High mountain regions including the Andean region are very sensitive to climate change. Farmers in the central Andes of Peru are increasingly being exposed to the impacts of climate variability. This transdisciplinary research uses field laboratories, combining the farming system and the sustainable livelihood approaches, to carry out social, ecological, and financial assessments so as to identify sustainable and resilient livelihood strategies for small-scale Andean farmers. The first research step studies and characterizes farm household systems, influenced by their biophysical and socioeconomic contexts, for which two vulnerability indices were elaborated. Focused on the climate variability, the five livelihood assets and the three IPCC’s vulnerability components, these indices show the highly sensitive conditions of most communities with poor health conditions, access to infrastructure and public services. Farmers’ capacity of response is often limited by the low on-farm diversity and lack of organization. Thereafter, sustainable livelihood strategies were identified. These include on-farm intensification and non-farm labor intensification for lowland and better-access communities. In the middle-access and highland communities, where temporary migration is a common coping strategy, sustainable scenarios include an increment in diversification strategies through agrobiodiversity and a larger share of tree-based production systems. Furthermore, research step II explores local strategies to cope with agricultural droughts and evaluates, by means of natural resource assessment methods, agroforestry systems as an alternative to reduce their negative effects. Mainly affected by the increasing variation in precipitation events, farmers identify off-farm and on-farm diversification as adaptive strategies against agricultural droughts that reduce the weather dependence and covariance between livelihood activities. Among the introduction of more resistant crop and pasture varieties, the incorporation of trees in their system is desired because of their positive influence in soil moisture and crop yields. Soil moisture in agroforestry systems with eucalyptus trees is 10-20% higher than in agricultural systems during the beginning of the wet season. Differences in the soil moisture during the end of the dry season and in the potato yield are not evident between these systems, although an area without sowing reduced the agricultural output in 13-17% in agroforestry systems. Research step III seeks to maximize the efficiency of resource allocation in farm household systems by developing a linear programming optimization model. This financial assessment underpinned the need of additional off-farm activities for resource-scarcer farmers. In addition, under interest rates below 15% the model includes tree-based production systems as part of the optimal solution. However, with increasing interest rates, a higher share of land is used to cover household’s basic needs and fewer resources are available for capital accumulation activities such as forestry. Variations introduced in the model show that pasture systems are more sensitive to changes in the production outputs, whereas variation in farm worker wages and tree prices affected less the optimal solutions, making farming systems less sensitive to these market changes. Finally, the incorporation of tree-based systems have proved to be a sustainable and resilient livelihood strategy against climate variability available for particular farm household systems of the study area.:1 Introduction - 1 - 1.1 Introduction and justification - 1 - 1.2 Objectives and thesis statements - 2 - 1.3 Outline - 3 - 1.4 Definition of terms - 5 - 1.4.1 Vulnerability - 5 - 1.4.2 Resilience - 7 - 1.4.3 Agroforestry systems - 8 - 1.4.4 Farming system approach - 9 - 1.4.5 Farm household system - 10 - 1.4.6 Sustainable livelihood approach - 10 - 2 Framework and study site - 14 - 2.1 Theoretical framework - 14 - 2.2 Methodological framework - 18 - 2.2.1 Field laboratories - 18 - 2.2.2 Methods - 19 - 2.2.3 Methodology applied in research step I: Vulnerability in Achamayo - 21 - 2.2.4 Methodology applied in research step II: Agroforestry systems and agricultural droughts - 29 - 2.2.5 Methodology applied in research step III: Modeling small farm production systems - 33 - 2.2.6 Selection of case studies - 34 - 2.3 Study area - 35 - 2.3.1 Soils and topography - 35 - 2.3.2 Weather - 37 - 2.3.3 Agro-ecological zones and vegetation - 38 - 2.3.4 Climate change - 40 - 2.3.5 Socioeconomic characteristics - 42 - 2.3.6 Population - 43 - 2.3.7 External determinants - 71 - 2.4 Case studies - 47 - 2.4.1 Lowland communities (L) - 49 - 2.4.2 Middle access communities (M) - 50 - 2.4.3 Highland communities (H) - 51 - 3 Vulnerability in Achamayo - 53 - 3.1 Results - 53 - 3.1.1 Sustainable Livelihood Vulnerability Index (S-LVI) - 53 - 3.1.2 IPCC Livelihood Vulnerability Index (LVI-IPCC) - 68 - 3.2 Discussion - 71 - 3.2.1 Climate variability and extreme events - 71 - 3.2.2 Human capital - 71 - 3.2.3 Social capital - 71 - 3.2.4 Natural capital - 71 - 3.2.5 Physical capital - 71 - 3.2.6 Financial capital - 71 - 3.2.7 Livelihood strategies following the S-LVI and LVI-IPCC indices - 86 - 3.3 Conclusion - 92 - 4 Agroforestry systems and agricultural droughts - 95 - 4.1 Results - 96 - 4.1.1 Farmers’ experience and perception on climate variability and agricultural droughts - 96 - 4.1.2 Agricultural droughts in the farm household systems - 97 - 4.1.3 Farming forestry systems and land-use decision-making - 102 - 4.1.4 Influence of trees in the soil moisture and yield - 104 - 4.2 Discussion - 110 - 4.2.1 Climate change and agricultural droughts - 110 - 4.2.2 Farm forestry systems and land-use decision-making - 115 - 4.2.3 Influence of trees in the soil moisture and yield - 117 - 4.3 Conclusion - 121 - 5 Modeling small farm production systems: optimization of resource allocation - 123 - 5.1 Methodology - 124 - 5.1.1 Optimization Model - 126 - 5.1.2 Plan of optimization - 128 - 5.1.3 Production systems - 131 - 5.1.4 Constraints - 134 - 5.2 Results - 138 - 5.2.1 Model - 138 - 5.2.2 Interest rates - 142 - 5.2.3 Sensitivity analyses - 146 - 5.3 Discussion - 151 - 5.3.1 Cash flows - 151 - 5.3.2 Model outcomes - 152 - 5.3.3 Interest rates - 155 - 5.3.4 Sensitivity analyses - 159 - 5.4 Conclusion - 169 - 6 Synthesis - 171 - 6.1 Lessons learned - 171 - 6.1.1 Research step I - 172 - 6.1.2 Research step II - 175 - 6.1.3 Research step III - 176 - 6.2 Conclusions & outlook - 179 - 6.2.1 General conclusions - 179 - 6.2.2 Outlook - 181 - References - 185 - Appendix - 199 - / Las zonas montañosas, incluyendo la región andina son muy sensibles al cambio climático. Los agricultores de los Andes centrales del Perú están cada vez más expuestos a los efectos de la variabilidad climática. Esta investigación transdisciplinaria utiliza laboratorios de campo (field laboratories), combinando los enfoques de sistemas agrícolas y de medios de vida sostenibles, para llevar a cabo evaluaciones sociales, ecológicas y financieras con el fin de identificar estrategias sostenibles y resilientes para los agricultores andinos de pequeña escala. La primera fase de la investigación caracteriza a los sistemas agrícolas familiares, influenciados por sus contextos biofísicos y socioeconómicos, para lo cual se elaboraron dos índices de vulnerabilidad centrados en la variabilidad del clima, los cinco activos de los medios de vida y los tres componentes de la vulnerabilidad del IPCC. Estos índices muestran las condiciones de alta sensibilidad de la mayoría de las comunidades, con malas condiciones de salud y poco acceso a la infraestructura y a los servicios públicos. La capacidad de respuesta de los agricultores es a menudo limitada por la baja diversidad en las actividades agrícolas y la falta de organización. Posteriormente se identificaron las estrategias de medios de vida sostenibles. Estas incluyen la intensificación en las actividades agrícolas y la intensificación del trabajo no agrícola en las comunidades de zonas bajas y con mejor acceso. En las comunidades con menor acceso y zonas altas la migración temporal es una estrategia de afrontamiento común. Los escenarios sostenibles en estas comunidades incluyen un incremento en las estrategias de diversificación p. ej. a través de un aumento de la biodiversidad agrícola y una mayor proporción de sistemas de producción asociados con árboles. Por otra parte, la segunda fase de la investigación explora las estrategias locales para hacer frente a las sequías agrícolas y evalúa, por medio de métodos de evaluación de recursos naturales, los sistemas agroforestales como alternativa para reducir sus efectos negativos. Afectados principalmente por el aumento en la variación de las precipitaciones, los pequeños agricultores identifican a la diversificación de actividades dentro y fuera de sus parcelas agrícolas como una estrategia de adaptación frente a las sequías agrícolas que reduce la dependencia climática y la covarianza entre las actividades de subsistencia. Dentro de la introducción de variedades de cultivos y pastos más resistentes, como parte de la solución, los agricultores desean la incorporación de árboles en su sistema debido a su influencia positiva en la humedad del suelo y en los rendimientos de los cultivos. La humedad del suelo en sistemas agroforestales con árboles de eucalipto es un 10-20% mayor que en los sistemas agrícolas durante el comienzo de la estación húmeda. Las diferencias en la humedad del suelo durante el final de la estación seca y en el rendimiento de los cultivos de papa no son evidentes entre estos dos sistemas. A pesar de esto, el espacio sin siembra dejado en los sistemas agroforestales redujo la producción agrícola en un 13-17%. La tercera fase de la investigación busca maximizar la eficiencia en la asignación de recursos en los sistemas agrícolas familiares mediante el desarrollo de un modelo de optimización de programación lineal. Esta evaluación financiera respalda la necesidad de actividades adicionales no-agrícolas para agricultores con recursos más escasos. Además, con tasas de interés por debajo del 15%, el modelo siempre incluye a los sistemas de producción forestales y/o agroforestales como parte de las soluciones óptimas. Sin embargo, con el aumento de las tasas de interés, una mayor proporción de tierra se utiliza para cubrir las necesidades básicas del hogar y menos recursos están disponibles para las actividades de acumulación de capital como la silvicultura. Las variaciones introducidas en el modelo muestran que los sistemas de pastoreo son más sensibles a los cambios en los condiciones de producción. Por otro lado, la variación en los salarios de los trabajadores agrícolas y en los precios de los árboles afectan en un menor grado las soluciones óptimas, proporcionando sistemas agrícolas menos sensibles a estos cambios en el mercado. Finalmente, la incorporación de árboles en los sistemas agrícolas ha demostrado ser una estrategia de vida sostenible y resiliente a la variabilidad climática disponible para determinados sistemas agrícolas familiares de la zona de estudio.:1 Introduction - 1 - 1.1 Introduction and justification - 1 - 1.2 Objectives and thesis statements - 2 - 1.3 Outline - 3 - 1.4 Definition of terms - 5 - 1.4.1 Vulnerability - 5 - 1.4.2 Resilience - 7 - 1.4.3 Agroforestry systems - 8 - 1.4.4 Farming system approach - 9 - 1.4.5 Farm household system - 10 - 1.4.6 Sustainable livelihood approach - 10 - 2 Framework and study site - 14 - 2.1 Theoretical framework - 14 - 2.2 Methodological framework - 18 - 2.2.1 Field laboratories - 18 - 2.2.2 Methods - 19 - 2.2.3 Methodology applied in research step I: Vulnerability in Achamayo - 21 - 2.2.4 Methodology applied in research step II: Agroforestry systems and agricultural droughts - 29 - 2.2.5 Methodology applied in research step III: Modeling small farm production systems - 33 - 2.2.6 Selection of case studies - 34 - 2.3 Study area - 35 - 2.3.1 Soils and topography - 35 - 2.3.2 Weather - 37 - 2.3.3 Agro-ecological zones and vegetation - 38 - 2.3.4 Climate change - 40 - 2.3.5 Socioeconomic characteristics - 42 - 2.3.6 Population - 43 - 2.3.7 External determinants - 71 - 2.4 Case studies - 47 - 2.4.1 Lowland communities (L) - 49 - 2.4.2 Middle access communities (M) - 50 - 2.4.3 Highland communities (H) - 51 - 3 Vulnerability in Achamayo - 53 - 3.1 Results - 53 - 3.1.1 Sustainable Livelihood Vulnerability Index (S-LVI) - 53 - 3.1.2 IPCC Livelihood Vulnerability Index (LVI-IPCC) - 68 - 3.2 Discussion - 71 - 3.2.1 Climate variability and extreme events - 71 - 3.2.2 Human capital - 71 - 3.2.3 Social capital - 71 - 3.2.4 Natural capital - 71 - 3.2.5 Physical capital - 71 - 3.2.6 Financial capital - 71 - 3.2.7 Livelihood strategies following the S-LVI and LVI-IPCC indices - 86 - 3.3 Conclusion - 92 - 4 Agroforestry systems and agricultural droughts - 95 - 4.1 Results - 96 - 4.1.1 Farmers’ experience and perception on climate variability and agricultural droughts - 96 - 4.1.2 Agricultural droughts in the farm household systems - 97 - 4.1.3 Farming forestry systems and land-use decision-making - 102 - 4.1.4 Influence of trees in the soil moisture and yield - 104 - 4.2 Discussion - 110 - 4.2.1 Climate change and agricultural droughts - 110 - 4.2.2 Farm forestry systems and land-use decision-making - 115 - 4.2.3 Influence of trees in the soil moisture and yield - 117 - 4.3 Conclusion - 121 - 5 Modeling small farm production systems: optimization of resource allocation - 123 - 5.1 Methodology - 124 - 5.1.1 Optimization Model - 126 - 5.1.2 Plan of optimization - 128 - 5.1.3 Production systems - 131 - 5.1.4 Constraints - 134 - 5.2 Results - 138 - 5.2.1 Model - 138 - 5.2.2 Interest rates - 142 - 5.2.3 Sensitivity analyses - 146 - 5.3 Discussion - 151 - 5.3.1 Cash flows - 151 - 5.3.2 Model outcomes - 152 - 5.3.3 Interest rates - 155 - 5.3.4 Sensitivity analyses - 159 - 5.4 Conclusion - 169 - 6 Synthesis - 171 - 6.1 Lessons learned - 171 - 6.1.1 Research step I - 172 - 6.1.2 Research step II - 175 - 6.1.3 Research step III - 176 - 6.2 Conclusions & outlook - 179 - 6.2.1 General conclusions - 179 - 6.2.2 Outlook - 181 - References - 185 - Appendix - 199 - / Hochgebirgsregionen einschließlich der Andenregion sind gegenüber dem Klimawandel sehr empfindlich. Die in den zentralen Anden von Peru lebenden Bauern sind mehr und mehr den Auswirkungen durch Klimaschwankungen ausgesetzt. Diese transdisziplinäre Forschung nutzt Feldlabore, die das System der landwirtschaftlichen Bewirtschaftung und Ansätze zur nachhaltigen Lebensunterhaltssicherung kombinieren, um soziale, ökologische und ökonomische Erhebungen durchzuführen, so dass nachhaltige Livelihood-Strategien für die Kleinbauern in den Anden aufgezeigt werden können. Der erste Forschungsschritt untersucht und charakterisiert die bäuerlichen Haushaltssysteme, die durch ihre biophysikalischen und sozioökonomischen Kontexte beeinflusst sind. Hierfür wurden zwei Vulnerabilitätsindizes herausgearbeitet, die Klimavariabilität und die fünf Güter des Sustainable Livelihood-Konzepts im Fokus haben, sowie die drei Vulnerabilitätskomponenten des Intergovernmental Panel on Climate Change (IPCC). Diese Indizes decken die hochgradige Sensitivität für die meisten Gemeinden auf, aufgrund des schlechten Gesundheitszustandes sowie dem Mangel an Infrastruktur und öffentlichen Dienstleistungen. Die Fähigkeit der Bauern damit umzugehen, ist zumeist begrenzt durch eine geringe Diversität und fehlende Organisation auf den Farmen. Anschließend werden nachhaltige Livelihood-Strategien aufgezeigt. Diese umfassen die Intensivierung der Arbeit in der Landwirtschaft und der Arbeitskraft außerhalb der Landwirtschaft für Gemeinden im Flachland sowie besser erreichbare Gemeinden. In Hochlandgemeinden und Gemeinden die schwer zugänglich sind, ist temporäre Migration eine geläufige Bewältigungsstrategie. Nachhaltige Szenarien in diesen Gemeinden beinhalten eine höhere Anzahl an Diversifizierungsstrategien wie die Steigerung von Agro-Biodiversität und dem Anteil an baumbasierten Produktionssystemen. Forschungsschritt II untersucht lokale Strategien, um die landwirtschaftliche Dürre zu bewältigen und bewertet – mit Hilfe von Naturressourcenbewertungsverfahren – Agroforstsysteme als eine Alternative, um die negativen Auswirkungen der Trockenzeiten zu verringern. Beeinträchtigt durch zunehmende Niederschlagsschwankungen, identifizieren Bauern die Diversifizierung von landwirtschaftlichen und nicht-landwirtschaftlichen Aktivitäten als Anpassungsstrategie bei landwirtschaftliche Dürre, wodurch die Abhängigkeit vom Wetter und die Kovarianz zwischen den Aktivitäten für den Lebensunterhalt reduziert werden kann. Neben der Einführung resistenterer Kultur- und Weidepflanzen, ist die Einbeziehung von Bäumen in das System wünschenswert, aufgrund ihres positiven Einflusses auf die Bodenfeuchte und Erträge. Die Bodenfeuchte in agroforstwirtschaftlichen Systemen mit Eukalyptusbäumen ist während der beginnenden Feuchtperiode 20% höher als in landwirtschaftlichen Systemen. Die Unterschiede der Bodenfeuchte am Ende der Trockenzeit und bezüglich des Kartoffelertrags sind zwischen diesen Systemen nicht markant, obwohl eine Fläche, auf der keine Saat ausgebracht wurde, den landwirtschaftlichen Ertrag in Agroforstsystemen um 13 bis 17% mindert. Forschungsschritt III versucht die Effizienz der Ressourcenzuordnung in Farmhaushaltssystemen zu maximieren, indem ein Optimierungsmodell mit Hilfe der linearen Programmierung entwickelt wird. Diese ökonomische Erhebung unterstreicht die Notwendigkeit zusätzlicher nichtlandwirtschaftlicher Aktivitäten für ressourcenärmere Bauern. Bei Zinsraten unter 15% umfasst das Model baumbasierte Produktionssysteme als einen Teil der optimalen Lösung. Mit steigenden Zinsraten wird jedoch eine größere Bodenfläche dazu verwendet, um die Grundbedürfnisse der Haushalte zu decken und es stehen weniger Ressourcen für Aktivitäten zur Kapitalanhäufung wie Forstwirtschaft zur Verfügung. Die in das Modell involvierten Variationen zeigen, dass Weidesysteme sensibler auf Veränderungen des Produktionsausstoßes reagieren. Schwankungen bei den Löhnen der Farmer und Veränderungen der Baumpreise beeinträchtigen hingegen die optimalen Lösungen weniger. Dadurch sind die landwirtschaftlichen Systeme gegenüber Marktschwankungen weniger anfällig. Abschließend erweist sich, dass – für bestimmte Farmhaushaltssysteme im Untersuchungsgebiet – die Einbeziehung baumbasierter Systeme als nachhaltige und resiliente Livelihood-Strategie angesichts von Klimaschwankungen nützlich ist.:1 Introduction - 1 - 1.1 Introduction and justification - 1 - 1.2 Objectives and thesis statements - 2 - 1.3 Outline - 3 - 1.4 Definition of terms - 5 - 1.4.1 Vulnerability - 5 - 1.4.2 Resilience - 7 - 1.4.3 Agroforestry systems - 8 - 1.4.4 Farming system approach - 9 - 1.4.5 Farm household system - 10 - 1.4.6 Sustainable livelihood approach - 10 - 2 Framework and study site - 14 - 2.1 Theoretical framework - 14 - 2.2 Methodological framework - 18 - 2.2.1 Field laboratories - 18 - 2.2.2 Methods - 19 - 2.2.3 Methodology applied in research step I: Vulnerability in Achamayo - 21 - 2.2.4 Methodology applied in research step II: Agroforestry systems and agricultural droughts - 29 - 2.2.5 Methodology applied in research step III: Modeling small farm production systems - 33 - 2.2.6 Selection of case studies - 34 - 2.3 Study area - 35 - 2.3.1 Soils and topography - 35 - 2.3.2 Weather - 37 - 2.3.3 Agro-ecological zones and vegetation - 38 - 2.3.4 Climate change - 40 - 2.3.5 Socioeconomic characteristics - 42 - 2.3.6 Population - 43 - 2.3.7 External determinants - 71 - 2.4 Case studies - 47 - 2.4.1 Lowland communities (L) - 49 - 2.4.2 Middle access communities (M) - 50 - 2.4.3 Highland communities (H) - 51 - 3 Vulnerability in Achamayo - 53 - 3.1 Results - 53 - 3.1.1 Sustainable Livelihood Vulnerability Index (S-LVI) - 53 - 3.1.2 IPCC Livelihood Vulnerability Index (LVI-IPCC) - 68 - 3.2 Discussion - 71 - 3.2.1 Climate variability and extreme events - 71 - 3.2.2 Human capital - 71 - 3.2.3 Social capital - 71 - 3.2.4 Natural capital - 71 - 3.2.5 Physical capital - 71 - 3.2.6 Financial capital - 71 - 3.2.7 Livelihood strategies following the S-LVI and LVI-IPCC indices - 86 - 3.3 Conclusion - 92 - 4 Agroforestry systems and agricultural droughts - 95 - 4.1 Results - 96 - 4.1.1 Farmers’ experience and perception on climate variability and agricultural droughts - 96 - 4.1.2 Agricultural droughts in the farm household systems - 97 - 4.1.3 Farming forestry systems and land-use decision-making - 102 - 4.1.4 Influence of trees in the soil moisture and yield - 104 - 4.2 Discussion - 110 - 4.2.1 Climate change and agricultural droughts - 110 - 4.2.2 Farm forestry systems and land-use decision-making - 115 - 4.2.3 Influence of trees in the soil moisture and yield - 117 - 4.3 Conclusion - 121 - 5 Modeling small farm production systems: optimization of resource allocation - 123 - 5.1 Methodology - 124 - 5.1.1 Optimization Model - 126 - 5.1.2 Plan of optimization - 128 - 5.1.3 Production systems - 131 - 5.1.4 Constraints - 134 - 5.2 Results - 138 - 5.2.1 Model - 138 - 5.2.2 Interest rates - 142 - 5.2.3 Sensitivity analyses - 146 - 5.3 Discussion - 151 - 5.3.1 Cash flows - 151 - 5.3.2 Model outcomes - 152 - 5.3.3 Interest rates - 155 - 5.3.4 Sensitivity analyses - 159 - 5.4 Conclusion - 169 - 6 Synthesis - 171 - 6.1 Lessons learned - 171 - 6.1.1 Research step I - 172 - 6.1.2 Research step II - 175 - 6.1.3 Research step III - 176 - 6.2 Conclusions & outlook - 179 - 6.2.1 General conclusions - 179 - 6.2.2 Outlook - 181 - References - 185 - Appendix - 199 - / Regiões altomontanas, incluindo os Andes são extremamente sensíveis aos impactos das mudanças climáticas. Pequenos agricultores da região central dos Andes Peruanos estão progressivamente sendo expostos aos impactos das variações climáticas. A presente investigação transdisciplinar utiliza “field laboratories”, combinando os enfoques de sistemas rurais e dos meios de subsistência sustentáveis, visando uma avaliação social, ecológica e financeira, com intuito de se identificar estratégias resilientes e sustentáveis para os pequenos agricultores Andinos. A primeira etapa do presente estudo investiga e caracteriza os sistemas rurais, influenciados por seus contextos biofísicos e socioeconômicos, para os quais foram elaborados dois índices de vulnerabilidade focados na variabilidade climática, nos recursos dos meios de vida (cinco capitais) e nos três componentes da vulnerabilidade do IPCC. Esses índices mostram as condições altamente sensíveis da maioria das comunidades, com más condições de saúde, acesso à infra-estrutura e serviços públicos. A capacidade de resposta dos pequenos agricultores é frequentemente limitada pela baixa diversificação de actividades na exploração agricola e falta de organização. Posteriormente, foram identificadas estratégias de subsitência sustentáveis. Estas incluem a intensificação tanto do trabalho rural, quanto do não-agrícola para as comunidades de terras baixas e mais acessíveis. Para as comunidades altomontanas e com menor acesso, a migração temporária é uma estratégia de enfrentamento comum. Cenários sustentáveis para essas comunidades incluem um incremento nas estratégias de diversificação p. ex. aumentando a agrobiodiversidade e a parcela dos sistemas de produção florestais. A segunda etapa da pesquisa explora estratégias locais para lidar com as secas agrícolas e investiga, por meio de métodos de avaliação de recursos naturais, sistemas agroflorestais como alternativa para reduzir os seus efeitos negativos. Afetado principalmente pelo aumento da variação da precipitação, os agricultores identificam a diversificação tanto no trabalho rural, quanto no não-agrícola, como estratégias adaptativas contra secas agrícolas que reduzam a dependência do clima e covariância entre atividades de subsitência. Entre a introdução de culturas e de pastagens de variedades mais resistentes, a incorporação de árvores em seu sistema é desejada por conta da sua influência positiva na umidade do solo e no rendimento das culturas. A umidade do solo em sistemas agroflorestais com árvores de eucalipto é de 10-20% maior do que em sistemas agrícolas durante o início da estação chuvosa. As diferenças na umidade do solo durante o final da estação seca e na produtividade da batata não são evidentes entre estes dois sistemas. Apesar disso, o espaço sem semeadura deixado em sistemas agroflorestais reduziu a produção agrícola em 13-17%. A terceira etapa da presente investigação visa maximizar a eficiência da alocação de recursos em sistemas agrícolas familiares por meio do desenvolvimento de um modelo de otimização de programação linear. Esta avaliação financeira sustenta a necessidade de atividades não-agrícolas adicionais para agricultores com recursos escassos. Ademais, sob taxas de juros abaixo de 15%, o modelo inclui sistemas de produção florestais como parte da solução ideal. Contudo, com o aumento das taxas de juros, uma parcela maior da propriedade é usada para garantir as necessidades básicas, e portanto, menos recursos do agregado familiar estão disponíveis para atividades de acumulação de capital, tais como a silvicultura. Variações introduzidas no modelo mostram que sistemas de pastagem são mais sensíveis a mudanças nas condições de produção. Ademais, variaçãoes nos salários dos trabalhadores agrícolas e nos preços de árvores afetam menos as soluções ótimas, tornando os sistemas agrícolas menos sensíveis a estas mudanças do mercado. Por fim, a incorporação de sistemas florestais provaram ser uma estratégia de subsistência sustentável e resiliente contra a variação climática para determinados sistemas de agricultura familiar da área de estudo.:1 Introduction - 1 - 1.1 Introduction and justification - 1 - 1.2 Objectives and thesis statements - 2 - 1.3 Outline - 3 - 1.4 Definition of terms - 5 - 1.4.1 Vulnerability - 5 - 1.4.2 Resilience - 7 - 1.4.3 Agroforestry systems - 8 - 1.4.4 Farming system approach - 9 - 1.4.5 Farm household system - 10 - 1.4.6 Sustainable livelihood approach - 10 - 2 Framework and study site - 14 - 2.1 Theoretical framework - 14 - 2.2 Methodological framework - 18 - 2.2.1 Field laboratories - 18 - 2.2.2 Methods - 19 - 2.2.3 Methodology applied in research step I: Vulnerability in Achamayo - 21 - 2.2.4 Methodology applied in research step II: Agroforestry systems and agricultural droughts - 29 - 2.2.5 Methodology applied in research step III: Modeling small farm production systems - 33 - 2.2.6 Selection of case studies - 34 - 2.3 Study area - 35 - 2.3.1 Soils and topography - 35 - 2.3.2 Weather - 37 - 2.3.3 Agro-ecological zones and vegetation - 38 - 2.3.4 Climate change - 40 - 2.3.5 Socioeconomic characteristics - 42 - 2.3.6 Population - 43 - 2.3.7 External determinants - 71 - 2.4 Case studies - 47 - 2.4.1 Lowland communities (L) - 49 - 2.4.2 Middle access communities (M) - 50 - 2.4.3 Highland communities (H) - 51 - 3 Vulnerability in Achamayo - 53 - 3.1 Results - 53 - 3.1.1 Sustainable Livelihood Vulnerability Index (S-LVI) - 53 - 3.1.2 IPCC Livelihood Vulnerability Index (LVI-IPCC) - 68 - 3.2 Discussion - 71 - 3.2.1 Climate variability and extreme events - 71 - 3.2.2 Human capital - 71 - 3.2.3 Social capital - 71 - 3.2.4 Natural capital - 71 - 3.2.5 Physical capital - 71 - 3.2.6 Financial capital - 71 - 3.2.7 Livelihood strategies following the S-LVI and LVI-IPCC indices - 86 - 3.3 Conclusion - 92 - 4 Agroforestry systems and agricultural droughts - 95 - 4.1 Results - 96 - 4.1.1 Farmers’ experience and perception on climate variability and agricultural droughts - 96 - 4.1.2 Agricultural droughts in the farm household systems - 97 - 4.1.3 Farming forestry systems and land-use decision-making - 102 - 4.1.4 Influence of trees in the soil moisture and yield - 104 - 4.2 Discussion - 110 - 4.2.1 Climate change and agricultural droughts - 110 - 4.2.2 Farm forestry systems and land-use decision-making - 115 - 4.2.3 Influence of trees in the soil moisture and yield - 117 - 4.3 Conclusion - 121 - 5 Modeling small farm production systems: optimization of resource allocation - 123 - 5.1 Methodology - 124 - 5.1.1 Optimization Model - 126 - 5.1.2 Plan of optimization - 128 - 5.1.3 Production systems - 131 - 5.1.4 Constraints - 134 - 5.2 Results - 138 - 5.2.1 Model - 138 - 5.2.2 Interest rates - 142 - 5.2.3 Sensitivity analyses - 146 - 5.3 Discussion - 151 - 5.3.1 Cash flows - 151 - 5.3.2 Model outcomes - 152 - 5.3.3 Interest rates - 155 - 5.3.4 Sensitivity analyses - 159 - 5.4 Conclusion - 169 - 6 Synthesis - 171 - 6.1 Lessons learned - 171 - 6.1.1 Research step I - 172 - 6.1.2 Research step II - 175 - 6.1.3 Research step III - 176 - 6.2 Conclusions & outlook - 179 - 6.2.1 General conclusions - 179 - 6.2.2 Outlook - 181 - References - 185 - Appendix - 199 -