H. Wattenbach and K.H. Friedrich, Agricultural Support Systems Division,
FAO, Rome, Italy
The paper presents some preliminary conclusions derived from a case study carried out in the West Usambara Mountains, Tanzania, for the Farm Management and Production Economics Service of FAO. The case study attempted to identify indicators for sustainable farming systems using a combination of quantitative and qualitative data. One key innovation of the sustainability discussion into the development paradigm is the increased emphasis on the time dimension in natural resource management. As natural resource parameters generally change slowly, particularly the soil-related ones, the time horizon for measuring of relevant parameters can easily reach a decade. A cheap and quick approach using proxy-indicators and relating them to existing historic data should reduce this time span.
The case study assessed the changes in the farming system through the use of historic, small questionnaire surveys in five villages in the West Usambara Mountains. These data were complemented with qualitative data on the reasons for change as well as strategies for nutrient management as one component of the farming system. As agricultural production is only one of many farmer activities, the agricultural system was put into the broad context of off-farm and non-agricultural components of the farming systems. Interview partners for the qualitative data collection were farmers, key informants such as village authorities, head teachers, project and church personnel. The paper presents selected examples of proxy indicators at the field, cropping system and farming systems level. Special emphasis is given to the possibilities of using the nutrient balance in cropping systems as an alternative to the measurement of soil fertility parameters.
There is a need to identify indicators at a higher level than the field level, but below the level of agro-ecological zones. The reasons are that farmers as the managers of the natural resources are the ultimate decision-makers about the implementation of any conclusions derived from indicators for unsustainable resource management. These indicators consequently have to consider the farmers' decision-making processes concerning their resource management. Indicators need to cover the natural resource domain, but simultaneously have to consider the socio-economic sphere shaping farmers' behaviour.
Concern about the possibilities for future generations to meet their basic needs was the main reason for the change in development paradigm in the late 1980s, the origins of which can be traced back to earlier studies such as Global 2000 or even the work of Robert Malthus in the 19th century. These publications highlighted the problems for feeding the world's population and maintaining the livelihood system under existing population growth rates and with limited natural resources. The underlying assumption of these studies of linear trends has been frequently criticized, but the concerns about the potential negative impacts of present growth on future generations have been revived by more differentiated analyses. These concerns have received worldwide attention and focus under the term "sustainable development".
"Sustainable development is development that meets the need of the present without compromising the ability of the future generations to meet their own needs. It contains within it two key concepts:
¤the concept of "needs", in particular the needs of the world's poor, to which overriding priority should be given; and
¤ the idea of limitations imposed by the state of technology and social organization on the environment's ability to meet present and future needs." (WCED 1987, p. 43).
The major conclusions at international level have been discussed and agreed upon at the Earth Summit held in Rio de Janeiro in 1992 with the adoption of the United Nations Programme of Action from Rio. However, despite the changes in the global and general development perspective from short-term improvements to a long-term perspective, the operationalization of the term "sustainability" is still lacking.
CONTEXT OF THE STUDY
Most natural resources of the world are presently still being used by farmers. Therefore, the successful implementation of any conclusions derived from the monitoring of indicators for sustainable resource management will depend on their acceptance by the farmers. It is believed that an understanding of the farmers' resource management system is indispensable to be able to implement sustainable resource management concepts. The indicators for sustainability needed to monitor and evaluate modified development concepts will consequently have to consider the farmers' decision-making process about the use of the natural resources at their disposition. Several names for approaches which aim at understanding these mechanisms are being used, the most prominent one of them being the farming system concept.
A farming system is a natural resource management unit operated by a farm household, and includes the entire range of economic activities of the family members (on-farm, off-farm agricultural as well as off-farm non-agricultural activities) to ensure their physical survival as well as their social and economic well-being. This broad definition is of importance, as the farm family takes decisions considering not only the farming possibilities, but also the off-farm employment opportunities. Within an agro-ecological zone, several farming systems will typically be found, with variations in resource endowment, preferences and socio-economic position of the respective family.
Based on these considerations, the case study tested an approach to determine indicators relating to the sustainability of farming systems. The approach should meet two additional criteria:
¤ yield a result after a relatively short period of time, i.e., the study should not be a long-term research project;
¤ be repeatable by independent local research teams, i.e., be based on moderate or nil external inputs.
This excluded from considerations the measurement of any indicators requiring repeated data collection over several years. Instead, the approach had to consider finding proxy indicators substituting these data collections by other appropriate means.
The important question: "What should be sustained?" was preliminarily answered as follows:
"In order to achieve sustainable use of natural resources, farmers must be able to ensure a non-negative trend in real (per caput) income."
This statement by itself does not yet provide an operationalized approach to derive indicators. In order to achieve this, decisions had to be made to define the appropriate level of assessment within the systems hierarchy, to determine the time dimension to be considered and to specify data sources.
The case study is based on two field visits of one month each by two international researchers 1 to the West Usambara Mountains and data collection by three local researchers for two to three weeks. The aim of the first visit of the international researchers was to identify the field study villages, to become familiar with the study area, to adapt and test the questionnaires, and to identify and review major sources of secondary data. The second visit was carried out after completion of the questionnaire survey and encompassed the review of secondary information as well as data collection using qualitative methods. The following includes a brief description of the approach and a discussion of those conclusions relevant in the context of land quality indicators. A more detailed presentation of the results of the study will be made in a forthcoming publication in the Farming Systems Management series of FAO.
1 W. Grisley and H. Wattenbach undertook the first mission and M. Hall and H. Wattenbach participated in the second mission.
Dealing with the "TIME DIMENSION"
The time dimension of development requires strong emphasis on dynamic variables. To meet the complementary criterion of avoiding long-term research studies, there were two options for dealing with the dynamics of the systems: a) through proxy indicators and b) comparison of data over time.
The latter was possible, as there exist historic farming system studies in the West Usambara Mountains. Two sets of data were available, one from a study carried out in three villages in 1965 (Attems, 1967; 1968) and the second study carried out in 1985 (Due et al., 1985) in two other villages. The original questionnaire forms for both studies were available. These were used in a modified version (not all questions were found suitable to be repeated) and a sample of farmers was interviewed by local consultants in four of these five villages. The questionnaire covered all aspects of farming systems, such as family composition, land and livestock ownership, cropping pattern, forest resources on the farm as well as sources of non-agricultural income, among others. The comparison of these sets of data was expected to describe changes in the farming systems over two periods of time.
The mere description of changes without information on the driving forces behind them would leave an incomplete picture and yield unsatisfactory indicators, if any. Therefore, the questionnaire data were complemented by secondary sources on changes in macro-economic parameters such as price, major policy shifts, net population numbers for the area, infrastructural development, reports about the ecological situation (forest, soil and water resources, as available) of the district, etc.
During the second field visit by the international researchers, elements of rapid rural appraisal were used to collect information on, among others, the farmers nutrient management strategies, and the farmers' interpretation of changes in their livelihood system within the last decade. These questions were directed to farmers' groups (men and women groups), key informants such as school head teachers as well as project, church and NGO staff. The specific intention behind these discussions was to assess the reasons for change in resource use and to receive information about the farmers' perspective of the future development of their livelihood system. The methods used to achieve this were the drawing of resource flow charts, discussions with farmers during transect walks, and inviting farmers' reactions to a presentation of the changes identified in the questionnaire interviews in combination with open questions on the subject.
Dealing with the "SYSTEMS HIERARCHY"
Potential aggregation levels for indicator testing include the individual soil erosion or soil structure measurement points, the field, the cropping system, the farming system, the watershed, an agro-ecological zone, administrative units or the agricultural sector of a country. As the case study aimed at defining indicators for agriculture-based systems, the biophysical resource base was the starting point of analysis. Farmers frequently make specific decisions on how to use certain parts of a given field, based on their long-term observation of soil type, insect infestation and other biophysical factors. However, for the purpose of this study, the minimum level for assessment was the field. The cropping system and the crop rotation at field level were assessed through formal questionnaires.
The results of indicator assessment at any of the above-mentioned levels may be aggregated to the next higher level, if information on the geographical distribution of the respective unit is known and the relative weights of the different values for that indicator have been established. For example, soil erosion status at plot level may be estimated on the basis of soil erosion measurement points, as long as their distribution satisfactorily describes the soil erosion status of the field. In addition, the weights given to the indicator levels of different measurement points need to be established. In other words: is there an absolute limit for an indicator, beyond which that unit could not be sustainable (according to that indicator)? Or, should the definition of a higher level indicator permit that the values of two measurement points compensate each other?
Erosion on one part of a slope, for example, is certainly a negative indicator for the long-term ecological sustainability of that soil unit. But, assessing the situation from a larger perspective could show that the soil is accumulated at a different part of the slope and being used by the farmer for producing valuable crops, for which the growth conditions might not have been suitable, without more fertile material accumulated through erosion. It may therefore happen that the soil erosion on that aggregate unit is negligible and considering the farm-household system, the shift of soil may stabilize the farmers' income possibilities. In other words, external factors at a lower level (erosion on site A) may become internal at a higher level (e.g. a broader geographical unit including points of soil erosion and soil accumulation).
For practical purposes, it was decided that the case study would:
¤ start at the field level and attempt to identify the nutrient balance at the cropping systems level through proxy indicators (assessed through discussions with individual farmers and small groups, complemented by information from the questionnaires);
¤ identify the importance of the cropping system in the case study villages and assess the changes in relative importance of the cropping systems in the study period (assessed through the questionnaire survey as well as one group discussion);
¤ determine changes in the farming system (relative importance of different crops, off-farm employment and plans for the future) through a combination of questions answered during the questionnaire interviews as well as rapid assessment methods.
The questions focusing on the farming systems referred to the relative importance of different sources of income, shifts in their relative importance and the reasons why these changes occurred. Major changes included the reduction in coffee cultivation and wattle trees over time, the expansion of vegetables (partly in combination with increased terracing) and beans in certain areas of the mountains.
Major changes at the village level were reported on the use of off-farm natural resources such as forests and communal grazing areas. These types of changes were most clearly identified through interviews with key informants in villages and with NGO representatives. The reasons given for these changes were changes in village and district level by-laws such as the abolition of free grazing systems, and changing attitudes towards forest use on hill-tops. Changes in the latter were particularly emphasized under the extension advice provided under different phases of the Soil Erosion Control and Agroforest Project (SECAP).
This information on the plant production subsystem was complemented by the information collected about the livestock and forestry subsystems. Livestock is important as a provider of manure and user of grasses and by-products as feed. In semi-arid and arid areas, livestock is frequently accused and sometimes positively identified as a major cause of resource degradation. During the case study, the sources of livestock feed, the grazing regime as well as the manure use strategies were assessed qualitatively. Given the relatively low livestock numbers, the extensive feeding regime and low intensity, its contribution to the plant nutrient balance (and also soil structure) is almost negligible on most farms. Nevertheless, it could reach considerable importance for an improved nutrient management system, given that the demand for livestock and particularly dairy products could increase substantially in the regional economy.
Even though it is undisputed that most tree species have a positive effect on the long-term sustainability of ecosystems, their specific contribution is difficult to put into context when a relatively large number of tree species occurs scattered in the farmers' fields. Deriving indicators on the sustainability of cropping systems from information on the numbers of different tree species will require further consultation with experts in that field.
The changes in the farming systems have taken and are taking place in the context of a great number of changes in the institutional and political setting of Tanzania. These include changes in the price structure, the marketing system for inputs and produce and the infrastructure of the area. These changes already led to a rapid expansion of the area under beans and vegetables and may lead to a major increase in coffee production in the future, reversing the decline of the industry in the last decades. Simultaneously the decline in prices for the bark of wattle trees and the recent closing of the wattle extraction plant may stimulate further land-use changes in parts of the mountains.
Examples of indicators based on rapid assessment methods
Examples of the reasoning followed in the case study for testing simple indicators at different aggregation levels are presented below:
¤ Field level indicators
The assessment of sustainability at the plot level mainly addressed the physical and ecological aspects of sustainability. The determination of the nutrient balance and its trend over time through proxy indicators was tested as an alternative to a long-term measurement of physical data.
Sustainability at the field level should reflect the capacity of the agricultural production system on that specific field to maintain a certain yield in the future. In a first step, sustainability at the field level shall be defined as the possibilities for continued cultivation of the same cropping sequence. There are two possible approaches to assess the production potential:
· to derive the trend in the production capacity from the physical properties of that field. This would require the assessment of these properties, their monitoring over a sufficient period of time and the estimation of the changes in production potential for certain crops over time. The most prominent physical parameters would be related to the soil (organic matter content, physical soil structure (pore structure), soil type, water retention capacity, nutrient content, erodibility of the topsoil, etc.). These data would be complemented with climatic data so that the crop or fodder production potential could be derived from the most important agro-ecological parameters. It should be noted that sustainability indicators would have to reflect the status of the variable over time. Two reasons led to the conclusion not to attempt this approach: even in a relatively favourable situation of data availability in the West Usambara Mountains, there are hardly any relevant soil data for the past, and even for the few soil research sites of the past the exact location of these sites would have been quite difficult. In addition, the combination of any such data with sufficiently detailed climatic data in an area with such difficult topographic and consequently site-specific climatic data is close to impossible.
Therefore, even the measurement of the required parameters in a short-term study would have been meaningless because of the lack of a reference system in the past. The only possible solution would be the direct measurement of relevant parameters and therefore a research project over a time span sufficiently long to allow the identification of changes of the parameters, probably exceeding a decade.
· to identify proxy indicators for the nutrient balance for the major crops cultivated in the research area; since the nutrient balance in a field will result in the more immediate effect on production potential than soil structural change, it should be given first priority during the assessment. The nutrient flows were divided into inflow and outflow: harvest, by-product removal, erosion, N-evaporation and nutrient leaching being the most important outflows and inflows comprising manure, fertilizer, mulch, by-products left or added to the field, mineralization, N-fixation and atmospheric addition of nutrients.
Several of these parameters would be almost as tedious to assess as the above criticized soil parameters. However, it was expected that assessing only the most prominent of these parameters would be sufficient to allow statements about the trend on that plot and would therefore qualify as a proxy indicators. Through interviews information was collected on assessment of the yield level, the use of byproducts from different crops, and applications of manure, fertilizer and mulch. This information was combined with information of the position on the slope to allow for adjustments for additional nutrient loss through erosion (or the addition of eroded nutrients in the valley bottom).
Two examples illustrate the argument: a field under 15-year old tree shrubs, which receive a relatively high level of annual fertilization is unlikely to be under a negative nutrient trend, even without measuring the nutrient loss through leaching and the nutrient input through mineralization. In contrast, a maize field on a relatively steep slope without leguminous crops, which is only cultivated one season per year and left fallow for the rest of the year, and which systematically does not receive manure or fertilizer, can be classified as an unsustainable production system. Attempting to measure the degree of erosion may help quantifying the total nutrient loss per year, but the negative trend is indisputable. The additional research budget which would be needed to measure nutrient inflows such as through mineralization in the soil is probably better invested in understanding the reasons why farmers accept the obviously negative nutrient balance.
Further dialogue is needed to determine under which situations such proxy indicators for the nutrient balance yield a satisfactory result and under which conditions more refined methods need to be applied.
¤ Cropping systems level indicators
Field level information was mainly assessed through the questionnaires, and their analysis resulted in information on the cropping system. As this type of information is generally also presented in historic farming system studies, changes in the cropping pattern can be assessed relatively easily through comparisons of older and new field studies. The questionnaires covered the following aspects: the cropping pattern per field from 1990 to 1995, distance of the field to the house, position of the field on the slope, steepness of the field and information about the farmers' assessment of change in productivity on that particular field compared to ten years ago.
The questionnaire information was complemented by discussions with the farmers about their strategies for nutrient transfer between fields. It was found that nutrient transfers occur in addition to crop specific strategies: the removal or addition of by-products depends in addition on considerations such as feed requirements for livestock, strategies for livestock grazing in the field after harvest, fertilizer application and application of manure to specific target crops and depending on the distance of the field from the house. It was again assumed that an identified negative nutrient balance in a cropping system based on the above assessment would not be reversed if other factors such as leaching and nitrogen fixation were to be assessed as well and that therefore the additional budget implications and time needed to do so should better be invested in researching other determinants of the farmers' resource management strategies.
¤ Farming system level indicators
The development of the farming systems takes place in the context of the village and watershed and of macro-economic trends. Farmers will adjust their cropping pattern and farming systems to these conditions, including off-farm employment opportunities, migration and concerns about the livelihood of their children. Any advice given to farmers on sustainable natural resource management will be put by them into the wider context of considerations about the development of their livelihood system.
Farmers therefore tend to give less weight to field-level, biophysical and ecological considerations than to sustainability of the whole farming system, which could include the acceptance of a transfer of nutrients between different fields, within the catena and from distant fields towards the homestead. And depending on the trends in the non-agricultural sector of a village, the farmers may even decide to shift emphasis from agricultural to non-agricultural activities (or vice versa). These shifts in importance were assessed in the formal questionnaires by asking ranking questions on the importance of different sources of income, which had also been determined - albeit by different means - in the previous studies. However, the estimation of the socio-economic sustainability of the farming system through qualitative means was more interesting in the present study.
Before addressing the farmers' perspective of their farming systems in the future, attempts were made to discuss with them their evaluation about the past development. In order to allow active contribution of the younger farmers, the time horizon for doing so was reduced to changes within the last decade. Such overall assessments will always depend on the socio-economic position of the interviewed person, and a long list of other factors. However, it was found useful to see which of the indicators used by the farmers would indicate changes in their overall well-being. Indicators listed by farmers themselves included corrugated iron sheets, increased installation of water pipes, increased availability of consumer goods in the villages and more meat available in the market.
Despite differences in the indicators chosen by men and women and by different individuals, a discussion is necessary on which of them directly indicates changes in the socio-economic well-being. Most items listed referred to consumption indicators. Among those, many reflect at the same time increased cash availability in the villages, but also changes in macro-economic import and marketing policies. These increased imports into the regions, but do not necessarily reflect the consumption potential before these changes occurred. It appears more relevant to review changes in the consumption of goods which where available already under the previous macro-economic system. Therefore, the availability of meat in the market appears to be a more promising indicator than corrugated iron sheets.
The need to assess the existing plans for changes in the farming system could potentially contribute further to the development of indicators for sustainability of the systems. Particularly during individual interviews, the importance of considering the social position of the farmer concerned became immediately obvious. The difference shall be illustrated through two examples:
Example 1: a poor widow owns only one plot of land, which covers her basic needs for cassava and banana with small quantities of coffee for sale. She does not see any future for her children in the area, as land is increasingly difficult to borrow. Three of her four children already migrated out of the West Usambara Mountains, and the widow assumes that they will never return due to the lack of a resource base to fall back upon. Despite the relatively diversified cropping pattern on that single piece of land, there is no prospect for investment in soil erosion control on the steep slope where the field is located.
Example 2: the son of a relatively large farmer returned from migration in the Lake Zone. He starts cultivation of a fairly large area with new technologies (vegetable cultivation on irrigated land, cultivation of sorghum for chicken feed and poultry production). His cropping system will be relatively more prone to erosion, but as the crops cultivated will be more profitable than those of the widow, he will have the resources to invest in soil erosion control should he wish to do so. In addition, the farmer is open to further modifications of his production technologies in case they meet his farming strategies.
These two examples illustrate a further dimension to the definition of indicators for the sustainability of farming system, which should guide assistance and interventions to help modify unsustainable systems. The resource availability in the family or the situation within the family life cycle has a strong influence on the farmers' strategies. Indicators on the farming system level may ultimately depend on the social and economic dynamic of an area. And more recently, new dimensions were added to the already complex determinants of the farming systems dynamics in the West Usambara Mountains. For example, the farmer returning from the Lake Zone stated that all of his plans for farming will depend on the development of his health situation, as his wife unexpectedly died in the Lake Zone and his own health situation has become quite unstable.
The case study yields conclusions on various issues. Some of them are related to difficulties arising from data collection through recruited staff, be it through questionnaire surveys or rapid or participatory appraisal techniques. Others are more specific to participatory methods, such as the creation of expectations by the farmers by the intensive work with them, which are not easily fulfilled in the context of case studies. The cooperation with an existing project made the execution of the case study relatively easy, but highlighted an important area for consideration for similar studies. Some more specific conclusions from the material presented above are the following:
¤ The usage of the terms "parameters" and "indicators" in the present discussion should be clarified. "Indicator" is the more widely used term and this was also the starting point for the case study. The intensive use of the logical framework method in the project cycle has led to extensive studies on the requirements which need to be fulfilled by indicators in order to qualify as a helpful tool for monitoring purposes. According to these, indicators have to be specified in terms of quantity, quality and time horizon.
The further complication of using indicators for sustainability is the inherent time dimension of the issue, in which a constant change in variables is expected. An intensive dialogue among the professional disciplines will be needed if practical conclusions are to be drawn from the indicators developed in the different disciplines. As the lead in this discussion will depend on the context of usage of the indicators, the weights given to certain types of indicators will vary. It can be expected, however, that the longer the lists of indicators, the more of them will be given relatively little attention in the aggregation process. An early and open dialogue across the technical disciplines in the matter would therefore increase the efficiency of research in the development of indicators.
¤ The bio-physical resource base has been the starting point of the discussion on sustainability and therefore holds a prominent position in attempts to identify indicators for sustainability. The minimum and maximum aggregation levels of data needed and/or permitted, however, need to be discussed in the context of the types of conclusions which should be drawn from these indicators.
¤ If the indicators should only be used for regional priority setting of agricultural and environmental policies, relatively high geographical aggregation levels and a relatively strong bias towards biophysical indicators might be acceptable. These indicators will then point out the regions or agro-ecological zones with the most urgent "average intervention need". However, this does not necessarily point to the most cost-efficient geographical or sectoral interventions, as these indicators may not allow drawing conclusions on the underlying problems leading to the need for intervention. The cost-efficiency of interventions will also depend on the monetary, social and administrative costs to implement programmes after identifying areas for intervention. These costs will depend on the magnitude and type of changes the farmers would have to make in their resource management system and is therefore a function of the farming system operated by the farmer.
¤ Similarly, should the indicators yield a detailed understanding of the soil physical and soil chemical mechanisms leading to ecologically unsustainable resource management practices, the development of indicators even below the field level may be necessary. In order to draw conclusions on the necessity for and acceptance of modified resource management systems, the farmers' decision-making process has to be taken into consideration. As the farmers differentiate their resource management not only on the basis of ecological variables (which will generally be included in data aiming at field level indicators of sustainability) but also on the basis of socio-economic factors such as accessibility of the field from the house, ownership security for different fields, etc. (which are usually not included in field level databases), the ecological variables by themselves are not sufficient for the development of intervention- and extension-oriented indicators.
¤ The discussion about the best stratification criteria of farming systems for the purpose of developing indicators for sustainable resource management is still at an initial stage. However, it is obvious that improved resource management can only be achieved if advice given to the farmers is based on an understanding of these systems. In times of increasingly restricted research budgets, proxy indicators should be developed.
Attems, M. 1967. Bauernbetriebe in tropischen Höhenlagen Ostafrikas - Die Usambara-Berge im Übergang von der Subsistenz- zur Marktwirtschaft. ifo-Institut für Wirtschaftsforschung München. Afrika Studien No. 28. Weltforum Verlag. München.
Attems, M. 1968. Permanent Cropping in the Usambara Mountains. In: Farming Systems in the Tropics, Ruthenberg, H. (ed). Clarendon Press, Oxford, UK. pp. 137-174.
Conway, G.R. 1994. Sustainability in agricultural development: trade-offs between productivity, stability and equitability. Journal for Farming Systems Research-Extension 4 (2): 1-14.
Due J., Ragumbisa J. et al. 1985. Beans in the Farming Systems in Tanga Region, Tanzania 1985. Department of Economics, University of Illinois, USA, and Department of Rural Economy, Sokoine University of Agriculture, Morogoro, Tanzania.
Dvorak, K. A. 1993. Economic Indicators. Draft working paper as part of a Sustainability Assessment Framework of the Agroforestry Subgroup of FAO. June 1993. FAO, Rome.
Geißen, V. 1984. Firewood Consumption and Related Aspects in Five Selected Villages in Lushoto District, Tanzania. Soil Erosion Control and Agroforest Project (SECAP).
Hailu Zegeyu, Runge-Metzger A. 1993. Sustainability of land use systems - the potential of indigenous measures for the maintenance of soil productivity in Sub-Sahara African agriculture. Tropical Agroecology No. 7. GTZ. Verlag Josef Margraf, Germany.
Iversen, S.T. 1991. The Usambara Mountains, Tanzania: History, Vegetation and Conservation. Uppsala universitet. Reprocentralen, HSC. Uppsala, Sweden.
Mnkeni, P.N.S. 1992. Role of soil management in enhancing Sustainability of small-holder cropping systems in some agro-ecosystems of Tanzania: a review. Ecology and Development Paper No 3. Ecology and Development Research Programme, The Agricultural University of Norway. Centre for Sustainable Development. Aas, Norway.
Pfeiffer, R. 1990. Sustainable Agriculture in Practice - the Production Potential and the Environmental Effects of Macro-contour lines in the West Usambara Mountains of Tanzania. PhD Dissertation. Institut für Pflanzenbau und Grünland. University Hohenheim, Germany.
Pitt-Schenkel, C.J.W. 1938. Some important communities of warm temperate rain forest at Magamba, West Usambara, Tanganyika Territory. J. Ecology 26: 50-81.
Prinz, D. 1982. Ökologisch angepasste Bodennutzung in den Usambara Bergen Tanzanias. In: Wiss. Zentr. Tropeninst. Giessen (Hrsg.): Tropisches Gebirge: Ökologie und Agrarwirtschaft. Giessener Beiträge zur Entwicklungsforschung Reihe I Band 18.
Redclift, M. 1987. Sustainable Development - Exploring the Contradictions. Methuen, London and New York.
Ruthenberg, H. 1964. Agricultural development in Tanganyika. ifo-Institut für Wirtschaftsforschung München. Afrika Studien No. 2. Springer Verlag, Berlin, Göttingen, Heidelberg, New York.
Ruthenberg, H. (ed.) 1968. Smallholder farming and smallholder development in Tanganyika. ifo-Institut für Wirtschaftsforschung München. Afrika Studien No. 24. Weltforum Verlag, München.
Sandhofer, K. 1979. Soil Erosion Control and Afforestation in the West Usambara Mountains. Prefeasibility Study. TIRDEP, Tanga, Tanzania.
Schoonmaker Freudenberger, K. 1994. New technology for rural women: paradoxes of sustainability. Development in Practice 4 (1): 13-22.
Taube, G. 1992. Wirtschaftliche Stabilisierung und Strukturanpassung in Tansania: Die Auswirkung des Economic Recovery Programme 1986-1989 im ländlichen Bereich. Fallstudie West-Usambara-Berge, Distrikt Lushoto. Hamburger Beiträge zur Afrika-Kunde 41. Institut für Afrika-Kunde Hamburg (PhD Dissertation (Economics) Free University, Berlin), Germany.
Tiffen, M., Mortimore, M. and Gichiki, F. 1994. More People, Less Erosion - Environmental Recovery in Kenya. Nairobi 1994. J. Wiley, Chichester, UK.
United Nations. 1992. Agenda 21: Programme of Action for Sustainable Development. Rio Declaration on Environment and Development. United Nations. New York 1992.
van der Pol, F. and van der Geest, N. 1993. Economics of the Nutrient Balance. In: Agriculture, Economics and Sustainability in the Sahel. Papers presented at the Third Sahel Seminar, May 1992. Royal Tropical Institute, Amsterdam, and Agricultural University, Wageningen. pp. 47-68.
Watson, J.R. 1972. Conservation problems, policies and the origins of the Mlalo Basin rehabilitation scheme, Usambara Mountains, Tanzania. Geogr. Ann. 54: 221-226.
WCED. [World Commission on Environment and Development]. 1987. Our Common Future. "The Brundtland Report". Oxford University Press, Oxford, UK.
Zegeye Hailu 1992. Methodologies to Assess the Sustainability of Farming Systems. Working paper prepared for the Farm Management and Production Economics Service, Agricultural Services Division, FAO. September 1992. FAO, Rome.