Citation
Annual report

Material Information

Title:
Annual report
Series Title:
Annual report. 1985.
Creator:
Soil Management CRSP. TropSoils. Humid Tropics: Indonesia.
Publisher:
University of Hawaii, Dept. of Agronomy and Soil Science
Publication Date:
Language:
English

Subjects

Subjects / Keywords:
Farming ( LCSH )
Agriculture ( LCSH )
Farm life ( LCSH )
Genre:
serial ( sobekcm )

Notes

Funding:
Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.

Record Information

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

Full Text
Soil Management CRSP
TROPSOILS
Humid Tropics: Indonesia
Annual Report
FY 1985
Submitted by
Dept. of Agronomy and Soil Science
University of Hawaii




Soil Management CRSP
TROPSOILS
Humid Tropics: Indonesia
Annual Report
FY 1985
Submitted by
Dept. of Agronomy and Soil Science
University of Hawaii




3
TABLE OF CONTENTS
I. Introduction . . . . . . . . . . . . . . 5
II. Contribution of TropSoils Research to Hawaii Agriculture . . . . 8
III. Annual Progress Reports. ................ ......11
A. Site Characterization
Soil Variability in Mechanically Cleared Forest Land ........ .13
B. Soil Management and People
1. Interpreting Existing Soil Management Systems
a. Indigenous Knowledge Systems Related to Soil Management .. o40 b. Intra-household Decision Making o . 0 . . . . 0 0 .49
c. Economic Evaluation of Soil Management Technology . . 55 d. Time Allocation Study . . . . . ......... .58
e. Nutrition/Income Survey .. . .. .. .. .. .. .. ..73
2. Modification of Soil Management Systems
a. Collaborative Research with Farmers on Upland Fields. . .77
b. Land Management with Farmers in Small Catchment Areas . .85
c. Collaborative Research with Farmers on Home Gardens
1) Effects of Different Sources of Waste Material on Soil
Properties and Horticultural Crops in Home Gardens . . .88
2) Pasture Trials 91
3) Nutrient Component . . . . . . . . . . 94
4) Fish Pond Component . . . . . . . . 101
d. Farmers' Perceptions of Constraints to Agricultural
Production . 104
e. Minang Tree Farming Study . . . . . . . 109
C. Improving and Sustaining Productivity in Farmers' Fields
l. Matching Crop Requirements of Rice, Maize, Soybean, and
Peanut to Soil Characteristics with Crop Simulation Models 113
2. Assessing Field Inoculation with Introduced and Indigeous VA
Mycorrhizal Inocula by Crop Growth and Yield on Soil Cleared of
Tropical Rain Forest . . . . . . 0 . .. . 127
3. Modeling Phosphorus and Lime Interactions. .-. . s . . 137
D. Land Reclamation: Soil Physics and Soil Conservation
1. The Role of Soil Physical Properties on Land Quality for
Sustained Agricultural Production in Land Cleared by
Heavy Equipment ..................0 .....191
2. Pasture Grass and Legumes for the Humid Tropics. . . . . 194
3. Management of Organic Material in Indonesian
Farming Systems *... . . . . .. .* . . 203
4. Conservation-Effective Rainfed Farming Systems for
the Sitiung Area, West Sumatra . ... . 218
5. Expert Systems to Develop and Transfer
Soil Management Research . .. . . . .. . .. 223







5
INTRODUCTION
The principal goal of the TropSoils program is to uncover principles which will enable resource-poor farmers to adopt soil management practices that will increase family income and farm productivity and, at the same time preserve land quality. The research strategy is designed to ensure that social, cultural, economic, and environmental factors that enhance adoption of a soil management innovation are made an integral part of the research plan. To achieve its goal, the project conducts a major portion of the soil management research with farmers and in farmers' fields using systems-based research and crop simulation models.
The Setting
A 100,000 hectare transmigration site in Sitiung, West Sumatra,
Indonesia serves as the project's research area. Six thousand transmigrant families and 1500 indigenous families live in the area. Large cultural and language differences between the Javanese and Sudanese transmigrants, and between the transmigrants and indigenous groups present unparalleled opportunities to study the responses of different ethnic groups to soil management innovations.
The soils of the region range in quality from moderately fertile
Inceptisols on river terraces to highly leached and impoverished Oxisols and Ultisols of the dissected peneplain. Mean annual rainfall is 2800 mm and mean annual air temperature is 26@C. The tropical rain forest is gradually giving way to rubber plantations and subsistence farming by new settlers.
The first large group of transmigrants settled in Sitiung in 1976. A modest home, 1.25 hectare of recently cleared land and a year's supply of food, fuel, other living essentials, seed and fertilizer awaited each family upon its arrival. Since then, five additional areas in Sitiung have been settled. Bulldozer crews continue to clear more land to accommodate new settlers from the densely populated islands of Java and Bali. The productive land on the river terraces has long been settled and the newest transmigrants are being placed on the less-desirable lands of the dissected peneplain.
Developing a Collaborative Project
A major aim of the Collaborative Research Support Program is to produce research outputs that are relevant to both the United States and the collaborating less-developed countries. To achieve its aim, the Indonesia TropSoils program is developing soil-crop-climate simulation models that predict effects of soil management on crop performance in a wide range of agroenvironments. The program is also developing expert systems that mimic human experts. These expert systems are able to represent scientific knowledge including "local" information known to farmers. An expert system on soil acidity developed under the Indonesian TropSoils program has been designed to make liming recommendations for several important food crops produced on all major soils of the humid tropics. The utility of crop models and expert systems are universal and can deal with problems that are common to the United States and developing countries.
Training:
Although Indonesia has quality soil scientists, it lacks the quantity needed to meet the nation's development needs. The Indonesia TropSoils




6
program is dealing with this issue through a combination of on-site training of local staff, degree training in U.S. institutions, and by developing methods to increase research effectiveness. It is in the latter area that the TropSoils program can make its greatest contribution. Indonesia not only lacks the number of scientists to conduct development research, but also lacks the financial support to give to scientist to do the work.
Between June 26 and July 5, 1985, the International Fertilizer
Development Center, in collaboration with the Centre for Soil Research and the University of Hawaii, conducted a training workshop on crop modeling in Bogor, Indonesia. The purpose of the workshop was to train Indonesian scientists to use crop models to simulate and predict outcomes of routine field research which now consumes much of researchers' time. The workshop participants were trained to use crop models to simulate the effect of nitrogen rates, irrigation frequency, planting dates, and row spacings on maize and rice production for several locations in Indonesia. The simulations were run for 30 years duration to ascertain the number of crop failure due to drought. The participants were taught to use crop models to evaluate land suitability for specific crops and to install a few key field experiments to test and validate the reliability of simulated results. These experiments will be conducted by Indonesian scientists in the coming year.
Factors Affecting Progress of Program: Communication, timely shipment of equipment to the research site,
cutback in host country research budget, finding suitable candidates to fill vacant positions, religious issues, and staff health are the major factors that affect program progress.
Communication. Owing to the isolation of the research site and
distance between homes of team members in the site, communication between Indonesia and Hawaii and among team members has been difficult. To rectify this problem, the team purchased radio telephones to link residences located in the research area. The closest telephone is a few kilometers away from the nearest home so that communication between the U.S. and the research site remains slow.
Timely Shipment of Equipment. The team has encountered difficulty
clearing shipment through Indonesian customs. Arrangements to ship project materials through the USAID Mission will be of great help to the project.
Cutbacks in Host Country Research Budget. Declining oil prices have seriously eroded the Indonesian's capacity to maintain research budgets. Little improvement in the current situation is anticipated. This situation re-emphasizes the need to develop research methods that enable researchers to produce more relevant results faster and with fewer resources. The development of expert systems and crop simulation models offers a way to partly compensate staff shortage and low budgets.
Finding Suitable Candidates for Vacant Positions. The project is
currently looking for a soil conservation specialist and a soil physicist. Our Indonesian counterparts have indicated a desire to restrict applicants to Americans. Unfortunately, most of the qualified applicants are




7
non-Americans. A soil physicist from India with excellent credentials who is a candidate for American citizenship is applying for one of the positions. The Indonesians have been asked to consider this individual.
Religion. A problem which was thought solved has reappeared in the form of a letter from the District Governor asking that one of the team members leave the district. This is the most serious problem to date and the future of the team member in question remains uncertain at this time.
Staff Health. Heavy workloads and lack of good medical facilities have taken their toll. Last year, one team member returned to the United States for medical examination after a long bout with an unknown ailment. This year, the team leader suffered a shoulder injury. With only three senior scientists on site, health plays a key role in team performance, but the isolation prevents timely corrective action, except for the reliance on annual examinations and bit of good luck.




8
CONTRIBUTION OF TropSoils RESEARCH TO HAWAII AGRICULTURE
How can soil research conducted in an isolated village in the
rain forest of Sumatra be made relevant to Hawaii agriculture? What are the constraints that now prevent us from achieving this aim? What can we do to overcome these constraints, and what benefits accrue from overcoming them?
A major purpose of the Collaborative Research Support Program is to
integrate research authorized under it with other Federal and State programs to maximize contributions to agriculture in the United States and the developing countries. To achieve this purpose, the University of Hawaii has made its TropSoils program an integral part of a larger effort which employs
systems analysis and crop simulation to combine State, Federal, and international projects into a program that produces outputs useful to local, national and international clients. The central concept of systems-based research is that the whole systems must be understood in order to evaluate benefits derived from introducing new crops, products and practices into an existing farming system. Systems based research compels scientists to seek out and understand key processes that regulate systems performance and enables users to predict and control outcomes of farm operations.
Crop Models and Expert Systems
An important ingredient of systems-based research is crop modeling. A crop model, properly formulated, can predict the performance of the crop in question in Hawaii, Sumatra, or within limits, anywhere in the tropics. A corn model obtained from another federally supported project has been tested and validated in Hawaii and Indonesia (see project on "Matching Crop Requirements of Rice, Maize, Soybean, and Peanut to Soil Characteristics with Crop Simulation Models"). The research in Hawaii and Sumatra are coupled to produce a product that performs equally well in Hawaii or Indonesia. It can be assumed that the model will operate equally well elsewhere in the world. The model is able to simulate effects of water stress and nitrogen shortages. It does not now have the capability to simulate phosphorus deficiency or aluminum toxicity. A TropSoils subproject (see project on Modeling Phosphorus and Lime Interactions) has been designed to incorporate phosphorus and lime interactions into crop models. This collaborative research between State, Federal, and host country agencies will be sufficiently general so that the model will be able to simulate water, nitrogen, phosphorus, and lime interactions not only for corn, but eventually for rice, soybean and peanut.
Two constraints prevent ready use of these models. The first deals with the highly spatially variable nature of soils and the highly variable nature of weather and climate. The deterministic crop models predict outcomes at a point in space. To deal with random soil spatial variability, geostatistical techniques developed under a Federally funded regional project has been refined and applied to map soils (see project on "Soil Variability in Mechanically Cleared Forest Land") in Sumatra. The new technique, refined through collaborative research with Indonesian scientists, enables soil properties at unsampled locations to be predicted.
The crop model is able to deal with temporal weather variability, but requires that a minimum data set of daily solar radiation, rainfall and temperature be recorded. A network of weather stations purchased from State and TropSoils funds has been installed in Hawaii and Sumatra to collect the minimum data set to test and validate crop models.




9
A more difficult problem with use of simulation models is the
variations in human preference for crops and the variations in farm resource characteristics. These socioeconomic characteristics are being studied under a group of projects (see projects on farming systems titled "Soil Management and People") dealing with time allocations, nutrition/diets, and incomes. Close collaborative work between faculty from universities in Indonesia, North Carolina State and Hawaii add to our understanding of the way farmers respond to agricultural innovations. But simulation models are not well suited to capture farmer knowledge and "local wisdom."
A decision support system for farmers must make use of (1) models that can simulate soil and weather effects on farm production, and (2) the large amount of local knowledge stored in the minds of farmers. This type of local knowledge can best be captured and represented in expert systems (see project on "Expert Systems to Develop and Transfer Soil Management Research").
A decision support system for Soil Management that includes
geostatistics and crop simulation models to deal with spatial and temporal variability and expert systems to deal with human differences and local knowledge has implications for farmers in Hawaii and Indonesia. The crop models and expert systems tested and developed under the TropSoils Project are sufficiently general to accommodate conditions in Hawaii and Indonesia and forms the basis for Hawaii's continuing involvement with collaborative research.
Farming Systems
A soil-climate project funded by the Hawaii Institute of Tropical Agriculture and Human Resources is applying, in Hawaii, principles and concepts of farming systems designed for the Indonesia TropSoils Project. The purpose of the project is to account for the high variability in crop productivity and quality in a major vegetable growing area of the State. The project hypothesis is that the variability is due to soil and climate variability along the slopes of Mt. Haleakala where vegetables are grown. A network of weather stations has been installed in the area, and key benchmark locations have been characterized for soils variability.
Before installing experiments to test this hypothesis, the Principal investigator arranged with county extension agents for the researchers to meet with farmers to obtain farmer perceptions of problems and priorities. The farmers were quick to indicate that they wanted reliable recommendation on farm production to obtain cabbage of uniform size, shape, and color. They also wanted more information on irrigation scheduling. Farmers, extension agents and researchers agreed that field experiments on cabbage would be conducted in three locations at 300, 700, and 1500 meter elevations. The experiments would assess the effect of planting density on size, shape and color of cabbage for different temperature regimes and seasons. The farmers insisted that the experiment be repeated at least every two months to enable the study to yield information on the effect of planting date on cabbage quality and yield.
The farmers are delighted that a research project designed by them to answer specific questions raised by them is in place. The farm foreman, who is himself a farmer, makes special effort to ensure that the field plots are properly monitored. The farmers are provided with summary reports of project status and many visit the experiments to observe how location, date




of planting, and planting density affect size, shape, and color of cabbage. They are now asking that the project look at a second cultivar.
Researchers involved in this study learned that fertilizer application among farmers varied greatly. This difference was clearly reflected in the soil test. The most variable nutrient was phosphorus which ranged from 20 to 40 ppm by a modified Truog test. Researchers are now beginning to suspect that the critical soil test value for phosphorus may be temperature dependent. Corn grown at three elevations seem to confirm the fact that corn grown in the highest elevation require higher soil test phosphorus.
Research on VA Mycorrhizal indicated that this group of organisms not only affects soil-plant phosphorus relations but the soil-plant-water relationship as well (see project on "Assessing Field Inoculation with Introduced and Indigeous Mycorrhizal Inocula by Crop Growth and Yield on Soil Cleared of Tropical Rain Forest"). Related work on a state-funded project showed that methyl bromide soil fumigation severely reduces yield in mycorrhizal plants, but can significantly increase yield in non-mycorrhizal plants such as cabbage.
By enabling researchers to look at soil management research in the context of the soil-plant-climate system and in the larger context of farming systems, the TropSoils program helps researchers to (1) understand why farmers choose to do things in ways that often contradict scientific knowledge, (2) predict how a particular crop will perform in any location and for any planting date under rainfed or irrigated conditions for different nitrogen rates, and (3) control production outcomes by choosing soil management inputs based on simulated results.




Annual Progress Reports







13
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: 10/1/85 9/30/86
Project Leader/Institution: R. S. Yost, University of Hawaii
Principal Collaborators/Institution:
G. Uehara, University of Hawaii
M. Wade, North Carolina State University
Research Site(s):
Sitiung II, Sitiung V, West Sumatra, Indonesia
Project Title:
Soil Variability in Mechanically Cleared Forest Land
Project Objectives:
i. Construct semi-variograms of soil properties to determine structure in the
variance of soil properties.
2. Use the structure in the variance of soil chemical and physical properties by
geostatistical methods to predict soil properties in unsampled locations.
3. Relate soil and crop variability patterns using geostatistical approaches to
match soil management inputs to spatially variable soils.
4. Analyze soil variability with the purpose of suggesting alternative means of
managing such variability in field research and in farmer's gardens and dry
land fields.
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
American Society of Agronomy meeting, 1984, Las Vegas, Nevada.
Manuscripts in Preparation or Published: (Author(s), title, journal)
Trangmar, B. B., R. S. Yost, and G. Uehara. 1985. Application of
Geostatistics to Spatial Studies of. Soil Properties. Vol. 38. Advances in
Agronomy, Academic Press.
(continued on separate page)
Work Planned for Next Year: (Report on attached form)
Signatures:
(Project leader) (Date) (Program Coordinator) (Date)
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




14
Manuscripts in Preparation or Published: (continued)
Uehara, G., B. B. Trangmar and R. S. Yost. 1985. Spatial variability of soil properties. In Soil Spatial Variability, Eds. D. R. Nielson and J. Bouma. Proceedings of a Workshop of the ISSA and SSSA, Las Vegas, NV 30 Nov-i Dec 1984. Pudoc, Wageningen.
Trangmar, B. B, R. S. Yost, and G. Uehara. 1985. Spatial Dependence and interpolation of soil properties in West Sumatra, Indonesia. I. Anisotropic variation II. Co-regionalization and co-kriging. Soil Sci. Soc. Am. J. (Summitted for publication).
Trangmar, B. B. 1984. Spatial variability of soil properties in Sitiung, West Sumatra, Indonesia. Ph.D. dissertation, University of Hawaii. 329 p.




15
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables of data and graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
Objectives 1, 2 and 3.
Research activties to achieve objectives 1, 2, and 3 include mapping and
display of regional soil variation as well as suggesting management techniques
to deal with tremendous spatial variability of soil chemical and physical
properties (Table 1). This activity had two parts:
1) Mapping and display of macroscale or regional variation in soil chemical
and physical properties which would provide some indication of the
representativeness of the experimental sites of the TropSoils Project.
2) Mapping and determining the cause of the considerable microscale soil
variability that has severely hindered field experimentation.
These two activities primarily differ in scale, the geostatistical
techniques applied were scale independent. In the first case an area of about 100,000 ha was considered (anisotropic kridging) while for the second case an
area of about 800 m2 (co-kridging)- was considered.
Anisotropic kriging of regional soil chemical variables.
Spatial and temporal interactions of the soil-forming factors and processes
determine the distribution of soil properties within a landscape. These spatial processes are sometimes more effective in some directions than others (i.e. they
are not isotropic) yet current methods of interpolation assume isotropy. The
purpose of this study was to determine the importance of considering directional
effects in making regional estimates of soil properties and in maps developed
from such estimates.
Results:
This study related anisotropic spatial dependence of particle size
fractions, pH and 25% HCl-extractable P to directional differences in the main
soil-forming factors in Sitiung, West Sumatra, Indonesia.
Previous studies of regional variation in Sitiung (Trangmar et al., 1984)
showed that sand content of soils was higher and silt content lower on the
peneplain compared to the Quaternary terraces and floodplains. There was no
clear pattern of differences in clay content among geomorphic units. Soil pH
and levels of HCI-P both decreased from younger to older surfaces as the
intensity of soil weathering increased. Changes in soil properties across geomorphic boundaries (Figure 1) were generally gradational and continuous.
Semi-variance usually increases with distance indicating a decrease in
spatial dependence with increasing distance. In isotropic conditions this relationship is independent of direction. With anisotropic variation the
direction must be included in the equation relating semi-variance with distance.
The description of anisotropy is given by the following equation:
gamma(h) = C(theta(i)) + w(theta(i))h ( 1 )
t E-e-8 o3
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




16
Progress and Achievements-continued, page 2
where gamma(h) is the semi-variance in distance h, C(theta(i)) is the nugget
variance of the semi-variogram in direction theta(i), and w(theta(i)) is the slope in direction theta(i), (Burgess and Webster, 1980).
The four directional semi-variograms of each property had similar nugget variances, but slopes that varied with direction. This indicated general presence of geometric anisotropy for each property (Journel and Huijbregts, 1978).
Geometric anisotropy gives an ellipsoidal neighborhood of spatial
dependence elongated in the sector of maximum variation and compressed in the sector of minimum variation (Burgess and Webster, 1980). The directional semi-variogram with the steepest slope (shortest range of spatial dependence) marks the sector of maximum variation while that of the least slope (longest range) indicates the sector of minimum variation.
The linear geometric anisotropic model fitted to the pooled semi-variances of four directions of topsoil sand is shown in Figure 2. Each experimental semi-variance value was plotted with a symbol indicating the 45 degree of the two-dimensional semi-variogram in which it lies. The solid lines are those of the maximum and minimum slope defining the envelope of the fitted anisotropic model.
The range of spatial dependence in any direction, theta, was assumed to be the distance (h) at which gamma equalled the sill value or general variance. The estimated range of sand was shortest (5.9 km) for the semi-variogram calculated in the sector of maximum variation (NE-SW) and longest (20.8 km) in the sector of minimum variation (NW-SE). Semi-variograms for the E-W and N-S sectors had intermediate ranges (13.6 km, 6.6 km, respectively). In contrast, the range would have defined a circular spatial dependence neighborhood of 10.0 km if topsoil sand had been assumed to vary isotropically.
The slopes estimated by w(theta) for the pooled semi-variance of topsoil sand and HCl-P in the four principal directions are listed in Table 2 for comparison with slopes (m(theta)) estimated directly for the individual directional semi-variograms. The agreement is generally good.
The parameter estimates of eq. (1) fitted to the pooled semi-variance of
the particle size fractions are listed in Table 3 and for pH and HCl-P in Table
4. The anisotropic model generally provided a good fit to each property as indicated by the significant R2 values. The model slightly underestimated slopes of clay, pH, and subsoil HCl-P in directions of minimum variation. Only for subsoil silt did the model provide a marginal fit.
Anisotropic semi-variogram modeling enables the identification of changes in spatial dependence with direction which, in turn, reflects soil forming processes. The estimated direction of maximum variation (phi) of all properties lies in a sector 7 to 28 degrees east of north. The direction of minimum variation occurs at right-angles in a sector 97 to 118 degrees (ESE-NNW).
Anisotropy of the textural component is largely caused by directional
effects of volcanic tuff fallout and deposition of alluvium by the major rivers. The SSW to NNE direction corresponds to the main axis of tuff deposition from the acidic volcanoes of the Braisan mountains to the southwest of the study region. Sand content is highest close to the tuff source along the southwest margin of the region and decreases to the north and northeast (Figure 3). Conversely, silt and clay content of soils are generally lowest in the southeast, but increases to the northwest.




17
Progress and Achievements-continued, page 3
The sector of minimum variation (97 to 118 degrees) occurred at right
angles to the direction of tuff fallout and also coincides with the main axis of the Batanghari River. The gentle gradient and low sediment load of the Batanghari River (Soil Research Institute, 1979) apparently has resulted in small downstream variation in particle size of suspended sediment. Alluvial materials of the low Quaternary terraces and floodplains deposited in this direction are relatively uniform in sand (Figure 3) silt and clay contents. Lateral variation of textural components over short distances in flood deposits away from the Batanghari River may also add to the anisotropy.
The E-W and SE-NW direction semi-variograms for sand, silt, and clay were generally more continuous and better structured than those in the NE-SW and N-S directions. Semi-variograms in the latter two directions showed a periodic or 'hole effect' pattern (Journal and Huijbregts, 1978) of variation at lags of 10 and 11 km. This was probably caused by low values of sand on the Quaternary terraces and Batanghari floodplain in the north and Jujuhan floodplain in the south with higher values on the peneplain in between (Figure 3). Complementary amounts of finer materials in these areas also explain similar periodic semi-variograms of silt and clay at this distance in the NE-SW and N-S directions. The NE-SW and N-S semi-variograms for topsoil (Figure 2) and subsoil sand exceeded the general variance, s, indicating the presence of a weak trend in these directions.
Linear geometric anisotropic models were fitted to pooled directional
semi-variances for each of these properties. Directions of maximum variation coincided with the main southwest to northeast axis of volcanic tuff fallout, deposition of alluvium and the general sequence of soil weathering in the region. Ranges of spatial dependence for each property were shortest (2.6 km for pH) in the direction parallel to this axis and longest (20.8 km for sand content) at right angles to it. Anisotropy ratios ranged from 1.5 for subsoil sand to 5.2 for subsoil HCI-P. Topsoil textural components and pH were more variable than in subsoils, having larger sample variances and larger anisotropy ratios. Punctual kriging of topsoil sand content using the linear geometric anisotropic model resulted in low estimation variances in densely sampled areas where weighting of neighbor samples by direction as well as distance took most effect.
Additional sampling in areas of high estimation variance should utilize the known anisotropy in relation to distance and direction from initial sites. An optimal sampling scheme in the presence of geometric anisotropy and a known semi-variogram is a rectangular or triangular grid with relative directional spacings in proportion to the anisotropy ratio and with the longer intervals aligned in the direction of least variation.
Structural analysis of soil variation using geostatistics can aid
understanding of the spatial distribution of soil properties and identifying the spatial effects of soil-forming factors and processes on soil genesis. In this study, geostatistical analysis of textural components, pH and HCl-extractable P indicated that anisotropy of these properties in Sitiung, West Sumatra was directly related to directional deposition of volcanic tuff materials and alluvium; and to the regional sequence of soil weathering.
Estimates of anisotropic spatial dependence can also be incorporated into kriging interpolation for optimal, unbiased spatial estimation with minimum variance.




18
Progress and Achievements-continued, page 4
Cokriging of regional soil chemical variables.
In soil science certain variables are much more easily measured than
others. In some cases there may be data available for certain variables but not for other variables which were subsequently determined to be necessary or critical to the major soil constraints in the area. The method of co-kriging permits making estimates of one, infrequently measured variable, based on spatial correlation with another variable which is more easily or less costly measured.
The spatial distribution of soil property values may often be closely
related to those of other soil properties with which they are highly correlated. Such properties are said to be co-regionalized or spatially-dependent on one another. The spatial covariance of co-regionalized properties may be positive or negative depending on the nature of the correlation. Spatially covariant properties are often those whose spatial distributions are caused by a common, interacting set of soil processes. Inferences about the spatial structure of a co-regionalized variable can be made from knowledge of the spatial structure of its other covariant properties. This information can be employed for optimal, unbiased interpolation of undersampled variables by using their spatial covariance with other more densely sampled covariables. This is achieved by the co-kriging approach in geostatistics.
Co-kriging extends the principle of optimal estimation using regionalized variable theory from that of a single property to situations where there are two or more co-regionalized ones. Co-kriging is most efficiently used where one variable may not have been sampled sufficiently (due to high cost, experimental difficulties, etc.) to provide estimates of acceptable precision. Estimation precision can be improved by utilizing the spatial correlation between the undersampled primary variable and the other more frequently sampled covariables.
Co-kriging was developed for ore reserve estimation in mining studies and has more recently been applied to spatial interpolation of soil-water properties and soil particle size fractions. It has been demonstrated that a grid system of covariate samples located geometrically around the primary variable is the optimal sampling scheme to exploit the spatial covariance of two variables by co-kriging. In many soil sampling programs, and particularly in those associated with soil surveys, samples are not always collected in a grid pattern because kriging is not usually the sampling objective.
Results:
In Indonesia P extracted with 25% HCL is commonly used as an indication of soil weathering status in unfertilized areas. Phosphorus extracted with 0.5M NaHCO3 is often regarded as a measure of plant-available P in soils making it more useful for agronomic interpretation of soils than HCI-P. During soil surveys in Sitiung, West Sumatra, Indonesia 0.5M NaHCO3-P was undersampled relative to HCI-P. Therefore 0.5M NaHCO3-P was selected as the primary variable and HCl-P was selected as the covariable. The respective semi-variograms and cross-variogram are shown in Figure 4.
The map of co-kriged values for NaHCO3-P showed a similar regional pattern but more local detail than that achieved by kriging from NaHCO3-P samples alone. Co-kriging reduced estimation variances by up to 40% in areas where sampling density of NaHCO3-P was lowest. Co-kriging variances exceeded those of kriging by up to 10% in areas where sampling density of NaHCO3-P was high.




19
Progress and Achievements-continued, page 5
In such cases, the covariate HCL-P had little effect on the interpolated value but still added a component to the estimation variance. Co-kriging could not be used to interpolate values for extreme locations where there were no NaHCO3-P sampled within the radius of the kriging neighborhood.
A prerequisite for interpolation using co-kriging is the presence of a well-structured cross semi-variogram with low nugget variance. The apparent difficulties of obtaining this indicates that co-kriging may be restricted to interpolation of very well structured variables.
Being a widely used measure of plant-available P, NaHCO3-P is of more
interest for agronomic interpretation of soil P status than the more densely sampled measure of HCI-P. The co-regionalization of the two P measurements was exploited using co-kriging to interpolate topsoil NaHCO3-P at 268 unsampled locations on a 2 km by 2 km grid across the study of area. Fifty-two samples of NaHCO3-P and 107 samples of HCl-P were used in the co-kriging operation.
The isarithmic map of co-kriged NaHCO3-P shows that topsoils in 84% of the study area have less than 15 ppm NaHCO3-P, a commonly used critical value of adequate P nutrition in many crops grown on acid, upland tropical soils. The sample areas with greater than 15 ppm coincide with young alluvial soils and areas recently fertilized prior to sampling. Soils in the east of the region appear very P deficient having less than 5 ppm NaHCO3-P.
Isarithmic map of NaHCO3-P obtained by auto punctual kriging from 52
samples was similar to that achieved by co-kriging, but local detail gained by utilizing the 107 covariate samples of HCl-P was absent.
The percent reduction in estimation variance achieved by co-kriging (Gamma) relative to that of auto kriging was determined as (1-SCK/SK) x 100 and is shown in Figure 5. Positive values indicate the percent improvement in estimation precision while negative values indicate the decrease in precision obtained by co-kriging. Co-kriging improved estimation precision by up to 40% for locations where there were few samples of NaHCO3-P but relatively more HCI-P samples. Smaller estimation variances were obtained by co-kriging in such areas because the spatial correlation with another better sampled variable was taken into account. The percent improvement in SCK was generally less than 10% and occurred for about 60% of the 260 kriged locations.
Estimation variances were not reduced by co-kriging in areas where
NaHCO3-P samples were concentrated. These locations are bounded by the 0 to
-10% contour lines in Figure 5. Each kriging location in these areas had several neighboring samples so NaHCO3-P received most of the weighting in solving the cokriging equations. Under these conditions, the covariable (HCI-P) added another variance component to the estimation without causing much or any improvement to the overall estimation precision. As a result SCK exceeded SK. The relative increase in SCK over SK was generally less than 10% but occurred over about 40% of the region. The large negative values along the south-west margin of the region mark one location for which there was only one sample of each P measurement causing instability of the co-kriging matrices and unreliable interpolation.
The co-kriging system requires at least one sample point of both primary NaHCO3-P and covariable HCI-P within the estimation neighborhood. The range of spatial dependence (6.3 km) of the NaHCO3-P auto semi-variogram defined the search radius for samples of NaHCO3-P and HC1-P to obtain the auto and cross semi-variances used in the weighting procedure of the co-kriging system. This radius resulted in between 1 and 13 neighbor samples of NaHCO3-P for 234 of the 268 kriging locations. Co-kriging could not be performed at the remaining




20
Progress and Achievements-continued, page 6
34 locations because there were no NaHCO3-P samples within 6.3 km of the kriging location. Twenty-seven of these 34 locations occurred in the eastern end of the region. The remaining 7 locations occurred along the southwest margin and in the central area. Estimates of NaHCO3-P at these locations could be just as reliably estimated using simpler techniques such as the sample mean, or by kriging values for HCI-P and then obtaining NaHCO3-P values by regression analysis. A radius of 6.3 km gave between I and 30 samples of HCI-P for all kriging locations.
These results demonstrate the sensitivity of co-kriging interpolation
precision to the relative spatial distribution of primary and covariable sample locations. The irregular, nongeometric sampling scheme of this study is clearly nonoptimal from a co-kriging standpoint. Others (Vauclin et al. (1983) and McBratney and Webster (1983) found that co-kriging consistently reduced estimation variances where primary and covariable properties were sampled in geometric patterns. Their results and the irregular improvement in estimation variance for this study indicate that a geometric scheme with samples of the primary variable regularly interspersed with covariable samples may be the optimum to gain maximum benefit from co-kriging. Clusters of the primary variable do not occur in such a scheme. As a result, the variance-reducing effects that accrue from using the better sampled covariable are not hindered in the interpolation. Design of such schemes requires prior knowledge of the two variables' spatial covariance, which can be determined from reconnaisance data. Such data may be most economically obtained from samples collected over a range of spacings along transects parallel to, and also at right angles to, the apparent direction of maximum variation.
Many existing data sets, such as those of soil survey, have been collected without geostatistical analysis in mind. Co-kriging of such data may yield additional information and can give acceptable interpolation for locations inadequately sampled areas even if the sampling pattern is nongeometric. However, insufficient samples in an area still makes interpolation unreliable regardless of the procedure used, and additional sampling may be the only way to improve estimation precision.
Despite the numerous qualifying conditions identified in this and other
studies co-kriging has potential to make considerable time and economic savings in mapping soil properties for which there are cheap surrogates.
Objective 4.
Mapping soil microvariability and determining factors causing yield
reduction of rice, cassava, and peanut are the primary activities initiated to achieve objective 4.
The identification of the causes of microscale (2 m or less) variability
and options for dealing with it are fundamental to the entire research effort in the Tropsoils/ Indonesia project. Previous field research has lead to inclusive results because of such variability (Dr. M. Sudjadi, personnal communication).
The approach adopted to examine this variability was based in part on
concepts of geostatistics and in part on concepts of the traditional uniformity experiment. A parcel of previously cleared land which is representative of the high soil variability was selected for a continuing study of the nature and sources of such variability.
A series of crops were planted across the unfertilized and untilled field. Plant growth and nutrient composition were monitored to identify the sources of variability. This design, in effect, utilizes the crop as the experimental material and the soil's potential to support plant growth becomes the




21
Progress and Achievements--continued, page 7
experimental treatments. The soil was also sampled intensively in order to identify the chemical and physical variation (Figure 6).
Results:
As indicated in the Table 1,5,6 and 7 and in various regression
relationships, extractable acidity (Al + H) and exchangeable cations accounted for most of the soil-chemical related variance in growth and yield components of the first rice crop. Lower levels of Al saturation (Figure 7) and higher concentrations of exchangeable cations in burned sites coincided with larger plants (Figure 8), higher stover and grain yields compared to those grown on exposed subsoil areas. The ranges of spatial dependence (3 to 4 m) for soil acidity and exchange characteristics coincided with about one-half the distance between burned sites. Spatial dependence of plant height and yield components was influenced by unidentified longer range sources of variation. Block kriged (for I square meter) rice grain yields ranged from 11 to 355 g/m2 with estimation standard deviations ranging from 9 to 24 g/m, depending on the distance of interpolated cell from the sample locations.
The effect of microvariability on precision of a field experiment was examined by laying out four replicates of seven potential treatments in a randomized complete block design across the site. Individual plot size was 4 m by 7 m. Sample values of grain yield and Al saturation from 1 m square cells in each plot were averaged to obtain means for each of the 28 plots. This is analogous to the common practice of bulking samples for analysis on a whole plot basis. A large experimental error (CV of 37%) resulted in nonsignificant differences among treatment means of yield estimates which ranged from 1.00 to 2.08 mg/ha with a Waller-Duncan LSD of 1.08 Mg/ha. Statistical analysis on a whole plot basis masked the large within-plot variability of grain yield which ranged from 0.11 to 3.55 Mg/ha when considered on a I m square basis. Such variability was not shown by analysis of treatment means because the variation was averaged with each plot and because of the large error term which was partly due to significant variation among replicates. Similarly, the large variability of Al saturation (CV of 58%) within plots was masked by averaging on a whole plot basis.
As an further illustration of within plot variability, a P experiment was conducted in Sitiung V by Dr. M. Wade. The preliminary results of yields with mung bean showed so much scatter that the interpretation of response due to P was inconclusive (Figure 1). When microplots were subsampled for both crop yield and soil analysis, two general groups were identified, those with low Ca + Mg levels and those with high Ca + Mg levels. As shown in Figure 9, the response to P then was quite clear and significant. Mung bean yield increased linearly throughout the range of applied P. Had there been no soil sample data upon which to segregate the low and high levels of Ca + Mg the conclusion of the analysis would certainly have been that no response was obtained to applied P. Observations such as this have become frequent in the field research efforts in the Sitiung region. This suggests therefore a strong need to sample both the crop and the soil when conducting field experiments in such highly variable soils.
Literature cited:
Burgess, T.M. and R. Webster. 1980. Optimal interpolation and isarithmic
mapping of soil properties. I. The semi-variogram and punctual kriging.
J. Soil Sci. 3 1:315-331.




22
Progress and Achievements-continued, page 8
Journel, A.G. and C.H. Huijbregts. 1978. Mining Geostatistics, Academic
Press, New York.
McBratney, A.B. and R. Webster. 1983. Optimal interpolation and isarithmic
mapping of soil properties. V. Co-regionalization and multiple sampling
strategy. J. Soil Sci. 34:137-162.
Soil Research Institute, 1979. Soil Survey of the irrigation project area,
Sitiung S. Jujuhan (West Sumatra). Ministry of Agriculture, Agency of
Agricultural Research and Development, Indonesia.
Trangmar, B.B. 1984. Regional variation in soil properties in the Sitiung
Transmigration area, Indonesia.
Vauclin, M., S.R. Vieira, G. Vachaud, and D.R. Nielsen. 1983. The use of
co-kriging with limited field observations. Soil Sci. Soc. Am. J.
47:175-184.




23
Progress and Achievements-continued, page 9
Table 1. Soil chemical properties of the various terrain units
within a 28m x 28m experiment.
Burn Surrounding Exposed
Soil sites soil subsoil Range in
Property (n=26) (n=87) (n=24) values
------------------------------------------------------------------pH 5.0 a 4.4 b 4.1 c 3.4 6.3
Organic C (%) 3.9 a 3.4 a 2.9 a 0.2 7.2
Total N (%) 0.27 a 0.24 a 0.22 a 0.07 0.49
NaHCO3)-P
(mg/kg) 13 a 9 b 7 c 2.0 29.2
Exchangeable
cations
(cmol(p*)kgl)
Ca 3.9 a 1.6 b 0.6 c 0.0 11.2
Mg 1.4 a 0.6 b 0.2 c 0.0 3.2
K 0.5 a 0.3 b 0.2 c 0.1 1.4
Al 1.3 b 2.4 a 2.8 a 0.0 6.3
Sum (ECEC) 6.9 a 4.8 b 3.9 c 2.8 15.3
Al saturation (%) 26 c 54 b 73 a 0.0 94
Cu mg/kg 2 a 2 a 2 a 1 6
Zn mg/kg 3 a 2 b 1 c 1 11
-------------------------------------------------------------a. Means within a row followed by the same letter are not
significantly different (P=0.05) according to Student's t
test. T tests were performed on log transformed values of
all properties except pH, Mg, Al and Al saturation.
Arithmetic means reported for pH, Mg, Al and Al saturation,
while means reported for log transformed properties are
those reexpressed in terms of the original data using
equations of Haan (1977).




24
Progress and Achievements-continued, page 10
Table 2. Slopes of semi-variograms of topsoil sand and HC1-P estimated for each direction separately, mia), and from pooled semi-variances after fitting a geometric anisotropic model, w(thetai)b. Direction, theta --------------- Slope---------------(degrees E of N) mi w(thetai)
Sand (%)
45 38.4 35.0
90 16.6 15.3
135 15.3 11.8
180 30.5 31.5
(HCI-P (mg kg-l)))
45 0.099 0.082
90 0.041 0.036
135 0.022 0.036
180 0.094 0.082
a. mi = slope of the linear model.
b. w(thetai) = Acos2(theta-phi) + Bsin2(theta-phi)
where theta is the direction in which the semi-variance
is estimated; phi is the direction of maximum slope A;
and B is the slope of the semi-variogram at 90) to phi.




25
Progress and Achievements--continued, page 11
Table 3. Parameter estimates of equation fitted to pooled semi-variance of particle-size fractions.
-------------------------------------------------------------------------- Property------------Parameter Sand (%) Silt (%) Clay (%)
----------------------------------------------------------Mean 16 26 58
Variance, s2 212.8 213.5 201.4
Nugget
variance, Co 6.4 31.0 64.9
% of sill 3 15 32
Max. gradient 37.5 58.2 37.9
Min. gradient 9.3 19.3 8.7
Anisotropy ratio 4.0 3.0 4.4
Direction of maximum
gradient, phi 27.6 7.3 29.3
R2 of model 0.89 0.81 0.69
degrees of freedom 58 19 26
-------------------------------------------------------------a. Semivariance = Co + Acos2(theta-phi) + Bsin2(theta-phi)*h
b. R2 = (SScorrected SSresidual )/SScorrected*
All R2 values were significant at P < 0.01, except for subsoil
silt which was significant at P < 0.05.




26
Progress and Achievements-continued, page 12
Table 4. Parameter estimates of the geometric anisotropic
equation fitted to pooled semi-variance of pH and
HCl-P (0 15 cm depth).
------------------------------------------------------------------------- Property--------------Parameter pH HCl P
--------------------------------------------------------------Mean 4.5 107
Variance, s2 0.19 0.59
Nugget
variance, Co 0.05 0.25
% of sill 26 42
Max. gradient 0.05 0.10
Min. gradient 0.02 0.03
Anisotropy ratio 2.4 3.5
Direction of maximum
gradient, phi 11.3 22.9
R2 of model 0.73 0.56
degrees of freedom 15 28
-------------------------------------------------------------a. Semivariance = Co + Acos2(theta-phi) + Bsin2(theta-phi)*h
where theta is the direction in which the semi-variance is
estimated; phi is tdirection of maximum slope A; and B is
the slope of the .semi-variogram at 90 to phi.
b. Semivariance for HCI-P determined on log transformed values.
cR2= (SS SSreida)/SSc)
c. R2 = (SScorrected Sresidual)/SScorrected')
All R2 values were significant at P < 0.01.




27
Progress and Achievements-continued, page 13
Table 5. Spatial dependence of soil properties highly correlated
with rice plant height and dry matter yield.
---------------------------------------------------------------------- Correlation ----Range of Stover Grain
Soil spatial Plant height dry dry
Property dependence at 60 days weight weight
-----------------------------------------------------------------meters
pH 4.1 0.51 0.44 0.38
Exchangeable
cations:
Ca no pattern 0.56 0.55 0.49
Mg 4.3 0.52 0.45 0.40
K no pattern 0.41 0.38 0.46
Al 4.0 -0.55 -0.47 -0.41
Al saturation 3.9 -0.56 -0.51 -0.41
(%)
Zn (mg/kg) < 0.5 0.42 0.34 0.33
Plant measures:
Plant height
at 60 days 19.9
Stover dry
weight 15.3
Grain dry
weight 18.5
-----------------------------------------------------------------All r values were significant at P < 0.01 unless otherwise
indicated. ns = nonsignificant at P = 0.05.




28
Progress and Achievements-continued, page 14
Table 6. Effects of soil variability on yield of upland rice,
peanut, and cassava among terrain units.
Burn Surrounding Exposed
Plant sites soil subsoil
Property (n=26) (n=87) (n=24)
Rice
Plant height
(60 days), cm 92.5 a 78.5 b 71.2 b
Stover weight
(g/m2) 328 a 234 b 193 b
Grain dry
(weight, g/m2) 227 a 166 b 146 b
Peanuts (Crop 1)
Plant dry wt,
g/m2 150 a 101 a 74.3 b
Nut weight,
g/m2 59.3 a 33.9 b 21.0 b
Number of pods
/m2 123 a 78.8 b 56.9 b
Number of empty
pods/m2 14.9 a 15.0 a 14.0 a
Cassava (Crop 1)
Tuber weight 2476 2459 2067
Stem weight 3013 a 2165 b 1945 b
Leaf weight 243 225 215
a. Means within a row followed by the same letter are not
significantly different (P = 0.05) according to Student's t
test




29
Progress and Achievements--continued, page 15
Table 7. Correlations of soil chemical properties of the various
terrain units within a 28m x 28m experiment on crop growth.
--------------------------------------------------------------------Soil Plant height Stover dry Grain dry
Property at 60 days (cm) weight (g/m2) weight (g/m2)
------------------------------------------------------------pH 0.51 0.44 0.38
Organic C (%) 0.07 ns 0.02 ns 0.02 ns
Total N (%) 0.09 ns 0.11 ns 0.14 ns
NaHCO3-P
(mg/kg) 0.06 ns 0.11 ns 0.11 ns
Exchangeable
cations
(cmol(p+kgl)
Ca 0.56 0.55 0.49
Mg 0.52 0.45 0.40
K 0.41 0.38 0.46
Al 0.55 0.47 0.41
Al saturation (%) 0.56 0.51 0.41
Cu mg/kg 0.07 ns 0.09 ns 0.12 ns
Zn mg/kg 0.42 0.34 0.33
--------------------------------------------------------------------a. All r values are significant at P < 0.01 unless otherwise
indicated. ns = nonsignificant at P = 0.05.




0
00
rt
ox
x SUMATRA /, .
2 -' '>
2xx x 4 ,SITIUNG I t 0
5 x xx Siting
x 2 x x X"
xUT Rx R xTGA
x xx 4
40 RCN4xOPAN
N x x x0
xx
xx
X --X x
LEGEND x x ,x SAMPLE SITE 1 x x x 1 M
1 DISSECTED PENEPLAIN x x
2 TRANSITION PENEPLAIN-TERRACES x
SQUATERNARY TERRACES xx x
4LECENT FLOPLI x
5 GRANITE x
Figure 1. Location and soil sampling pattern in Sitiung, West Sumatra, Indonesia.




P1
0
CL
6oo00- DIRECTION
n
+ NE-SW M
0 E-W 500- 0 SE-NW
rt'
t S-N
o
+ ~0
400- a
+
o 300+0
-- 2 OOE +
a 0~
> + +
0O
3 00
+0
U) A 1 10 .J2a0
Ioo- 1h M a 0
O
0j 0
I I I I I I I
0 3 6 9 12 15
Average Distance (kin)
Figure 2. Linear geometric anisotropic model for semi-variances of topsoil sand
for four directions.




0 O1
cn
M 30 M
0
0
C3 20 020
Figure 3. Isarithm map of sand content(% made by anisotropic punctual krigin-, 0 to 15 cm denthI.




33
Progress and Achievements-continued, page 19
Na HCO 3- P, HCI-P
0.8
S0.64
E
cn 0.40.0-1
0369. 2 15
Distance (kin) 250/ HCI-Extr. P
0.8
0.6
0.4
00
c~ 0.60.0
Distance (kin)
0.5M NaHCO3 -Extr. P
0.8
~0.6 0
> 0.4002
0.0-11
0 3 6 9 12 15
Distance (kin)
Figure 4. Isotropic cross-variogrami (upper grph and auto serni-variograms (lower two
graphs) for lo- transformed values (mg.L-l) of
25/%. H.Cl-extractable P and- 0.5M NaT-C03 extractable
P, 0 to 15 cm depth.




0
IL 0
at
00
9 20
00
x0
Fi.-re Prcet rducion n etimtio vaiane byco-rigng op oil0.511 NAC0-exracabl P create tokriing Crosesindcatesamle lcatonsof HIIC3-0




35
Progress and Achievements-continued, page 21
N
+ + /Z% + + +_+
+ IZ7 + +
ASH
SHS
+ ++ ;
RS
+ ++ ASH + +
+ as
es ++ ++ ASH + ASH + +
+ +
ss+ +
8<7 +.H4 ss
+ + + + + + + +s
+ SSs
as + + +s
+ BS
+ + +. + +
LEGEND
Tilage ridge RS Rice Straw
Tree stump BS Baked Soil
S+ and plant SS Exposed Subsoil
+ ASH+
SURFACE ASH
SS +
sample site
Figure 6. Sampling for soil and plant SS Exposmeasurements in relation
sample site
Figure 6. Sampling for soil and plant measurements in relation
to surface features of the site.




36
Progress and Achievements-continued, page 22
N
AL SATURATION (%)
80
60 60
40 .40
20
20
0 0
& +
Figure 7. Three-dimensional diagram of Al saturation (%) block.
kriged over 1m2 cells, 0 to 20 cm depth.




37
Progress and Achievements-continued, page 23
N
GRAIN YIELD (G/M2)
G/M2 G/M2
300 300
200 200
100 100
0
0
Figure 8. Three-dimensional diagram of rice grain yield block
kriged over 1m2 cells, units are (2/12).




38
Progress and Achievements-continued, page 24
a
900
Y =-107 14P'78 Ca Mg
R = 0.74**
a a
a
o 600 a
CQ
b
a b
01b
en 300-b b
2 b
b
Obb
0 1b 2o 30
Extractable P (ppm)
Figure 9. Mung bean yields as affected by extractable P at
low (b's) and high (a's) Ca and Mg levels.




39
TROPSOLLS
Management Entity, BOX 71 13, Raligh, North Carolina 27695-71 13
SWork Plan for Next Year:
(Include: /Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
During the initial phase of this study soil variability was characterized
and mapped both for on the macro scale for the 100,000 ha study area as well as
on a microscale of 784 m Those results are included in this report. In the
current phase of the study alternative methods of dealing with high soil
variability are being developed and compared. These include but are not limited
to:
1. Individual spot applications of fertilizer compared to traditional
blanket applications. This approach is one that is consistent with the small,
hand-fertilized and tilled home gardens (pekerangan) or dryland fields (ladang).
Such applications may be a particularly efficient way of applying costly
fertilizer to the precise location where it is needed and also be an operation
that the farmers, who best know their land, can do themselves.
2. Techniques of preparing for standard field experiments. As experience
has often unhappily shown in the past, soil variability can destroy otherwise
useful experiments because of excessive variation both between replications and
within plots. Current techniques of first conducting a uniformity trial and using the data to select replications has proven useful although frequently
insufficient to achieve suitably uniform experimental material. In addition to
such methods it is frequently necessary to select localized harvest areas within
the plot in which both the crop and soil are sampled carefully. Such localized and coordinated sampling permits establishing regression relationships between the soil properties and the crop yield or nutrient content. Experience so far
suggests such quantitative relationships are, in fact, preferable to standard
analysis of variance and mean separation tests. The advantage of these
quantitative relationships is that they are directly suitable for incorporation
into expert systems for soil management information transfer.
3. Further monitor difficulties encountered with field-based
experimentation caused by undue soil and crop spatial variability. Look for
practical solutions and opportunities to better deal with soil variability.
4. Application of geostatistics to the determination of optimal plot size.
A comparison will be made between classical treatment of uniformity trial data
(Gomez and Gomez, 1985) and a geostatistical approach.
5. Manuscripts will be prepared as appropriate.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




40
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: 10/1/85 9/30/86
Project Leader/Institution: Carol J. Pierce Colfer, University of Hawaii
Principal Collaborators/Institution:
Herman Agus, Center for Soil Research; Dan Gill, North Carolina State
University; Barbara Newton, University of Hawaii
Research Site(s):
Koto Padang, Pulai, Sitiung I, Sitiung V, West Sumatra
Project Title:
Indigenous Knowledge Systems Related to Soil Management (Galileo, indigenous
soil classification systems)
Project Objectives:
Identify and map indigenous concepts related to soil.
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
None
Manuscripts in Preparation or Published: (Author(s), title, journal)
None
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) (Date)
WIE- : -8 3-a
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




41
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-71 13
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables ot data and graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
Planned Research Activities and Procedures
The first part of this ongoing project was planned, recognizing that
differing ethnic groups have differing views of soil management and different
perceptions of the relevant problems related to the soil (see e.g., Moran 1983,
Sherman 1981). The "Galileo" method was chosen as an efficient and reliable way
to characterize people's views related to soil in a quantitative manner (See
Woelfel and Fink 1981, for more detailed discussion).
This study began with the taping of open-ended interviews with ten
individuals from each of the three main ethnic groups in the area (Sundanese,
Javanese, and Minangkabau). Every effort was made to influence the respondents
as little as possible as they responded to the question "What is the
relationship between soil and people?" The interviews were conducted in the
respondents' native language, and tape recorded.
In each ethnic group respondents were selected such that the sexes were
equally represented. Additional considerations included age, income,
educational level, and in the case of the Minang, clan affiliation. These interviews were then content analyzed to ascertain the recurrent concepts, within each ethnic group. The Minang interviews have been translated into
English for use by the American team members.
For two of the three ethnic groups, a survey instrument was constructed
based on these findings. This instrument is simply a pairing of every concept
identified as important in the content analyses with every other concept so
identified. There were of course differences in the important concepts, based
on ethnicity, and the instrument was expanded to reflect the concepts that were
important for one group and not for another. Twenty-one concepts were deemed sufficient to document and measure inter-ethnic differences in perceptions of
soil as it relates to people.
Respondents are then asked to "measure" the distances between the concepts,
using a cognitive measuring stick. In this case the measuring stick was the
distance between black and white, set at 100 units (Table i). Thus a respondent
is asked, "If black and white are 100 units apart, how far apart are soil and water?" "soil and tree crops?", "Soil and rice?", etc. The distances between
concepts are averaged across a given population to provide a "cognitive map" of
the domain of soil-people relations. This instrument has been administered to
100 Javanese and 100 Minang respondents. Half the respondents were male and half female; an attempt was made in the selection of respondents to represent the diversity that exists in each of the locations (Koto Padang and Sitiung I
Blok B). Interviews were conducted by two trained native speakers of Javanese
and Minang, respectively. The data are currently being entered on disk, and
will be sent to UH for analysis, when the Sundanese instruments and interviews
are completed.
The most fundamental and useful results from this analysis take the form
of a triangular means matrix which shows the average distances between each
concept and every other concept. Thus we will be able to compare variation in
%%E- P 4.8 e8
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




42
Progress and Achievements-continued, page 2
people's perceptions based on ethnicity and based on sex. The Galileo program uses these data as well to create eigenvectors, from which 3- dimensional plots can be constructed to make these "cognitive maps" more appealing visually for policymakers and people from other disciplines. There is also an "automatic message generator" included in the Galileo program which can help with soil-related extension efforts at a later date, or in some other context.
Results, Interpretations, Potential Application and Use
Most immediately, the interviews provided us with the 21 most frequently mentioned concepts related to soil and people. These were: soil, smallholder plantation/garden, upland field, wet rice field, home garden, rubber, fruits, rice, field crops, vegetables, water, fertilizer, pests, yields, cultivation, good, enough, I, male, female, government. The identification of these concepts alone is of interest, providing us with a firmer sense of people's normal assumptions related to soil.
The tendency for the Minang to discuss tree crops was in striking contrast to the transmigrants' propensity to discuss field crops and vegetables. Similarly, smallholder plantation/garden (kebun) was commonly mentioned among the Minang, whereas upland fields (ladang) were a mainstay for the transmigrants.
The Minang interviews were translated first, since the Tropsoils team has worked most intensively with the transmigrants, and therefore has a fuller intuitive sense of transmigrant farming systems than of the Minang system. Some of the Minang experience with trying to grow field crops duplicated our own problems with unpredictable rain, insects, plant diseases, and poor soils. Their solution seems to be to grow tree crops and wet rice. The Minang experience with extension efforts was likewise of interest. We learned of problems with the hybrid rice varieties introduced by the government (e.g., a tendency for the young rice stalks to collapse into the surrounding mud when planted at the government-recommended spacing, whenever anyone walked through the field). Some bitterness was expressed about the fact that the government supplied nearby transmigrants with an irrigation system (and other projects, such as TropSoils), without adequate regard for the indigenous people.
The transmigrants similarly recounted woeful tales of their experience with field crops in Sitiung; but their previous experience was with these crops, and until June 1985 the government required that transmigrant fields be used exclusively for food crops.
Literature Cited
Moran, E. 1983. "Government-Directed Settlement in the 1970's: An Assessment
of Transamazon Highway Colonization" In The Dilemma of Amazonian
Development (E. Moran, ed.). Boulder: Westview Press.
Sherman, George. 1981. The Culture-Bound Notion of "Soil Fertility:" On
Interpreting Non-Western Criteria of Selecting Land for Cultivation. In
Blowing in the Wind: Deforestation and Long-Range Implications.
Williamsburg, VA: Studies in Third World Societies. pp. 487-512.
Woelfel, J. D. and E. L. Fink. 1981. The Measurement of Communication Process:
Galileo Theory and Method. New York: Academic Press.




43
Progress and Achievements--continued, page 3
Table 1. Galileo Survey Form
GALILEO: ENGLISH
DATE:
NAME:
SEX:
ETHNI CITY:
RESIDENCE:
IF BLACK AND WHITE ARE 100 UNITS APART, hOW FAR APART ARE EACH OF THE CONCEPTS BELOW? concept
# card # CONCEPT DISTANCE
(duplicate 1-6)
Card #01 7-8
102 09-17 soil and garden
103 18-26 soil and unirrig field
104 27-35 soil and wet rice field
105 36-44 soil and home garden
106 45-53 soil and rubber
107 54-62 soil and fruits
108 63-71 soil and rice
109 72-80 soil and other fld crop
(dupl icate 1-6)
Card #02 7-8
110 09-17 soil and vegetables
ill 18-26 soil and water
112 27-35 soil and fertilizer
113 36-44 soil and pests
114 45-53 soil and yield
115 54-62 soil and cultivation
116 63-71 soil and good
117 72-80 soil and enough
(duplicate 1-6)
Card #03 7-8
118 09-17 soil and I/me
119 18-26 soil and male
120 27-35 soil and female
121 36-44 soil and government
203 45-53 garden and unirrig field
204 54-62 garden and wet rice field
205 63-71 garden and home garden
206 72-80 garden and rubber
(duplicate 1-6)
Card #04 7-8
207 09-17 garden and fruits
208 18-26 garden and rice
209 27-35 garden and other fId crop
210 36-44 garden and vegetables
211 45-53 garden and water
212 54-62 garden and fertilizer
213 63-71 garden and pests
214 72-80 garden and yield




44
Progress and Achievements--continued, page 4
(duplicate 1-6)
Card #05 7-8
215 09-17 garden and cultivation
216 18-26 garden and good
217 27-35 garden and enough
218 36-44 garden and I/me
219 45-53 garden and male
220 54-62 garden and female
221 63-71 garden and government
304 72-80 unirrig. field and wet rice field (duplicate 1-6)
Card #06 7-8
305 09-17 unirrig. field and home garden
306 18-26 unirrig. field and rubber 307 27-35 unirrig. field and fruits
308 36-44 unirrig. field and rice
309 45-53 unirrig. field and other fld crop
310 54-62 unirrig. field and vegetables
311 63-71 unirrig. field and water
312 72-80 unirrig. field and fertilizer (duplicate 1-6)
Card #07 7-8
313 09-17 unirrig. field and pests 314 18-26 unirrig. field and yield
315 27-35 unirrig. field and cultivation
316 36-44 unirrig. field and good
317 45-53 unirrig. field and enough 318 54-62 unirrig. field and I/me319 63-71 unirrig. field and male
320 72-80 unirrig. field and female (duplicate 1-6)
Card #08 7-8
321 09-17 unirrig. field and government 405 18-26 wet rice field and home garden
406 27-35 wet rice field and rubber 407 36-44 wet rice field and fruits
408 45-53 wet rice field and rice
409 54-62 wet rice field and other fld crop
410 63-71 wet rice field and vegetables
411 72-80 wet rice field and water (duplicate 1-6)
Card #09 7-8
412 09-17 wet rice field and fertilizer
413 18-26 wet rice field and pests 414 27-35 wet rice field and yield
415 36-44 wet rice field and cultivation
416 45-53 wet rice field and good
417 54-62 wet rice field and enough
418 63-71 wet rice field and I/me 419 72-80 wet rice field and male (duplicate 1-6)
Card #10 .7-8
420 09-17 wet rice field and female
421 18-26 wet rice field and government
506 27-35 home garden and rubber 507 36-44 home garden and fruits




45
Progress and Achievements--continued, page 5
508 45-53 home garden and rice
509 54-62 home garden and other fld crop
510 63-71 home garden and vegetables
511 72-80 home garden and water
(duplicate 1-6)
Card #11 7-8
512 09-17 home garden and fertilizer
513 18-26 home garden and pests
514 27-35 home garden and yield
515 36-44 home garden and cultivation
516 45-53 home garden and good
517 54-62 home garden and enough
518 63-71 home garden and I/me
519 72-80 home garden and male (duplicate 1-6)
Card #12 7-8
520 09-17 home garden and female
521 18-26 home garden and government
607 27-35 rubber and fruits
608 36-44 rubber and rice
609 45-53 rubber and other fld crop
610 54-62 rubber and vegetables
611 63-71 rubber and water
612 72-80 rubber and fertilizer
(duplicate 1-6)
Card #13 7-8
613 09-17 rubber and pests
614 18-26 rubber and yield
615 27-35 rubber and cultivation
616 36-44 rubber and good
617 45-53 rubber and enough
618 54-62 rubber and I/me
61? 6_3-71 rubber and male
620 72-80 rubber and female
(duplicate 1-6)
Card #14 7-8
621 09-17 rubber and -- ernment
708 18-26 fruits and rice
709 27-35 fruits and other fld crop
710 36-44 fruits and vegetables
711 45-53 fruits and water
712 54-62 fruits and fertilizer
713 63-71 fruits and pests
714 72-80 fruits and yield
(duplicate 1-6)
Card #15 7-8
715 09-17 fruits and cultivation
716 18-26 fruits and good
717 27-35 fruits and enough
718 36-44 frui ts and I/me
719 45-53 fruits and male
720 54-62 fruits and fema le
721 63-71 fruits and government
809 72-80 rice and other l d crop
(dupl icate 1-6)
Card #16 7-8




46
Progress and Achievements--continued, page 6
810 09-17 rice and vegetables
811 18-26 rice and water
812 27-35 rice and fertile izer
813 36-44 rice and pests
814 45-53 rice and yield
815 54-62 rice and cultivation
816 63-71 rice and good
817 72-80 rice and enough
(duplicate 1-6)
Card #17 7-8
818 09-17 rice and I/me
819 18-26 rice and male
820 27-35 rice and female
821 36-44 rice and government
910 45-53 other field crops & vegetables
911 54-62 other field crops & water
912 63-71 other field crops & fertilizer
913 72-80 other field crops & pests (duplicate 1-6)
Card #18 7-8
914 09-17 other field crops & yield
915 18-26 other field crops & cultivation
916 27-35 other field crops & good
917 36-44 other field crops & enough
918 45-53 other field crops & I/me 919 54-62 other field crops & male
920 63-71 other field crops & female
921 72-80 other field crops & government (duplicate 1-6)
Card #19 7-8
1011 09-17 vegetables and water
1012 18-26 vegetables and fertilizer
1013 27-35 vegetables and pests
1014 36-44 vegetables and yield
1015 45-53 vegetables and cultivation
1016 54-62 vegetables and good
1017 63-71 vegetables and enough
1018 72-80 vegetables and I/me
(duplicate 1-6)
Card #20 7-8
1019 09-17 vegetables and male
1020 18-26 vegetables and female
1021 27-35 vegetables and government
1112 36-44 water and fertilizer
1113 45-53 water and pests
1114 54-62 water and yield
1115 63-71 water and cultivation
1116 72-80 water and good
(dupl icate 1-6)
Card #21 7-8
1117 09-17 water and enough
1118 18-26 water and I./me
1119 27-35 water arid male
1120 36-44 water and +femal e
1121 45-53 water and government
1213 54-62 fertilizer and pests




47
Progress and Achievements--continued, page 7
1214 63-71 fertile izer and yield
1215 72-80 ferti l izer and cultivation (dupl icate 1-6)
Card #22 7-8
1216 09-17 fertilizer and good
1217 18-26 fertilizer and enough
1218 27-35 fertilizer and I/me
1219 36-44 fertilizer and male
1220 45-53 fertilizer and female
1221 54-62 fertilizer and government
1314 63-71 pests and yield
1315 72-80 pests and cultivation
(duplicate 1-6)
Card #23 7-8
1316 09-17 pests and good
1317 18-26 pests and enough
1318 27-35 pests and I/me
1319 36-44 pests and male
1320 45-53 pests and female
1321 54-62 pests and government
1415 63-71 yield and cultivation
1416 72-80 yield and good
(duplicate 1-6)
Card #24 7-8
1417 09-17 yield and enough
1418 18-26 yield and I/me
1419 27-35 yield and male
1420 36-44 yield and female
1421 45-53 yield and government
1516 54-62 cultivation and good
1517 63-71 cultivation and enough
1518 72-80 cultivation and I/me
(duplicate 1-6)
Card #25 7-8
1519 09-17 cultivation and male
1520 18-26 cultivation and female
1521 27-35 cultivation and government
1617 36-44 good and enough
1618 45-53 good and I/me
1619 54-62 good and male
1620 63-71 good and femal e
1621 72-80 good and government
(dupl icate 1-6)
Card #26 7-8
1718 09-17 enough and I/me
1719 18-26 enough and male
1720 27-35 enough and female
1721 36-44 enough and government
181' 45-53 ./me and mal e
1820 54-62 I/me and femal e
1221 63-71 I./me and government t
1920 72-:0 male and femal e
(duplicate 1-6:)
Card #27 7-8
1921 0'-17 male and governmen t
2021 18-26 female and government




48
TROPSOLLS
Management Entity, Box 7-113, Raleigh, North Carolina 27695-7113
Work Plan for Next Year:
(Include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
The objectives, outlined in Golfer's revised "Soil Management and People"
proposal (June 1985), include the following:
1. to compare the concepts and "maps" among the relevant ethnic groups and
the soils cientists, to ascertain signficant differences,
2. to discover the nature of the classification systems that relate to
soil in this region,
3. to ascertain whether there is a demonstrable basis from a scientific
point of view, for such folk classification systems.
4. to trace the flow of indigenous and introduced agronomic information,
for subsequent use in extension.
The accomplishment of objective (1) above requires completion of the
Galileo instrument for the Sundanese population, and submission of the data to our collaborator at UH, Dr. Barbara Newton. A Galileo instrument will also be
prepared for administration to soil scientists, to ascertain important
differences in their views vis-a-vis farmers' A joint paper reporting on our
findings to date is anticipated. No problem is anticipated with completing this
work within the upcoming year.
Objective (2) will be a major focus during the upcoming year. Objective
(3) is dependent on our understanding of folk classification systems (2), and
will be addressed as we are able.
TeMy current plan is to postpone attention to Objective (4) until 1986-87.
Interest of Dr. Vickie Sigman (UH) and a UR agronomy and soils graduate
student (Pamela Mills) in investigating the flow of indigenous and introduced agronomic information in more depth provides an opportunity for collaboration i that we must consider seriously.
The open-ended interviews conducted as part of the Galileo can be used as
baseline data for continued study of indigenous knowledge systems. Using these
30 recorded interviews as a body of ethnolinguistic data, hints about the
indigenous soil classification system and its "distinctive features"~ will be
sought. These "hypotheses" will then be further investigated using the methods
of ethnoscience. If the people have elaborate systems of classification related
to soil or other agronomic matters, folk taxonomies will be collected. The
categories in the classification systems will then be analyzed to determine the defining characteristics of the concepts. Triadic sorting methods will be used
to ascertain differences that the people use to differentiate cells in the
taxonomies. Interesting concepts, such as the hot-cold distinction (already encountered repeatedly) will then be pursued in greater detail. Building on
these findings we will do laboratory analyses of soils to ascertain whether or
not the native classification systems have predictive potential for crop
productivity.
M.N83-C
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




49
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: 10/1/85 9/30/86
Project Leader/Institution: Carol J. Pierce Colfer (UH)
Principal Collaborators/Institution:
Evi Martha and Mira Elfina, Andalas University, Padang; Fahmuddin Agus, Center
for Soil Research; Vickie Sigman and Stacy Evensen, University of Hawaii
Research Site(s):
Sitiung V Blok C, and Koto Padang, West Sumatra
Project Title:
Intra Household Decisionmaking
Project Objectives:
1. Ascertain the patterns of intrahousehold decisionmaking among Sitiung
farmers, particularly relating to soil and crop management
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
Vickie Sigman, 9-12 June 1985, Population ouncil/FSSP Case Studies Project
Training Workshop, Boston, Massachusetts (supported by Population Council/FSSP
and UH Strengthening grant).
Manuscripts in Preparation or Published: (Authors), title, journal)
Evi Martha, "Peranan Wanita Dalam Proses Pengambilan Keputusan dan
Kaitannya dengan Pertanian." (Women's Role in the Decisionmaking Process and its
Link to Agriculture); publication outlet not yet decided.
(continued)
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) (Date)
'EWhite: Management Entity Yellow: Pro-ram Coordinator Pink: Proect Leader
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




50
Manuscripts--continued page 2
Mira Elfina, "Tingkat Pendidikan dan Hubungannya dengan Proses
Pengambilan Keputusan Serta Kaitannya dengan Pertanian di desa Koto Padang" (Level of Education and its Relationship with the Decisionmaking Process as Linked to Agriculture in Koto Padang); no plans for publication.
Vickie Sigman, Carol J. Pierce Colfer, and others, "Case Study of
TropSoils Project: Sitiung"; being prepared for publication by the Population Council and the Farming Systems Support Program.




51
TROPSOILS
Management Entity, Box 7113. Raleigh, North Carolina 27695-71 13
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables of data and graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
Research Planned and Performed
The study of intrahousehold decisionmaking, though recognized as important,
is in its infancy. No good methods have, in our estimation, really been
developed. To date, we have used two approaches. The first was to closely
monitor the income and expenditures/consumption in four families for a period of
two and a half months, in Sitiung V. These families, with whom we were also
conducting on-farm trials, were interviewed every three days, with a structured interview relating to specific expenses and consumption. During the interview,
an attempt was made to discuss their intrahousehold decisionmaking in an informal way, to better understand the process. Although the structured
interview data were acceptable, the open-ended component was not adequate. We hypothesized that part of the problem was that our interviewer, though diligent
and intelligent, was not trained in social science.
We therefore requested assistance from Andalas University, which supplied
us with two undergraduate sociology majors. These women lived and worked in Koto Padang, a Minang community within the Sitiung area, from mid-June to the
end of August 1985. They were trained in participant observation methods, and
they developed questionnaires, focusing on intra-household decisionmaking, after
they had been in the community for one month. The intent was to combine the open-endedness of participant observation (useful in an area where little is known), with the advantages of quantification offered by formal surveys. In
addition to the formal survey, the students were requested to develop mini-cases
(following White 1984) of decisions within the Koto Padang families they worked
with.
These students were being asked to undertake research for which a higher
level of training is normally required; however, more highly trained personnel
were simply not available. Mira Elfina undertook a study which ultimately
focused on educational matters. While it was a reasonably good sociological study, its utility for TropSoils is only marginal at this time. However, Evi Martha's work was well worth the "experiment" in using Andalas students. She
basically undertook two surveys in one: A random sample of 30 households,
stratified by the three neighborhoods in Koto Padang, was chosen. Although
Evi's work focused on women's status, she also interviewed 18 husbands (6 in
each neighborhood) on topics relating to the division of labor and
decisionmaking. Neither student dealt with the request for mini-cases of
decisions made, unfortunately. It appears that this method requires a higher
level of training than is available at this time.
Results
In the first attempt (January March 1985), which included two Javanese
and two Sundanese families in Sitiung V, we found a full spread of
decisionmaking within the family. The wife in one of the Javanese families
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




52
Progress and Achievements-continued, page 2
was clearly the major decisionmaker. Her husband was fearful of interacting with outsiders and she seemed fully in charge. The husband in one of the Sundanese families was quite dominant, basically deciding everything, including details of food consumption. The other two families appeared to manifest a joint decisionmaking pattern.
One point of interest is that the meagreness of resources in these
families appears to reduce the decisionmaking potential. Money and even time frequently simply go for some obvious need, without anyone particularly choosing or deciding what to do with that resource. Subsistence requirements and health requirements are sometimes very dramatic, too dramatic for decisions to need to be made. If the money is there, it goes for that purpose.
The report from Koto Padang was just submitted at the end of August and has yet to be translated into English. However, some of its most interesting findings follow (subject to revision on closer inspection of the data):
- 75% of the women interviewed reported being involved in some kind of
income generating activity (farmer, unskilled laborer, or petty
merchant)
- The reported involvement of Minang women in each of the 9 steps in
local wet rice production, from ground preparation through carrying the yields home, surpasses men's, except for spraying and carrying the rice
home from the fields. Women's involvement in home gardening far
surpasses men's in the planting and weeding activities, and slightly
surpasses it in harvesting.
- Men, on the other hand are more active in smallholder plantation
activities, both in connection with the ADP rubber replanting project,
and in private holdings with other tree crops.
- Women expressed general ignorance about new seeds, fertilizer and
pesticide management (though in other interviews with men, the same
ignorance has frequently been expressed). Unfortunately, only women
were queried on this topic in this survey.
- When queried on decisionmaking regarding seeds, fertilizer and
pesticides, responses were fairly even distributed among the following family actors: wife (5), husband (4), couple (4), parents (5), and "do
what others do" (4).
- A group of questions on intra-family and inter-family decisions reveals
a pattern of generally joint decisions, with slight male dominance
(within the general Indonesian context of not forcing others).
One important reason for conducting this research comes from global
research concerns about the role of women in agriculture. Although the fact that women are active farmers is widely accepted now, many maintain that their labor is irrelevant, since the decisionmaking power remains with men. One purpose of this research is to establish whether this is in fact the case here. Our preliminary findings, reported above, suggest that women do have an active voice in agricultural decisionmaking in both locations. Our plan to implement some agroforestry trials on both transmigrant and local Minang fields requires better understanding of Minang human conditions. To match a new technology with the people who will use it, we need to know understand who will be making the decision to use or not to use the technology. Decisionmaking data are important, as well, in the creation of the farming systems expert system we are working on.




53
Progress and Achievements--continued, page 3
Literature Cited
White, Benjamin. 1984. Measuring Time Allocation, Decision-making and Agrarian Changes Affecting Rural Women: Examples from Recent Research in Indonesia. Institute of Development Studies Bulletin 15, No 1:18-31.




54
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
Work Plan for Next Year:
(Include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
Objectives remain the same for this study, and further effort will be made
to collect the mini-cases of decisionmaking discussed earlier. TropSoils needs
to have information on decisionmaking among each of the three major ethnic
groups. Evi Martha will be writing her thesis this coming year, and her
continued participation will be sought. Revision of Evi Martha's paper is
anticipated, for publication.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




55
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: 10/1/85 9/30/86
Project Leader/Institution: Carol J. Pierce Colfer, University of Hawaii
Principal Collaborators/Institution:
Stephenie Kan, University of Florida
Research Site(s):
Sitiung V Blok C, West Sumatra
Project Title:
Economic Evaluation of Soil Management Technology
Project Objectives:
To be developed.
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
None
Manuscripts in Preparation or Published: (Authors), title, journal)
None
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) (Date)
We-08 o g3-A
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




56
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables of data and graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
This study is in the process of being planned. However, Dr. Peter
Hildebrand has been working with Stephenie Kan on a proposal to do her research
for her masters degree in agricultural economics in Sitiung. She anticipates coming in January 1986 to begin the study, and has already requested and been
sent the data from the previous TropSoils time allocation study. She hopes to
develop a preliminary model for further analysis of those data, using linear
programing.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




57
TROPSOLLS
Management Entity, Box 113, Raleigh. North Carolina 279-11
Work Plan for Next Year:
- I (Include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
Although a precise plan has not yet been provided, it will be consistent,
in broad outlines, with the proposal Colfer submitted in June 1985. Ms. Kan
anticipates working closely with the agricultural scientists in evaluating
proposed agricultural technologies from an economic standpoint; and she will work on the difficult question of placing an economic value on essential but
usually not priced activities (a major problem for economists in partially
subs is tence economies).
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




58
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: None for 1985-86
Project Leader/Institution: Carol J. Pierce Colfer, University of Hawaii
Principal Collaborators/Institution:
Russell Yost, University of Hawaii
Research Site(s):
Sitiung I Blok A and Sitiung V Blok C, West Sumatra
Project Title:
Time Allocation Study
Project Objectives:
I. Quantify changes in the allocation of time in Sitiung, over the course of
the project,
2. Help in evaluating the impact of project activities on the lives of farmers
in the immediate vicinity of our research.
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
Carol J. Pierce Colfer, 4 October 1984, University of Hawaii, Honolulu,
Hawaii, "Variability of Agricultural Labor Allocations Among Village Sites:
Men, Women and Children." (continued)
Manuscripts in Preparation or Published: (Author(s), title, journal)
Carol J. Pierce Colfer, Russell Yost, Veronica Kasmini, and Atin Kurdiana,
"Time Allocation Studies: A Methodology for Food Production Systems"; submitted
to the Bulletin of Indonesian Economic Studies, spring 1985.
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) (Date)
\AE-09-83-A
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




59
Meetings Attended--continued page 2
Carol J. Pierce Colfer, 16 October 1984, Cornell University, Ithaca, New
York, Symposium on Household Food Production, Keynote Address: "Time Allocation Studies: A Methodology for Food Production Systems."
Carol J. Pierce Colfer, 19 October 1984, University of Minnesota, St. Paul, Minnesota, "Gender-Related Values in Development: Culture-Specific Concerns and Examples."




60
TROPSOLLS
management Entity, Box 7113, Raleigh, North Carolina 27695-71 13
Progress and Achievements:
- I (Include: Description of research initiated or performed, results obtained, tables of data and
graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
The approach and implementation of the time allocation study were described
in the previous annual report; however data collection was completed in
mid-October 1984. This year's activities have focused on analysis of the data,
and preparation for presentation to others. We had identified a number of
project-related information needs which we expected the time allocation study to
answer. These included the following:
- Division of labor by sex
- Division of labor between agricultural and other income producing
activities
- Division of labor among the various agricultural activities possible
- Seasonal variation in labor allocation
- Leisure time available
- Time spent on providing feed for animals
- Variation in the different locations within Sitiung
Agriculture, off-Farm Labor, and Leisure by Season Sitiung I and V
Looking at the attached graphs, the first three (Figures 1-3) show seasonal
variation in Sitiung I and V, with regard to allocation of time to agriculture,
paid or contract labor (including home industry), and leisure. The Y axis
represents average hours per person per month (with average family size being
about 5); and the X axis represents the months of the year (1=January).
* The most striking pattern was the greater amount of leisure time in Sitiung
V, relative to Sitiung 1. The time allocation study was begun during the first
year of settlement of Sitiung V. These data suggest that opportunities for
productive activity increase with the length of settlement. However, the obvious need for further land clearing activities during that time (since
settlers' fields were littered with logs when they arrived from Java) suggests
* that the many changes and disappointments on arrival may have demoralized people to such a degree that their productivity was affected. The subsidy provided to
transmigrants during their first year may also have adversely affected their
* productivity, necessary though that subsidy is. The availability of sawah (wet
* rice fields), the necessity to be fully self- supporting, and the expectable payoff (in yields) for agricultural activity may similarly have contributed to
the greater productivity in Sitiung I.
Not surprisingly peak agricultural times roughly follow the seasonal
variation, with heavy periods at the onset of the rains and rice harvest times.
The Sitiung V peak later in the year (April) undoubtedly relates to the marginal
harvests, and subsequent planting of legumes for a second crop.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




61
Progress and Achievements-continued, page 2
From the standpoint of off-farm work, there is again a rough
complementarity with the agricultural seasons, the main point being the low level of non-agricultural activity at the onset of the rains (sometime between August and October). However, the high levels of both off farm and on farm labor during March and May in Sitiung I suggest that people respond with increased work when there are multiple opportunities, rather than substituting one kind of work for the other. In Sitiung V, October represented a similar situation.
The relatively high levels of productive activity and leisure in October suggest that the remaining major category of activity (what we are calling broadly, "reproductive" activity) may have suffered neglect.
Production, Reproduction, and Leisure
Besides off-farm labor, household activities necessary for human survival can constitute a significant drain on human resources which might otherwise be available for productive activity. People are not available for productive work if they do not eat, bathe, dress, repair things, and ensure that children survive for a future workforce. This first bar graph (Figure 4) shows the division of adult labor between activities which result in the creation of value ("Productive activities"), those devoted to the maintenance of the work force ("Reproductive activities") and the remainder ("Leisure").
The fact that reproductive activity requires more time than productive activity, in Sitiung V (in contrast to the Sitiung I situation) is partially related to a higher proportion of children under 15 in Sitiung V than in Sitiung I. The easier availability of markets in Sitiung I may also be a major factor.
Kinds of Fields, Sex, and Seasonality
The next three graphs (Figure 5-7) show seasonal variation in the
allocation of labor to the upland fields, home gardens, and wet rice fields, by sex, in Sitiung I. The Y axis again reflects average hours per person per month; and the X axis represents the months. Females are coded 1 and males are coded 2.
One of the most interesting findings reflected here is the relatively high proportion of labor devoted to the home garden (accounting for 1/4 ha, compared to the 1 ha devoted to other fields combined). Our previous attempts to work with women farmers were not particularly successful, and the involvement of women in home gardening activities, as well as the significant amounts of time devoted to home garden production, as shown here, were important in our decision to pursue more research in the home garden context. That upland fields (ladang) reflect the seasonal variation more than does wet rice (sawah) is not surprising, since the sawah schedule is determined by the government irrigation project.
In Sitiung V there is no sawah (Figure 8 and 9). The complementarity of
busy times in the home garden and the ladang is quite clear. Rice is the major ladang crop, and it must be planted at the time the rains come. In the early years of a settlement, the home garden appears to be cultivated intensively throughout the year, for subsistence purposes. Men and women are more equally involved in home gardening activities in Sitiung V than in Sitiung I.
The second bar graph shows hours per day per adult devoted to sawah, ladang and gardens in Sitiung I and V, averaged over the whole year.




62
Progress and Achievements-continued, page 3
Frequencies of all activities in the 5,635 observations of individuals remain a valuable resource data set and available to us as we develop new research questions relating to people's behaviour and its impact on soil management practices.




0
n
3
Ft
C
0
FF
1 2 -m
E
L
0
2- '11 "-U
IIIIIiII I I T FI
01 02 03 04 05 06 07 08 09 10 11 12
MONTH
LEGEND: LOCATION --- 51TV -- SIT1
Figure 1. Seasonal variation in Sitiung I and V with regard to allocation
of time for agriculture.




0
'3
M
0
2.0
1. 5- t' ,
I 1. \
0.50
0 0
I' I" I'
/ I 3.
I iI
/ I /13
E V
I \ I / (1
I Is I a
I I /
/ I I."?
III III ll l.
0. 5- L
.0
01 02 03 04 05 06 07 08 09 10 11 12
MONTH
LEGEND: LOCATION SITV ..-.--.. SITI
Figure 2. Seasonal variation in Sitiung I and V with regard to allocation
of time for hiring paid or contract labor.




'1
0
OQ
CD
4
n
3
L 00g
E H
I CD
U V O
B 2- ,O
E m
\ j I
1 I I I I I | I i i I I I
01 02 03 04 05 06 07 08 09 10 11 12
MONTH
LEGEND: LOCATION SITV ----- SIT1
Figure 3. Seasonal variation in Sitiung I and V with regard to allocation
of time for leisure.




Progress and Achievements--continued, page 7
10
8
] Sitiung I
6 Sitiung V
o 0
4
2
O '
0-
Activity
Figure 4. The division of adult labor among activities.




'1
OQ
0
I
1 1
,~ /- ,, I \
\ I 1*I
%/ 0
--- j 'I 'ir
LI
A 0.50- ,
II
G1 1
S0. 25I1
I1
\ ...
#, ,
N \\
0.25-,
0.00
01 02 03 04 05 06 07 08 09 10 11 12
MONTH
LEGEND: SEX 1------- 2
Figure 5. Seasonal variation in the allocation of labor to the
upland fields (ladang) of Sitiung I.




0
04
01
0.6
0
0.6 4B C O
00
0.0-.
01 0.4 4 50 7 80 1 11
MONTH
LE E D SE --Figue 6 Seaona var~aton i th allcaton o laor t th
NoegresinStug1




0
IOCATION=SITI
El
/'1
0
0.75-,
I '- 1 ,
I \./ O'
S 0.50- ,
Wn
H m
I..
A /
0.25-,
I
I
0. 0001 02 03 04 05 06 07 08 09 10 11 12
MONTH
LEGEND: SEX 1 ------- 2
Figure 7. Seasonal variation in the allocation of labor to the
wet rice fields (sawah) of Sitiung I.




-1
0
LOCAT I ON=S IV/
m
0. 750
A 5N
L I I
/ I 0
G
/ /I
0.25- 0-,
0.00
01 02 03 04 05 06 07 08 09 10 11 12
MONTH
LEGEND: SEX 1 ------- 2
Figure 8. Seasonal variation in the allocation of labor in the
upland fields (ladang) of Sitiung V.




LOCATION=SITV
/1
G. 4- ,
1 1
I I
/ I
0.3-'
GI
AA
R\ /
L 0 Cii ON=SIT
Bn
S 0.2- '
E \s /I / \ I
N \ 'I \ !
\ I 1 1
0. 1- '
' \!
II
0. 0
IIIIIIIII I
01 02 03 04 05 06 07 08 09 10 11 12
MONTH
LEGEND: SEX 1 ------- 2
Figure 9. Seasonal variation in the allocation of time for home gardens
in Sitiung V.




72
TROPSOILS
Management Entitv, Box 7113, Raleigh. North Carolina 27695-7113
Work Plan for Next Year:
(Include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
None planned until 1987.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




73
TROPSOILS
management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: none for FY 86
Project Leader/Institution: Carol J. Pierce Colfer, University of Hawaii
Principal Collaborators/Institution:
Barbara Chapman, University of Hawaii
Research Site(s):
Sitiung I, II, V, West Sumatra
Project Title:
Nutrition/Income Survey
Project Objectives:
1. Provide further descriptive analyses of 1984 nutrition income survey,
focusing on crop diversity, production, and income.
2. Provide the team with a summary of literature on nutritional status on
Java, for purposes of comparison.
3. Do a multivariate analysis of the factors related to better diets among
these settlers.
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
None
Manuscripts in Preparation or Published: (Author(s), title, journal)
Barbara Chapman, "Diet and Production Survey of Sitiung, West Sumatra:
Research Memos 2 and 3"; perhaps The Ecology of Food and Nutrition, but not yet
submitted.
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) (Date)
M E-Eo -8 n-:
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




74
TROPSOILS
Management Entity. Box 7113, Raleigh, North Carolina 27695-7113
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables of data and graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
The analyses reported here are based on a survey undertaken in April 1984
in Sitiung I, II and V. The survey was planned by Carol Colfer and Barbara
Chapman, conducted by Bartholomeus Weid, Harry Apriadji and Liek Irianti, and
analyzed by Barbara Chapman.
Description
A variety of descriptive data are available in the Chapman report, (1984).
One of the most important findings was the low level of agricultural
productivity of Sitiung's farms. The major crops with the percentages of
farmers producing particular crops worth less than $100 ranged from 70% for rice
to 100% for cassava and banana. Of 77 families in the three locations, 36 had
total farm incomes (from agricultural produce, including the value of produce
consumed by the family) of less than $100; another 36 had agricultural incomes
between $101 and $500; and only 5 had farm incomes over $500.
The diversity of crops, important from a nutritional standpoint, is most
evident in the home garden---though on site experience suggests that the
diversity in home gardens is considerably greater than that reflected in the
survey.
One concern of the team has been the degree to which settlers are (and can
be) dependent on agriculture for their livelihood.
The availability of cash is an important concern for those interested in
enhancing agricultural productivity by increasing inputs. Annual cash incomes
ranged from under $100 to over $2,000. Thirty families had under $100 in annual cash income, 29 had between $100 and $500, and 18 had cash incomes of over $500.
Although perhaps higher than cash incomes would be in Java, they are still low
for significant allocation of funds to fertilizers, pesticides, and other
agricultural inputs.
Java Compared to Sitiung
Chapman's previous research had been in Gunung Kidul, an area in Central
Java, not far from Wonogiri (the home area of inhabitants of Sitiung I and II).
Drawing on her own experience as well as the published literature, she
identifies the following probable nutritional difficulties in the settlers' sending communities: calorie/protein malnutrition, Vitamin A deficiencies,
anemia, and goitre. She points out that since 1) many of the people come from an unusually poor part of Java, and 2) transmigrants tend to be from among the
poorer segments of any particular local community, the probability of their
having experienced these nutritional difficulties is high.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




75
Progress and Achievements-continued, page 2
In Sitiung, the dietary pattern is not very different from patterns found earlier in Java by Chapman and others. Meals are starch and vegetable centered with most of the protein derived from rice and the bean/peanut family. Meat, eggs, fish and milk products are rare. The overall conclusion though is that settlers are probably considerably better off nutritionally than they were in Java. Still roughly 50% of the households in Sitiung appear to be getting less than their nutritional needs in calories, protein and vitamin A.
Factors Related to Better Diets among Settlers
Chapman postulated that seven household characteristics should have an
influence on nutritional status, as measured by the three indices, per capita calories, protein and vitamin A. These eight household characteristics were ethnicity, length of residence, education, total income, and diversity indices of agricultural production, of vegetable and fruit production, and of consumption. However, these characteristics only accounted for 28% of the variation in per capita calories, 31% of the per capita protein, and 19% of the per capita vitamin A, using a multivariate analysis.
The diversity index of consumption had the highest explanatory value in regard to calorie and protein intake. Educational attainment of the highest educated member of the household contributed strongly to these same nutrients. Ethnicity was important in per capita protein (the Minangkabau eat more meat).
A significant, and frequently found conclusion, is that household income is singularly unimportant in explaining a good diet in Sitiung. For vitamin A,
the strongest factor was the diversity index of vegetables and fruits, with length of residence in Sitiung and educational attainment right behind.
Some specific recommendations emerging from this study include the following:
- Research on home gardens should encourage the planting of a number of
non-commercial plants high in vitamin A and C (e.g., amaranth, edible
pod-beans, and tubers of a deep color).
- Increased dietary diversity through planting a greater variety of crops,
since Sitiung families are primarily dependent on their own production
for their diets, and thus cannot afford to take a narrowly commercial
view of cropping strategies.
- Investigation of diet and cropping patterns in area of origin, in search
of more species of plants which can be tried in Sitiung.
- Increased incomes only result in moxdest improvements in diets. Income
generating activities that directly increase the variety of foods available, such as garden plantings of vegetables and fruits now
unavailable or household processing of agricultural products into salable
snacks at the markets and school yards are recommended. Both of these
activities could predominantly involve women, since they are active
entrepreneurs in these activities in Java.
Literature Cited
Chapman, Barbara. 1984. Diet and Production Survey of Sitiung, West Sumatra.
Research Memos 2 and 3.




76
TROPSOLLS
Management Entity, BoX 7113, Raleigh. North Carolina 27695-7113
Work Plan for Next Year:
(Include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
None until 1988.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




77
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: 10/1/85 9/30/85
Project Leader/Institution: Carol J. Pierce Colfer, University of Hawaii Mike Wade, North Carolina State University
Principal Collaborators/Institution:
Carl Evensen, University of Hawaii; I Putu Gedjer Widjaja-Adhi and Fahmuddin
Agus, Center for Soil Research; Dan Gill, North Carolina State University
Research Site(s):
Sitiung, West Sumatra
Project Title:
Collaborative Research with Farmers on Upland Fields
Project Objectives:
1. Gain access to the knowledge and experience of the farmers in a
practical, hands-on fashion,
2. Provide a testing ground for agricultural technologies that appear
appropriate for this area,
3. Determine in an iterative manner the most important limiting factors
affecting these farmers' production.
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
Carol J. Pierce Colfer, 3 October 1984, University of Hawaii, "Adapting
Field Research Designs to Incorporate Farmers' Technologies".
(continued)
Manuscripts in Preparation or Published: (Author(s), title, journal)
Carol J. Pierce Colfer, Mike Wade, Atin Kurdiana and Suwandi,
"Incorporating Farmer Conditions and Preferences into Farming Systems Research";
submitted to Agriculture and Human Values.
(continued)
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) IDate)
ME-09-83-A
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




78
Meetings Attended--continued
Carol J. Pierce Colfer, 10 October 1984, World Bank, Washington, D.C., "Incorporating Farmer Preferences into Farming Systems Research".
Carol J. Pierce Colfer, 17 October 1984, Cornell University, Ithaca, New York, "Farming Systems Research in Indonesia: The TROPSOILS Example".
Carol J. Pierce Colfer, 19 October 1984, Department of Soil Science, University of Minnesota, "The TropSoils Project, Sitiung, West Sumatra".
Carol J. Pierce Colfer, 12 August 1985, Workshop on Subsistence, Stability and Sustainability for the Small Farmers of Asia, Environment and Policy Institute, East-West Center, Honolulu, Hawaii. "The TropSoils Project, Sitiung, West Sumatra".




79
Manuscripts in Preparation or Published-continued
Carol J. Pierce Colfer, "People Factors in Farming Systems: An
Anthropological Perspective;" Revised at the request of the WID Working Papers Series, Michigan State University.
Fahmuddin Agus, Mike Wade, "The Incorporation of Information from
Collaborator Farmers into a Research Strategy: A Till-No- Till Example from Sitiung." Paper being prepared for presentation at the International Farming Systems Meeting to be held at Sukarami, West Sumatra in December 1985.




80
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables of data and graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
In response to the farmers' complaint that the control plots did not
produce well, and the fact that our plan to plant mucuna as a green manure crop as intended the first year was not successful, we harvested Calapagonium from a
nearby rubber plantation and distributed it to be used as green manure (10 T
fresh/ha) on the control plots. The other treatments had only inorganic
fertilizer. The government treatment was limed according to the government's
liming program (3 1/2 T/ha) and continued to get 100 kg TSP and urea per ha.
The rock phosphate treatment was unchanged, also getting the standard 100 kg TSP
and urea per ha, in addition to the base application of 800 kg of rock
phosphate/ha that was applied at the beginning of the trial. The lime treatment
got a maintenance lime application of 0.5 T/ha as well as 200 kg TSP, 100 kg
urea and 50 kg KCl/ha.
Rice seed was again obtained from West Java, a local "red rice for upland
fields. The overall farmer averages by treatment are as follows: Treatment kg/plot (200m2)
1. "control 25
2. gov't 20
3. rock phosphate 24
4. lime+NPK 27
There is no significant differences among these yields and thus no real response
to the lime and fertilizers. However the "control", with its green leaf manure
(GUA) application, performed remarkably well and produced interesting results.
Not all farmers handled their GLM in the same manner. Some farmers (4)
buried the material in furrows (plowed under), others (5) spread the material
and hoed it in (thereby mixing it with the soil but with partial exposure),
still others (4) merely mulched the soil with the GLM, and two farmers failed to
use it at all. When the GLM was not used the control yielded only 65% of the
fertilized plots. Mulching or mixing of the GLM helped some, but the buried GL1M
was dramatically better than the other methods as well as better than the
inorganically fertilized plots.
Peanuts was the second crop in the season and responded as follows to the
fertilized treatments:
Treatment Yield (kg/plot)
1. "control" 11.0
2. gov' t 15.5
3. rock phosphate 14.0
4. lime +NPK 16.2
LSD.05=1.9
Both of the limed treatments (#2 and 4) gave the best results, which may be
expected due to peanuts' specific requirement for Ca during pod fill.
(continued)
W E-04-8 o8
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




81
Progress and Achievements--continued, page 2
Again however organic matter management was critical. This time
utilization of the rice straw had a strong impact on the peanut response to the fertility levels. Most farmers (7) burned the straw (on the plots), while three removed the straw and two incorporated it. When straw was removed, burned or incorporated, the control plots yielded 40, 71 and 110% respectively of the lime plots.
The rambutan fruit trees were measured for height and diameter one year
after transplanting. There has been no treatment effect on the initial growth of these young trees, as tree development seems to be more or less independent of fertility treatment and farmer practice.
The value and management of organic materials, either as GLM or crop
residue is apparent. These initial results strongly indicate there is a real possibility of alleviating the need for lime and fertilizers by proper management of plant material. The need for more research as to how and why these relatively modest inputs can produce (and hopefully maintain) food crop yields is obvious. Also, methodology or systems to produce this positive effect that would be feasible and practiced by farmers must be developed.




82
TROPSOLLS
Management Entity. BOX 711 3. Raleigh. .\orth Carolina 2.7695-7-1 13
Work Plan for Next Year:
(include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
All of the Minangkabau (indigenous) farmers (5) have dropped from the
program as they have basically abandoned their fields and are living in their
* original village of Koto Padang. One of the Javanese farmers has withdrawn due to poor health. Thus we currently have 13 active cooperator farmers. We had planted Centrosema on 1/3 of each plot of all farmers to establish a source of
GLM for successive crops. However this planting was done lust before the dry season and did not establish well. The intention was to leave this 1/3 of the
plots as a permanent source of GLM but all farmers hoed it under (what little
* there was) in preparation for rice planting. Current plans in Sitiung V are to try to re-establish a source of glm on the plots during this wet season and also
take another look at the effect of rice straw management on the subsequent
peanut crop, with the remaining 13 active cooperators.
The following additional on-farm research and validation of component trial
results is anticipated, if the person power is available on site to carry it
out.
Subtitle: Systems Management of Annual, Cover and Tree Crops
Objectives:
1. To evaluate recommended fertilizer and lime package on annual crop
performance in farmer managed fields,
2. To study the effect of using green leaf manure on crop production in
the cropping system,
3. To study the effect of intercropping annuals and cover crop management
on tree crop production,
*4. To evaluate production, and farmer acceptibility and preference of the
various management combinations.
Several factors have surfaced as pertinent for additional research. These
include (a) the need for long-term systems research, studying management
* alternatives in going from recent clearing to a stable, productive system, (b) the need to test our initial two years of component fertility research, drawing
the results together as a recommended package to the farmer, (c) the need for
research on tree crops in a cropping system, as perennial tree crops have
historically been treated as monoculture plantation crops and because they are
so obviously well-suited to the acid, sloping land and insect and disease
pressures of a humid, tropical climate, and (d) the need to evaluate our
research across a wider spectrum of farmers and conditions in Sumatra.
This experiment is an attempt to consolidate these factors into a series of
* farmer-managed locations in several Outer Island locations. We hope to maintain fairly close contact with the farmers however in order to minimize deviation
from the intended plan. The best and ultimate evaluation of any system is the
farmers preference and adoption of the management practices, so we feel that (continued)
White: Management Entity Yellow: Program Coordinator Pink: Proiect Leader




83
Work Plan for Next Year--continued, page 2
that this "systems" research should be done there, on farmer fields. This is not, however, simply a farmer adoption experiment. Serious data collection will be made to evaluate the various treatments on the annual, cover and tree crop yields.
This can perhaps be called second generation research for the
TropSoils-Indonesia program. This experiment will serve at least two very important functions. One, it is, as "second generation" implies, an outgrowth of initial research results and identified needs for future research, and is a critical step towards developing tried and true recommendations for both farmers and policy makers (3rd generation). Two, it should provide important feedback information to the researchers as regards (a) new areas to be researched, (b) confirmation of or adjustments needed in current research and (c) better understanding of farmers' preferences.
Generalized Procedures:
i. Systems (main plot)
a. annual food crops without green leaf manure (rice-soybeans)
b. annual food crops with green leaf manure harvested from system c
c. tree crop (cocount) with cover crop (mixture of Pueraria, Calapagonium
and Centrosema) that is harvested as glm for system b
d. tree crop with cover crop (managed as per usual plantation practices)
e. tree crop with annual food crops (same as a and b) intercropped for
first two years of tree establishment
2. Fertility Packages (split-strip plot, on systems a and b only)
a. Ff-farmer practice,
b. Fr-recommended; lime and fertilizer to critical levels of macro
nutrients-lime to <20% acid saturation
N-50 kg urea/ha (rice only) P-40 kg P/ha initial-10 kg/ha/crop K-50 kg K/ha/crop
Mg-8 kg Mg/ha (soybeans only) c. Fh-high rates, well onto yield plateau
lime to 0% acid saturation N-100 kg urea/ha/rice crop P-160 kg P/ha initial-20kg/ha/crop K-120 kg K/ha/crop
Mg-16 kg Mg/ha/crop
micronutrients to be added if evidence warrants.
Principal Site for Experimentation
Six locations (five transmigration and one indigenous village) will be selected with four participating farmers at each location (24 farmers). The target villages are:




84
Work Plan for Next Year-continued, page 3
Sitiung I a/b, West Sumatra
Sitiung II e, Jambi
Sitiung III c, West Sumatra
Sitiung IVb/d
Kuamang Kuning, Jambi
Koto Padang, West Sumatra (near Sitiung, indigenous)
Duration
The trial should continue until stable tree crop production is achieved, which would probably be 8 years for hybrid coconut.




85
TROPSOILS
,Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: 10/1/85 9/30/86
Project Leader/Institution: Fahmuddin Agus, Center for Soil Research
Carol J. Pierce Colfer, University of Hawaii
Principal Collaborators/institution:
Samir El-Swaify, University of Hawaii; Soleh Sukmana, Center for Soil Research
Research Site(s):
Sitiung, West Sumatra
Project Title:
Land Management with Farmers in Small Catchment Areas
Project Objectives:
i. Evaluate various erosion control methods on farmers' fields, for
effectiveness and potential for use by farmers,
2. Ascertain best feasible systems (including possible pasture crops, food
crops, tree crops, and fish and livestock components) to utilize on the
various land types represented.
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
None
Manuscripts in Preparation or Published: (Author(s), title, journal)
None
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) (Date)
ME i09-- 5-A
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




86
TROPSOLLS
Management Entity, BOX 711 3, Raleigh, N~orth Carolina 27695-7113
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables of data and graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
Activities to achieve objectives have been carried out in close
collaboration with four families whose fields comprise a small catchment area.
The farmers currently have four experimental plots on their fields, examining
different erosion control methods. Currently bench terraces, grass strips with
rice, grass strips with legumes, and no terrace/no strip, each with coconuts
interplanted, are being compared for their erosion control capacity and effects
on crop growth on the steep areas of each farmer's field.
Progress has been marred by a variety of unavoidable, but troublesome
events: (1) the coconut seedlings we began seeking in August 1984 were not available until June 1985 (2) One of the four farmers returned to Java for
several months with his family; and another became involved in a village
controversy which resulted in several of the other trees on his field being cut
down and (3) the arrival of a pilot project under another government agency to
* terrace the fields in the area of our experiment. One of the families wound up terracing their field consistent with the pilot project plan. Now none of the
farmers has experimental plots that are comparable with any other.
Although we remain committed to the objectives outlined for this project,
* after considerable intra-team discussion of potential benefit to be derived from continuing working in this catchment area, we have decided to redesign this effort in concert with the soil conservation project of Dr. Samir El-Swaify.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




87
TROPSOILS
Management Entirv, Box 7113, Raleigh. North Carolina 27695-71 13
Work Plan for Next Year:
(Include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
The objectives for this study remain the same. Since Samir El-Swaify's
erosion work will be beginning shortly, there are important advantages to
working closely with him. He has already expressed interest in developing a
plan to compare several catchment areas (as total experimental units), managed according to principles developed within soil science linked to the experience
of the team working in the Sitiung area. It is probable that the research
proposed here would include tree crops planted along the slopes of the catchment
area, intercropped with food crops (particularly in the early years), and with
various food and pasture cropping systems covering the flatter areas. Wet
rice/fish culture systems may be appropriate for the areas along the streambed.
Erosion under such circumstances could be profitably compared to erosion under
natural forest cover, under the indigenous patterns of land use in such areas
("wild" rubber growing in the forest, with wet rice in the swampy areas), and
under the typical Javanese patterns of planting these areas to food crops (consistent with government policy, until June 1985). Most of the effort
anticipated for this year will be planning. The exact kinds of management
strategies that are most important to look at, the question of replications, the
difficulty of finding roughly comparable catchment areas will all require
considerable thought and planning.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




88
TROPSOILS
iManagement Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: 10/1/85 9/30/85
Project Leader/Institution: Carol J. Pierce Colfer (University of Hawaii)
Principal Collaborators/Institution:
Edi Joniarta and Fahmuddin Agus, Center for Soil Research,
Stacy Evensen, University of Hawaii
Research Site(s):
Sitiung I and Sitiung V, West Sumatra, Indonesia
Project Title:
Collaborative Research with Farmers on Home Gardens (1) Effects of Different
Sources of Waste Material on Soil Properties and Horticultural Crops in Home
Gardens
Project Objectives:
To measure the effects of different sources of waste material on soil chemical
and physical properties, and on horticultural crop production
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
None
Manuscripts in Preparation or Published: (Author(s), title, journal)
None
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) (Date)
Whit: Mi Y3-a
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




89
TROPSOLLS
Management Entitv, Box 711 3, Raleigh, North Carolina 27695-71 1 3
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables of data and graphs as appropriate, interpretations, potential application and utilization, Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
Twelve families were to be selected (six in each location) whose home
gardens were to provide the location for an experiment on the use of various
waste materials as soil amendments. After a survey of available materials for i composting was completed, a "compost formula" was to be developed from a representative source of materials. One farmer was to be selected to
manufacture the compost according to the researcher-determined formula, in order to minimize variability among farmer-participators. Each farmer is to have four
plots: compost; barnyard manure; a complete N, P, and K treatment; and a
control. The crops currently planned are vegetable crops, to be planted in
succession over a one year cycle. The specific crops planted and the timing
will be determined collaboratively with farmers.
This year a compost formula was created, based on available materials, and
a farmer was hired in each location to make the compost, beginning in June. In
September, 12 families were selected. Interest in incorporating a fishpond
sludge component resulted in our choosing 8 home gardens with fishponds. On
those locations, a plot fertilized with the sludge from the fishponds will be
added to the other four plots, for comparative purposes. Chili was planted in
September 1985.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




90
TROPSOILS
Management Entity, Box 7113, Raleigh. North Carolina 27693-7113
Work Plan for Next Year:
(Include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
There are no changes in the objectives. A series of three crops are
planned to be planted in the experimental plots of the twelve home gardens over
the course of one year. Soil analyses of the plots will be done prior to any treatment, and then the treatments will be re-applied prior to each planting.
Food production and soil chemical and physical parameters will be measured
following each harvest.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




91
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: 10/1/85 9/30/86
Project Leader/Institution:
Carol J. Pierce Colfer and John Thompson, University of Hawaii
Principal Collaborators/Institution:
Carl Evensen, University of Hawaii
Research Site(s):
Sitiung 1, West Sumatra, Indonesia
Project Title:
Collaborative Research with Farmers on Home Gardens (2) Pasture Trials
Project Objectives:
To begin assessment of the productivity and appropriateness of forages within
home gardens
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
None
Manuscripts in Preparation or Published: (Author(s), title, journal)
None
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) (Date)
WhEie nt9-A
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




92
TROPSOILS
Management Entity, Box 7113, Raleigh. North Carolina 27695-7113
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables of data and graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
Ten farmers familiar with the forage trials underway in Sitiung I requested
and were given forages (Brachiaria dictyoneura and B. decumbens grasses) for planting in their home gardens. No attempt was made to choose farmers; they
were entirely self-selecting, and no one was refused. They were instructed in
planting methods, and given the materials. Subsequent inspection was made of
five home gardens to ascertain success.
The principal result of this project has been some rethinking on the part
of project personnel regarding the acceptability, in the Sitiung area, of
pasture in the conventional sense. Although we have had some questions about
the willingness of people to plant pasture from the start, our evidence
regarding the amount of time spent on cutting and carrying, as well as the
erosion prevention potential of pasture were sufficient to convince us to devote
some project effort to this area. However, continued interaction with farmers
has increased our concern that they may indeed be unwilling or have insufficient
land to plant pasture crops. Since rainfall during the period immediately
following dispersal of the cuttings was almost non-existent, an adequate
agronomic "test" on farmers' fields has yet to be accomplished. Only one of the
farmers had significant establishment of the pasture crops, although with the
start of consistent rains, some plantings previously thought dead, have begun to
grow.
Some scope for planting of productive "bunch type" grasses(e.g., Setaria
ancepts, Andropogon gayanus, etc.) is envisioned as feasible and desired by the
farmers for their home gardens. Also a potentially productive research area
might be pasture/cover crop plantings under tree crops. However, the main
limitation to further research on these topics is insufficient researcher time.
The team is currently rethinking our commitment to pasture.
\ E)Q-8 5-8
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




93
TROPSOILS
V1anagement Entitv, Box 71 13, Raileigh. North Carolina 27695-71 13
Work Plan for Next Year:
(Include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
Given the shortage of personnel, no specific plan has been formulated nor
is any budget being set aside for this activity. However, insofar as continued
informal collaboration of this kind with farmers is possible in the ordinary performance of other duties, C. Evensen intends to continue working with the
farmers.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




94
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
ANNUAL PROGRESS REPORT AND WORK PLAN
Progress Report Period: 10/1/84 9/30/85 Work Plan Period: 10/1/85 9/30/86
Project Leader/Institution: Stacy Evensen, University of Hawaii
Principal Collaborators/Institution:
Carol J. Pierce Colfer, University of Hawaii
Research Site(s):
Sitiung I, Sitiung V and Koto Padang, West Sumatra
Project Title:
Collaborative Research with Farmers on Home Gardens (3) Nutrition Component
Project Objectives:
1. Ascertain the variety of home gardening patterns that exist in
Sitiung,
2. To characterize Sitiung's home gardens in terms of type of crops grown,
soil management practices, annual yields and use of crops.
Progress and Achievements: (Report on attached form)
Meetings Attended: (Name, date, location, title of report given)
Carol J. Pierce Colfer, 2 March 1985, Andalas University, Padang,
"Pekarangan Transmigran: Kesempatan Penelitian Yang Tertinggal"
(continued)
Manuscripts in Preparation or Published: (Author(s), title, journal)
C. J. P. Colfer, C. Evensen, S. Evensen, Fahmuddin Agus, D. Gill, A. Wade,
"Transmigrants' Gardens: A Neglected Research Opportunity," Published in Pertemuan Teknis Proyek Penelitian Tanah Bogor: Center for Soils Research,
Volume 1. (March 1985).
(continued)
Work Planned for Next Year: (Report on attached form)
Signatures:
G. Uehara 12/31/85
(Project leader) (Date) (Program Coordinator) (Date)
White: Management Entity Yellow: Program Coordinator Pink: Proiect Leader




95
Meetings Attended--continued
Stacy Evensen, Carol J. Pierce Colfer, 18-20 March 1985, Cipayung,
Indonesia, Center for Soils Research Annual Technical Meeting, "Transmigrants' Gardens: A Neglected Research Opportunity."
Stacy Evensen, 29-30 July 1985, Andalas University, Padang "Symposium on "Future Prospects for Food Processing in West Sumatra": No report given.




96
Manuscripts in Preparation or Published--continued
Murni Rita, "Pola Usaha Tani Tanaman Pekarangan di Daerah Transmigrasi
Sitiung I, Sitiung V, dan Desa Koto Padang" ("Cropping Systems for Home Garden Crops in the Transmigration Areas Sitiung I, Sitiung V and the village of Koto Padang"); no plans for publication.




97
TROPSOILS
Management Entity, Box 7113, Raleigh, North Carolina 27695-7113
Progress and Achievements:
(Include: Description of research initiated or performed, results obtained, tables of data and graphs as appropriate, interpretations, potential application and utilization. Provide sufficient
details for analysis and evaluation. Use additional sheets if necessary.)
Planned Research Activities and Procedures
Two steps in this ongoing study have been completed to date. The first
involved the conduct of a modified "sondeo." A sondeo is a technique by which people from different disciplines pair off, and circulate in a given community, interviewing the inhabitants, to get a quick overview of a situation. In this
case, the method was used to determine how best to plan research on home
gardens. In January-February 1985, Carl Evensen (agronomist), Dan Gill (soil scientist), Stacy Evensen (nutritionist), Ann Wade (social worker), and Carol
Colfer (anthropologist) alternated partners and interviewed 12 farm households
about their home gardens. At the same time, we made maps of their home gardens
to identify variation in home gardens in the area.
The second step involved collaboration from Ita, a 4th year sociology
student from Andalas University. Ita worked with Stacy Evensen, first getting a
general idea of the diversity, and then constructing an interview schedule to
administer to 30 families in Sitiung I and V and Koto Padang. The plan for this two month study was to survey farmers in Sitiung I and V as well as Koto Padang
to see what crops are planted in home gardens, for what purposes, with what kind
of inputs, soil management practices, and constraints to production. It was
also planned to draw simple maps of home gardens outlining the number of
specific trees planted, plot size of food crops, orientation to house, stable,
fishponds, etc.
The study began with a series of casual observations of gardens in late
June. After this initial sweep, a survey tool was developed specifically asking
about farmers' plans for the next planting and about use and source of inputs, as well as the questions mentioned above. It was decided to survey a total of 30 farmers---13 each from Sitiung I and V (5 each from the Sundanese and East Javanese communities; 3 from the Minang transmigrants) and 4 from Koto Padang
(in order to observe possible differences among Minang farmers by location).
Research Results and Interpretations
Since the first round of data collection has only recently been completed,
the following must be taken as preliminary observations only. It had been noted
in previous years that planting usually ceased in the upland fields during the
dry season. From the literature it is known that often home gardens are used during these times of scarcity to supplement sources of family food. However,
the first clear results from the interviews and mapping were that no new
planting occurs during the dry season. Instead, farmers in Sitiung I and V
devote their time to soil preparation (hoeing, fertilizing, plot preparation)
and general yard cleaning (raking and burning of leaves). In Sitiung I farmers
use this time to repair fences and make improvements or additions to their
(continued)
WgE-hy-8 3-8
White: Management Entity Yellow: Program Coordinator Pink: Project Leader




98
Progress and Achievements-continued, page 2
houses or stables. This occurs to a much lesser extent in Sitiung V, perhaps because their fences were recently constructed anyway, and they do not yet have need of stables.
A wide variety of soil management practices were observed in both Sitiung I and V. While some farmers practiced rigorous hoeing of their fields and incorporation of manure and fertilizer, others simply broadcast a light application of fertilizer without incorporating it at all (visual observation--no specific quantities noted). Composting seems to be recognized as a valuable management practice but not all farmers engage in this, activity. "Composting" in Sitiung V often means collecting leaf litter, kitchen waste and crop residue which is either allowed to rot or is burned and used as a soil amendment in the garden. The government has not yet supplied farmers in Sitiung V with cows so no manure is available. In Sitiung I a similar process occurs with the addition of manure (both from cows and goats). This compost mixture is placed around perennial trees and incorporated into garden plots. Quite a lot of stump burning was observed in Sitiung V; when burned, this stump ash is used extensively in the home garden along with government-provided TSP and urea. Conversely, in Sitiung I, manure and some ash are the main soil amendments used. TSP and urea are used minimally.
During the dry season much of the available planting area in home gardens in Sitiung I and V remains bare except for perennial tree crops such as jackfruit, rambutan, guava, papaya, coconuts, bananas, cloves, stinkbean, etc. However, small plantings of certain crops were observed. Cassava was seen at various stages of maturity; new plantings occured at the end of June while older plantings (9 months 1 year old) were seen throughout the study period. Small areas of sweet potato, corn, mungbean and pigeon pea were seen in some home gardens in Sitiung I, and to a lesser extent, in Sitiung V. A fair number of families in Sitiung I and V planted small areas of their home garden to cowpeas for home consumption. When asked if this was a good crop to grow during the dry season the most common response was "It's ok, but soybeans are better." However, soybeans were not planted until late July.
The constraints to production mentioned by the farmers in Sitiung I and V included pests ("hama") and worms ("ulat"), pigs and unavailability of seeds. Another constraint mentioned frequently in Sitiung V was the lack of a market nearby at which excess production can be sold. Currently, farmers in Sitiung V must travel approximately 15-20 km at a substantial cost to a market to sell their surplus. Most farmers seemed to think this effort was not worth it. However, whether farmers feel disinclined to produce more than their family's food needs because of the lack of adequate market facilities, is yet to be determined.
A few informal observations of home gardens in Koto Padang indicate
considerable variation from home gardens in Sitiung I and V. Perennial tree crops dominate the plantings in Koto Padang home gardens. Sugarcane and some spice plants (lemon grass, basil, laos, turmeric, ginger, etc.) are also frequently planted. Much less care is taken of these gardens compared to those in Sitiung I and V. The use of soil amendments appears to be minimal. The food crops grown in Koto Padang home gardens seem to be used for home consumption, bartering, and sale at local markets.
Several interesting questions resulted from the interviews in Sitiung I and V. When asked "when will you begin planting?" farmers responded "when the rains come" or "in September (bulan 9)". They assumed the question referred to the time for planting rice although each interview focused on the home garden where




99
Progress and Achievements--continued, page 3
little rice is grown. Soybeans, peanuts and cowpeas were planted extensively in both Sitiung villages around July 28 because there had been a good rain at that time. This was done even though everyone interviewed said that the wet season did not begin until September. Why should farmers plant three important legume crops a full month before consistent rains normally come? Was the upland field (ladang) also planted to the legume crops at the first rain?
Farmers' attitudes toward various crops emerged as an important variable in crop selection. Whenever asked an open question such as "when will you begin to plant?" transmigrant farmers almost invariably answered with respect to rice, as in the above example. And, although some farmers did have cowpeas in their garden, they seemed to downplay their importance. These observations result in the following hypothesis: There exists a hierarchy within food crops based on social, economic and production value, and to a much lesser degree, nutritional value. In Sitiung rice is king followed by perhaps peanuts and soybeans. For whatever reason, cowpea is further down in the hierarchy. Attitudes of this sort will be investigated further as this research proceeds.
A final observation is that most food grown in the home gardens in Sitiung V is for home consumption. If there is extra produce it is sold or traded with neighbors. Farmers in Sitiung I appear less dependent on their home gardens for the family's food supply and have more available for sale. In Sitiung I the home gardens are a distinct system from the family ladang whereas the Sitiung V home garden appears to be used as an extension of the ladang. Often the same soil management practices and inputs are used. Traditionally, much of the family's fresh fruits, vegetables and other staple foods come from their home gardens. By using a large part of home garden space for ladang crops, families in Sitiung V may be eliminating access to fresh fruits and vegetables (most families shop at the market once or twice a month, so purchasing these commodities is not a viable alternative). This may have considerable negative impact on family nutrition and warrants additional study.
It became clear during the interviewing of these 30 families that, while a part of each home garden was usually planted to perennial crops, food crops in the rest of the home garden could, at best, only characterize the situation during that planting cycle. In order to better characterize the home garden system on an annual basis it was decided to intensively interview and observe selected farmers in Sitiung I and V for a one-year period. Gardens in Koto Padang will be mapped at least four times, with less intensive monitoring (largely because perennials dominate in Minang home gardens, and therefore seasonal variation should be less).




100
TROPSOLLS
Management Entity, Box '71 13, Raleign, North Carolina 27695-71 1 3
Work Plan for Next Year:
(Include: Any revision in objectives, experiments to be performed, projected meetings to be attended, reports to be prepared, major changes in budget, intended equipment purchases.
Provide sufficient detail for analysis and evaluation. Use additional sheets if necessary.)
The characterization of Sitiung's home gardens in terms of types of crops
grown, soil management practices) annual yields and use of crops will continue.
Progress is anticipated in the assessment of food production, income value and
nutrient contribution of the home garden compared to the overall farming system.
A series of questions is currently being developed, with input from the
entire TropSoils research team for use in intensive interviews with farmers to
take place every three months. The questions will pertain to seasonal
variations in soil management practices, choice of crops to grow, access to
agricultural inputs, limitations to production and source of knowledge about
agriculture. At the time of these interviews, maps will be drawn of the home
gardens to visually describe any seasonal or structural changes that occur
there. In addition, a method for collecting food production data will be
developed and tested based on the on-farm record-keeping suggested in Fanning
Systems Research and Development, Chapter 5. Family food consumption patterns
will also be characterized via quarterly dietary recall surveys. Periodic,
informal "drop-in" visits will also be made with the families to maintain good
rapport and to ensure that critical events in the yearly cycle are not being
missed by the-interview schedule.
The criteria for selection of the families to be monitored are being
finalized; but will include at least the following considerations: actively
working both a ladang and a home garden, willing to submit to a series of
intensive interviews throughout the next year, willing to communicate openly
about their farming activities, and able to understand and speak Bahasa
Indonesia. Additional matters under consideration include whether or not to
include farmers with wet rice fields as well, and whether to maximize diversity
or seek generalizability in a quite diverse context.
Stacy Evensen will attend an intensive language course in Java for 4 weeks
to enhance her ability to communicate with these families with whom she must
work. Effective communication is essential in conducting this kind of work. By
attending the course in January 1986, she, Carl Evensen, Stephenie Kan (new
agricultural economics graduate student), and the new soil physicist could go
together.
Both Stacy Evensen and Carol Colfer expect to attend an international
workshop on farming systems to be held at the Food Crops Institute's Sukarami Experiment Station in December 1985. Stacy anticipates presenting a paper on
the research described here.
White: Management Entity Yellow: Program Coordinator Pink: Project Leader