• TABLE OF CONTENTS
HIDE
 Front Cover
 Title Page
 Copyright
 Table of Contents
 Abstract
 Acknowledgement
 List of Tables
 List of Figures
 Acronyms and abbreviations
 Introduction
 Methodology and classification
 The production system (farm...
 The maize-bean intercropping system...
 Economics of the maize-bean intercropping...
 The adoption of conservation...
 Summary and conclusions
 Reference
 Appendix A. Visual aids used in...
 Appendix B. Field prices and local...
 Appendix C. Sensititivity analysis...
 New papers from the Natural Resources...






Group Title: Paper - Natural Resources Group - 97-01
Title: The adoption of conservation tillage in a hillside maize production system in Motozintla, Chiapas
CITATION PDF VIEWER PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00077522/00001
 Material Information
Title: The adoption of conservation tillage in a hillside maize production system in Motozintla, Chiapas
Series Title: NRG paper
Physical Description: vii, 51 p. : ill. (some col.) ; 28 cm.
Language: English
Creator: Erenstein, Olaf
Cadena Iñiguez, Pedro
Publisher: CIMMYT
Place of Publication: Mexico D.F
Publication Date: 1997
 Subjects
Subject: Conservation tillage -- Mexico -- Motozintla de Mendoza   ( lcsh )
Soil conservation -- Mexico -- Motozintla de Mendoza   ( lcsh )
Crop residue management -- Mexico -- Motozintla de Mendoza   ( lcsh )
Corn -- Mexico -- Motozintla de Mendoza   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Bibliography: Includes bibliographical references (p. 45).
Statement of Responsibility: Olaf Erenstein and Pedro Cadena Iñiguez.
General Note: "INIFAP-CIMMYT Collaborative Project on Natural Resource Management."
Funding: Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
 Record Information
Bibliographic ID: UF00077522
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 37789749
issn - 1405-2830 ;

Downloads

This item has the following downloads:

nrgp97_01 ( PDF )


Table of Contents
    Front Cover
        Front cover
    Title Page
        Page i
    Copyright
        Page ii
    Table of Contents
        Page iii
    Abstract
        Page iv
    Acknowledgement
        Page iv
    List of Tables
        Page v
    List of Figures
        Page vi
    Acronyms and abbreviations
        Page vii
    Introduction
        Page 1
        Page 2
    Methodology and classification
        Page 3
        Page 4
        Page 5
    The production system (farm level)
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
    The maize-bean intercropping system (field level)
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
    Economics of the maize-bean intercropping system
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
    The adoption of conservation tillage
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
    Summary and conclusions
        Page 42
        Page 43
        Page 44
    Reference
        Page 45
    Appendix A. Visual aids used in the farmer survey
        Page 46
        Page 47
        Page 48
    Appendix B. Field prices and local units of measure
        Page 49
    Appendix C. Sensititivity analysis of farm budgets
        Page 50
    New papers from the Natural Resources Group
        Page 51
Full Text



II
CIMMYT
Sustainable Maize
and Wheat Systems
for the Poor


The Adoption of Conservation

Tillage in a Hillside Maize

Production System

in Motozintla, Chiapas

Olaf Erenstein and
Pedro Cadena Ifiiguez


Natural Resources Group


Paper 97-01









IIMY
CIMMYT


The Adoption of Conservation
Tillage in a Hillside Maize
Production System
in Motozintla, Chiapas

Olaf Erenstein and
Pedro Cadena Ifiiguez*



Natural Resources Group
Paper 97-01




INIFAP-CIMMYT Collaborative Project on Natural Resource Management


At the time this research was conducted, Olaf Erenstein was an associate expert with the CIMMYT Natural Resources
Group. Pedro Cadena Ifiiguez is a researcher with the Instituto Nacional de Investigaciones Forestales y
Agropecuarias (National Institute of Forestry, Agriculture, and Livestock Research), Mexico.





































CIMMYT is an internationally funded, nonprofit scientific research and training organization.
Headquartered in Mexico, the Center works with agricultural research institutions worldwide to improve
the productivity and sustainability of maize and wheat systems for poor farmers in developing countries. It
is one of 16 similar centers supported by the Consultative Group on International Agricultural Research
(CGIAR). The CGIAR comprises over 50 partner countries, international and regional organizations, and
private foundations. It is co-sponsored by the Food and Agriculture Organization (FAO) of the United
Nations, the International Bank for Reconstruction and Development (World Bank), the United Nations
Development Programme (UNDP), and the United Nations Environment Programme (UNEP).

Financial support for CIMMYT's research agenda currently comes from many sources, including the
governments of Australia, Austria, Belgium, Canada, China, Denmark, the European Union, the Ford
Foundation, France, Germany, India, the Inter-American Development Bank, Iran, Italy, Japan, the Kellogg
Foundation, the Republic of Korea, Mexico, the Netherlands, Norway, the OPEC Fund for International
Development, the Philippines, the Rockefeller Foundation, the Sasakawa Africa Association, Spain,
Switzerland, the United Kingdom, UNDP, the USA, and the World Bank

Further information on CIMMYT is available at: www.cimmyt.mx.

Responsibility for this publication rest solely with CIMMYT.

Printed in Mexico.

Correct citation: Erenstein, O., and P. Cadena. 1997. The Adoption of Conservation Tillage in a Hillside Maize
Production System in Motozintla, Chiapas. NRG Paper 97-01. Mexico, D.F.: CIMMYT.

ISSN: 1405-2830
AGROVOC descriptors: Mexico; Chiapas; conservation tillage; farming systems; cropping systems; maize;
Zea mays; Phaseolus vulgaris; intercropping; technology transfer; innovation adoption; economic viability;
profitability; sloping land; highlands.
AGRIS category codes: E14, E16
Dewey decimal classification: 338.162











Contents

Page

iv Abstract
iv Acknowledgments
v Tables
vi Figures
vii Acronyms and Abbreviations
1 Introduction
3 Methodology and Classification
3 The Study Area
4 Methodology
4 Classification of Adopters
6 The Production System (Farm Level)
6 The Land Resource and Its Use
7 Livestock Resources
8 Family Characteristics, Labor, and Outmigration
10 Links with the External Environment
11 The Maize-Bean Intercropping System (Field Level)
11 Field Characteristics
12 Crop Establishment
13 Seed
13 Fertilizer Use
14 Weed Control
15 Harvest
15 Yields
20 Residue Management
22 Economics of the Maize-Bean Intercropping System
22 Resources Dedicated to the Cropping System
22 Valuation of Production Factors
27 Budgets
31 The Adoption of Conservation Tillage
31 Adoption
32 Adoption over Time
35 Factors Influencing Adoption
40 Farmers' Opinions of Conservation Tillage
42 Summary and Conclusions
45 References


Appendices


46 Appendix A. Visual Aids Used in the Farmer Survey
49 Appendix B. Field Prices and Local Units of Measure
50 Appendix C. Sensitivity Analysis of Budgets










Abstract


Data from a 1994 survey of 82 farmers who grow maize on steep hillsides in Motozintla, Chiapas,
provided information on agricultural practices, including the adoption of conservation tillage
practices; the profitability of the local maize-bean intercropping system; and factors affecting
diffusion of conservation tillage practices. Adoption of conservation tillage appears promising:
farmers no longer burn crop residues but leave them in the field as mulch, and 66% of survey
farmers had adopted the no-tillage component of the technology. At present, however, only 29%
of farmers are true adopters of both components of conservation tillage. Farmers who adopt both
components obtain more favorable yields and farm budgets. Adopters of the mulch component of
the technology appear to be less exposed to production risks. Results of a multivariate logistic
model indicate that adoption of the mulch component can largely be explained by the slope of the
maize field, which affects access of livestock to the field for grazing on crop residues. Adoption of
the no-tillage component was explained by the availability of cash and farm size. Communal
livestock pressure had a significant effect on adoption of both components, as did the availability
of family labor. State agricultural policy also stimulated adoption, particularly the distribution of
incentives, in combination with the local law against burning. However, because farmers still use
local varieties, system productivity remains low. In addition to improving the productivity of the
system, the use of improved varieties could also increase the availability of residues for forage or
mulch.

Acknowledgments


The authors greatly appreciate the valuable comments of Gustavo Sain, Larry Harrington, Jorge
Bolafios, Rob Tripp, Walter L6pez Baez, Jeff White, and Jim McMillan; the editorial assistance of
Alma McNab, Adriana Maldonado, and Kelly Cassaday; and the work of CIMMYT's production
and design staff, headed by Miguel Mellado.










Abstract


Data from a 1994 survey of 82 farmers who grow maize on steep hillsides in Motozintla, Chiapas,
provided information on agricultural practices, including the adoption of conservation tillage
practices; the profitability of the local maize-bean intercropping system; and factors affecting
diffusion of conservation tillage practices. Adoption of conservation tillage appears promising:
farmers no longer burn crop residues but leave them in the field as mulch, and 66% of survey
farmers had adopted the no-tillage component of the technology. At present, however, only 29%
of farmers are true adopters of both components of conservation tillage. Farmers who adopt both
components obtain more favorable yields and farm budgets. Adopters of the mulch component of
the technology appear to be less exposed to production risks. Results of a multivariate logistic
model indicate that adoption of the mulch component can largely be explained by the slope of the
maize field, which affects access of livestock to the field for grazing on crop residues. Adoption of
the no-tillage component was explained by the availability of cash and farm size. Communal
livestock pressure had a significant effect on adoption of both components, as did the availability
of family labor. State agricultural policy also stimulated adoption, particularly the distribution of
incentives, in combination with the local law against burning. However, because farmers still use
local varieties, system productivity remains low. In addition to improving the productivity of the
system, the use of improved varieties could also increase the availability of residues for forage or
mulch.

Acknowledgments


The authors greatly appreciate the valuable comments of Gustavo Sain, Larry Harrington, Jorge
Bolafios, Rob Tripp, Walter L6pez Baez, Jeff White, and Jim McMillan; the editorial assistance of
Alma McNab, Adriana Maldonado, and Kelly Cassaday; and the work of CIMMYT's production
and design staff, headed by Miguel Mellado.










Tables

6 Table 1. Farm area (ha) by adopter group and land use, Motozintla, Chiapas
7 Table 2. Livestock indicators, Motozintla, Chiapas
9 Table 3. Family characteristics and family labor, Motozintla, Chiapas
11 Table 4. Crop sales immediately after harvest, Motozintla, Chiapas
12 Table 5. Characteristics of the field selected for the farmer survey, Motozintla, Chiapas
14 Table 6. Labor for fertilizer application, maize-bean intercropping system (days/ha),
Motozintla, Chiapas
15 Table 7. Labor for weeding, maize-bean intercropping system (days/ha), Motozintla, Chiapas
16 Table 8. Yields, maize-bean intercropping system, Motozintla, Chiapas
17 Table 9. Factors affecting yields in poor years, maize-bean intercropping system,
Motozintla, Chiapas
19 Table 10. Factors affecting maize yields in the maize-bean intercropping system, 1993
summer cycle, Motozintla, Chiapas
19 Table 11. Factors affecting bean yields in the maize-bean intercropping system, 1993
summer cycle, Motozintla, Chiapas
23 Table 12. Labor use (days/ha) by operation and type of adopter, maize-bean intercropping
system, Motozintla, Chiapas
28 Table 13. Budgets for the maize-bean intercropping system, 1993 summer cycle, Motozintla,
Chiapas
31 Table 14. Adoption matrix for conservation tillage in the manual hillside production
systems of Motozintla, Chiapas
36 Table 15. Dependent variable, logit analysis of adoption of conservation tillage practices in
Motozintla, Chiapas
36 Table 16. Independent variables in the logit analysis and their hypothetical effect s on adoption
of each component of the conservation tillage technology in Motozintla, Chiapas
38 Table 17. Factors affecting adoption of conservation tillage in Motozintla, Chiapas
(multivariate logistic model, normalized on nonadoption)
39 Table 18. Probabilities of adoption of conservation tillage for different groups of farmers,
Motozintla, Chiapas
49 Table B1. Field prices, Motozintla, Chiapas
50 Table C1. Sensitivity analysis of farm budgets, hillside maize-bean intercropping system,
Motozintla, Chiapas










Figures


3 Figure 1. Location of the Motozintla study zone, Chiapas, Mexico
5 Figure 2. Venn diagram of groups of adopters and nonadopters of conservation tillage
practices (not to scale).
18 Figure 3. Relationship between maize yield and amount of mulch.
21 Figure 4. Treatment of residues by farmers in Motozintla, Chiapas.
23 Figure 5. Factors and interactions affecting the likelihood that sufficient crop residues will
be conserved in accessible fields, Motozintla, Chiapas.
31 Figure 6. Relationship between relative erosion and soil cover.
33 Figure 7. Diffusion of fertilizers, herbicides, and the abandonment of burning, Motozintla,
Chiapas.
34 Figure 8. Diffusion of the practice of not burning residues, taking into account the 1985 law
against burning, Motozintla, Chiapas.
34 Figure 9. Incentives for adoption of conservation tillage, Motozintla, Chiapas.
40 Figure 10. Farmers' opinions of potential disadvantages of conservation tillage, Motozintla,
Chiapas.
41 Figure 11. Farmers' opinions of the potential advantages of conservation tillage, Motozintla,
Chiapas.











Acronyms and Abbreviations


AU
c.c.
CIMMYT

FIRCO
FOSOLPRO

INIFAP

masl
Mx$

n
na
ns
prob.
PL
PRONASOL
SAG

SAGAR

SARH

sd
SDRE


Animal unit
Correlation coefficient
Centro Internacional de Mejoramiento de Maiz y Trigo [International
Maize and Wheat Improvement Center]
Fideicomiso de Riesgo Compartido [Shared Risk Trust]
Fondos de Solidaridad para la Producci6n [Solidarity Funds for Farm
Production]
Institute Nacional de Investigaciones Forestales y Agropecuarias
[National Institute of Forestry, Agriculture, and Livestock Research]
Meters above sea level
Mexican pesos; average exchange rate for 1993 was M$ 1 = US$ 3.1
(source: International Monetary Fund)
Number of cases
Not available
not significant
Probability
Potential labor
Program Nacional de Solidaridad [National Solidarity Program]
Secretarfa de Agricultura y Ganaderfa [Ministry of Agriculture and
Livestock, State of Chiapas]
Secretarfa de Agricultura, Ganaderfa y Desarrollo Rural [Federal
Ministry of Agriculture, Livestock, and Rural Development]
Secretarfa de Agricultura y Recursos Hidrdulicos [Federal Ministry of
Agriculture and Water Resources; now SAGAR]
Standard deviation
Secretarfa de Desarrollo Rural y Ecologia [Ministry of Rural
Development and Ecology, State of Chiapas; now SAG]










The Adoption of Conservation Tillage
in a Hillside Maize Production System
in Motozintla, Chiapas

Olaf Erenstein and
Pedro Cadena Ifiiguez


Introduction

The degradation of the soil and water resources
dedicated to agriculture in the State of Chiapas,
Mexico, is especially serious in hillside areas.
The declining quality of these resources, caused
by increased erosion and diminished soil
fertility, is alarming, for it threatens the present
and future productivity of the region's
agricultural systems. The hillside areas are also
a strategic zone in Chiapas for capturing water.
Resource degradation in this zone has already
seriously affected the supply of potable water,
state infrastructure, and the potential for
hydroelectric energy. Continued degradation
of the resources devoted to agriculture will
reduce the well-being of people throughout the
region over the short and long term.

Soil degradation is particularly severe when
farmers sow annual crops on hillsides using
traditional production practices. Under these
conditions, annual crops offer little protection
to the highly erodable soils, which are more
exposed to the elements that cause erosion.
Perennial crops such as coffee offer greater
protection for the soil, because they form a
more complete and permanent canopy. But in
Chiapas, as in many other places, annual crops
such as maize and beans meet the greater part
of the local population's demand for food.
Resource-poor farmers are reluctant to stop
planting the crops that make them self-
sufficient in food production, even if other
crops may yield better returns or degrade the


soil less. If these farmers continue to plant
annual crops to meet their food needs, they will
have to attempt to reduce the resource
degradation that this practice implies or face
the consequences of a declining resource base.

Several analysts agree that one of the most
efficient ways to reduce the soil degradation
that results from producing annual crops on
steep slopes is to increase soil cover (Shaxson et
al. 1989; Lal 1989; Hudson 1995). An
appropriate cover protects the soil against the
direct impact of rainfall and improves the
infiltration of water, which in turn reduces
runoff and soil erosion. In conservation tillage,
farmers use the residues of the previous crop as
a protective mulch on the soil and practice
minimal cultivation of the soil to avoid
destroying the mulch. Conservation tillage
technology conserves water as well, which
reduces the risk of drought. An additional
advantage of the technology is that farmers
must stop burning crop residues to prepare
land for planting, thereby reducing the risk of
forest fires. At the beginning of the 1990s in
Chiapas, agricultural activities caused an
estimated 55% of forest fires (Sandoval
1994:55).

Various institutions in Chiapas have experience
in developing and promoting conservation
tillage practices. The Sierra Madre of Chiapas
was one of the first areas where this technology
was extended on a large scale: the first pilot
program to extend conservation tillage










practices was conducted in this mountainous
area in 1983. The number of programs
promoting the technology grew in subsequent
years, peaking between 1989 and 1993
(Cadena 1995:3-7). Among the institutions
that promoted conservation tillage were the
Secretarfa de Agricultura y Recursos
Hidrdulicos (SARH, the Federal Ministry of
Agriculture and Water Resources); the
Secretarfa de Desarrollo Rural y Ecologia
(SDRE, the Ministry of Rural Development
and Ecology, State of Chiapas); and the
Fideicomiso de Riesgo Compartido (FIRCO,
the Shared Risk Trust). By the end of 1992, a
state law that regulated agricultural burning
came into effect, which facilitated (or perhaps
forced) the adoption of conservation tillage.

The Sierra Madre of Chiapas is characterized
by steeply sloping hills, which prevent
mechanization of the maize-bean
intercropping system. Previously, farmers
prepared land and controlled weeds with a
hoe; before planting, farmers burned the
residues left over from communal grazing by
livestock. The promotion of conservation
tillage in these cropping systems encouraged
farmers not to burn residues and to replace
the manual weed control with herbicides.
Incentives such as herbicide sprayers and
credit were provided to reduce the cost of
switching from traditional practices to
conservation tillage, which required a sprayer
for applying herbicide and some means to
purchase equipment and inputs. Farmers
could obtain these items only on the
condition that they stop burning residues. At
the same time, the benefits of conserving
residues rather than burning them were
strongly emphasized to farmers.


Despite the great effort to promote conservation
tillage in the Sierra Madre of Chiapas and other
areas in the state, information on the adoption of
the technology is relatively scarce.' What is the
current level of adoption? What practices
changed with the adoption of conservation
tillage, and which remained nearly the same?
Which factors encouraged farmers to adopt the
technology, and which ones limited adoption?
This study seeks to answer these questions for
one specific study area in the Sierra Madre of
Chiapas.2

The specific objectives of the study were to:
* Describe the agricultural production systems,
giving special attention to the production
system characteristic of the study area;
* Identify which factors influenced the adoption
of conservation tillage in these systems; and
* Quantify those factors in economic terms from
the farmers' point of view.

This information can help develop useful
recommendations for policy makers and
research and extension managers interested in
assessing the potential advantages and
disadvantages for poor, semisubsistence farmers
to adopt conservation tillage in areas highly
subject to resource degradation.

In the next section of this paper, we describe the
study area, the methods used to conduct the
study, and the typology of the data. The third
section of the paper discusses the farm-level
production system, particularly cropping and
livestock production practices, labor use, and
links between the farm and the external
environment. The fourth section of the paper
discusses the maize-bean intercropping system
at the field level, and the fifth section focuses on


1 See van Nieuwkoop et al. (1994) for one of the few studies examining the adoption of conservation tillage in another
area of Chiapas.
2 Preliminary results of this study are presented in Cadena (1995).










the economics of maize-bean intercropping.
The sixth section gives closer attention to
factors influencing the adoption of different
components of the conservation tillage
technology. The last section consists of a
summary and conclusions.



Methodology and Classification

The Study Area
The study area is located in the easternmost
part of the Sierra Madre of Chiapas in the
municipality of Motozintla, near Mexico's
frontier with Guatemala (Figure 1). The
municipality encompasses 325 km2 of mostly
mountainous terrain and belongs to Rural
Development District No. 7 (Distrito de
Desarrollo Rural No. 7 Sierra). Despite its
mountainous topography, 90% of the area in
the municipality is dedicated to agriculture,
7% to natural pasture, and only 3% to forest.
Coffee, the main crop, occupies 52% of the
arable area, especially in relatively lower areas.
Maize occupied 19% of the arable area in the
spring-summer cycle3 of 1991; beans, 5%; and
perennial crops other than coffee, 7%. The
remaining 17% of the arable area was not
cultivated in the 1991 summer cycle. Of the
total area in the municipality, 59% was ejido
land,4 34% belonged to smallholders, and the
rest was communal or public land. The
municipality has 40 ejidos (INEGI 1994).

The study area was restricted to the ejidos of El
Carrizal and Tuixcum, which are located,
respectively, about 7 km directly northwest
and 5 km directly southeast of the city of
Motozintla. Motozintla, the main city in the


municipality, has about 50,000 persons and is
located at the bottom of a valley which lies at
about 1,300 masl. The elevation of the two
ejidos is about 2,000 masl (1,800-2,200 masl),
near mountains of about 3,000 masl. The
villages of both ejidos are located on the
relatively flat tops of the mountains. Both ejidos
are mainly agricultural, and farmers' fields are
located on hillsides with slopes ranging from
40% to 100%. Both ejidos can be reached by dirt
roads, although it takes about an hour to drive
to them from Motozintla.

The higher part (>2,000 masl) of the ejidos is
classified by K6ppen as C(m)(w): a humid
temperate climate with abundant rain from


Figure 1. Location of the Motozintla study
zone, Chiapas, Mexico.


3 The spring-summer cycle, which we refer to as the "summer cycle" or "summer season," is the principal growing
season for maize in the study area.
4 Ejido land is agricultural land whose distribution, use, and sale were previously heavily controlled by the Mexican
government. Ejido members had usufructuary rights to ejido land, but the government retained ownership. Recent
constitutional reforms conferred ownership of the land on the individual ejido members.


Mexico










May to November; an average temperature of
12-180C; and an average temperature in the
coolest month between 3C and 180C. The
lower part of the ejidos is classified by K6ppen
as A(C)m(w): a semihot, humid climate with
abundant rain from May to November; an
average temperature of 18-220C; and an
average temperature in the coolest month of
>180C. Because the zone is mountainous,
rainfall is quite variable, reaching an annual
total of 1,500-3,000 mm (SPP 1981). The
combination of low temperatures and high
humidity makes the climate quite favorable
for maize production. However, solar
radiation can be relatively low because of
cloudiness. The low temperature also reduces
the speed at which organic matter decomposes
in the soil (Bolafos, pers. comm.).

The ejido of El Carrizal was established in 1935,
whereas that of Tuixcum was founded in 1928.
El Carrizal is relatively larger than Tuixcum,
with a population of 1,639 persons belonging
to 302 families, who occupy 229 residences.
The ejido of Tuixcum has only 677 inhabitants
(142 families living in 129 residences)
(Cadena 1995).

It is important to note at the outset that this
case study was done in a small part of a highly
variable mountainous area and that
information generated by the study cannot
necessarily be extrapolated to other
communities in the Sierra Madre. For example,
the two ejidos are characterized by their steep
slopes and relative accessibility two
characteristics that may not be found together
in other communities of the Sierra.

Methodology
Motozintla was one of the pilot areas where
conservation tillage was promoted in the
Sierra Madre of Chiapas. El Carrizal was one


of the first ejidos to participate in the extension
campaign for conservation tillage, partly
because many of the ejidatarios produced maize
on steep slopes and because the ejido was
relatively accessible. Tuixcum was chosen for
this study to avoid limiting the study to just
one ejido. Tuixcum is agroecologically and
socioeconomically similar to El Carrizal.

The data presented in this paper were
gathered through a formal survey conducted
at the beginning of the 1994 summer cropping
cycle. The sample was drawn from a list of the
farmers belonging to each ejido; from a total of
443 farmers (El Carrizal, 318; Tuixcum, 125), 82
were selected for the survey (El Carrizal, 52;
Tuixcum, 30). The sample was stratified by
ejido and the average sampling fraction was
18.5%.

Farm-level as well as field-level data were
obtained for each of the sample farmers. The
field was selected on the basis of two criteria:
size (the largest) and cropping system (maize
intercropped with beans or, if the farmer did
not intercrop, monocropped maize). Most of
the survey questions focused on current
practices, although some retrospective data
were collected. Aside from requesting specific
information through the questionnaire,5 the
enumerators also used visual aids (Appendix
A) to estimate the slope of the selected fields
and soil cover after planting.

Classification of Adopters
"Conservation tillage" is a rubric given to a
wide number of agricultural practices, and
many definitions of the term exist, ranging
from extremely general to quite specific.
However, conservation tillage generally refers
to a reduction in tillage operations, combined
with the conservation of crop residues, to

5 See Cadena (1995) for the complete questionnaire.










preserve and/or improve soil.6 The hillside
production system used by farmers in this
study was classified as a conservation tillage
system if it met two criteria:

1. The reduced tillage criterion: Soil is tilled
only for sowing; land preparation and weed
control exclude any cultivation of the soil.
Given the steep slopes of the fields in the
study area, the criterion of reduced tilled was
limited to no-tillage only. Adoption of this
component of the conservation tillage
technology was determined based on land
preparation and weed control practices
reported by farmers for the 1993 summer
cycle in the selected field.

2. The soil cover (mulch) criterion: At least 30%
of the soil surface had to be covered by crop
residues immediately after sowing (this level
corresponds to 2 t/ha of maize stover used as
mulch; see Tripp and Barreto 1993).7 To
determine adoption of this component of the
technology, the enumerator estimated the
amount of mulch on the selected field at
planting in the 1994 summer season, using a
device developed by Tripp and Barreto (1993)
(see Appendix A). It is assumed that the
amount of mulch on the field at the start of
the 1993 summer cycle was similar.

If farmers are to benefit from conservation
tillage, they must meet both criteria, for the use
of one component of the technology without the
other can generate unexpected results. The
farmer can meet one or both criteria or neither,
so it is possible to distinguish four categories of
adopters: nonadopters, adopters of either of the


components by itself, and adopters of both
components (Figure 2). "Nonadopters" are
farmers who meet neither of the two
conservation tillage criteria, whereas "adopters
of both components" have met both criteria and
thus are considered to have completely adopted
conservation tillage technology. We also refer to
"adopters of the mulch component" (farmers
who have met the mulch criterion) and
"adopters of the no-tillage component" (farmers
who have met the no-tillage criterion).

The distinctions between these three categories
is are important for understanding the adoption
of conservation tillage technology. In this paper,
we will present information for the entire
sample of farmers and for the different
categories of adopters when the differences
between them are significant.


Adopters, N
mulch only /

/ / Adopters, \
Mulch and no-till ge

Adopters,
no-tillage only /

\ /
Nonadopters \ /



Figure 2. Venn diagram of groups of
adopters and nonadopters of conservation
tillage practices (not to scale).


6 In simplified terms, the soil improves when organic matter is added at a greater rate than it decomposes (Bolafos,
pers. comm.).
7 The 30% threshold was developed initially in the United States for mechanized production systems (CTIC 1994).
Some researchers contend that in tropical and/or hillside production systems a higher threshold must be considered,
although several sources suggest that the response in terms of reduced erosion is similar in these environments (see,
for example, Shaxson et al. 1989). Even so, a higher degree of soil cover would imply less soil degradation.










The Production
System (Farm Level)


The Land Resource and Its Use
The average farm size among sample farmers
was 2.8 ha. Of this area, only 2.2 ha were
cultivated during the 1993 summer cycle
(Table 1) and 0.6 ha were left fallow. Maize
occupied nearly all of the cultivated area and
was usually intercropped with beans (an
average of 2 ha); occasionally maize was
monocropped (0.2 ha on average; 27% of the
sample). On the remaining 0.05 ha of
cultivated area, beans were solecropped (by
13% of the sample) or the land was planted to
other crops such as cabbage, potatoes, and faba
beans (13% of sample). Farmers have grown
virtually these same crops over the past
decade. Some producers also have a few fruit
trees, such as peaches, apples, citrus, or plums,
in their fields.

The different groups of adopters differ
significantly in farm size and cultivated area.
Adopters of both components of the
conservation tillage technology (that is,
complete adopters of conservation tillage) have
significantly larger farms and cultivated area
than nonadopters and those who only use


mulch. Farmers who have adopted only the no-
tillage component of the technology lie between
the two extremes, although the size of their
farms and cultivated area is still significantly
greater than that of the nonadopters. These
differences appear to indicate that adoption of
the no-tillage component (alone or in
conjunction with the mulch component) is
related to the size of a farmer's cultivated area.
This conclusion seems reasonable if we consider
that the no-tillage component saves labor.

Most farmers (72%) have the same amount of
cultivated land as they did ten years ago (data
obtained through retrospective questioning).
One-fourth (24%) of the sample farmers
increased their cultivated area in relation to the
availability of labor and herbicides. Only 4%
had reduced their cultivated area because they
transferred part of their land to someone else.

Fallowing continues to be a common practice:
55% of the sample farmers leave part of their
land fallow for an entire year, so that on
average one-fifth (21%) of the farm area was in
fallow in the 1993 summer season.
Nevertheless, fallow fields also serve as
pastures for livestock during the cropping
cycle.8 The 51 farmers who fallowed (or had


Table 1. Farm area (ha) by adopter group and land use, Motozintla, Chiapas
Total sample Type of adopter
Average Standard Non- Mulch No-tillage Both
(ha) deviation adopter only only components Probability
Total area 2.76 1.26 2.11 a 2.30 ab 2.85 bc 3.34 c .01
Cultivated area, 1993 summer cycle 2.18 + 0.90 1.68 a 1.81 ab 2.10 b 2.82 c .00
Maize (solecropped or
intercroppped with beans) 2.13 + 0.89 1.62 a 1.81 ab 2.04 b 2.79 c .00
Other crops (including
bean solecrop) 0.05 + 0.10 ns
Fallow area, 1993 summer cycle 0.59 + 0.74 ns
Note: Figures followed by different letters are significantly different (Duncan 0.1, row comparison); ns = not significant.

8 After the cropping season, all area dedicated to producing annual crops is left fallow and serves as animal pasture
until the subsequent cropping cycle.










done so in the past) cited the following reasons
for doing so:
* recovery of the soil (37%);
* the need to pasture livestock during the
cropping season (27%);
* lack of labor or herbicide (24%); and
* prevention of soil diseases (12%).

Fallowing appears to have been slightly more
common ten years ago. It is interesting to note
that most farmers who reported that they had
completely abandoned fallowing in the
previous decade were short of labor and had
adopted the no-tillage component.

Livestock Resources
Most farmers (87%) have at least one head of
cattle, while the average size of the livestock
herd is 5.2 head per farm. The herd is
composed chiefly of sheep and horses, with a
few goats, pigs, and cows (Table 2). Although
on average half of the herd consists of sheep,
only 29% of the producers have sheep.
Ownership of horses is relatively more
common (81% of the survey farmers have at
least one horse), mostly because horses are


used as pack animals in the ejido, especially to
carry inputs to and remove the harvest from
- relatively inaccessible fields.

During the past decade, the livestock herd has
decreased by an average of 2.0 head per farm
(prob.: .04). This reduction has probably
occurred because of the increasing human
population, the limited land area, and the use
of land to produce food for home consumption.
Shrinking livestock herds are the result of two
opposing forces: on the one hand, the numbers
of sheep and cows have declined; on the other,
the numbers of horses and pigs have grown.
Fluctuations in livestock numbers per farm
were different among the groups of adopters.
Most reductions occurred among adopters of
both components of the technology, whereas
livestock herds increased mostly for farmers
who did not adopt the mulch component a
result that one would expect.

To obtain an aggregate number of livestock for
each farm, the herd was converted into animal
units (AU) (adapted from Gittinger 1982).9 On
average, each farm had 2.6 AU, resulting in a


Table 2. Livestock indicators, Motozintla, Chiapas
Total sample Type of adopter
Standard Non- Mulch No-tillage Both
Average deviation adopter only only components Probability
Herd composition (head/farm) 5.15 6.12 ns
Sheep 2.59 + 5.25 ns
Horses 1.72 1.15 1.38a 1.86 ab 1.57a 2.17b .10
Goats 0.30 + 2.76 ns
Pigs 0.30 + 1.12 ns
Cattle 0.23 + 0.84 ns
Livestock pressure/farm
(AU/cultivated ha) 1.27 1.11 1.47 1.28 1.30 1.06 ns
Communal livestock pressure
(AU in ejido/cultivated area in ejido) 1.13 + 0.45 1.26 a 1.16a 1.26 a 0.85 b .00
Note: Figures followed by different letters are significantly different (Duncan 0.1, row comparison); ns = not significant;
AU = animal units.


9 Where the number of animal units (AU) per farm is calculated as:
AU = (# cattle) + 0.17 (# sheep + # goats) + (# horses) + 0.5 (# pigs).










pressure of 1.3 AU per cultivated hectare or
1.0 AU per hectare of farm area. The AU per
cultivated hectare is the best indicator of
livestock numbers if one considers that the
fallow area usually serves as pasture during
the cropping season. By the end of the
summer cropping cycle, little biomass is left
in these fallow fields. During the dry season,
animals graze in fields that were cultivated in
the cropping season and thus retain a
substantial amount of crop residues and
weeds. Furthermore, in the study area there
are few alternative sources of forage to
substitute for maize residues during the dry
season. Although there are no significant
differences in livestock ownership among the
categories of adopters, there is a tendency for
livestock pressure to be lower among
adopters of the mulch component (alone or
with the no-tillage component; see Table 2).

Nevertheless, communal grazing during the
dry season continues to be common in both
ejidos, which means livestock other than those
owned by the farmer may graze on the crop
residues in his fields. To take this into
account, communal livestock grazing
pressure was used as an additional indicator
of livestock pressure. This indicator was
calculated based on livestock pressure for the
ejido as a whole, corrected for the fencing of
fields (free grazing is limited when parcels are
well enclosed).10 Calculated in this way,
communal livestock pressure tends to be less
for adopters of the mulch component than for


those who did not adopt mulch, although the
difference is only significant among those
who adopted both components of the
technology (Table 2).

Family Characteristics,
Labor, and Outmigration
Average family size11 is 5.9 persons,
composed of 2.3 adult men, 1.9 adult women,
and 1.6 children (Table 3). In general, men
perform most of the agricultural work.
Women help with the field work in 60% of
the cases, while children help occasionally.
Women participate mostly in sowing,
fertilization, and harvest operations (+90%)
and less (35%) in weed control operations.
Women's contribution to chemical weed
control is usually limited to carrying water.
In most instances (70%), women's
participation in agricultural activities has
remained relatively unchanged over the past
ten years. The remaining 30% of farmers
reported that women's involvement had
generally declined. Changes in women's
participation were related to changes in
family structure (for example, child care
responsibilities, growth of children and their
participation in agriculture) and in no case
were they related directly to changes in
agricultural production.

The head of the family is generally male,
literate, 44 years old on average, and has
farmed for 30 years. Farmers who adopted
the mulch component of the technology


10 Communal livestock pressure for the ejido is calculated as follows:
Pressure (AU)f,
I (Area)f,
(I forf 1,...,n in ejido e),
where AUf is the number of animal units for farmf and Areaf is the total area of farm
Communal livestock pressure for the farm is calculated as:
Pressure = Pressuree Factorf,
where Factorf is the correction for fencing on farm f(1 not fenced; 0 = fenced; and 0.5 = partially fenced).
11 A family is defined as the group of persons living on the farm and depending on it for their livelihood.











(alone or together with the no-tillage
component) usually have much more farming
experience (an average of ten years) and are
older (by an average of five years) than
nonadopters.

The amount of family labor available on the
farm is estimated based on the composition of
the family and the participation of the family
members.12 On average, there are 3.4 units of
potential labor (PL) available per farm (1.8 PL
per cultivated hectare or 1.5 PL per hectare per
farm). Labor availability per hectare of
cultivated area tends to be less among adopters
of the no-tillage component, although
differences in labor availability are significant
only among nonadopters and adopters of both
components of the technology (Table 3).

Half (51%) of the survey farmers reported that
they hired labor to complement family labor.
About 15% of the labor dedicated to maize-


bean intercropping each season is hired.
Hired labor (expressed as a percentage of
total labor) is negatively correlated to the
availability of family labor per unit of
cultivated area (c.c.: -.32; prob.: .00). Adopters
of the no-tillage component (alone or in
conjunction with the mulch component) hire
in relatively more labor (about 20% of the
total) than nonadopters (less than 7% of the
total; Table 3). It is probable that the
diminished availability of labor, along with
the need for hiring in additional labor, have
encouraged farmers to adopt the no-tillage
component, because this implies a reduction
in the labor needed to produce the maize and
bean crops.

Outmigration of at least one family member is
quite common (79%) in the study area, and
outmigrants generally provide economic
support to family members remaining on the
farm. Forty-one percent of sample farmers


Table 3. Family characteristics and family labor, Motozintla, Chiapas

Total sample Type of adopter

Standard Non- Mulch No-tillage Both
Average deviation adopter only only components Probability

Family size (per farm) 5.85 1.79 5.29 5.57 5.90 6.38 ns
Men 2.30 1.13 1.90a 2.14 ab 2.17a 2.88b .02
Women 1.91 1.01 ns
Children 1.62 1.41 ns
Head of household:
Years as farmer 30.0 12.0 25.3 a 37.4 b 27.4 a 35.0 b .01
Age 44.4 10.9 40.7 a 49.4 bc 42.9 ab 48.2 c .06
Literacy (% cases) 96.3 na 100 85.7 93.3 100 na
Availability of family labor
(potential labor/cultivated ha) 1.80 + 0.92 2.16a 1.97ab 1.78ab 1.47b .09
Use of family labor (% of total) 85.3 + 20.4 94.2 a 93.3 ab 82.1 b 79.3 b .05


Note: Figures followed by different letters are significantly different (Duncan 0.1,
na = not applicable.


row comparison); ns = not significant;


12 Where potential labor (PL) per farm is calculated as:
(PL)= (# men) + (# women 0.5 factor) + (# children 0.3).
The factor corrects for women's assistance with field work (1 yes, helped with field work; 0 = did not help).
Weights (0.5 for women, 0.3 for children) were applied to account for potential participation in field work, based on
other activities and characteristics of women (cooking, caring for children) and children (school, age).










reported that it was the head of the household/
farmer who left, whereas for 47% of the sample
it was another family member who left (alone or
with the household head). There is a marked
difference in the areas to which these two
groups migrate in search of work. Most heads
of households (73% of those who leave the
farm) remain relatively nearby, working along
the coast of Chiapas. Most (82%) of the other
family members who leave in search of work
travel relatively far away (to Mexico City, the
United States, or port cities outside of Chiapas);
only 8% work somewhere along the Chiapas
coast. This difference is related to the form of
outmigration, which generally is seasonal in the
case of the head of household and more
extended in the case of other family members
who emigrate. Heads of households are
normally responsible for establishing the crop
in the field (including fertilization and weed
control), and afterwards they are free from field
work to a certain extent until harvest time. It is
during this period that many farmers leave for
off-farm work, especially to pick coffee in areas
along the Chiapas coast.

Outmigration by household heads appears to be
related to the adoption of the mulch component
but not the no-tillage component of the
technology. Fewer farmers who adopted the
mulch component (24%) emigrated (prob.: .05)
compared to nonadopters (46%). This is
understandable if one recalls that the head of
the household has primary responsibility for
agricultural activities and that his absence can
limit the control of communal grazing during
the dry season. On the other hand, the financial
resources obtained through outmigration can
enable a family to acquire the herbicides needed
to adopt the no-tillage component of the
conservation tillage technology. However,
outmigration is so common that it is not a
sufficiently distinguishing characteristic among
groups of adopters.


Migration is strongly related to the availability
of labor. Families whose members do not
emigrate have less available labor (2.6 PL per
farm) than those whose members leave (3.6 PL
per farm; prob.: .00). The type of migration is
also related to the availability of labor. A family
in which only the household head emigrates
has less available labor (3.2 PL per farm) than a
family in which another family member
emigrates (3.9) or in which several members
emigrate (4.0).

Links with the External Environment
Some 13% of the survey farmers were visited
by an agricultural extension agent before or
during the 1993 summer cycle. Most of these
visits were made by agents of SARH in the
1993 summer cycle. Nonadopters were visited
less (5%) than adopters of just one component
(13-14%) or both components (21%) of the
technology.

Nine percent of the survey farmers are
members of the Uni6n de Ejidos, and all of them
are adopters of the no-tillage component (alone
or in conjunction with the mulch component),
possibly because this organization, among
others, markets inputs. Almost 20% of the
producers occupied some post within the ejido
(for example, on the ejido council or the security
committee).

About 36% of the maize production is sold
immediately after harvest, while the rest is
used primarily for home consumption on the
farm, to feed livestock on the farm (especially
chickens, horses, and pigs), or occasionally
- sold later in the year. The groups of adopters
differ markedly in their maize marketing
practices. Adopters of both technology
components sell 53% of their production on
average, whereas nonadopters sell only 15%
(Table 4). Those who adopted only one
component of the technology are in the middle










of these two extremes, with sales fluctuating
around the general average. Differences in
maize marketing are related mainly to
cultivated maize area, which in turn determines
total production. Average consumption levels
are similar for the different groups of
adopters.13 The differences among adopters
suggest that the production of maize for sale
encourages the adoption of the components of
conservation tillage. This is not surprising,
given that the adoption of both components
generally requires a greater amount of cash (for
inputs for the no-tillage component, for
example, or for fencing materials for the mulch
component).

In the case of beans, about 31% of the
production is sold immediately after the
harvest, whereas the rest is chiefly used for
home consumption. Differences among groups
of adopters are similar to those for maize
production.


The Maize-Bean Intercropping
System (Field Level)

Field Characteristics
Maize is the most important food crop in the
study area, and the survey gathered detailed


information about farmers' management of the
maize crop. This information was obtained by
focusing on a single maize field belonging to
each survey farmer. The overwhelming
majority of sample farmers (96%) produce a
maize-bean intercrop in the selected field,
whereas the remainder grow a maize solecrop.
In general, the same crop had been sown on the
selected field during the previous cycle and
selected fields had been cultivated for more
than 48 years on average.

As noted earlier, these fields were located
mostly on steep slopes. The average slope was
71%, but it is interesting to observe the great
differences in the slopes of the fields cultivated
by different groups of adopters. Adopters of
the mulch component (alone or with the no-
tillage component) had fields with a slope of
about 85%. The slope of nonadopters' fields
was slightly less, at about 65% (Table 5), which
seems to indicate that adoption of the mulch
component is closely related to the slope of the
field. This relationship makes sense when one
considers that steep slopes make it difficult for
livestock to gain access to the land for grazing.
Adoption of mulch on the steepest fields may
also be related to the fact that conserving crop
residues on such fields can markedly reduce
soil erosion.


Table 4. Crop sales immediately after harvest, Motozintla, Chiapas

Total sample Type of adopter

Standard Non- Mulch No-tillage Both
Average deviation adopter only only components Probability

Maize sales (% of production) 36.0 + 28.7 14.7 a 35.1 bc 37.5b 53.2c .00
Bean sales (% of production) 31.4 + 29.5 9.1 a 21.4 ab 35.0 b 49.2 c .00

Note: Figures followed by different letters are significantly different (Duncan 0.1, row comparison).


13 On average, 3-4 t of maize are stored per farm. This implies a hypothetical per capital consumption of 1.5-1.9 kg per
day (hypothetical because it includes consumption by animals and sporadic sales). This figure corresponds to similar
calculations of home consumption of 1.5 kg per capital per day for the State of Chiapas (L6pez Baez, pers. comm.).










Crop Establishment
Virtually all of the farmers surveyed (99%)
prepared their land and sowed maize by
hand, using a hoe. In general, these activities
are combined in a form of reduced tillage,
known locally as sowing "cajeteada" (a cajete is
a shallow clay bowl). A small area, about 20 x
20 cm (a cajete), is cleared and 3-5 seeds are
sown in the middle of the area. These cajetes
are prepared in such a way that each mound
forms a small terrace. In clearing the cajete, all
vegetation is removed, along with the
residues of the previous crop. This ensures
that the germinating plants are practically
free of weeds and facilitates the capture of
water. Land preparation and sowing require
an average of 24 days/ha.

Given that producers already practice some
form of reduced tillage, no significant
differences exist in the number of days that
each group of adopters needs for land
preparation and sowing. Because all farmers
have stopped burning crop residues, all of
them can be considered potential adopters of
conservation tillage, although many do not
leave enough residues on the field to create
an effective mulch. On the other hand,
despite the fact that farmers practice a form of


reduced tillage, a considerable number
continue to till the soil during the crop cycle. In
most cases (82%), however, land preparation
and sowing have not changed over the past
decade, with the exception that farmers have
stopped burning crop residues.

More than ten years ago, farmers used other
land preparation practices, some of them in
combination, such as:
* Clearing ("slashing") with a machete: This
practice was used to clear land, particularly
land that had been left fallow and become
quite overgrown.
* Burning: This practice, once an important
component of land preparation, included the
burning of crop residues from the previous
harvest or of bushy fallow regrowth.
Depending on the amount of biomass in the
fallow field, the cleared residues and
vegetation were sometimes piled up to make
them easier to burn.
* Scraping with a hoe: The surface of the
entire field was cleared by scraping
("shaving") it with a hoe to leave it barren of
vegetation.
* "Plowing" with a hoe: In this intensive land
preparation practice, a hoe was used to turn
over all of the soil in the field.


Table 5. Characteristics of the field selected for the farmer survey, Motozintla, Chiapas

Total sample Type of adopter

Standard Non- Mulch No-tillage Both
Average deviation adopter only only components Probability

Area intercropped with beans,
1993 summer cycle (%) 94.2 + 20.9 ns
Field planted to maize-bean intercrop,
1992 summer cycle (% of cases) 96 na 90 100 100 96 na
Years field cultivated >47.8 12.9 >45.7 >44.1 >46.1 >52.9 .14
Average slope of selected field 71.3 + 23.1 61.4 a 87.1 b 64.6 a 83.6 b .00

Note: Figures followed by different letters are significantly different (Duncan 0.1, row comparison); ns = not significant
and na = not applicable.










Seed
Most farmers (85%) grow yellow maize; the
remainder grow white varieties. All of the
yellow maize varieties and most of the white
ones (92%) are local materials. The yellow
varieties generally originated in the region,
where they have been grown since before
anyone can remember. Apparently the yellow
varieties are better adapted to the very high
locations in the study area (>1,500 masl) and are
more resistant to diseases and pests, whereas
the white varieties are more common in the
lower locations in the study area.14 The absence
of improved varieties in farmers' fields results
partly from a lack of improved materials suited
to the zone, which has probably occurred for
several reasons. First, the local preference for
yellow materials is unusual in a country where
white materials predominate. Second, the
inherently high agroecological variability of this
mountainous zone, where elevation, rainfall,
temperature, soils, wind, and other factors can
vary widely, makes it quite difficult to adapt as
well as disseminate materials for each
agroecological niche. Indeed, maize
improvement in Mexico has traditionally
focused on the lowland tropics (<1,200 masl)
(L6pez Baez, pers. comm.).

In the higher parts of the study area, maize is
sown in May; in the lower parts, in June.
Farmers use about 16 kg/ha of seed. Generally,
beans are sown a month after the maize so that
the bean plants do not become entangled with
the maize seedlings. According to some
farmers, this delay also diminishes damage to
the beans from Diabrotica balteata (the banded
cucumber beetle), a chrysomelid beetle that
feeds on bean leaves but also on other species


such as maize. Unlike maize, beans are sown
with a dibbling stick. An average of 9 kg/ha of
seed is used. All of the bean varieties are local;
the most common one, sown by 74% of the
survey farmers, is called "negro de vara."

Fertilizer Use
Almost all farmers (98%) used chemical
fertilizers in the 1993 summer cycle, and 4%
used chemical fertilizers in combination with
organic ones. Chemical fertilization consists
mainly of nitrogen (N) applications farmers
applied an average of 79 kg/ha of N in the 1993
summer cycle. Phosphorus (P205) was applied
only by 12% of the farmers, and potassium
(K20) only by 11%, resulting in an average rate
of only 4 kg P205/ha and 3 kg K20. Adopters of
the no-tillage component applied more N on
average (87 kg/ha) than nonadopters (65 kg/ha;
prob. = .05). Similarly, the use of P205 and K20
was more common among adopters of the no-
tillage component (no-tillage alone or with
mulch). Recent research results from other areas
of Mesoamerica suggest that the levels of N used
by the farmers in the study area are probably
low for conservation tillage in relation to the
immobilization of N by the mulch (Zea et al.
1997).15

Most farmers (61%) applied fertilizer in two
doses, whereas 38% applied fertilizer only once.
Two applications were more common
(prob.= .08) among adopters of the no-tillage
component (alone or with the mulch
component) than among nonadopters. On
average, 58% of all N applied was applied in the
first dose. The most common source of N was
ammonium sulfate (21-0-0), followed by urea
(46-0-0). The chief source of P205 and K20 was
Triple 17 (17-17-17).


14 White maize is more commonly grown by adopters of the no-tillage component (either alone or with the mulch
component).
15 According to Zea et al. (1997), under low levels of N there is a negative interaction between the net effect of the mulch
and grain yield, with the equilibrium for N application at a level near the average in the study area.










Most farmers (62%) applied fertilizer by hand
at the base of the maize plant ("mateada"). The
rest of the farmers who applied fertilizer
incorporated it with a hoe to reduce the risk of
runoff. This practice has the additional
advantage of reducing volatilization of N, but
has the disadvantage of requiring about 50%
more labor than is needed to apply N at the
base of the maize plant (eight days) (Table 6).
Recent research elsewhere in Mesoamerica has
shown that incorporating N can raise maize
yields by 140 kg/ha over simply applying N at
the base of the plant, by increasing the N use
efficiency of the maize plant (Larios et al. 1997).
Assuming a similar response in the study area,
the practice of incorporating N would be
economic for survey farmers, providing a
marginal rate of return of 190%.16 No
relationship appears to exist between the two
kinds of fertilizer application and the different
groups of adopters. In any event, the number of
days needed to apply fertilizer is high, mainly
because it is difficult to access steeply sloping
fields.

A local alternative to chemical fertilizers is the
use of organic fertilizers, which are basically
composts of animal dung (particularly sheep
dung) and vegetative waste (such as stover and
dead leaves). About one-third of farmers (28%)
have prepared such fertilizers in the past,


although their use is not very common at
present. Farmers maintain that the principal
advantages of organic fertilizers are the savings
on chemical fertilizers (64%) and higher yields
(34%). Nevertheless, these advantages do not
seem to compensate for the main
disadvantages of organic fertilizers, which
include the considerable amount of labor
needed to prepare them (58%) and the limited
availability of compost in relation to cropped
area (40%).

Weed Control
Most survey farmers (96%) weed twice (either
chemically or manually) after the maize crop
emerges, and the remainder weed once. Only
one farmer applied a herbicide before
emergence. Most farmers (67%) relied on
herbicides alone, applied with a pump sprayer,
to control weeds; one-fourth (24%) of survey
farmers used both chemical and manual weed
control; and 9% weeded their maize only by
hand, using a hoe.

In general, the use of herbicide was limited to
paraquat (Gramoxone), which is one of the
most toxic herbicides but is popular because of
its cost (it is cheaper than other herbicides) as
well as its apparent efficiency (effects are
visible after application). The high use of
paraquat is a cause for concern when one


Table 6. Labor for fertilizer application, maize-bean intercropping system (days/ha),
Motozintla, Chiapas

First Second
application application Both
Avg. sd n Avg. sd n Avg. sd n

Fertilizer incorporated 11.2 4.8 30 12.1 4.7 18 24.2 +9.7 18
Fertilizer at base of plant 7.7 5.0 51 7.9 4.6 32 15.8 +9.3 32
Probability .00 .00 .00
Note: Avg. = average; sd = standard deviation; n = number of cases.


16 Based on a yield adjustment of 20% (CIMMYT 1988) and field prices (Appendix B).










considers the costs of health risks for the
farmer.17 Farmers generally took no additional
precautions when they applied herbicide, and
this can be quite problematic because it is
difficult to apply herbicide safely on steeply
sloping fields. On average, farmers applied
2.9 l/haf1.14; n = 73) of commercial product
in each application, and there was no
significant difference between the two
applications.

Chemical weed control provides a significant
labor savings, amounting to a 43% reduction
over the 22-23.5 days required to weed by hand
(Table 7). Just as with fertilizer application, the
time needed for weeding is high, primarily
owing to the steeply inclined fields. The need
to transport water for herbicide applications to
relatively inaccessible fields further
complicates chemical weed control.

Controlling weeds manually with a hoe is the
main factor that distinguishes adopters from
nonadopters of the no-tillage component. For
this reason, it is not surprising that the groups
of adopters differ significantly in the number
of days needed to perform the first weeding
(see Table 12). Nevertheless, this difference is


not important for the second weeding, given
that the use of herbicides for the second
weeding is common in all groups of adopters.
Ten years ago, all farmers controlled weeds
manually using a hoe, whereas 20% used a
machete as well as a hoe.

Harvest
Harvesting usually takes place in January or
February and the maize is not doubled over
beforehand. Farmers harvest by hand, taking
an average of 21 days/ha, which generally
includes transporting the harvest to the house.
There are no significant differences in harvest
practices among the groups of adopters. Most
farmers (63%) used animals to transport the
harvest, mostly their own animals (90%). One-
third of the survey farmers (34%) used a local
form of manual transport known as a
"mecapal."18 Only 2% of farmers transported
the harvest in a pickup truck.

Yields
Average yields19 in the 1993 summer season
were 2.67 t/ha for maize and 270 kg/ha for
beans. There is a nonsignificant tendency for
maize yields to be higher among adopters of
conservation tillage than for other groups of


Table 7. Labor for weeding, maize-bean intercropping system (days/ha), Motozintla, Chiapas

First Second
weeding weeding Both

Avg. sd n Avg. sd n Avg. sd n

Hoe only 23.5 4.9 27 22.0 6.8 6 44.2 a 11.2 5
Hoe and herbicide application 33.0 b +5.1 21
Herbicide only 10.0 4.1 55 9.4 3.5 73 19.6 c 7.9 53
Probability .00 .00 .00
Note: Avg. = average; sd = standard deviation; n = number of cases. Note: Figures followed by different letters are
significantly different (Duncan 0.1, column comparison).

17 Paraquat is immediately immobilized in the soil upon being absorbed by inert colloids (Tasistro 1989).
18 A mecapal is a band of cloth or tanned leather, which enables a person to carry a 75 kg bag of maize cobs on his or her
back. The band is used to support the bag of maize and passed around the wearer's forehead to counterbalance the
weight of the maize on the back.
19 From farmers' estimates of yield per field.










farmers (Table 8). To facilitate comparisons, we
obtained estimates of maize yields in "good,"
"normal," and "poor" years (respectively, 3.2,
2.3, and 1.5 t/ha on average). Significant
differences were found among groups of
adopters. Adopters of both components of the
conservation tillage technology obtained higher
yields than nonadopters of the mulch
component, regardless of whether the year was
a good, normal, or poor one for the maize crop.
Conservation tillage appears to offer a yield
advantage of 25% over the yields of
nonadopters, regardless of the type of crop
year. Farmers who adopt only the mulch
component obtain yields that are intermediate
between those of full adopters and
nonadopters, suggesting that it is the mulch
component that has the greatest influence on
maize yield levels.

Yields for good and poor years can also be
expressed in relation to yields in a normal year.
On average, the maize yield in a good year is
140% of the yield in a normal year, and only
66% in a poor year. The 1993 summer cycle was
thus a good year, with a relative yield of 120%.
According to the literature, farmers who adopt
conservation tillage practices can expect a


reduction in risk as well as a reduction in yield
variability. However, this potential reduction
in risk has primarily been found to be related to
the effect of moisture conservation, and
moisture is not much of a problem in the study
area. In the study area, risk seems to be mostly
related to the incidence of wind (see below).
Nevertheless, the relative yields obtained in the
1993 summer cycle by adopters of the mulch
component versus other groups of adopters
suggest that there has been a reduction in risk
for mulch adopters. In relative terms, the maize
yield in the 1993 summer cycle was 11-15%
better than the yield in a normal year for
adopters of the mulch component (alone or
with the no-tillage component), but 25-26%
higher relative to yields obtained by
nonadopters in a normal year. In absolute
terms, however, the yields obtained by
adopters of the mulch component were slightly
higher than yields obtained by nonadopters,
suggesting a reduction in risk that could be an
additional advantage for mulch adopters.

With regard to bean yields in the 1993 summer
cycle, groups of adopters differ significantly.
Adopters of both components obtained higher
bean yields than farmers who did not adopt the


Table 8. Yields, maize-bean intercropping system, Motozintla, Chiapas

Total sample Type of adopter

Standard Non- Mulch No-tillage Both
Average deviation adopter only only components Probability

Maize yield (t/ha)
1993 summer cycle 2.67 + 0.61 2.61 2.66 2.55 2.87 ns
Good year 3.17 0.80 2.91 a 3.2 ab 2.95 a 3.66 b .00
Normal year 2.26 + 0.52 2.12a 2.37 ab 2.06a 2.61 b .00
Poor year 1.49 + 0.46 1.40 a 1.54 ab 1.38 a 1.71 b .04
Bean yield (kg/ha)
1993 summer cycle 267 + 88 239a 213a 272ab 301 b .04
Good year 318 +107 ns
Normal year 212 + 90 208ab 152a 201 a 248b .05
Pooryear 115 70 104ab 50a 113b 148c .01

Note: Figures followed by different letters are significantly different (Duncan 0.1, row comparison); ns = not significant.










no-tillage component. Average yields in good,
normal, and poor years were 320, 210, and 120
kg/ha, respectively, and differences in yields
obtained by groups of adopters were found
only for normal and poor years. In these sorts
of years, adopters of both components
generally obtained higher yields, whereas those
who adopted only the mulch component
obtained lower yields. In relative terms, the
bean yields in the 1993 summer cycle were
129% (33) of bean yields in a normal year. In
good and poor years, bean yields were on
average 151% and 51%, respectively, of yields
in a normal year. Bean yields were more
variable than maize yields.20

Of all the factors that negatively affect yields in
poor years, wind seems to be the most
important, as much for maize (88% of farmers)
as for beans (85% of farmers) (Table 9). Given
the height of the maize plants and the
characteristics of the study zone, it is not
surprising that maize is vulnerable to damage
from strong winds. Maize fields are generally
found on the higher parts of the hillsides in the
study area, where they are exposed to the
elements. The vulnerability of beans to wind
damage can be deduced from the fact that

Table 9. Factors affecting yields in poor years,
Motozintla, Chiapas


beans are not solecropped but instead
intercropped with maize. Wind seems to be a
fairly frequent problem: in most cases (77% of
farmers affected by the problem), it seems to
occur each year. Compared to farmers who did
not adopt the no-tillage component, adopters of
that component (alone or in conjunction with
mulch) reported more frequent wind damage
to their crops. This finding is in agreement with
the assumption that maize grown under
reduced or no tillage is more susceptible to
lodging because it has more shallow roots.

Another factor that depresses yields in poor
years is diseases and pests, as much in maize
(26% of cases) as in beans (15%). The chief
maize pest was white grub (gallina ciega)
(Phyllophaga spp.), cited by 79% of the farmers
who reported disease and pest problems.
Farmers rarely apply any kind of control
measures for these problems, which are more
common among nonadopters than among
adopters of both components of the technology.
This finding contradicts the assumption that
under conservation tillage there is greater
potential for soil pests such as gallina ciega,
because farmers work the soil less and birds
have fewer opportunities to control pests

maize-bean intercropping system,


Type of adopter

Total sample Non- Mulch No-tillage Both
(average) adopter only only components Probability

Maizea
Wind (% cases) 87.8 76.2 71.4 90.0 100 na
Pests (% cases) 25.6 42.9 28.6 26.7 8.3 na
Beans
Wind (% cases) 85.4 85.7 57.1 86.7 91.7 na
Pests (% cases) 14.6 14.3 42.9 13.3 8.3 na

Note: na = not applicable.
a Figures do not sum to 100 because in some instances both factors affected yield.

20 Coefficient of variation for bean yields was 33-61%; for maize, 23-31%.










naturally (Bolafios, pers. comm.). Nevertheless,
this finding is consistent with information
reported by Ortega (1989) in reviewing other
studies. Bean pests are mostly foliar pests
(mentioned by 83% of farmers affected by
disease problems), especially D. balteata

The yield-mulch relationship The findings
described in the previous section seem to
indicate that the mulch component of the
conservation tillage technology has a greater
effect on yield than the no-tillage component.
This section describes the effects of mulch on
yield in more detail. Although farmers were
considered to have adopted the mulch
component only if they had at least 2 t/ha of
residues on their fields, the survey gathered
data for four levels of crop residues (Appendix
A): two below the threshold and two above it.
Figure 3 shows the positive relation between
maize yield and amount of mulch. This
relationship is very significant and presents
significant correlation coefficients of 0.3 for the
1993 summer season and up to 0.4-0.5 for the
different kinds of years. Although there are not
great differences in average yields for the first


4-



S3 L


< 1 t/ha 1-2 t/ha 2-4 t/ha >4 t/ha
Amount of mulch


S Poor year
EE Good year


3 Normal year
= 1993 summer cycle


Figure 3. Relationship between maize yield
and amount of mulch.


two levels of mulch applications (level 1 is <1 t/ha
and level 2 is 1-2 t/ha of mulch), the average
yield for level 3 (2-4 t/ha) does differ
significantly from the first two and, in turn,
differs from level 4 (>4t/ha). This seems to
indicate that below the threshold level of 2 t/ha
it is difficult to discern any effect of mulch on
yield. Over the threshold level there is a positive
response to incremental increases in the amount
of mulch applied. This response may have
something to do with the steep slopes in the
study area, and shows that for areas with similar
slopes it may be useful to raise the threshold
level of mulch to 4 t/ha of crop residues.
Nevertheless, the viability of such a high
threshold level depends greatly on the
availability of crop residues for mulch and the
opportunity cost.

It is worth noting a few points about the positive
interaction between mulch and maize yields. In
the first place, the interaction can be mutual: a
greater amount of mulch not only raises yields
(through improving the soil), but higher yields
can also increase the availability of residues to be
used as mulch, all things being equal. However,
the condition of "all things being equal" implies
a similar management of residues, and, as we
will see later in this paper, this is not the case in
the study zone. Earlier we described the positive
correlation (c.c.: 42; prob.: .00) between the
degree of slope and the conservation of residues
(for example, steeply sloping fields are grazed
less), whereas there is no interaction between
slope and yield (although one would expect a
negative relationship). However, the interaction
between bean yield and mulch is difficult to
disentangle from the available data.

For beans there is also a positive interaction
between yield and mulching, although generally
less significant and of lesser magnitude. For
example, for the 1993 summer cycle, the
correlation coefficient reaches 0.2 (prob. = .10).











Multiple regression analysis A linear
regression was done to analyze the factors
influencing maize and bean yields in the 1993
summer cycle. The equations with the best fit
explain approximately one-third of the
variation (see Table 10 for results for maize and
Table 11 for beans). Most of the variables in
both equations are related to the current
management of the field.

In the case of maize, several variables related to
current management practices are significant in
explaining observed variability in maize yields.
The application of N seems to increase yield by


nearly 12 kg for each kilogram of N applied.
Even so, because N application interacts
strongly with the incidence of lodging, the use
of N is subject to an interaction with wind.
When there are strong winds, the interaction
with N becomes negative. It is also important
to note that conservation tillage practices
increased yields in the 1993 summer cycle by
about 350 kg/ha on average. This effect seems
to be related primarily to the mulch component
of the technology, since an adequate mulch (>2
t/ha of residues at sowing) produces a similar
yield effect.21 Fallowing also has beneficial
effects on production: producers who fallow


Table 10. Factors affecting maize yields in the maize-bean intercropping system, 1993
summer cycle, Motozintla, Chiapas

Variable Description Coefficient t-value Probability

Dependent
MZYIELD93 Maize yield (kg/ha), 1993 summer cycle
Independent
NITROGEN Amount of nitrogen (kg/ha) 12.0 2.2 .029
N*WIND Nitrogen-wind interaction -14.6 -2.9 .006
ADOPCT Dummy for adoption of conservation tillage 347 2.6 .013
EARLYWEED Dummy for early weeding 121 0.9 ns
FERTINCORP Dummy for incorporating fertilizer -192 -1.5 ns
FALLOW Dummy for fallow 248 2.0 .051
Constant 2,580 15.1 .000
Note: R multiple = 0.50; R2 = 0.25; adjusted R2 = 0.19; degrees of freedom, 73; "enter" method; ns = not significant.

Table 11. Factors affecting bean yields in the maize-bean intercropping system, 1993
summer cycle, Motozintla, Chiapas

Variable Description Coefficient t-value Probability
Dependent
BNYIELD93 Bean yield (kg/ha), 1993 summer cycle
Independent
BNSEED Quantity of bean seed (kg/ha) 9.2 5.4 .000
ADOPCT Adoption of conservation tillage 29.7 1.9 .07
EARLYWEED Dummy for early weeding 24.8 1.5 ns
EARLYFERT Dummy for early fertilizer application 56.7 3.9 .000
NITROGEN Amount of nitrogen (kg/ha) 0.24 1.6 ns
WINDPROB Dummy for wind problem -4.2 -0.2 ns
Constant 103.2 3.4 .001
R multiple = 0.67; R2 = 0.46; adjusted R2= 0.41; degrees of freedom, 69; "enter" method; ns = not significant.

21 The two components of conservation tillage were not considered separately because there is an interaction between
them, such as the interaction between the no-tillage component and some other variables, which makes it difficult to
consider them simultaneously in the equation.










obtain an average of 250 kg/ha more yield than
those who do not.

The other two variables (early weeding and the
practice of incorporating fertilizer) were not
significant. Other variables of interest were not
very discriminatory (for example, the use of
improved versus local varieties) or were closely
interrelated. The interaction among some
variables made it difficult to interpret results of
the equation, so they were left out of the
equation. For example, the variables related to
fertilizer use (quantity, elements applied, and
number of applications) were closely
interrelated and were also related to herbicide
use (and thus to adoption of the no-tillage
component). On the other hand, some
characteristics of the field (such as slope) were
closely related to the management of the crop.

For beans, three variables related to current
management practices help explain the
variation observed in yields. The first is related
to the amount of bean seed used; the others are
related to fertilizer use and conservation tillage.
The equation demonstrates that there is a
strong correlation between seed rate and bean
yield. For the 1993 summer cycle, each
additional kilogram of bean seed raised average
yield by about 9 kg/ha. Planting density
therefore seems to be a limitation on the
production of the intercrop. The early
application of fertilizer to maize (at the three-
leaf stage at the latest) raised average bean
yields in the 1993 summer cycle by almost
60 kg/ha. This increase may be related to good
establishment of the intercrop. It is interesting
to note that in the 1993 summer cycle the use of
conservation tillage raised bean yields 30 kg/ha
on average. This effect could be related to the


no-tillage component, as it produces a similar
effect. Apparently the beneficial effects of the
technology on the intercrop (soil and water
conservation, for example) compensate for
many of the deleterious effects (such as the
"burning" caused by herbicides).

The three other variables in the equation -
early weeding, the amount of N applied to
maize, and wind damage were not
significant. Even so, it is important once again
to emphasize that other variables of interest
interact strongly among themselves or with
other variables in the equation, and so were
excluded from consideration.

Residue Management
All farmers left crop residues in the field after
harvest. Almost half of the farmers (49%)
mentioned that the residues prevented the soil
from eroding, and most reported that the
residues served as an organic fertilizer (84%).22
However, maize residues generally have a high
ratio of carbon to nitrogen (C:N), so the effect
of maize residues as an organic fertilizer is
rather limited in terms of a greater availability
of nutrients in the short run (in fact, it is more
likely that N will be immobilized). On the other
hand, the bean and weed residues (especially if
they are still green) have higher C:N ratios and
thus act as more of an organic fertilizer.

Although all producers leave residues on the
field, the amount that remains by planting time
varies considerably. The varying quantity of
mulch is reflected in the groups of adopters,
given that adopters of the mulch component
(alone or with the no-tillage component) by
definition have more than 2 t/ha of residue in
the field at sowing. Two factors that influence


22 The number does not add to 100% because 43% of farmers mentioned the organic fertilizer as well as prevention of
erosion, and 8% mentioned other reasons for leaving residues on the field.










this quantity of residue are the "treatment" of
residues23 and grazing by livestock.

Treatment of residues Farmers leave
residues on the field in different ways. The
most common practice (60% of farmers) is to
leave the maize plant standing after the
harvest without any special management. The
rest of the farmers chop the residues with a
machete and leave them on the field (21%) or
double the maize stalks at a height of about 1
m (20%). There are advantages and
disadvantages for each treatment, although
most farmers who leave the residues standing
in the field did not perceive any. One-third,
however, commented that their animals
consume less stover when it is left standing in
the field. The principal advantage of not
leaving the maize stalks standing appears to
be that they decompose better and prevent soil
erosion better. The main disadvantage of
chopping the residues is that more labor is
needed to do so.

The groups of adopters differ markedly in the
frequency with which they leave the stover
standing or chopped up. Most farmers (75%)
who did not adopt the mulch component
(alone or with the no-tillage component) left
their stover standing, whereas fewer than 5%
chopped it up (Figure 4). Most farmers (86%)
who adopted only mulch chopped their
residues, and no-one left it standing in the
field. Among adopters of both components,
42% chopped residues and 38% left them
standing. The frequency of the practice of
doubling the maize stalks is similar among all
groups of adopters.

Grazing by livestock Most fields (60%)
were grazed by livestock, either owned by the
farmer or by others, during the cropping


season. The fields of adopters of the mulch
component (alone or with the no-tillage
component) were relatively less exposed to
grazing (<45% of farmers) than those of
nonadopters of this component (>65%, prob.
.00). None of the adopters of the mulch
component (alone or with the no-tillage
component) considered crop residues to be an
important source of forage for animals,
whereas more than 40% of the nonadopters of
this component thought they were.

The great majority (95%) of farmers dislike
having their fields grazed, and the only
advantage they saw in this practice was that it
satisfied the demand for animal forage. The
main disadvantages of grazing included the
fact that the soil remains unprotected (45% of
farmers) and that the soil becomes loose or falls
apart (32%). Most farmers (63%) considered
that fencing their fields was a good option for
discouraging grazing. Twenty percent
mentioned the possibility of creating local
prohibitions to communal grazing (for
example, a law within the ejido to prohibit free


100-

80-
c0
o,
60-
3 40-
20
20-

0.


Non-adopter Mulch only No-tillage Both
only components


E Standing 1 Chopped = Doubled over

Figure 4. Treatment of residues by farmers
in Motozintla, Chiapas.


23 The "management" of residues is the more common term. Nevertheless, management of residues generally includes
grazing, and here we refer to management other than grazing.










grazing of livestock, or a rule that each person
must care for his or her own animals). With
regard to these options, there is not much
difference among the different classes of
adopters.

Nine percent of farmers had fenced their fields
before the 1993 summer cycle harvest, whereas
12% of farmers were thinking about doing so.24
Most farmers who had already enclosed their
fields were adopters of both components of the
technology. Only one farmer among those who
had fenced their land had adopted only the no-
tillage component and in fact he had not
enclosed his field sufficiently.25 Of the farmers
who had enclosed their fields adequately, 83%
had more than 4 t/ha of mulch. This indicates
once again that livestock belonging to others,
and not so much the farmer's own stock, limit
the availability of mulch. It is also useful to
recall that the slope of the field is another factor
that restricts the access of livestock to the field.

Most of the farmers who are considering
fencing their fields have adopted only the no-
tillage component of the technology.
Nevertheless, the fact that they are
contemplating fencing their fields does not
mean that they will actually do so in the short
or medium term. Various factors make it
difficult for farmers to enclose their fields,
including the limited availability of fence posts
(owing to the state law that prohibits the
cutting of trees) and of financial resources (to
purchase wire). In relation to this last factor,
44% of the farmers who considered fencing
their land mentioned that they needed credit to
do so.


Figure 5 summarizes some of the factors that
influence whether farmers can conserve
sufficient residues in fields accessible
to livestock.


Economics of the Maize-Bean
Intercropping System

Resources Dedicated to the
Cropping System
Table 12 summarizes crop operations for the
entire sample and for the different types of
adopters. On average, farmers spend 84 days/ha
on the maize-bean intercropping system,
although there are significant differences among
the groups of adopters. Adopters of the no-
tillage component (alone or with mulch) use less
labor than nonadopters. The difference can be as
much as 11-16 days/ha, and it is significant in
comparison with nonadopters (of any
component). This difference is related primarily
to weed control practices, because for the other
crop operations there is no significant difference
among groups of adopters. The difference is
even more pronounced (14-16 days) and
significant if we compare weed control among
groups of adopters. The only significant
difference in input use among the different
groups of adopters is also related to weed
control.

Valuation of Production Factors
In the sections that follow we present some of
the issues related to valuing factors of
production. 26 Because the valuing of residues is
more problematic, this issue will be discussed
separately.


24 This includes one farmer (an adopter of the no-tillage component alone) who had just finished fencing his field at the
start of the 1994 summer cycle.
25 The farmer had only fenced the field using two rows of barbed wire. Although the field was less accessible, sheep
could still enter to graze.
26 In 1993 Mexican pesos (Mx$). The average exchange rate for 1993 was Mx$ 3.1= US$ 1 (source: International Monetary
Fund).
















































Figure 5. Factors and interactions affecting the likelihood that sufficient crop residues will
be conserved in accessible fields, Motozintla, Chiapas.


Table 12. Labor use (days/ha) by operation and type of adopter, maize-bean intercropping
system Motozintla, Chiapas

Total sample Type of adopter

Standard Non- Mulch No-tillage Both
Average deviation adopter only only components Probability

Land preparation and sowing 23.7 6.5 ns
Fertilizer application (total) 14.7 +10.0 ns
First 8.9 + 5.3
Second 5.9 + 6.0
Weeding 24.4 10.7 34.1 a 34.2 a 18.5 b 20.4 b .00
First 14.4 + 7.7 22.3 a 24.5 a 9.8 b 10.5 b .00
Second 10.0 + 5.4 ns
Total pre-harvest labor 62.8 +16.5 73.9 a 66.9 ab 58.5 b 57.3 b .00
Harvest 20.8 7.3 ns
Total post-harvest labor 83.6 20.2 94.6 a 89.5 ab 78.9 b 78.2 b .02

Note: Figures followed by different letters are significantly different (Duncan 0.1, row comparison); ns = not significant.


SDifficult to conserve
2 t/ha residues or more
in accessible fields










Inputs and outputs Purchased inputs have
a relatively visible cost, although there are
slight differences depending on the point of
sale (for example, depending on whether the
input was purchased at the Uni6n de Ejidos, in
the market, and so on). It is important to take
transport costs into account when calculating
the farm price of an input, given that the study
area is relatively distant from the market and
that transport costs are considerable (for
example, they add about 12% to the cost of
ammonium sulfate). Appendix B presents
farm-level prices of the most common
purchased inputs. Maize and bean seed is
usually seed of local varieties, retained from
the previous harvest. The opportunity cost was
calculated using the most common sale price
for local seed within the ejidos.

Output prices are based on the sale price,
adjusted for transport costs and the cost of
shelling maize. It is important to recall that on
average more than half of the production is
destined for home consumption. For that
reason, transport price has a net positive effect
on the output price (in other words, owing to
the relative importance of maize for home
consumption, the value of the output is closer
to the value of the purchase price than it is to
the sale price).

Labor It is relatively common for farmers to
hire labor. Most hired laborers are paid
Mx$ 8/day and provided with food (reported
by 72% of farmers). If food is not provided, the
daily rate for labor rises to Mx$ 10 (19%). There
are other, less common, forms of payment,
such as payment in kind. This includes
arrangements whereby one farmer will work
for another in exchange for similar assistance
on another day (7%), or loans of land in


exchange for labor (2%). In this study we value
farm labor at Mx$ 10/day, for hired labor as
well as family labor (opportunity cost).

Land Arriving at an adequate valuation of
land presents some difficulties in places where
land markets are not very developed. In the
study area, land rental is not a common
practice; in only five instances (5%) did farmers
rent part of their land. However, in each case
the rent paid was similar, and Mx$ 250/ha/
cycle was chosen as a reasonable
approximation of the opportunity cost of land.
This rent represents 12% of the average gross
benefit.

Capital The cost of capital is reflected in the
interest rates charged for loans. However, these
rates vary considerably and depend greatly on
the source of the loan.27

We did not obtain reliable data on the cost of
capital, but we estimate that a 2.5% monthly
rate is an adequate approximation of the
average cost of capital. The real cost for the
farmer depends on many factors, and for this
reason we included a sensibility analysis on the
farm budgets for different interest rates.

Residues An adequate valuation of residues
is difficult in areas where there is hardly any
market for crop byproducts. Residues are a
secondary product of maize production, and
during the study we did not encounter any
farmer who had marketed residues either
selling the harvested residues or selling the
residues that were still standing in the field.

To solve the problem of valuing crop residues
in the study area, it is important to consider the
nature of the product. Residues in a field can be


27 For example, bank loans at commercial interest rates; state credit provided without interest; and loans from other family
members or from middlemen.











an exclusive or nonexclusive product. In an
enclosed field, crop residues are an exclusive
product: the person who is not prepared to
pay for their use is easily excluded from using
them.28 In an unfenced field, residues are not
an exclusive product. It is difficult to exclude
users who do not want to pay, especially when
the practice of free grazing is common, as in
the study area. In the study area, residues can
be considered a quasi-private product (a
nonexclusive and divisible product see
Turner et al. 1993:77), for which valuations
based on the market price do not work well.
The next sections of this paper will explore in
greater detail how exclusivity or
nonexclusivity influences the valuation of
residues.

Residues as an exclusive product There is
no market for crop residues in the study zone,
which means that no-one is prepared to pay
for residues in present circumstances.
Therefore, if we consider residues to be an
exclusive product, we would only need to
consider supply and demand for residues
within the farm itself. Although residue yields
were not estimated during the survey, they
can be estimated from harvest index
(estimated at 35%) and grain yield (Erenstein
1996). Residue yields were estimated at 5 t/ha
on average.


We estimate that demand for residues as forage
to feed the farmer's own livestock is 1.5 t of
residues per hectare per year.29 If weathering of
residues during the dry season does not
surpass 10%,30 3 t/ha of residues will be
available for other uses on the farm. Presently
the only additional use of residues is as mulch,
so 3 t/ha of residues are available for mulch,
amply exceeding the 2 t/ha required for an
effective mulch. If residues are an exclusive
product, there is no conflict in demand for their
use as forage and as mulch. In addition, given
that residues used to form an effective mulch
have no alternative use, the opportunity cost is
practically zero.31

Residues as a nonexclusive product When
residues are a nonexclusive product, it does not
matter that there is no market for residues in
fact, the nonexclusivity of the product partly
explains why there is no market. However,
there are many users of the product, although
under present circumstances they are not
inclined to pay (why pay for something that
you can get for free?). For the nonadopters of
the mulch component, this does not have great
consequences. The pressure of livestock owned
by others is still limited, so there are still
sufficient resources for the farmer's own
livestock. And even if the residues on the
farmer's field run out, there will be enough in


28 At least where it is socially acceptable for fields to be enclosed and where the enclosure is respected. This seems to be the
case in the study area.
29 Based on the following assumptions: energy requirements of 45 MJ/day/AU; an energy value of 8 MJ/kg dry matter of
residues (Euroconsult 1989: 603-607); a 13% moisture content in the residues; residues as the only source of energy during
the dry season; consumption of residues limited to the dry season (six months); and an average livestock density of 1.3 AU
per hectare of maize.
30 Estimate based on the following assumptions: the average yield for producers in level 4 for mulch for the 1993 summer
cycle was 3.07 t/ha of grain, which is equivalent to an estimated yield of 5.7 t/ha of residues at harvest time. The great
majority (89%) of these farmers do not graze livestock on their fields and had an estimated average 5 t/ha of residues at
sowing in the 1994 summer cycle. This implies a weathering of 0.7 t of residues (12%).
31 Recall that beyond threshold level for mulch of 2 t/ha there is still a favorable yield response as the amount of mulch
increases (Figure 3). However, to obtain more than 3 t/ha of residues it is necessary to reduce the use of residues for forage.
In this case there is obviously a conflict between the use of residues as mulch and as animal forage, which substantially
raises the opportunity cost of the residues. At the same time, the use of residues as forage also has an opportunity cost: by
using residues for forage, the farmer cannot obtain the benefits of applying more than 4 t/ha of mulch.










neighboring fields, since under present
circumstances the supply of residues still
exceeds demand for forage.

However, for farmers who have adopted the
mulch component, there is a problem. How can
a farmer ensure that enough residues remain in
a field to meet the threshold level needed for a
good mulch? There are several options, each
with its own cost and implications.

* Each farmer cares for his own animals. All
responsibility for conserving residues lies
with the livestock owners. For this reason,
this option does not appear very viable
unless there is some kind of positive or
negative inducement, such as severe
consequences for those who disobey this
option; a limited number of politically weak
livestock owners; or a strong social
organization in which people are motivated
to follow the rules for the good of the
community. In the Tuixcum ejido, the ejido
committee tried to impose a local law in 1993
that required everyone to care for his or her
own animals. However, the law was not
enforced. Several ejido members who owned
livestock violated the law. At present this
option does not seem to be a good means of
controlling the use of crop residues.

* Each producer takes care of his/her crop
residues. All responsibility for conserving
crop residues lies with the owner of the
residues. If the cost is not too high, the user
may even be willing to pay to conserve his
own crop residues. There are several
alternatives here:
1. Personally ensure that livestock belonging to
others do not graze the field. This method
could be effective for fields that are close


to farmers' homes, but it is not a very
reliable or cheap option, especially for
distant fields. A daily visit to the field can
take an hour. Over six months, at an
opportunity cost of labor of Mx$ 10/day
and a field of one hectare, this would
come to Mx$ 228/ha. The need to visit
the field daily also limits the farmer's
options for working outside the farm,
which further raises the opportunity cost
of labor for checking fields. Even a daily
visit will not necessarily ensure that
enough residue remains ungrazed in the
field. The option of permanently
watching the field would be safer but
even more expensive.
2. Store the residues. This option does not
appear very feasible given the amount of
work required (collecting, transporting,
storing, and then returning considerable
quantities of residues to the field).
3. Enclose the field. This method is quite
effective but, as observed earlier, requires
a considerable investment, driven up
further by the law prohibiting people
from cutting trees. The cost of fencing a
field is estimated at about Mx$ 850/ha.32
A useful life of ten years and an
opportunity cost of capital of 30% implies
fixed costs of depreciation at Mx$ 213/
ha/yr. The real cost per hectare, of
course, depends on the characteristics of
the field. It is relatively cheaper to fence
large fields or fields that border on fields
that have already been fenced.

A final option would be to purchase residues
to meet the threshold needed for mulch, but it
generally is not cost effective to do so (Lal
1989:94). Such a cost could be prohibitive in the
study zone: approximately Mx$ 450 would be


32 Using a price of barbed wire of Mx$ 0.4/m, three rows of wire for the enclosure, Mx$ 2/post/2.5 m, and 5 days to
install the fence.










needed to cover one hectare with 2 t of
residues.33 Even if 1 t/ha of residues remains
from the previous harvest, the cost of
conserving enough residues is lower than the
cost of buying them.

The opportunity cost of the residues needed to
satisfy the threshold of 2 t is assumed to be
reflected in the annual cost of fencing the field.
Once the field is enclosed, the residues can be
considered an exclusive product and their
availability for forming an effective mulch will
be assured, at virtually no opportunity cost.
However, the opportunity cost could increase
in the future if the number of enclosed fields
grows considerably. This would limit the
availability of residues for community grazing
and would facilitate the creation of a market
for residues.

Budgets
Table 13 presents budgets for different groups
of adopters. To facilitate interpretation and
comparisons, the inputs that did not differ
significantly among adopters were kept
constant. Components of the budgets are
discussed in the following sections.

Income Income (gross benefits) is similar
for nonadopters and adopters of only one
component of the technology, reaching nearly
Mx$ 2,000/ha. However, income is
considerably higher (12-14%) for adopters of
both components. The budgets reveal the
economic importance of the bean crop, which
provides about 20% of the gross benefit.

Variable costs For variable costs, we
distinguish between costs of inputs and labor.
Input costs are much higher for adopters of the
no-tillage component (alone or with the mulch


component) owing to the higher cost of the
herbicides (Mx$ 70-80/ha plus Mx$ 10-12 in
interest per cycle). Labor costs are substantially
higher for those who did not adopt the no-
tillage component, owing to the greater labor
requirement for controlling weeds (almost
Mx$ 150/ha). As a result, total variable costs
are slightly lower for adopters of the no-tillage
component. It is important to recall that the
cost of residues used for mulch (in fact a
variable cost) was calculated as part of the cost
of fencing (a fixed cost), so that residues could
be considered an exclusive product with a
minimal opportunity cost.

Fixed costs Fixed costs comprise the
opportunity cost of land and the fixed cost of
capital. The opportunity cost of land is
assumed to be equal for all groups of adopters,
whereas the fixed cost of capital varies for the
different groups. The cost of capital for
nonadopters of the mulch component (alone or
with the no-tillage component) includes the
costs of depreciation and interest on the
equipment used (sprayer, hoe, machete, and
file). The cost of capital for adopters of the
mulch component is substantially higher and
includes the cost of depreciation and interest
on the enclosure as well as the equipment
used.

Returns The budgets present indicators of
the returns received by different groups of
farmers. The value added is the gross benefit
minus the expenses for inputs (that is, the
return to cover the investment in land, labor,
and capital resources). Value added is
relatively similar for nonadopters and
adopters of only one component of the
technology but is considerably higher for
adopters of both components (11-16%),
reaching Mx$ 1,760/ha for the latter group.


33 Assuming Mx$ 5/bale (including the cost of transport) and 22 kg/bale.












The net benefit is the gross benefit minus the
cost of all resources invested, that is, the profit
for the farmer. On average, all farmers
(regardless of their classification as adopters)
obtained a small profit in the 1993 summer
cycle, although this varied from a minimum of
Mx$ 100/ha for adopters of mulch only to a
maximum of Mx$ 430/ha for adopters of both
components. However, it is important to point
out that the bean crop makes an important
contribution to this profit. In all cases, the
gross benefit of the bean crop exceeds the net


benefits from the maize-bean intercropping
system. For this reason, assuming that there is
no interaction between the maize and bean
crops, the cultivation of maize alone would
give an approximate net benefit of zero (from
negative to slightly positive). The low
profitability of the maize crop can be related
to the production of maize for home
consumption and the sale of any surplus. It
could be that farmers give a higher value to
maize for home consumption than its
estimated opportunity cost.


Table 13. Budgets for the maize-bean intercropping system, 1993 summer cycle,
Motozintla, Chiapas


Type of adopter


Nonadopters

Units/ Units/


Mulch only

Units/


Unit (MX$) ha ha Mx$/ha ha Mx$/ha


No-tillage only

Units/
ha Mx$/ha


Both components

Units/
ha Mx$/ha


Maize
Beans
Total gross benefit


Fertilizer

Herbicide
Credit (6 mo.)
Total inputs


Maize
Beans
Sulfate
Triple 17
Gramoxone


Land preparation/sowing
Fertilizer application
Weeding
Harvest
Total labor
Total variable costs


Land
Capital (depreciation, interest)
Total fixed costs


PI--eun


Value added [A-B1]
Net benefit [A-B-C]
Costs per kg maize
Labor productivity
Return per day


Unit
price


2,610 1,566
239 387
1,953


2,550
272


1,530
441
1,971


kg
kg
50 kg
50 kg
lit
%/mo.


16
27
207
20
1.97 41
311 47
358


2,660 1,596
213 345
1,941


16
27
207
20
2.46 52
321 48
370


237
147
34.2 342
208
93.4 934
1,304


2,870 1,722
301 488
2,210


16
27
207
20
5.64 118
388 58
446


237
147
20.4 204
208
79.6 796
1,242


16
27
207
20
5.78 121
391 59
450


237
147
18.5 185
208
77.7 777
1,227


237
147
34.1 341
208
93.3 933
1,291


ha 250
30%


Mx$/ha
Mx$/ha
Mx$/kg
kg maize/day
Mx$/day


1,595
336
0.609
28.0
13.6


1,571
99
0.682
28.5
11.1


1,521
418
0.598
32.8
15.4


1,763
428
0.611
36.1
15.4


. Beefit


31. nput


. Fxed ost










The low profitability of the maize crop can also
be seen by calculating the costs per kilogram of
maize, excluding costs and benefits of the bean
intercrop.34 With the exception of farmers who
adopted only the mulch component, the costs
per kilogram of maize are similar, quite close to
the field price of maize, indicating that the profit
for the farmer is practically zero. Those who
adopted only the mulch component actually
incurred losses their costs were greater than
the price of the output.35 For adopters of the no-
tillage component alone, costs were slightly
lower and yields similar to those obtained by
nonadopters. Adoption of the mulch component
substantially raised costs, which were only
recovered if the farmer also adopted the no-
tillage component and benefited from the
increased maize yields. However, it is important
to note the low international competitiveness of
maize production in the study area: for all
farmers, the costs of producing a kilogram of
maize are substantially higher than the
international maize price.

The budgets also include labor productivity in
physical and monetary terms. In physical terms,
the productivity of labor, expressed as
kilograms of maize per day of labor, is similar
for farmers who did not adopt the mulch
component and considerably higher for those
who adopted both components. Labor
productivity for farmers who adopted only the
no-tillage component is between the two
other groups.

These relationships change slightly when labor
productivity is expressed in monetary terms
(Mx$/day). Labor productivity is the lowest
(Mx$ 11/day) for those who only adopted the
mulch component. Labor productivity for
adopters of the no-tillage component, alone or


with the mulch component, is the highest,
about Mx$ 15/day. Nonadopters are in second
place, at Mx$ 14/day. However, it is important
to note that labor productivity is higher than
the local wage for hired labor.

Sensitivity to cost of capital The budgets
assume an interest rate of 2.5% per month as
the cost of capital. However, this is one of the
most variable costs. In the following
paragraphs we summarize results of the
sensitivity analysis for different interest rates
(see also Appendix C).

Interest rates influence the cost of working
capital and fixed capital. A low interest rate
would be advantageous for components of the
conservation tillage technology that require
more spending on inputs (as in the no-tillage
component) and/or more investment in fixed
capital (as in the mulch component). One can
expect that budgets for adopters of both
components of the conservation tillage
technology would be most sensitive to changes
in interest rates, followed by budgets for
adopters of a single component, and that the
budget for nonadopters would be least
sensitive to such changes.

If we consider a cost of capital of only 1% per
month, the budget for adopters of both
components becomes even more attractive.
The net benefit increases by more than
Mx$ 100 to nearly Mx$ 580/ha; the cost of
producing one kilogram of maize falls to less
than Mx$ 0.56/kg and labor productivity rises
to more than Mx$ 17/day. The situation also
improves for adopters of the mulch
component alone. Most of these improvements
are related to the lower cost of the mulch
component, given that the fixed cost of the


34 Assuming that there is no interaction between the two crops. This indicator sums all costs (minus the costs of bean
seed) and divides them by maize yield.
35 The loss implies that the return on the factors of production invested does not cover their opportunity cost.










enclosure now is Mx$ 136/ha. However, in
relative terms the situation does not change
much: indicators for those who adopted only
the mulch component remain low, whereas
nonadopters and adopters of only the no-
tillage component occupy second place.

If one assumes an interest rate of 5% per
month, the economic advantages for adopters
of both components disappear. The net benefit
for this group of farmers falls by Mx$ 150/ha
to Mx$ 280/ha, the cost of producing one
kilogram of maize rises to more than
Mx$ 0.66/kg, and labor productivity falls to
Mx$ 13.5/day. As a result, the budget for
adopters of only the no-tillage component
becomes relatively more attractive. The
situation for adopters of mulch alone becomes
worse: the net benefit becomes negative and
labor productivity falls below the wage rate for
hired labor.

As noted earlier, the interest rate of 2.5% is an
approximation of the average rate. Various
factors specific to the individual farmer can
influence this rate. For example, access to state
credit or some other kind of subsidized credit
scheme can substantially reduce the interest
rate for the farmer. In the same way, cash
income (obtained through outmigration, for
example) can reduce the cost of capital if the
alternatives for investment are limited. On the
other hand, falling behind on loan payments
can raise the interest rate by limiting possible
sources of credit and increasing the risk of
providing credit.

Sensitivity to changes in yield The budgets
represent an approximation of costs and
benefits in the 1993 summer cycle. However,
this was a relatively good cropping season.
What would the budgets look like at lower
yield levels? Results of a sensitivity analysis
done for yield levels reported for good,


normal, and poor years are described briefly
here (see also Appendix C). Note that all
production costs remain constant (that is,
independent of yield levels). The levels
reported for the 1993 summer cycle are
assumed to be representative for all years.

In all types of years, adopters of both
components obtained the best economic
results. The opposite occurred for farmers who
adopted only mulch. Those who did not adopt
the mulch component are in the middle of the
range. Only adopters of the no-tillage
component surpassed the nonadopters (of
either component).

Compared with the budgets for the 1993
summer cycle, the budgets for a normal year
are not very attractive for the different classes
of adopters, with the exception of the adopters
of both components. For nonadopters and
adopters of no-tillage alone, the net benefits are
approximately zero; the cost per kilogram
surpasses Mx$ 0.7 and labor productivity
comes close to the daily wage rate for hired
labor. The situation is worse for adopters of
mulch alone. However, for adopters of both
components the situation is much more
advantageous: a net benefit of approximately
Mx$ 225/ha; a production cost of Mx$ 0.66/kg;
and a return on labor of almost Mx$ 13/day.

In a bad year, the situation is difficult for all
farmers, with losses of Mx$ 500-800/ha,
production costs of Mx$ 1.00-1.16/kg, and
labor productivity of only Mx$ 1.4-4.0/day. In
such years it is very difficult for farmers to
meet their financial obligations and avoid
falling behind on their loan payments.

In a good year the situation is relatively good
for all farmers, with a considerable advantage
for adopters of both components.











The Adoption of
Conservation Tillage

Adoption
Table 14 shows the distribution of the survey
farmers by class of adopter. Of the 82 farmers
surveyed, 29% had adopted both components
of the technology (that is, they were full
adopters of conservation tillage). Twenty-six
percent of farmers had adopted neither
component; 37% had adopted only the no-
tillage component, and 9% had adopted the
mulch component alone.

The distribution of cases in the adoption
matrix also shows how relatively more farmers
have adopted the no-tillage component (66%)
than the mulch component (38%). This
suggests that it is more easy or more attractive
to adopt the no-tillage component. It is
important to note that although 34% of the
sample farmers had not adopted the no-tillage
component, all farmers practiced some form of
reduced tillage. In addition, the great majority
of farmers used as least one application of
herbicide to control weeds. For this reason, the
difference between farmers who adopted the
no-tillage component and those who did not is


Table 14. Adoption matrix for conservation
tillage in the manual hillside production
systems of Motozintla, Chiapas

Adopted no-tillage component

No Yes
No Nonadopter Adopted 62.2%
(25.6%) no-tillage only
Adopted (36.6%)
mulch
component Yes Adopted Adopted both 37.8%
mulch only components
(8.5%) (conservation
tillage)
(29.3%)

I 34.1% 65.9% 100%


almost hypothetical, for the benefit of adopting
this component is minimal.

The situation is somewhat similar for adoption
of the mulch component. Although 62% of the
sample farmers did not adopt the mulch
component, all producers had stopped burning
crop residues. The fact that they did not adopt
the mulch component of the technology does
not necessarily mean that these farmers left no
residues as mulch on their fields, completely
exposing their fields to the elements, but that
the quantity of residues remaining on the field
at planting was not enough to be considered an
effective mulch. Furthermore, there is a
continuous soil conservation response as the
amount of soil cover increases, although this is
subject to the law of diminishing returns (Figure
6). All of these factors directly influence the
effects of adopting mulch.

To illustrate this more clearly, we present three
hypothetical farmer cases (X, Y, and Z) in Figure
6. Farmer Z is an adopter of mulch, for 35% of
his field is covered with maize residues.
Farmers X and Y are not adopters of the mulch
component, given that their fields have a 0%
and 15% coverage, respectively. However, the


100 -X(nonadopter)

80-
S- d(Z-X)
60

S.......\Y (nonadopter)
240 .



0-
- 204 .......a...... Z ()adopter)


0 10 20 30 40 50 60 70 80 90 100
Soil cover (%)
Figure 6. Relationship between relative
erosion and soil cover.
Source: Adapted from Shaxson et al. (1989).










fact that Farmer Y has some mulch on his field
influences the level of erosion in the field. For
that reason, the difference in the relative levels
of erosion in these farmers' fields is much less
between the fields of Z and Y (only 30%) than
between those of Z and X (87%).

This problem partly explains why some of the
differences between groups of adopters are less
marked than one would expect. However, as
this paper shows, there are even differences in
the relevant indicators of technology adoption,
which can be explained by various factors that
influence the farmer. The effects of the no-tillage
component as well as the mulch component
would have been much greater if they had been
compared to farmers' "traditional" (original)
practices, before farmers stopped burning their
fields. However, times have changed, and what
is considered "traditional" practice has changed
with them. The impact of adopting conservation
tillage technology may in fact be higher than
results of this study would suggest.

Adoption over Time
Conservation tillage technology can be thought
of as a package of technology consisting of two
basic components: no-tillage and mulch. As
Byerlee and Hesse (1986) documented in the
Mexican highlands, different components of a
technology package can have independent
patterns of adoption. This has occurred in the
study area, where some farmers have adopted
one, two, or none of the components of
conservation tillage, and the components were
not adopted at the same time. However, it is
problematic to determine the real level of
adoption of the mulch component, and it is
even more complicated to try to determine
when farmers first began using mulch. Our
ability to determine the historical framework for
the adoption of the no-tillage component is
similarly limited, but we can use other variables
as approximate indicators of the adoption of


both components. The decision to stop burning
residues can be seen as a proxy for mulch
adoption; similarly, adoption of herbicide is a
proxy for adoption of the no-tillage component.

The actual component is usually adopted at the
same time or after the indicator has been
adopted. It is obvious that mulch cannot be
adopted until the farmer stops burning residues,
so disadoption of burning is a necessary though
not sufficient condition for adoption of the mulch
component. The relationship between herbicide
use and adoption of the no-tillage component is
similar. For comparative purposes in our
discussion below, we also include data on
fertilizer adoption.

Disadoption of burning Farmers stopped
burning crop residues for land preparation about
ten years ago (data from retrospective
questioning). Adopters of only the mulch
component stopped burning even longer ago
than that (14 years on average). Most farmers
(62%) stopped burning after local regulations
forbade the practice, whereas the remainder
stopped burning once they saw the advantages of
abandoning burning. Most farmers (67%) first
heard about the practice of not burning from
government extension agents.

Herbicide adoption On average, sample
farmers have used herbicides for eight years.
Most (70%) began to use herbicides because of the
labor savings involved herbicides enabled
farmers to cover more area in a shorter time.
Most farmers learned to use herbicides from
neighbors or family members (84%) and bought
their herbicide in the market the first time they
used it (99%). This suggests that farmers
generally taught themselves how to use
herbicides (learning by doing), which indicates
that not all farmers may know about the risks
inherent in using these chemicals. This is
especially problematic when one considers that










the most frequently used herbicide, paraquat,
is relatively toxic.

Fertilizer adoption On average, farmers
have 11-12 years of experience in using
fertilizers.36 Most farmers (80%) began to use
fertilizers to raise yields of maize grown in
exhausted soils. The majority learned to use
fertilizer from neighbors and family members
(83%); the first time they used fertilizer, they
purchased it in the market (95%).

Figure 7 shows the initial diffusion of the three
technologies (not burning; using herbicide;
using fertilizer) in the study zone. The pattern
of adoption reflects a logistic curve based on
observed data (CIMMYT 1993):


Not burning: Y,


1/(1 + e+[867.7-0.4376*t])
[R2: 0.88];


Herbicide: Y = 1/(1 + e+[1426- 0.7184*t])
[R2: 0.92];

Fertilizer: Yt = 1/(1e ++[1339-0.6757*t])
[R2: 0.94];


where Y, is the cumulative percentage of
farmers who adopted the technology in time t.

Figure 7 indicates that the practice of not
burning residues diffused less rapidly than the
other two practices. The figure also shows that
60% of farmers stopped burning between 1981
and 1987 (7-13 years ago). Fertilizer adoption
was about three years ahead of herbicide
adoption and the rate of adoption was similar
for the two technologies. Between 1981 and
1985, 60% of farmers started to use fertilizer
(9-13 years ago), whereas 60% started to use
herbicides between 1984 and 1988 (6-10
years ago).


One notable feature of Figure 7 is that the
adoption curve for the no-burning practice
flattens out near the 50% adoption level
around 1985 and that after 1985 adoption grew
more rapidly as it neared 100%. The local
regulation against burning was promulgated
in 1985, pre-dating the state law, which may
explain the sudden jump in adoption of the no-
burning practice. It is also useful to remember
that estimation of the logistic curve is based on
certain assumptions about the diffusion of
technology and that the fixed parameters
estimated for the curve imply that the
environment has remained constant in the
period to which the curve is fitted (CIMMYT
1993:13). Thus it is more precise to present a
logistic curve for the years up to 1985, and
another for the years after 1985:

To 1985: Yt = 0.5/(1 +e+[923.7 0.4664*t)
[R2: 0.96];


After 1985:


Yt = 1/(1 +e+[2411- 1.215*t])
[R2: 0.97].


100 ..L J

80- 4O **0

60- O/
/
- 60

40 / 4.
0 0 /
S20
0 ... .- A
1977 1980 1983 1986 1989 1992

0 Fertilizer A Herbicide S No burning
Figure 7. Diffusion of fertilizers, herbicides,
and the abandonment of burning, Motozintla,
Chiapas.


36 Farmers who apply fertilizer around the base of the maize plant have been using fertilizer longer (12.3 years on
average) than farmers who incorporate fertilizer (10.6 years; probability = .03).











Figure 8 shows the results if the local law
against burning is taken into account. Before
the local law entered into effect, the diffusion
of the no-burning practice was slower than that
of fertilizer and herbicide use. Once the law
was promulgated, the no-burning practice
spread at a more rapid rate than that achieved
by the other technologies. According to Figure
8, 60% of farmers stopped burning between
1979 and 1986 (8-15 years ago).

After the local law against burning came into
effect, diffusion of the no-burning practice was
not a matter of choice but of compulsion, given
that burning residues would have serious
consequences. Presently a farmer who is
caught burning residues without permission is
arrested. He is required to pay the municipality
of Motozintla a fine equaling 50 days of the
minimum wage in the Federal District of
Mexico City or to serve 50 days in jail.
Occasional burning permits are issued only so
farmers can clear land that has been fallowed
for more than two years (Cadena 1995:5). The
data from this study suggest that the law was
fairly effective at discouraging burning,
probably because it was really enforced at the
local level. The data also suggest that the state


law forbidding burning, passed at the end of
1992, had relatively little effect in the study
area, for by that time adoption of the no-
burning practice was complete.

Specific incentives for the adoption of
conservation tillage To promote
conservation tillage practices, and particularly
to encourage farmers to stop burning residues,
the federal and state governments offered
various incentives (backpack sprayers, cash
credit, and inputs) to farmers in Chiapas,
including the study area. Altogether, 96% of
the sample farmers received one of these
incentives (Figure 9).

The great majority of farmers (94%) received a
backpack sprayer, and most of these sprayers
were distributed by SARH in 1990 (according
to 92% of the survey farmers who received
them). Most farmers (81%) thought that the
sprayers were given to them as a way of
encouraging them not to burn crop residues.
Because nearly all farmers received a sprayer,
ownership of a sprayer was not a
discriminatory variable among the different
groups of adopters. However, it is important to
emphasize that farmers who did not receive


1977 1980 1983 1986 1989 1992

Figure 8. Diffusion of the practice of not
burning residues, taking into account the 1985
law against burning, Motozintla, Chiapas.


80

S60
0,

2 40

3 20

0
0


Nonadopter Mulch
only


No-tillage Both
only components


SSprayer M Credit Inputs


E Other


Figure 9. Incentives for adoption of
conservation tillage, Motozintla, Chiapas.










this incentive are concentrated among the
groups who did not adopt the no-tillage
component.

Slightly more than half of all farmers surveyed
(55%) have received a state credit (called
credito de palabra") at least once as an
incentive. This credit is awarded in cash and no
guarantees are needed: the farmer's word
(palabra) is sufficient to guarantee repayment. If
farmers repay the credit at harvest, they are
offered a similar credit for the next cropping
season. Most farmers (91% of those who
received such credit) obtained it starting in
1992 through the FOSOLPRO program (Fondos
de Solidaridad para la Producci6n Solidarity
Funds for Farm Production) of the SDRE. It is
important to note that adopters of the mulch
component (alone or with the no-tillage
component) received credit more frequently
(>80% of farmers) than nonadopters (<40%,
prob. = .00). This difference is probably related
to the conditional nature of the credit. Most
farmers believed that they were awarded the
credit so that they would leave crop residues in
the field and not till the soil, whereas the
remainder believed that credit was awarded
primarily to discourage farmers from burning
residues (36%).

Sixty percent of the farmers who said they
received credit obtained it during three
cropping cycles (the 1992 to 1994 summer
cycles); 27% received it in only two cycles (1992
and 1993); and the remainder in one cycle.
Most producers who did not obtain credit over
the three cycles had been unable to pay their
debts. In the summer 1994 cycle, the amount of
credit varied from Mx$ 125/ha to Mx$ 400/ha.
The proportion of farmers who received credit
three times varies in the different groups of
adopters. Among adopters of both components
of the technology who obtained credit, 75%
obtained it for three years, compared to 50-55%


of adopters of only one component and 38% of
nonadopters. Compared to nonadopters,
adopters of both components apparently have
fewer problems paying their debts.

A small percentage of farmers (6%) received
inputs as a stimulus, including herbicides and/
or fertilizers. Although so few farmers received
inputs, it is clear that adopters of the mulch
component (alone or with the no-tillage
component) received inputs more often (>10%)
than nonadopters (<5%).

Factors Influencing Adoption
In the previous sections of this paper we
described several factors that appear to be
related to the adoption of different components
of the conservation tillage technology. Some of
these factors are a consequence of adoption,
whereas others have favored the adoption
process. For a more detailed analysis of the
factors influencing adoption, we developed a
multivariate logistic regression model which
predicts the probability that a farmer will adopt
only one component (partial adoption) or both
components (total adoption) of the
conservation tillage technology, based on a
series of farm or farmer characteristics
(CIMMYT 1993; Nagy and Ahmad 1993; and
Sain and Herrera 1996). The model is described
below.

Dependent variable: partial or total
adoption -The dependent variable in the
equation is a qualitative variable that classifies
farmers into four groups: nonadopters; the two
groups of farmers who adopted only one
component of the technology; and farmers who
adopted both components. Table 15 summarizes
the characteristics of the dependent variable for
these categories. The calculation of the
probability that a farmer will adopt only one or
both components is normalized on
nonadoption.










Independent variables Table 16 lists the
independent variables used in the model, as
well as the expected effect of each variable on
adoption of both components of the
conservation tillage technology. By definition,
adoption of conservation tillage requires the
no-tillage component and the mulch
component to be adopted at the same time.
"Partial adoption" refers to adoption of only
one component of the technology, whereas
"total adoption" refers to adoption of both
components. In general, different variables are
expected to influence adoption of each
component, whereas a combination of all of
those variables would influence total adoption.
For that reason, it is expected that some
variables that help explain the probability of
adopting one component will also help explain


Table 15. Dependent variable, logit analysis
of adoption of conservation tillage practices
in Motozintla, Chiapas

Value Percent of
(Yi) Category Mulch Tillage sample

0 Nonadopter X X 26
1 Adopter,
mulch only X 9
2 Adopter,
no-tillage only X 37
3 Adopter, both
components 29


the probability of adopting the package of
technology in which it is included. Next we will
list the independent variables and explain the
rationale behind them.

* Slope (SLOPE): All of the fields in the study
area are quite steep, but even so, it is expected
that ease of access, for livestock and farmers,
will negatively affect the probability that
farmers will adopt both components of the
technology. Access for livestock is related to
grazing intensity, and thus with the
probability of adopting the mulch component.
Livestock prefer to graze level fields rather
than sloping fields (if there is sufficient forage
in accessible fields, why graze steep fields?).
Access for farmers is related primarily to the
ease of tillage, and thus to the probability of
adopting the no-tillage component. It is much
easier to work the soil of a level field than of a
sloping one; on such steeps slopes it is difficult
to move across the field at all. The no-tillage
component considerably facilitates land
preparation and weed control. Slope can also
be expected to affect soil conservation: the
greater the slope, the greater the erosion, and
the greater the soil conservation effect if mulch
is used. However, given the difficulty of
maintaining enough residues without fencing
the field, it is probable that the effect of slope is
more closely related to access to the field.


Table 16. Independent variables in the logit analysis and their hypothetical effects on
adoption of each component of the conservation tillage technology in Motozintla, Chiapas

Predicted effect on adoption

Variable Description Mulch No-tillage

SLOPE Average slope of field (%) + +
COMLIVE Communal livestock pressure (AU per planted ha,
modified for fencing) +/-
FSIZE Total farm area, 1993 summer cycle (ha/farm) +/- +
FAMLAB Availability of family labor (potential labor/farm) +/- +
DMONMAIZE Dummy variable if farmer grew a maize monocrop +/- +
DBUSINESS Dummy variable if farmer had another business +/- +










* Farm size (FSIZE): Farm size is expected to
contribute positively to the probability of
adopting the no-tillage component. Farm size
is an indicator of the amount of work
required on the farm. When there is more
work on the farm, there is a greater
probability that the farmer will adopt a
component that saves on labor and permits
him to complete farm operations more
rapidly.

Farm size is also a good indicator of income
and the other resources on which the farmer
depends. For example, there is a positive
relationship between farm size and several
indicators of cash flow on the farm, such as
maize sales (c.c.: .42; prob.: .00), sales of other
agricultural products (c.c.: 19; prob.: .09), and
the farmer's use of state credit (c.c.: .21; prob.:
.06). A greater flow of cash on the farm makes
it easier for the farmer to purchase the inputs
that are required. There is also a positive
correlation between farm size and resources
such as information (for example, visits from
extension; c.c.: .27; prob.: .01) and the
livestock herd (c.c.: .52; prob.: .00).

Farm size also influences the hiring of labor.
There is a positive correlation between farm
size and the hiring of day laborers (c.c.: 20;
prob.: .07), and a negative correlation between
farm size and the farmer's need to work as a
day laborer in the fields of other producers in
the zone (c.c.: -.21; prob.: .06). One would
expect that hiring of day laborers would favor
adoption of the no-tillage component, given
that hired laborers are generally paid in cash.
One cash expense (hired labor) would be
replaced with another (inputs for the no-
tillage component), and the potential savings
would determine the potential for adoption.
On the other hand, it is expected that the need
to go out and work on other farmers' fields
would limit adoption of the no-tillage


component. This need is closely related to the
scarcity of income and thus with difficulty in
acquiring the inputs needed for the no-tillage
component.

* Communal livestock pressure (COMLIVE):
Communal livestock pressure is expected to
contribute negatively to the probability of
adopting the mulch component. The
likelihood that a farmer can conserve
sufficient residues depends greatly on the
pressure on the residues as forage during the
dry season (and on whether fields are
enclosed, which eliminates external pressure).

Communal livestock pressure reflects only the
external pressure on a farmer's residues
related to free grazing. In addition, the
pressure of the farmer's own livestock reflects
the internal needs of the production system.
However, the size of the farmer's herd is
linked as much to farm size (see earlier) as to
the availability of family labor (c.c.: .27; prob.:
.01). The pressure of the farmer's own
livestock (per cropped hectare) is also closely
linked to communal livestock pressure (c.c.:
.31; prob.: .00). Thus these variables related to
internal livestock pressure cannot be included
in the same model.

* Availability offamily labor (FAMLAB): It is
expected that the availability of family labor
will influence the probability of adopting the
no-tillage component. When more family
labor is available, there is less need to adopt
labor-saving technologies, but as we have seen
earlier, availability of family labor is closely
linked to outmigration (c.c.: .36; prob.: .00). In
other words, there are opportunities outside
the farm to use the labor that has been saved.
This raises the opportunity cost of family
labor, which can encourage the decision to
replace labor with herbicides for weeding. On
the other hand, it is expected that










outmigration will raise cash resources:
through seasonal outmigration for off-farm
work, the family can sell some of its labor for
cash, thereby making it easier to buy the
required inputs.
* Monocropped maize (MONMAIZE): Maize
monocropping should have a positive
influence on the probability that farmers will
adopt the no-tillage component.
Monocropping is related closely to the
production of maize for sale, which in turn is
related to cash flow on the farm. If maize is
produced chiefly for sale, more maize will be
sold at harvest time, and it will be easier for
the farmer to purchase inputs with cash. On
the other hand, if maize is produced entirely
for home consumption, the farmer must
obtain cash from other sources to buy inputs.
Furthermore, monocropping makes it easier
to apply herbicide, as a single crop of maize
requires less care compared to the maize-
bean intercrop.
* Nonagricultural business (BUSINESS): Side
businesses (that is, businesses aside from
agricultural production) are expected to have
a positive influence on the adoption of the
no-tillage component. This is related once
again to cash flow on the farm, given that


someone with a side business is likely to
have more cash on hand. The presence of a
business is also likely to raise the
opportunity cost of family labor.

All of the variables that raise cash flow on the
farm can positively influence adoption of the
mulch component. As seen earlier, one of the
most effective means of conserving crop
residues is to enclose the field. Nevertheless,
this represents a considerable investment,
which might be easier to make if the farmer has
more access to cash (for example, through
outmigration or grain sales). However, up to
the time of this study, few farmers had fenced
their fields, so the effect of fencing is unlikely to
emerge clearly from the analysis.

Results Table 17 presents results of the
model. Most of the relevant variables for each
component of the conservation tillage
technology have the expected signs and are
statistically significant. Several statistical tests
indicate the goodness of fit of the model.
However, it is important to point out that very
few farmers adopted only the mulch
component, which weakens this part of the
model. On the other hand, it is expected that


Table 17. Factors affecting adoption of conservation tillage in Motozintla, Chiapas
(multivariate logistic model, normalized on nonadoption)
Adoption
Variable Mulch only No-tillage only Both components
SLOPE 0.0793 (.0276) *** 0.0108 (0.0183) 0.0687 (0.0232) ***
COMLIVE -0.773 (1.67) -0.997 (0.961) -3.62 (1.23) ***
FSIZE 0.627 (0.548) 0.995 (0.388) ** 1.48 (0.448) ***
FAMLAB 0.546 (0.530) 0.344 (0.374) 0.885 (0.432) **
DMONMAIZE 1.53 (1.12) 0.0614 (0.812) 1.48 (0.958)
DBUSINESS -0.183 (1.55) 2.17 (0.889) ** 1.11 (1.17)
Constant -9.56 -3.11 -8.27
Sample size 82
X2 for importance of education 63.5 degrees freedom: 15 prob.:.000
Cases predicted correctly 62%
Note: Values in parenthesis indicate asymptotic standard errors; ***, **, and indicate significance at 1%, 5%, and 10%,
respectively.











the differences among those who adopted only
the no-tillage component and those who
adopted both components were principally
related to the adoption of the mulch
component.

The model presented in Table 17 includes the
postulated variables. This model is slightly
better than others that did not include farm
size, for example, and in its place includes
variables closely related to farm size (for
example, hiring of day laborers, credit, and
extension).

Slope appears to be the variable that best
explains the adoption of the mulch component
alone. Its coefficient is very significant and of
the expected sign. Slope is also significant in
explaining total adoption. Aside from slope,
communal livestock pressure is also significant
in explaining total adoption of the technology
(especially the mulch component).

The main variables that explain adoption of
only the reduced tillage component include
farm size and having a side business. Both
variables are significant and have the expected


signs. However, only farm size is important in
total adoption. Along with the variables that
have been mentioned, the availability of family
labor through its effect on the adoption of the
reduced tillage component also helps explain
total adoption.

Table 18 presents the probability that a typical
farmer will adopt each component alone or
together, along with some farmer variations. The
probabilities for the typical farmer are based on
the average value of the continuous variables
and the most common value of the discrete
variables. Based on these values, the typical
farmer has a greater probability of adopting only
the reduced tillage component (which is also the
most common group of adopters in the sample,
at 37%). A substantial change in the slope of field
cultivated by the typical producer has a
particularly strong influence on adoption of the
mulch component. A farmer who is essentially
typical, except for the extreme slope of his field
(90%), has a greater likelihood of adopting both
components of the technology.

The table also presents the probabilities of
adoption based on other variations in


Table 18. Probabilities of adoption of conservation tillage for different groups of farmers,
Motozintla, Chiapas
Adoption
Variable Value Mulch only No-tillage only Both components
Average or most
Typical farmer common value 23 % 61 % 40 %
A slope 50% 5% 55% 14%
90% 57 % 66 % 71 %
A communal Tuixcum 17% 49 % 11 %
livestock pressure Carrizal 27 % 67 % 64 %
fenced 42 % 83 % 98 %
A total farm size 1 ha 9% 21% 5%
5 ha 55% 94% 95%
A labor 1 person 8% 41% 8%
6 people 56% 79% 87%
A side business Yes 20 % 93 % 67 %
a A typical farmer has a field with a slope of 71%, communal livestock pressure of 1.13 AU/ha, a total farm size of 2.76
ha, potential family labor of 3.38 persons, does not monocrop maize, and has no business other than farming.










characteristics of the typical farmer. When the
level of communal grazing reflects the level
inherent in the ejido of Tuixcum, the typical
farmer is likely to adopt neither component of
the technology, whereas the same farmer in
Carrizal is likely to adopt only the reduced
tillage component. However, if the farmer
encloses his field he will probably adopt both
components.

Farm size strongly influences the adoption of
the reduced tillage component. A typical farmer
with only one hectare of land will probably
adopt neither technology component, but a
farmer with 5 ha is likely to adopt both
components. The results for family labor are
similar. Having a side business considerably
increases the probability that the farmer will
adopt only the reduced tillage component.

Farmers' Opinions of
Conservation Tillage
Farmers were asked their opinions about
various statements reflecting the potential
advantageous and disadvantageous effects of


100 1


the conservation tillage technology (specifically,
leaving residues on the field) (summarized in
Figures 10 and 11).

The great majority of farmers (93%) thought
that conservation tillage practices did not
increase soil compaction, probably because
compaction is not a serious problem in these
unmechanized hillside production systems.
Most farmers (85%) also thought that
conservation tillage practices did not increase
the number of days needed for land preparation
and planting. Apparently the presence of the
residues does not impede these operations.

According to the majority (93%) of sample
farmers, conservation tillage practices do not
increase weed problems. This is contrary to
what one would expect because a greater
incidence of perennial weeds, which are hard to
control, would be likely under conservation
tillage. However, informal information
provided by survey farmers seems to link the
presence of some problem weeds (such as
Cynodon dactylon) to herbicide use.


Increases Increases Increases Increases costs Increases Increases
soil labor for land weeds of weed control soil pests lodging
compaction preparation/sowing
Figure 10. Farmers' opinions of potential disadvantages of conservation tillage,
Motozintla, Chiapas.










Opinion was divided with respect to whether
conservation tillage increased the cost of weed
control. Half of the farmers (49%) though that it
did, whereas the remainder thought it did not.
Economic data presented earlier indicate that
the cost of weed control was in fact reduced.
The fact that half of the farmers did not
perceive this reduction is probably related to
the "visibility" of cash costs for herbicides
compared to costs in kind for family labor.
Farmers' evaluation also probably did not
include the costs/health risks of using
herbicides, which are invisible costs.

Opinion was also divided with respect to
whether conservation tillage practices
increased soil diseases. More than half (51%) of
farmers thought that they did; one-third (29%)
thought that they did not; the remainder
thought that they made no difference. With
mulch, limited tillage, and no more burning of
residues, a greater incidence of soil diseases
might be expected. However, disease pressure
apparently continues to be low and is less of a
problem than the strong, crop-damaging
winds. The wind problem and resulting


lodging of maize plants do not seem to be
worse under conservation tillage. Nearly all
farmers claimed either that the problem had
not increased or that it had stayed the same.

With regard to the advantages of conservation
tillage, farmers unanimously said that
conservation tillage reduced soil erosion. Their
response is probably related to the fact that the
reduction in soil erosion is highly visible when
mulch is used on steeply sloping fields.
Greater amounts of mulch than those
presently used by farmers in the study area
could reduce erosion even more.

The great majority of producers (92%)
observed that conservation tillage practices
helped the soil retain more moisture. Like the
reduction in soil erosion, this effect is also
highly visible. Many farmers have observed
that the soil under the mulch is generally
damper than exposed soil.

Most farmers (84%) thought that conservation
tillage practices increased soil fertility. The
source of this perception is probably farmers'


Reduces Better soil
erosion moisture retention
Figure 11. Farmers' opinions of the potential
Motozintla, Chiapas.


Increases Increases
soil fertility crop yields
advantages of conservation tillage,










awareness that crop residues turn into soil. It is
important to recall that the residues have a high
C:N ratio and that their effect as an organic
fertilizer is limited in terms of a greater
availability of nutrients in the short run. Their
effect on soil fertility is related more to the
dynamics of organic matter and can be
considered a long-term effect.

Opinion is divided over whether conservation
tillage practices increase maize yields. Less than
half of the farmers surveyed (44%) thought that
they did, one-fourth (23%) thought that they did
not, and one-third (32%) thought that yields
remained unchanged. This agrees with observed
maize yields; in the 1993 summer cycle, there
were no significant differences among groups of
adopters. However, there is a difference in
relation to the adoption of the mulch
component. Those who adopted the mulch
component, alone or together with the tillage
component, had more favorable opinions about
whether conservation tillage increased yields
(for example, 55% of adopters thought that yield
improved, versus only 38% of nonadopters).


Summary and Conclusions

The adoption of conservation tillage practices in
the unmechanized hillside production systems
in the Motozintla area appears relatively
promising. Farmers no longer burn crop
residues, and 66% of the survey farmers have
adopted the reduced tillage component as well.
However, only 38% of farmers leave enough
residue on their fields to produce an effective
mulch. Thus only 29% of the farmers surveyed
can be considered true adopters of conservation
tillage.

To benefit from conservation tillage, farmers
must adopt both components of the technology:
limited tillage and the use of mulch. The results


obtained by the different groups of farmers
provide evidence of this. Farmers who adopted
both components of conservation tillage in the
maize-bean intercropping system that
predominates in the study area obtained more
favorable yields and farm budgets. For various
economic indicators, adopters of both
components surpass both nonadopters and
partial adopters of the technology. The value
added by adopters of both components
surpassed that of other groups of adopters by
approximately 11-16%, reaching Mx$ 1,770/ha
for the 1993 summer cycle. This implies a net
benefit of Mx$ 470/ha for these farmers.
Production costs are relatively high for all
producers, reaching Mx$ 0.60 per kilogram of
maize for adopters of both components. This
implies that the profit of the maize-bean
intercrop principally derives from the bean
intercrop. Returns from maize cropping are
probably low because maize is mainly
produced for home consumption and only the
surplus is sold. Finally, labor productivity
reached Mx$ 16/day for adopters of both
components. All of these indicators were less
attractive for the other groups of adopters. The
clear advantages of adopting both components
only disappear when higher interest rates are
assumed.

It is important to emphasize that the 1993
summer season was apparently a good year for
producing maize and beans. In normal years,
and especially in poor ones, the economic
indicators are much less favorable for all
farmers. However, there are indications that
adopters of the mulch component are less
exposed to production risks and that they could
benefit from a favorable yield response to levels
of mulch higher than the conventional
threshold level of 2 t/ha.

The adoption of conservation tillage practices in
the study zone particularly influenced weed










control and residue management practices.
Adoption of the reduced tillage component
implied a greater use of herbicides and
generated a slight reduction in production costs.
Adoption of the mulch component is more
complex. The farmer must make a substantial
investment to protect the crop residues, and this
investment is recovered only through the yield
increases obtained when the farmer adopts both
components of the technology.

There have been no great qualitative or
quantitative changes in other production
practices used in maize-bean intercropping.
Farmers still use local varieties, use reduced
tillage for land preparation, apply mostly
nitrogenous fertilizer, and generally use no pest
and disease control measures. Wind continues
to be one of the principal natural problems
affecting the system, partly because fields are
located on slopes that are quite exposed to the
elements.

This study demonstrates that patterns of
adoption were different for the different
components of conservation tillage technology.
Compared to the adoption of herbicides,
adoption of the practice of conserving rather
than burning crop residues took longer,
although the adoption process began earlier. A
local law prohibiting burning encouraged
adoption of residue conservation in the study
area; by the time a similar law was promulgated
at the state level, it produced few changes in the
study area. The study also shows that the
adoption of the no-burning practice is a
necessary but not sufficient condition for
adoption of the mulch component. Burning is
only one of the factors influencing the
availability of residues for mulch.

The study reviewed various farmer
characteristics that explained the differential
adoption of the two conservation tillage


components and used a multivariate logistic
model to analyze how the set of variables
affected adoption of the different technology
components. Adoption of the mulch component
can largely be explained by the slope of the
field, which determines the access of livestock
to the field in the dry season. In turn, adoption
of the no-tillage component can be explained by
the availability of cash (which is related to
ownership of a side business) and farm size.
There are various possible explanations for the
relation between adoption of the no-tillage
component and farm size, such as the amount of
labor required for maize-bean intercropping
and the amount of income (being able to sell a
greater share of the maize production, for
example).

The adoption of conservation tillage is the result
of the simultaneous adoption of both
components of the technology. For this reason,
factors that help explain the adoption of one
component also help explain adoption of the
entire technology. However, aside from the
variables we have already mentioned,
communal livestock pressure had a significant
effect on total adoption (in relation to the mulch
component), as did the availability of family
labor. Labor availability is closely related to
seasonal outmigration by some member of the
family, which in turn facilitates adoption of the
reduced tillage component.

Forage demand from the farmer's own livestock
does not seem to limit adoption of the mulch
component. However, livestock belonging to
other people can limit the possibility that there
will be sufficient residue left on the field to form
an effective mulch. This considerably increases
the cost of adoption because crop residues must
be protected in some way. Enclosing the field
seems to be the most economically attractive
means, aside from being the most effective, of
achieving this goal. However, the resources










needed to fence a field can be a limitation,
given the size of the investment required. The
state law limiting the cutting of trees does not
help in this respect.

The current study summarizes some of the
factors that help explain the diffusion of
conservation tillage practices on the hillside
farms of Motozintla. It is important to
emphasize that one of the factors that
stimulated adoption was the state agricultural
policy, particularly the policy of the
development district. There is no doubt that the
promotion and distribution of incentives, in
combination with the local law against
burning, facilitated the adoption of these
practices. However, results of the present study
show once again that these efforts are much
more effective when they fit with actual
production systems and farmers' priorities. The
new technologies proposed, such as
conservation tillage, are viable only if they
bring benefits to the producer without an
excessive increase in costs. To determine those
costs, it is essential to account for the true
opportunity costs confronted by these
producers with their limited resources.

The fact that the components promoted in the
study area fit well with current production
systems raises a question about the nature of
adoption (Tripp, pers. comm.). Adoption of the
practices that were promoted in the study area
supposedly reflects a new understanding on
the part of farmers and its subsequent
application. However, some of the technology
adoption observed in the study area is related
to changes in production practices that
originated in another way. The adoption of
herbicides may have occurred even without an
extension program, simply because farmers
needed to free up labor to perform more
remunerative work outside of the study zone.
The adoption of mulch on steeply sloping fields
seems to be related mainly to the practice of


free grazing and to farmers' decision to stop
burning crop residues. This more or less
spontaneous adoption process differs from the
adoption decision of farmers who purposely
fenced their fields to conserve crop residues.

Regardless of the process by which farmers
came to adopt the technology, the proposed
changes were relatively easy to adopt. Most
practices, such as sowing cajeteada, did not need
to be adapted for farmers to start using the
components of conservation tillage technology.
Nor did adoption of the mulch component
generate conflicting demands for use of the
residues, whereas it did offer the potential for
increasing yields and reducing production
risks. In fact, the results suggest a positive
relationship between yield and the amount of
mulch left on the field even higher than the
threshold of 2 t/ha proposed. The data also
indicate that it was the relatively better-off
farmers and commercial farmers who adopted
the components that were promoted.

The study illustrates that the law against
burning truly helped promote the residue
conservation practice, chiefly because the law
was applied at the local level. Clearly laws
related to the management of natural resources
will succeed only if they are actually enforced.
On the other hand, massive government
campaigns that promote soil conservation
practices are only successful if there is sufficient
follow-up and if the practices suit the needs of
farmers.

The use of conservation tillage practices in the
Motozintla area shows what can happen when
farmers with few resources adopt several
external inputs (fertilizers and herbicides) but
fail to adopt improved varieties at the same
time. Because farmers still use local varieties,
the productivity of the system has remained
relatively low despite the use of conservation
tillage practices, and potential cash flow












problems remain. However, the production
potential in the area is high, and it can be
realized more adequately through the
introduction of improved varieties adapted
to the zone. In addition to improving the
productivity of the system, the use of
improved varieties could also have the
positive effect of increasing the availability
of residues for forage or mulch.



References

Byerlee, D., and E. Hesse de Polanco. 1986.
Farmers' stepwise adoption of technological
packages: Evidence from the Mexican
Altiplano. American Journal of Agricultural
Economics 68.
Cadena, P. 1995. Del azad6n a la labranza de
conservaci6n: La adopci6n de la labranza de
conservaci6n en dos comunidades de la Sierra
Madre de Chiapas. Montecillos, Mexico:
Colegio de Postgraduados.
CIMMYT. 1988. From Agronomic Data to Farmer
Recommendations: An Economics Training
Manual. Mexico, D.F.: CIMMYT.
CIMMYT. 1993. The Adoption of Agricultural
T. -1I .l. 1.. A Guide for Survey Design.
Mexico, D.F.: CIMMYT.
CTIC. 1994. Definition Conservation Tillage.
Conservation Impact 12(6).
Erenstein, 0. 1996. ZLabranza de conservaci6n o
conservaci6n de residues? Una evaluaci6n del
manejo de los residues en Mexico. En Memorias:
Labranza de conservaci6n 4oforo international
96. Mexico, D.F: FIRA. Pp. 105-14.
Euroconsult. 1989. Agricultural Compendium: For
Rural Development in the Tropics and
Subtropics. Amsterdam, the Netherlands:
Elsevier.
Gittinger, J.P. 1982. Economic Analysis of
Agricultural Projects. 2nd edition. Baltimore,
Maryland: John Hopkins.
Hudson, N. 1995. Soil Conservation. 3rd edition
Ames, Iowa: Iowa State University.
INEGI. 1994. Resultados Definitivos VII Censo
Agricola Ganadero Chiapas. Aguascalientes,
Mexico: INEGI.
Lal, R. 1989. Conservation tillage for sustainable
agriculture: Tropic versus temperate
environments. Advances in A. ..,,..,..
42: 85-197.


Larios L., R. Gordon, R. Obando, M. Osorio, G. L6pez, and J.
Bolafos. 1997. Eficiencia de uso de nitr6geno en el cultivo
de mafz. Sintesis de resultados experimentales del PRM
1993-95, Vol. 5. Guatemala, Guatemala: PRM.
Makeham, J.P., and L.R. Malcolm, 1986. The Economics of
Tropical Farm Management. Cambridge, U.K.:
Cambridge University Press.
Nagy, J.G., and Z. Ahmad. 1993. Adoption and Dir-l. of
Agricultural Interventions: A Primer on the Use of
Statistical and Qualitative Response Model Estimation.
Islamabad, Pakistan: Pakistan Agricultural Research
Centre (PARC).
Nieuwkoop, M. van., W. L6pez Baez, A. Zamarripa Moran,
R. de la Piedra C., F.J. Cruz C., R. Camas G., and J.
L6pez M. 1994. La adopci6n de las tecnologfas de labranza
de conservaci6n en La Fraylesca, Chiapas. Mexico, D.F.:
INIFAP-CIMMYT.
Ortega, A. 1989. Importancia relative de plagas en labranza
traditional y de conservaci6n: Una revision de la literature
In H. Barreto, R. Raab, A. Tasistro, and A.D. Violic
(eds.), Labranza de conservaci6n en mafz. Mexico, D.F.:
CIMMYT. Pp. 91-118.
Sandoval, J.A. 1994. Regulaci6n de las quemas agropecuarias
en Chiapas y labranza de conservaci6n. En SARH-
INIFAP-INI-UNAM-SEDAP-CIMMYT-SEDESOL
(eds.), Taller sobre las polfticas para una agriculture
sustentable en la sierra de los Tuxtlas y Santa Marta,
Veracruz. Mexico, D.F.: CIMMYT. Pp. 53-69.
Shaxson, T.F., N.W. Hudson, D.W. Sanders, E. Roose, and
W.C. Moldenhauer, 1989. Land Husbandry: A Framework
for Soil and Water Conservation. Ankeny, Iowa: Soil and
Water Conservation Society.
SPP (Secretarfa de Programaci6n y Presupuesto). 1981.
Mapas de Precipitaci6n total annual y climas.
Villahermosa 1:1,000,000. Mexico, D.F.: Direcci6n
General de Geograffa del Territorio Nacional,
Secretarfa de Programaci6n y Presupuesto.
Sain, G., and F. Herrera 1996. Adopci6n y uso de semilla
mejorada de mafz entire pequenos agricultores de Guatemala.
Unpublished.
Tasistro, A. 1989. Factores que afectan a los herbicidas. In H.
Barreto, R. Raab, A. Tasistro, and A.D. Violic (eds.),
Labranza de conservaci6n en mafz. Mexico, D.F.:
CIMMYT. Pp. 133-44.
Tripp, R., and H. Barreto, 1993. Estimaci6n aproximada de la
cantidad de rastrojo de mafz sobre el suelo.
Unpublished training material. Mexico, D.F.: CIMMYT.
Turner, R.K., D. Pearce, and I. Bateman, 1993. Environmental
Economics: An elementary Introduction. Baltimore,
Maryland: Johns Hopkins.
Zea, J.L., M. Osorio, and J. Bolafos. 1997. Uso de rastrojo de
mafz como cobertura superficial y sus implicaciones en la
economfa del nitr6geno en el cultivo de mafz. Sintesis de
resultados experimentales del PRM 1993-95, Vol. 5.
Guatemala City, Guatemala: PRM.










Appendix A. Visual Aids Used in the Farmer Survey


Diagram with degrees of incline to estimate slope (CIMMYT, unpublished material)


0 10












Level of residue coverage:
The photos represent an area of approximately 2 m x 2 m, with dry stover distributed
uniformly on the soil surface. The amount in each photo represents the higher limit of the
range for each level.


Level 1
0-0.5 t/ha
Approximate coverage (10-20%)


Level 3
2-3 t/ha
Approximate coverage (30-60%)


Level 2
1-2 t/ha
Approximate coverage (20-30%)


Level 4
4-5 t/ha
Approximate coverage (>60%)


Source: Tripp and Baretto (1993) and unpublished training documents, CIMMYT.



















































































7





ft--


cn


a>

po
0)
-c



C
0
La
E
lU











Appendix B. Field Prices and Local Units of Measure


Table B1. Field prices, Motozintla, Chiapas

Type Unit Farm cost Specification

Labor

Any task day 8 Food included.
day 10 Food not included.
Harvest costal 1.5-3 Includes transport.
Animal transport carga 5

Seed Local price for seed of local varieties.

Maize kg 1
Beans kg 4

Fertilizer Commercial price plus transport. Transport price
calculated as Mx$ 3/bulto (50 kg).a

Ammonium sulfate (21-0-0 NPK) 50 kg 28 Includes transport.
Urea (46-0-0 NPK) 50 kg 38 Includes transport.
Triple 17 (17-17-17 NPK) 50 kg 48 Includes transport.
Diammonium phosphate (18-46-0) 50 kg 45 Includes transport.

Chemicals Commercial price plus transport. Transport price
calculated at Mx$ 1.0/1.b

Gramoxone 1 21
Azinotox 22.5

Products Cost of transport per costal of grain is calculated to
equal that of a costal of fertilizer (see above).

Yellow maize kg 0.60 Commercial price (Mx$ 0.63/kg), minus the cost of
shelling, minus the cost of transport. Shelling cost is
based on 200 kg/day (Mx$ 0.05/kg). Transport cost
calculated as the difference between the amount
of maize sold and the amount retained for
consumption on the farm, multiplied by Mx$ 0.06/kg.
Beans kg 1.62 Commercial price (Mx$ 1.60/kg) minus the cost of
transport. Transport cost calculated as the difference
between the amount of beans sold and the amount
retained for consumption on the farm, multiplied by
Mx$ 0.60/kg.

Note: Local units: almud = 7-8 kg maize; cuerda = area 20 m x 20 m = 0.04 ha; costal = 75 kg of maize ears = 50 kg
maize grain; and a carga = 2 costales.
a Transport price calculated as cost of transporting fertilizer from input supply center to the ejido (Mx$ 2.5/bulto), plus the
cost of transport for the farmer (Mx$ 0.5/bulto, assuming a round trip at Mx$ 5.0 to buy 10 bultos).
b Transport price calculated as the cost of transport for the farmer (Mx$ 1.0/1, assuming a round trip at Mx$ 5.0 to buy 5 I).












Appendix C. Sensitivity Analysis of Farm Budgets

Table C1. Sensitivity analysis of farm budgets, hillside maize-bean intercropping system, Motozintla, Chiapas

Units Nonadopter Mulch only No-tillage only Both components

Sensitivity to cost of capital
Capital: 1%/mo.
Total gross benefit Mx$/ha 1,953 1,941 1,971 2,210
Total inputs Mx$/ha 327 338 411 408
Total labor Mx$/ha 933 934 777 796
Total variable costs Mx$/ha 1,260 1,272 1,188 1,204
Total fixed costs Mx$/ha 299 426 299 426
Return
Value added Mx$/ha 1,626 1,603 1,559 1,802
Net benefit Mx$/ha 394 243 484 579
Costs per kg maize Mx$/kg 0.587 0.628 0.573 0.559
Labor productivity kg maize/day 28.0 28.5 32.8 36.1
Return per day Mx$/day 14.2 12.6 16.2 17.3
Capital: 5%/mo.
Total gross benefit Mx$/ha 1,953 1,941 1,971 2,210
Total inputs Mx$/ha 389 402 489 486
Total labor Mx$/ha 933 934 777 796
Total variable costs Mx$/ha 1,322 1,336 1,266 1,282
Total fixed costs Mx$/ha 354 651 354 651
Return
Value added Mx$/ha 1,564 1,539 1,481 1,724
Net benefit Mx$/ha 277 (47) 350 277
Costs per kg maize Mx$/kg 0.632 0.737 0.625 0.664
Labor productivity kg maize/day 28.0 28.5 32.8 36.1
Return per day Mx$/day 13.0 9.5 14.5 13.5

Sensitivity to yield
Yield: good year
Total gross benefit Mx$/ha 2,251 2,364 2,279 2,760
Total inputs Mx$/ha 351 362 441 437
Total labor Mx$/ha 933 934 777 796
Total variable costs Mx$/ha 1,284 1,296 1,218 1,233
Total fixed costs Mx$/ha 319 511 319 511
Return
Value added Mx$/ha 1,901 2,002 1,838 2,323
Net benefit Mx$/ha 649 558 742 1,016
Costs per kg maize Mx$/kg 0.542 0.546 0.512 0.469
Labor productivity kg maize/day 31.2 34.9 38.0 46.0
Return per day Mx$/day 17.0 16.0 19.5 22.8
Yield: normal year
Total gross benefit Mx$/ha 1,609 1,668 1,562 1,968
Total inputs Mx$/ha 351 362 441 437
Total labor Mx$/ha 933 934 777 796
Total variable costs Mx$/ha 1,284 1,296 1,218 1,233
Total fixed costs Mx$/ha 319 511 319 511
Return
Value added Mx$/ha 1,258 1,306 1,121 1,531
Net benefit Mx$/ha 6 (138) 25 224
Costs per kg maize Mx$/kg 0.743 0.751 0.733 0.658
Labor productivity kg maize/day 22.7 25.4 26.5 32.8
Return per day Mx$/day 10.1 8.5 10.3 12.8
Yield: poor year
Total gross benefit Mx$/ha 1,008 1,005 1,011 1,266
Total inputs Mx$/ha 351 362 441 437
Total labor Mx$/ha 933 934 777 796
Total variable costs Mx$/ha 1,284 1,296 1,218 1,233
Total fixed costs Mx$/ha 319 511 319 511
Return
Value added Mx$/ha 658 643 571 829
Net benefit Mx$/ha (594) (802) (526) (478)
Costs per kg maize Mx$/kg 1.126 1.156 1.094 1.004
Labor productivity kg maize/day 15.0 16.5 17.8 21.5
Return per day Mx$/day 3.6 1.4 3.2 4.0











New Papers from the Natural Resources Group


Paper Series
95-01 Land and Livelihoods: Patterns of Rural Development in Atlantic Honduras
D. Buckles and G. Sain

96-01 Meeting South Asia's Future Food Requirements from Rice-Wheat Cropping Systems: Priority Issues
Facing Researchers in the Post-Green Revolution Era
P. Hobbs and M. Morris

96-02 Soil Fertility Management Research for the Maize Cropping Systems of Smallholders in Southern
Africa: A Review
J.D.T. Kumwenda, S.R. Waddington, S.S. Snapp, R.B. Jones, and M.J. Blackie

96-03 Genetic Diversity and Maize Seed Management in a Traditional Mexican Community: Implications
for In Situ Conservation of Maize
D. Louette and M. Smale

96-04 Indicators of Wheat Genetic Diversity and Germplasm Use in the People's Republic of China
N. Yang and M. Smale

96-05 Low Use of Fertilizers and Low Productivity in Sub-Saharan Africa
W. Mwangi

96-06 Intensificaci6n de sistemas de agriculture tropical mediante leguminosas de cobertura: Un marco
conceptual
D. Buckles and H. Barreto

96-07 Intensifying Maize-based Cropping Systems in the Sierra de Santa Marta, Veracruz
D. Buckles and 0. Erenstein

96-08 In Situ Conservation of Crops and Their Relatives: A Review of Current Status and Prospects for
Wheat and Maize
G. J. Dempsey

97-01 The Adoption of Conservation Tillage in a Hillside Maize Production System in Motozintla, Chiapas
O. Erenstein and P. Cadena I 15/..:

97-01Es La adopci6n de la labranza de conservaci6n en un sistema de cultivo en ladera en Motozintla, Chiapas
0. Erenstein y P. Cadena Iit ,.:

97-02 Farmer Assessment of Velvetbean as a Green Manure in Veracruz, Mexico: Experimentation and
Expected Profits
M. J. Soule


Reprint Series
96-01 Evaluating the Potential of Conservation Tillage in Maize-based Farming Systems in the Mexican
Tropics
O. Erenstein




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