become risk avoidance tools. For example, in Palo Isla

it would not be recommendable to raise cotton in a "poor year" or even in a "fair

year" because of the low yields expected. In QuilmanA zone, where the worst

results were reported, it is perhaps recommendable not to raise cotton, even in

"good years."

Linear Programming


1. Statistical comparisons of the Cafiete linear programming models with the real

data allowed validation of the models, indicating that the models adequately

simulate the population sampled. This statistical validation process is an

innovation used for first time in this investigation to the author's knowledge.

2. The diversity of the household systems of Cafiete community requires individual

approaches. Use of an "average household" or a "representative household" is not

appropriate in drawing conclusions for the whole community. The validated

Cafiete linear programming models can be used to project production, income and

consumption in any household in the Cafiete small farmer community,

maintaining the system diversity, after inputting them with household specific

information. Both, the one-year and the six-year models are "interactive working

models" and they can be used with confidence and in a quick and efficient manner

through Visual Basic embedded programming.

3. The linear programming models (one and six years) could be applied in two

different perspectives. First, the small farmer perspective, in which the linear








programming may model any particular household in any price and activity

scenario to better assess household possibilities and to generate individual

household recommendations. In addition, a community perspective, in which the

developing agency may aggregate groups of individual small farmer models to

project community responses in different scenarios.

4. Based in the linear programming simulations, household land to person index -

land in hectares and number of people in household- can explain the household

potential risk or potential windfall characteristics in different price scenarios.

Potential risk and potential windfall are understood as the variability in incomes

found in a household after simulating it in different price scenarios. Higher

indexes determined higher potential risk in "bad" price years as well as higher

potential windfall in "good" price years.

5. Grape is a crop recommended to all small farmers by the development agencies.

This activity was analyzed in all different scenarios of the six-year linear

programming model and it was demonstrated that no small farmer would be able

to raise grapes and consequently this activity should not be recommended.

6. Asparagus is also a crop highly recommended to all small farmers by the

development agencies. The simulation of the six-year models found out that,

according to scenarios, only a relatively small segment of the population would be

able to raise asparagus. The recommendation of this crop should be on an

individual basis, after solving the appropriate model in the appropriate scenario.

Aggregation with market purposes should also be based on the individual models

rather than the model taken from the survey "average farm."









Extension Programming


Introduction

The subsequent programming information is proposed as recommendation for

Valle Grande Rural Institute extension work. After the previous chapter analyses along

with extension analysis, the major needs of the Cafiete's small farmers were identified.

These needs were the baseline for programming.

Identifying Target Publics

According to its mission and philosophy, Valle Grande has identified its target

public as limited resource farmers, who have less than 12 ha of land. This includes 4,800

families. The average household consists of seven people. In total, the general target

public is about 33,600 people. These farmers own 80% of the land in the Valley (18,080

ha). Most of the family members work to support the agricultural production system.

Collaborative Identification. Assessment, and Analysis of Needs Specific to Target
Publics

Based upon data collected by the researcher (sondeo, survey), supplemental data

(Valle Grande Rural Institute records, records maintained by the city government of

Cafete, and Peru's Ministry of Agriculture), and the researcher's knowledge and

experience in the region, the following nine major extension programs are proposed:

1. Traditional Crop Management. Pest control, fertilizer application, weed control,

and pesticide applications of traditional crops.

2. Adoption of New or Improved Crops. New crops with some economic advantage,

when compared to traditional crops.









3. Credit and Land Ownership. Farmers must know how to obtain credit. In addition,

they need to understand the relationship between land credit and ownership.

4. Commercial Marketing. Farmers need to analyze marketing elements to make

wiser decisions in order to obtain higher incomes.

5. Farm Management. Farmers need to follow a sequential orderly production

process based upon decisions made from clear and accurate records. The farmers

must generate such data through record keeping and budgeting.

6. Legal Issues. Small farmers need to know about agricultural policies and tax law.

7. Farming Associations. Small farmers need to know how to organize themselves in

associations in order to undertake common goals such as the advantage of

purchasing products by scale, labor efficiency, diffusion of information and

optimization of financial resources. One critical point of the associations should

be to decrease crime related to the theft of crops and assets.

8. Healthy Diets. There is a need to modify the diets of small farmers to include

more farm-raised products in their diets.

9. Pesticide Use and Environmental Conservation. Farmers need to decrease their

dependence upon chemical pesticides and incorporate Integrated Pest

Management (IPM) practices.

Design and Implementation

In order to accomplish the program macrobjectives and to achieve behavioral

change it is necessary to create a general plan of work that consists of the following

activities: Lectures, workshops, farm trials, formal education, informal education, mass

media delivery, bulletins and brochures, seminars, contests, and educational fairs.








The planned program seeks the overall well being of the Cafiete community with

sustainable development reflecting individual behavior change. More specifically, the

outcomes will increase the quantity and quality of the production of the main crops,

better farm management practices, and successful introduction of new crops. Farmers

will obtain fair prices for their commodities in the market place. The farmers will be able

to make informed decisions about investments and returns. They will increase the use of

bank loans. The farmers will be able to fulfill rights and responsibilities in commercial

and tax aspects. The small farmers will be able to balance better their diet and they will

be aware of the current environmental concerns.

Through farmers associations, they will be able to better negotiate costs, prices,

obtain technical assistance, fight crime, and receive mutual benefits. They will be able to

build a sustainable community through the development of common values, norms, and

rules without affecting the freedom to express themselves.

Implementing the Planned Program

The specific planned programs are presented below. For each of them, there is a

descriptive title, the program purpose, the program justification, and the program

significance. Target audiences and specific objectives are also included.

Developing Plans of Action

Cafiete first major program: improvement of traditional crop management (cotton, sweet
potato, and maize).

This program will help Cafiete's farmers to manage adequately the traditional

crops. Annually, according the sondeo and survey, Cafiete produces 10,000 ha of cotton,

3,800 ha of sweet potato, and 3,600 ha of maize. About 80% of this production is by

small farmers who produce on average 60 cwt. of cotton, 15 ton of sweet potato, and 4









ton of maize per hectare. These yields could greatly be improved by the adoption of more

appropriate management practices.

The preferred situation would be higher in both quality and quantity of the

traditional crops yields that permit small farmers a fair return. The previous analysis of

cotton production function showed the need to make fertilization recommendations on an

individual basis (geographic zone) using the production functions, according to the

anticipated environmental factors to obtain higher yields. More production function

development is proposed to analyze yields in maize and sweet potato and to analyze other

zones in cotton.

The traditional crop activities should be recommended in a household after

analyzing this, in different scenarios, on an individual basis through linear programming;

it will allow optimum resource use in each household.

The diffusion of appropriate crop management practices will allow higher levels

of productivity.

The targeted audience includes around 960 growers who have, on average, less

than US$ 800 annual income. They represent about 5,000 ha. Eighty percent of them are

males and the 20% are females. These farmers are 40 years and older.

The small farmers will reach, in a four-year period, an average production for

cotton, 70 cwt., sweet potato, 20 ton, and maize, 5 ton. According to researcher

experience, the harvested quality of these three crops could be 10% better. The goal is to

provide growers with information about pesticide application, weed control, pest control,

and fertilizer application. The following activities are proposed:








a) Weekly radio course for crop management. It will be broadcast, for one hour,
every Sunday morning on South Star radio and it will include specific topics
in pest control, crop nutrition, weed control, etc.
b) Seminars of cotton management. Two seminars, one in August and the other
in February. The first seminar will deal with the planting, irrigation, first
stages of pest control, and particular fertilization practices according the
cotton production function. The second seminar will deal with Pectinophora
gossipiella (pink worm) control and harvesting recommendations.
c) Seminars of maize management. There will be two maize seminars, one in
summer (January), and the other in winter (June). The summer seminar will
refer exclusively to the maize variety Cargill 404 and the winter seminar will
refer to the maize variety Cargill 606. The seminars will include soil
preparation, planting, pest and disease control, and crop nutrition.
d) Workshops for weed control. Two workshops, one in October, the other in
March. The October workshop will emphasize weed identification and
mechanical and cultural weed controls. The March seminar will emphasize
chemical weed control.
e) Fertilization demonstration in cotton and sweet potato. Follow-up between
August and April in three different farms one in the upper zone, another in the
middle zone and the last in the lower zone of the Valley. The demonstrations
will be held the last Saturday of the month.
f) Course on plant nutrition (cotton, sweet potato, and maize). Sixty hours
(twenty per crop) in a course that will include lectures, visits, practices,
demonstrations, and laboratory. It will be held any time between May to July.

Cafiete second major program: introduce new crops (horticultural, fruits, and annual
crops).

This program will promote new crops to Cailete's farmers that will better suit the

farmer's environments. Cafiete's traditional crops, cotton, sweet potatoes, and maize may

not be the most economical crops for all the small farmers. Cotton is a secure but low

profit crop, which requires about nine months to be harvested. Cotton production

functions showed for example that in some zones this crop should not be recommended

because of very low anticipated yields. Maize and sweet potatoes are high-risk crops

because of pests and prices. The small farmers' income may be greatly improved by the


adoption of different crops.








The low resource farmers preferred situation would be to raise new crops in order

to obtain higher benefits, but as displayed by the linear programming analysis, such

decisions are on an individual basis. Some of the new crops might be asparagus, chilies,

broccoli, cabbage, squash, and pumpkin. New varieties of tomato, cotton, maize, and

potato should also be considered. Fruits crops such as seedless grapes, annona

cherimolia, lucuma, and avocado are also alternatives. Before recommending any of

these new crops, the extension office will analyze the multiple options in each household

linear programming model in order to find the best combination for each farmer's

conditions.

The new crops to be recommended should be first tested. Local research must be

conducted through on-farm trials. The diffusion of new crops will allow production of

more diverse commodities in the community that may result in higher profits for all

farmers because of less risk in decreasing prices. Decreases in prices are commonly

associated with over-production of traditional crops.

The targeted audience includes around 600 farmers who have, on average, more

than US$ 3,600 annual income. They represent about 4,500 ha. Seventy-five percent of

them are male and the 25% are female. These farmers are 55 years and older.

These small farmers will adopt at least one new crop on their farms within a four-

year period. The following activities are proposed:

a) Workshops for new crops. Three workshops -September, February, and May.
The first workshop will be about new vegetable crops, the second will deal
with annual crops, and the May workshop will be about fruit crops. Each
workshop will include guest speakers, lectures, classes, group discussions, and
on-farm visits.
b) Field days. Every two months, on the first Saturday of the month, groups of
farmers will visit neighboring Valleys (Chincha, Ica) and interview innovative
farmers about the introduction of new crops.









c) Newsletters. On a monthly basis, newsletters will be produced with specific
information of new crops suitable for the Valley. The newsletters will be
delivered directly to the target audience together with invitations to the other
activities.
d) Learning trip. Once a year, in the summer, one trip to the third region of Chile
will be organized. There should be at least twenty participants and two
facilitators. The round trip should take about one week. It will be video-
recorded and one week after the trip a meeting will be held with the target
audience to discuss the findings.
e) Individual assistance. The office of entrepreneurial development of Valle
Grande will analyze each household with the linear programming model
before starting any new crop program. This office will be established on a
permanent basis.

Cafiete third major program: credit administration

The overall program purpose is to increase farm credit for production activities.

According to the sondeo and survey, less than 70% of Cafiete growers receive formal

credit. In many cases the greatest barrier is the lack of property titles (15% of the farmers

do not have property title).

The preferred situation would be that 100% of the clientele obtain property titles

and at least 90% of the farmers receive credit from banks. One critical limited resource

found in the linear programming analysis was the lack of credit. In fact, the simulation

showed that eight out often small farmers have credit constraints. Credit will improve the

farmer's financial status, increase productivity, and generate more wealth to the whole

system.

The targeted audience will be small farmers without farm property title and

farmers who do not work with bank credit. This population represents about 3,000

hectares (600 farmers). Of this group, 65% are male and 35% are female, with an average

age of 48.








The major program objective (4 years) will be to assist three hundred small

farmers in obtaining property titles and the whole targeted audience in obtaining bank

loans, resulting in at least 90% of the targeted audience obtaining formal credit.

In order to undertake the program objectives the following activities are proposed:

a) Property title obtainment workshops. On a monthly basis, these workshops
will be held at the Valle Grande facilities. They will be programmed for the
second Tuesday afternoon of the month and will include both private and
governmental legal assistance.
b) Establish an Office of Property Title assistance. It will operate in Cafiete one
day a week (Fridays). An attorney will be hired to provide legal services for
that office.
c) Agricultural credit workshops. Two workshops per year at the start of each
crop season (August and April). They will be held in a banking institution and
conducted by bank employees. Valle Grande would be only a facilitator.
d) Credit application assistance. Small farmers will be provided assistance in
completing credit applications in the office of entrepreneurial development
(Valle Grande).

Cafiete fourth major program: marketing for Caiiete growers

Prices for traditional crops range from, cotton S/. 68 to 121.5 /qq., sweet potato

S/. 0.08 to 0.45/kg, and maize S/. 0.43 to 0.90/kg as analyzed in the linear programming,

based upon information obtained in the sondeo and survey.

According data gathered in the survey, acceptable prices perceived for the small

farmers should be, at least: S/. 90 for a qq of cotton, S/. 0.18 for a kg of sweet potato, and

S/. 0.55 for kg of maize to compensate the production costs and obtain reasonable

benefits.

In order to obtain fair prices for these products, the farmers must gain marketing

skills. Such skills include avoiding high risks by predicting prices, interpreting and

managing marketing information, and negotiating with intermediaries.








The targeted audience will be small farmers with an annual income less than US$

2,000. Growers with this income represent about 11,000 hectares (1,600 farmers). Sixty-

five percent is male and 35% female; the average age is 58 years.

The major program objective will be to increase annual income by 15%.

In order to undertake the program objectives the following activities are proposed:

a) Marketing short courses. They will be held four times a year (October,
January, April, and July) in the Valle Grande facilities. Each course will deal
with a specific topic. The first course will be on price and area planted
analysis; the second course will teach the different commercialization chains;
the third course will be about the farmer-intermediary negotiation options; and
the last course will include commercialization opportunities for the farmers.
b) Visit to marketplaces. Four visits to four markets will be coordinated. These
visits will take a day on the following months November, February, May, and
August. These visits will be to the horticultural market in Lima, the fruit
market in Lima, the market in Ica, and the market in Chincha. In each visit,
there will be two lecturers: the market chair and the intermediaries'
association president.

Cafiete fifth major program: apply farm management techniques

Data from the sondeo and survey revealed that less than 10% of low resource

farmers keep records on their daily activities; less than 5% of small farmers make

decisions based upon budgeting, and less than 1 % of them analyze their own data for

decision-making purposes.

Information is important. For example, more production functions (other crops or

cotton in other zones) could have been generated if more information were available. One

evaluation objective for this program is that at least 20% of the farmers keep accurate

records and make budgeting decisions based upon such records. This program will result

in improved participants, who will ultimately benefit the entire valley.

The targeted audience will be all small farmers in the Cafiete community. They

represent about 22,600 hectares (4,800 farmers). A four-year goal is that a thousand








farmers will keep records on their daily labor, and be able to effectively apply budgeting

in advance, and two hundred farmers will use management analysis on their farms.

In order to undertake the program objectives the following activities are proposed:

a) Administration workshops. These workshops will be held each Thursday
afternoon in the Valle Grande facilities with the participation of the extension and
entrepreneurial offices. The workshops will be similar in content and they will
emphasize the need to keep daily records for all household activities. The
workshops will include group work to apply the knowledge.
b) Publication of daily records. Valle Grande will help the small farmers by
publishing real samples of daily records of representative farmers. The
publications should include detailed listing of items to include and instructions of
how to manage the system.
c) Radio promotion. On the local radio station, South Star, there will be
advertisements three times a day to encourage the farmers to keep daily records.
These advertisements will emphasize the need to keep records and publicize Valle
Grande record keeping services.
d) Billboards. Three strategic billboards will be installed on major roadways. They
will show the benefits to keeping records.

Cafiete sixth major program: agricultural legal issues

The main purpose is to educate small farmers about government agricultural

policies and tax issues. The survey pointed out that only 5% of the small farmers are

familiar with essential agricultural policies, and less than 2% are registered taxpayers.

The farmers should understand essential agricultural policies and apply them in

their daily work. Small farmers should start the unavoidable process to register as

taxpayers. Currently all the farmers are encouraged to register as taxpayers by the

government, although they will be exempt from taxes. Farmers who fail to register are

not allowed access to certain markets.

The improved knowledge and skills in agricultural policy will permit better

decisions along with more adequate commercial transactions. Becoming aware about tax








topics will prepare farmers with tools and skills to deal with tax issues that will be critical

in the subsequent years.

The targeted audience will include all Cafiete small farmers (4800). In four years,

20% of small farmers will make business decisions based on current agricultural policy,

and 10% will be registered as taxpayers.

In order to undertake the program objectives the following activities are proposed:

a) Weekly newspaper stories about agricultural policy issues. They will be
delivered in the entertainment section of the most diffused newspapers in the
community (Day to Day Cafiete and Cafiete News). They will be multiple-part
stories of some fictitious household about its experiences with legal topics. In
addition to a legal focus, agricultural policies and tax topics will be discussed.
b) Daily radio program on legal topics. A program called "Smart Farmers" will
be produced in a daily basis in the South Star station. It will include
interviews with government representatives and it will be interactive with a
question and answer segment.
c) Biannual exhibitions at regional fairs. The two major festivities of the Cafiete
community, Cafete Foundation Celebration (August) and Cafiete Christmas
(December) will be used to display vast information about agricultural
policies, commercial agricultural laws, and tax issues. The fair exhibitions
will be coordinated with the official government agencies.
d) Handouts and brochures. These will include a summary of successes in the
topics of agricultural policies and tax issues.
e) Monthly newsletters. There will be a database of the small farmers. It will be
used to deliver the newsletters. The newsletters will include summary
information of the monthly handouts.

Cafiete seventh major program: organize farmers in strong associations

The purpose of this program is to promote farmers' associations of cooperation

that enhance the well being of their household activities. The survey data revealed that

fewer than seven percent of the small farmers participated in organizations or

associations of any type in the community. Consequently, the small farmers have very

little power in the community decision-making process. Small farmers do not have many

organizational groups, although some organizations that do exist are not successful.








Because of that, farmers have to deal individually when purchasing farm inputs or selling

farm products.

The establishment of strong community organizations will permit a series of

advantages:

1. Obtain low cost of production through scale economies.
2. Better production by mutual help among members.
3. Better prices in commercial transactions.
4. Reduce the amount of theft in the community.

With the establishment of organizations, farmers will obtain stronger positions to

influence government policies, deal with intermediaries, purchase commercial services,

and the like. The organizations will improve the small farmers' economies and therefore

improve the quality of living. In fact, the linear programming analysis found critical

binding factors that could be resolved through strong associations.

The targeted audience will be all the small farmers (4,800). The four-year

objective of this program is that at least 20% (960 small farmers) of the target audience

will participate in at least in one purposive organization.

Farmers will receive information on how to establish an organization. There will

be four organizations established to serve as models. Each organization will have at least

20 members. The first group will exist to promote the production of a certain crop. The

second organization will be established to prevent crop theft. The third association will

exist to bargain for lower production inputs. The last one will negotiate better prices for

common products on the market.

In order to undertake the program objectives the following activities are proposed:

a) Workshops about Community Organizations. There will be 12 workshops,
one per month. They will be held the first Wednesday morning of each month
in the Valle Grande facilities. The topics will include the different









organizations' policies, the successful organizations in agriculture, purposes
of the organizations, etc. The Extension office will organize and deliver those
meetings with some government help.
b) Short courses about community organizations. There will be two short
courses, one in February and the other in April. They will be conducted
specifically to establish new organizations. The motivated people of the
workshops will be invited to participate in these short courses.
c) Monthly demonstration results bulletin. A monthly bulletin will be published
containing exclusive information on current organizations.
d) Monthly farmers' newsletters and brochures. They will contain mostly
motivational aspects to promote the creation of new organizations.


Cafiete eighth major program: learn healthy and affordable nutrition practices

This program will teach Cafiete's farmers to practice healthy and affordable

nutrition practices. The survey revealed that, in general, the small farmers spend a large

part of their budget to purchase food. It is common to consume industrialized products

such as wheat derivatives, canned milk, and the like and not their own farm products.

These industrialized products are more expensive than farm-produced products. It is

common knowledge that they do not eat a balanced diet.

The preferred situation is a higher consumption of farm-produced products, such

as maize, sweet potatoes, potatoes, milk, etc. contributing to a more healthy diet.

The targeted audience includes the more deprived small farmers, around 960

growers who have, on average, less than US$ 800 annual income. They all represent

about 5,000 ha.

These small farmers will realize a 40% decrease in food household expenses, in a

four-year period.

In order to undertake the program objectives the following activities are proposed:

a) Daily advertisement in radio. It will promote Cafiete products. The radio will
be South Star and the campaign will emphasize in the benefits of the own









products. The general idea will be revalue the local products in the minds of
farmers.
b) House nutrition counseling. A team of five women, specifically hired for that
purpose, will deliver knowledge directly to the households. They will visit
four households a day and they will talk mostly with the household women.
They will explain and demonstrate how to prepare nutritious meals with local
products.
c) Handouts of recipes. The Extension office will publish on a monthly basis a
handout of balanced meals with local products. The recipes will be delivered
taking advantage of other correspondence of Valle Grande and its clientele to
five hundred small farmers each month.


Cafiete ninth major program: decrease pesticide use and increase environmental
awareness

According to the sondeo and the survey, small farmers control pests primarily by

chemical applications. There is a need to make the farmers more environmentally

sensitive. There is an overuse of pesticides. Many small farmers expressed a concern with

increasing pest resistance to chemical controls. About forty percent of small farmers seek

pesticide assistance from chemical retailers.

The preferred situation would be to decrease the dependence of chemical use and

adopt Integrated Pest Management (IPM) practices. It is also desirable to promote

environmental conservation not only in the pesticide rationale but also in many other

aspects such as water management, soil conservation, air pollution, and the like.

To increase the small farmers' awareness of environmental issues is critical in the

sustainable development of the whole community. The knowledge, attitude, and behavior

toward environmental protection will allow the small farmers to continue in business for

subsequent generations.

The targeted audience will be all the small farmers (4,800). A four-year goal is to

decrease pesticide use by 25%.






84


In order to undertake the program objectives the following activities are proposed:

a) Video presentations. On a monthly basis, Valle Grande will present a video of
ecological disasters. After the presentations, there will be a discussion and a
dramatization of the case applied to the Cafiete community. It will be expected
that two hundred small farmers will attend each presentation.
b) Television advertisements. They will be coordinated and funded by the
government through the Agricultural Ministry in their conservation campaign.
The advertisements will encourage the farmers to decrease pesticides use and
conserve resources.
c) Integrated pest management program. Valle Grande will create an agreement
with the La Molina Agrarian University and the Agricultural Ministry to
promote integrated pest management in the Valley. The program will include
courses, workshops, demonstrations, and direct counseling.














APPENDIX A
CAfNETE VIDEO

There is a short video showing images of the Cafiete community and it can be

downloaded by clicking.

video.avi, size: 5,828 KB














APPENDIX B
SURVEY QUESTIONNAIRE

The survey questionnaire is a Microsoft World 97 SR-1 document. It can be

downloaded by clicking.

questionnaire.doc, size: 84 KB




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs