Management Practices for Dealing with Uncertainty and Change:

Social-Ecological Systems in Madagascar and Tanzania

Maria Tengö

Doctoral Thesis in Natural Resource Management

Department of Systems Ecology Stockholm University

106 91 Stockholm Sweden

Stockholm 2004

Doctoral dissertation 2004 Maria Tengö Department of Systems Ecology Stockholm University SE-106 91 Stockholm Sweden

© 2004 Maria Tengö ISBN 91-7265-970-X Printed in Sweden by Intellecta Docusys AB, Sollentuna Cover design by Simon Alexanderson. Photos by Simon Alexanderson and Maria Tengö

2

3

To

The people of the Mbulu highlands, Tanzania, and Androy, Madagascar, who above all taught me about being human

and

Simon

4

Abstract The development of human societies rests on functioning ecosystems. This thesis builds on integrated theories of linked social-ecological systems and complex adaptive systems to increase the understanding of how to strengthen the capacity of ecosystems to generate services that sustain human well-being. In this work, I analyze such capacity in human-dominated production ecosystems in Tanzania and Madagascar, and how this capacity is related to local management practices. Resilience of social-ecological systems refers to the capacity to buffer change, to re-organize following disruption, and for adaptation and learning.

In Papers I and II, qualitative interview methods are used for mapping and analyses of management practices in the agroecosystem of the Mbulu highlands, Northern Tanzania. Practices such as soil and water conservation, maintenance of habitats for pollinators and predators of pests, intercropping, and landscape diversification, act to buffer food production in a variable environment and sustain underlying ecological processes. The practices are embedded in a decentralized but nested system of institutions, such as communal land rights and social networks, that can buffer for localized disturbances such as temporary droughts. Paper II compares these findings with practices in a farming system in Sweden, and suggests that similar mechanisms for dealing with uncertainty and change can exist in spite of different biophysical conditions.

In Papers III and IV, interviews are combined with GIS tools and vegetation sampling to study characteristics and dynamics of the dry forests of Androy, southern Madagascar. Paper III reports on a previously underestimated capacity of the dry forest of southern Madagascar to regenerate, showing areas of regeneration roughly equal areas of degenerated forest (18 700 ha). The pattern of forest regeneration, degradation, and stable cover during the period 1986-2000 was related to the enforcement of customary property rights (Paper III). Paper IV reports on a network of locally protected forest patches in Androy that is embedded in a landscape managed for agricultural or livestock production and contributes to the generation of ecosystem services and ecosystem resilience at a landscape scale. Forest protection is secured by local taboos that provide a well-functioning and legitimate sanctioning system related to religious beliefs.

In Paper V, two spatial modeling tools are used to assess the generation of two services, crop pollination and seed dispersal, by the protected forest patches in southern Androy. The functioning of these services is dependent on the spatial configuration of protected patches in the fragmented landscape and can be highly vulnerable to even small changes in landscape forest cover.

In conclusion, many of the identified practices are found to make ecological sense in the context of complex systems and contribute to the resilience of social-ecological systems. The thesis illustrates that the capacity of human-dominated production ecosystems to sustain a flow of desired ecosystem services is strongly associated with local management practices and the governance system that they are embedded in, and that, contrary to what is often assumed, local management can and does add resilience for desired ecosystem services. These findings have substantial policy implications, as insufficient recognition of the dynamics of social-ecological interactions is likely to lead to failure of schemes for human development and biodiversity conservation.

5

Table of content

Abstract ................................................................................................................................. 5 List of papers......................................................................................................................... 7 Introduction ........................................................................................................................... 8

Generation of ecosystem services for human wellbeing................................................. 10 Managing complex adaptive ecosystems ........................................................................ 11 Scope of the thesis........................................................................................................... 12

Presentation of study areas and summary of papers ........................................................... 14 Management of agroecosystem production: Mbulu highlands in northern Tanzania..... 14 Summary of results of Papers I and II............................................................................. 17 Local taboos and dryland forest dynamics: Androy, southern Madagascar ................... 18 Summary and conclusions of Papers III, IV and V......................................................... 21

Discussion ........................................................................................................................... 22 Maintaining capacity for internal renewal ...................................................................... 22 Sustaining landscape processes....................................................................................... 25 Practices, knowledge, and governance systems ............................................................. 27

Conclusions ......................................................................................................................... 30 Acknowledgements ............................................................................................................. 32 References ........................................................................................................................... 27

6

List of papers

I Tengö, M. and M. Hammer. 2003. Management practices for building adaptive capacity: a case from northern Tanzania. in F. Berkes, J. Colding, and C. Folke, editors. Navigating Social-Ecological Systems: Building Resilience for Complexity and Change. Cambridge University Press, Cambridge.

II Tengö, M. and K. Belfrage. 2004. Local management practices for dealing with change and uncertainty: a cross-scale comparison of cases in Sweden and Tanzania. Ecology and Society 9(3): 4. [online] URL: http://www.ecologyandsociety.org/vol9/iss3/art4

III Elmqvist T., M. Pyykönen, M. Tengö, F. Rasoarisela, E. Rabakonandrianina, C. Radimilahy 2004. Large-scale regeneration and the role of social institutions: tropical dry forests in southern Madagascar. Manuscript.

IV Tengö, M., K. Johansson, F. Rasoarisela, J. Lundberg, J.-A. Andriamaherilala, E. Andersson, J.-A. Rakotoarisoa, T. Elmqvist 2004. Local protection of tropical dry forest: taboos and ecosystem services in southern Madagascar. Manuscript submitted to Conservation Biology.

V Bodin Ö., M. Tengö, A. Norman, J. Lundberg and T. Elmqvist 2004. The value of small size: loss of forest patches and threshold effects on ecosystem services in southern Madagascar. Manuscript.

7

Introduction Addressing the sustainability of natural resource management requires integrated thinking across disciplines (e.g. Scoones 1999, Kinzig and al. 2000, Gunderson and Holling 2003). In this thesis I will build on integrated theories of social-ecological systems (Berkes and Folke 1998, Berkes et al. 2003) and on recently developed theories on complex adaptive systems (CAS) (Levin 1999). Complex systems theories are emerging in parallel in several disciplines e.g. economy, human geography, ecology, and archaeology (Manson 2001). They have in common that they portray systems not as deterministic, predictable and mechanistic, but as process dependent, organic, and always evolving (e.g. Arthur 1999). An important aspect of the complex systems approach is the recognition that many different kinds of systems include self-regulation, feedback or adaptation in their dynamics and thus may have a common underlying structure despite their apparent differences. In addition to developing deeper understandings of specific systems, such interdisciplinary approaches should help elucidate the general structure and behavior of complex systems, and move us toward a deeper appreciation of the general nature of such systems. This thinking is starting to be used to bridge social and biophysical sciences to understand, for example, climate, history and human action (McIntosh et al. 2000), assessments of regions at risk (Kasperson et al. 1995), syndromes of global change (Steffen et al. 2004), and how to link social and ecological systems for sustainability (Berkes and Folke 1998, Scoones 1999, Gunderson and Holling 2003). It underpins many of the new integrative approaches, such as ecological economics (Costanza et al. 1993, Arrow et al. 1995) and sustainability science (Kates et al. 2001). New perspectives, concepts and tools developed within complex systems theory holds much promise for a deeper understanding and improved management of living systems (Holling 2001).

Development of human societies rest on functioning ecosystem services that sustain human wellbeing (Daily 1997, Millennium Ecosystem Assessment 2003). The ecosystems generating these services are increasingly altered by global changes and human impact at all scales (Bellwood et al. 2004, Folke et al. 2004). To secure development of human societies we need to increase our understanding of the behavior of complex ecosystem including human interactions. van der Leeuw et al. (2000) argue that for understanding the

8

history and future development of complex systems, we need to assume change and explain stability, rather than assuming stability and explaining change. This is captured in the heuristic model of ecosystem dynamics developed by Holling (Holling 1986, Gunderson and Holling 2003), where periods of gradual change (termed the frontloop) are disrupted by rapid transformation (the backloop), see Figure 1. During the re-organization of the backloop, the system may develop according to a different trajectory and emerge into an alternative stable state. CAS theory and the model in Figure 1 point at the role of disturbance to create and reorganize diversity. Resilience (Holling 1973, 2001) is an essential property of CAS. Resilience is the capacity of a system to absorb disturbance and reorganize while undergoing change so as to retain essentially the same function, structure, identity, and feedbacks (Carpenter et al. 2001). Resilience of a social-ecological system is further related to the degree of which the system is capable of self-organization (versus a lack of organization, or organization forced by external factors), and the degree to which it expresses a capacity for learning and adaptation.

Figure 1. Holling’s heuristic model of dynamic renewal in for example ecosystems (Holling 1986, Gunderson and Holling 2003). Starting in the lower left corner, systems go through periods of gradual change, referred to as the frontloop, encompassing the phases of exploitation (r) and conservation (K). Resource accumulation and increased connectedness during the frontloop lead to increased brittleness for disturbances in the K-phase and the frontloop is followed by periods of fast, disruptive change in the backloop (Ω and α). After a disturbance there is phase where resources are released (Ω) followed by renewal (α), where resources re-assembly through self-organization processes. Depending on the outcome of these processes, the system may enter into a new cycle of front-and backloop remaining within the same stability domain, or develop into a new state, illustrated by the x in the figure (e.g. a state shift, see further in the text). One example of these dynamics are a forest ecosystem, where forest succession in the frontloop accumulates and consolidated resources into biomass. As fuel is abundant the effects of a lightning can cause a forest fire that takes the system into the backloop, releasing resources such as nutrient and space that can be remobilized in the α-phase. Depending on e.g. intensity of fire, remaining stumps, seeds and faunal community, connections to other ecosystem, or perhaps incidents of additional disturbances, a new forest succession may start with a similar community of species, or a different ecosystem e.g. a grassland, may emerge.

9

Generation of ecosystem services for human wellbeing

Understanding and managing the capacity of ecosystems to sustain a flow of ecosystem services, i.e. the resilience of the ecosystem, are essential for sustainable development (Daily 1997, Millennium Ecosystem Assessment 2003). In the rural areas of many poor countries people are directly dependent on local ecosystems for the generation of food, fuel, construction wood, medicines and for securing livelihoods during periods of crisis such as temporary drought (e.g. Fabricius and Koch 2004, Paper I and IV). Local ecosystems also have important cultural values (Paper IV). Generation of these ecosystem goods is dependent on underlying ecosystem processes such as water purification, flood control, local climate regulation, soil formation, nutrient assimilation, pollination, and seed dispersal (Daily 1997).

This thesis addresses local ecosystems that are directly essential for sustaining local livelihoods through production in agro- and forest ecosystems. These ecosystems are human-dominated, and managed to be productive. They generate a bundle of ecosystem services, primary of them being food production. I make attempts to analyze the capacity of ecosystems to generate desired ecosystem services in such production landscapes, and how this capacity is interrelated with local management practices. In particular, I am investigating practices that may contribute to the resilience of ecosystem functioning for human wellbeing.

It is increasingly recognized that we need to address ecological functions in managed ecosystems, to keep them productive and in a state that generates ecosystem services desired for human wellbeing (Bawa et al. 2004, Palmer et al. 2004). As described above, an essential component of the capacity of ecosystem to generate services is related to the ability to cope with disturbance and change, i.e. its resilience. Many ecosystems exhibit thresholds and qualitative shift in the generation of ecosystem services. Ecosystems such as shallow lakes or savannas, have been found to exhibit more than one stability domain, such as turbid or clear states of a lake, or grass vs. woody species domination of a savanna (Scheffer et al. 2001, Folke et al. 2004). A transition between two states can be triggered by a fast, external event, but is often preceded by an internal loss of resilience (see Figure 1). For example, a period of drought may be the triggering factor for a savanna ecosystem to shift from a grass to a woody dominated state, but the vulnerability to a shift can be increase with stocking densities (e.g. Walker 1993). Regime shifts are increasingly common as human activities affect processes at all scales (Folke et al. 2004).

A shift to an alternate state may generate a different set of ecosystem services. From a management perspective, a certain state may be more desirable. Savannas are one example, where a grassland state has higher utility in terms of livestock production, compared to a

10

woody state (which on the other hand can provide services such as fuelwood production). There may also be conflicting views on preferred ecosystem states, as in the highlands of Madagascar, where the local herders prefer an ecosystem state, maintained through regular burning, that is highly productive as pasture for livestock. On the other hand, the government and associated conservationists prefer an ecosystem where prohibition of fire leads to succession with secondary forests (Kull 2004).

Diversity plays a key role in the resilience of ecosystem functions, and recent findings point at factors such as how species diversity is organized in functional groups (e.g. pollinators, decomposers, predators) and how the landscape is structured (Peterson et al. 1998, Diaz and Cabido 2001, Loreau et al. 2001, Nyström and Folke 2001, Kinzig et al. 2002, Bengtsson et al. 2003, Elmqvist et al. 2003).

Thus, ecosystems are complex and adaptive, with non-linear dynamics and thresholds between qualitatively different states (Levin 1999). Further, human interactions with ecosystems, e.g. land transformation, resource harvesting, emission, are affecting the underlying capacity of ecosystems to generate of services at different scales, from field level to global climate (Folke et al. 2004). Thus, surprises and limited predictability of ecosystem behavior needs to be taken into account when managing natural resources or ecosystems.

Managing complex adaptive ecosystems

Facing complex co-evolving systems for sustainability requires the ability to cope with, adapt to, and shape change without losing options for future adaptability (Folke et al. 2003). The understanding of how to deal with disturbances and improve regeneration after disturbance, e.g. build resilience, has previously received relatively little attention in resource management science (Holling et al. 1998).

It is argued that when dealing with complex resource management issues, one needs to draw on all kinds of knowledge for understanding ecosystem dynamics to design better management (e.g. Berkes et al. 2003). The role of local ecological knowledge1, resource-use practices, and institutions in the management of natural resources and ecosystems has received increasing attention in recent years (Ostrom 1990, Gadgil et al. 1993, Fairhead and Leach 1996a, Olsson and Folke 2001, Davis and Wagner 2003).

In this thesis, I use a theoretical framework for evaluating local management practices in relation to resilience and management of complex adaptive ecosystems that is based on the model of ecosystem development in Figure 1 (Berkes and Folke 1998, 2002). Backloop management refers to practices where natural disturbances are an integrated part of

11

manipulating and modifying the natural resource base, and managers adaptively respond to episodic or rare events using flexible institutions and management practices that reduce the risk that large scale ecological crisis will occur (Colding et al. 2003a). Local management practices and social mechanisms have been shown to improve the capacity for dealing with the disturbances and changes inherent in complex adaptive systems (Berkes et al. 2000, Berkes and Folke 2002, Colding et al. 2003a). Berkes and Folke (2002) propose that traditional resource-use practices can be complementary to conventional resource management science by monitoring and management through qualitative measures and indicators during the frontloop and by building capacity to deal with disruptive change during the backloop (Figure 1). One example is practices that trigger small-scale disturbances that allow for internal renewal of the ecosystem and thus prevent the buildup of larger scale crises. Another component found in many local or traditional management systems, is the nurturing of sources of renewal, e.g. through setting aside certain areas that can alleviate ecosystem regeneration (Colding et al. 2003b).

The development and implementation of local management practices is framed by institutions, i.e. the formal and informal rules at different levels of organization that guide human behavior (Ostrom 1990). This includes local customs, property right regimes, social norms and conventions, national legislation and procedures for decision-making and conflict resolution (e.g. Hanna et al. 1996). The institutional framework sets the conditions for human interaction with the ecosystem, e.g. which management practices are permitted, and by whom. An institutional framework that is flexible and allows for adaptive responses in accordance with ecosystem feedback can improve social-ecological resilience (Berkes and Folke 1998, Dietz et al. 2003).

Scope of the thesis

Strengthening ecosystems’ capacity to provide ecosystem services requires at least three levels of analysis and understanding: One regards the ecosystem and its dynamics, including the processes, elements, and links to other ecosystems that can contribute to resilience. Another regards the management of the ecosystem, learning and generation of ecological knowledge, and how it is applied in management practices, including monitoring and responses to ecosystem dynamics. A third regards the institutional dynamics, the governance system, that underlies management strategies, and that can allow for adequate responses to uncertainty and change (Berkes and Folke 1998, Berkes et al. 2003).

Figure 2 shows a framework for analyzing social-ecological systems. This thesis focuses on management practices adjusted through ecosystem feedback as a key link between

12

local ecosystems

larger ekosystems

regional ecosystems

nested institutions

management practices

institutions

ecological knowledge

Figure 2. A conceptual framework for analyzing social-ecological systems adapted from Berkes et al. (2003). In focus of this thesis are management practices as an interaction between social and ecological system, developed in response to ecosystem feedback. I also address aspects of the local ecosystem, ecological knowledge as reflected in management practices, and local institutions related to ecosystem management.

ecosystem feedback

social and ecological system. I also address the dynamics of local ecological knowledge as expressed in practices, and local institutions that frame management practices.

I investigate management practices in productive landscapes/ecosystems that has high levels of human impact and direct reliance on ecosystem services. Papers I and II looks at agroecosystems, the underlying services that sustain food production, and discusses an array of related management practices and social mechanisms. Papers III to V addresses the dry forest ecosystem of southern Madagascar, land-cover dynamics during the last two decades associated with social and ecological change, and how the local practices of assigning taboos to certain portions of forest are related to the capacity of the ecosystem to generate services.

The thesis reports on management practices that enhance the ecosystem capacity to sustain a flow of desired ecosystem services, that appreciates ecosystem feedback at several scales, and that build buffer capacity for uncertainty. These practices are found to enhance ecosystem processes that support ecosystem production in spite of disturbances and allow for ecosystem regeneration within a desired domain after disruptive change. Further, they create or sustain landscape structures that contribute to the generation of ecosystem services and spatial resilience in the landscape. The identified practices are not unique for the case studies in this thesis. The thesis rather illustrates that there are similarities across cases concerning the mechanisms applied to deal with change. Paper II illustrates that local ecological knowledge can also be relevant ina farming system in a highly mechanized and high-income society.

13

The management practices identified in the thesis were found to be embedded in and enabled by institutional frameworks, such as common property regimes and taboos that provide legitimate and well functioning enforcement of local management. Local institutions for managing complex ecosystems is often overlooked, although it has the potential both to reveal important understanding of ecosystem-human interactions, and improve the capacity to manage ecosystems for human wellbeing.

In Papers I and II, qualitative interview methods are applied for mapping and analyzing management practices, and literature studies are used to discuss their potential implication for generation of ecosystem services. In Papers III and IV, interviews are combined with GIS tools as well as vegetation sampling and analyses to study landscape patterns and land cover changes in southern Androy. In Paper V, two modeling tools are used to assess the generation of ecosystem services in the landscape and how it responds to loss of forest habitat.

The findings of the thesis have substantial policy implications. External interventions to improve local conditions have often had a narrow focus, e.g. increasing agricultural production, or preserving biodiversity (Scott 1998, Sayer and Campbell 2002, Kull 2004). A key message is that even to achieve such narrow goals, one has to use a system perspective, address feedback mechanisms across scales, non-linear dynamics, and social and ecological interactions. This thesis contributes to the understanding of such mechanisms and linkages between social-ecological systems, with the aim of development towards sustainable management of ecosystems for human wellbeing.

Presentation of study areas and summary of papers

Management of agroecosystem production: Mbulu highlands in northern Tanzania

Iraqw’ar Da/aw in the Mbulu highlands have a comparatively long history of cultivation of permanent fields combined with extensive cattle herding (Thornton 1980, Börjeson 2003). The farming system is intensive in terms of land and labor requirements, and has been referred to as an “island” within a matrix of less intensive land use (Börjeson 2004). Compared to other well documented cases of agricultural intensification in East Africa, such as the Machakos of Kenya (Tiffen et al. 1994), Mount Meru (Spear 1997) and Kilimanjaro (Tagseth 2000) in Tanzania, the population density in Iraqw’ar Da/aw is not as high (around 100 persons/km2). Further, while market forces are described as having a key role as a driving force for intensification in these near urban cases, such processes seem to have been of limited importance in the more isolated Mbulu highlands (Börjeson 2003).

14

The prevailing hypothesis explaining why intensive agriculture developed in the Mbulu highlands has been that of a siege, where isolation caused by geographical factors and warfare with neighboring pastoralist forces intensified land use practices (Thornton 1980). This hypothesis has recently been questioned by Börjeson (2004), who argue that the process of agricultural intensification has largely been its own driving force, based on self-reinforcing processes of change, and not a consequence of land scarcity.

Iraqw’ar Da/aw is located at the rim of the Rift Valley Escarpment in Northern Tanzania, in Mbulu District in Arusha Region, about 20 km southeast of Mbulu town. The inhabiting Iraqw apply a mixed crop-livestock farming system using hand hoe to cultivate the multitude of slopes and valley bottoms in the undulating Mbulu highlands. Generally farms are small, the farmers in the study cultivated 0.36-1.1 ha. Production levels are fairly high for a low input farming system (e.g. for maize 1.8 t ha-1, unpublished data based on farmers’ estimation, or 0.7-2.0 t ha-1 for maize, Meindertsma and Kessler (1997), cf. Rockström and Jonsson (1999)). Markets for farm produce such as fruit, tobacco or surplus of food-crops are not easily available, as roads are bad and means of transport limited. In the studied households, the harvests in a normal year was said to be sufficient to feed the families. In bad years, they rely on the networks for sharing of resources, on off farm work, or remittances (interviews, Loiske 2003). The farmers perceive declining yields on sloping field and common responses to this is increased cultivation of valley-bottom land and tree planting for fruit and timber on the slopes (Snyder 1996).

Conditions for cultivation is good in the Mbulu highland in terms of rainfall, with an average annual precipitation of more than 1000 mm, fairly well distributed over two main rain seasons. During the drier part of the year, the high water table in the valleys allow for an additional crop season and the total growing period stretches more or less over the whole year (NSS 1994). However, rainfall may vary considerably within and between years, and onset and duration is difficult to predict. Further, the dissected topography makes fields exposed to erosion. The underlying bedrock is gneissic, soft and highly weathered and exhibits great spatial variability of mineral composition. Papers I and II describe and analyze management practices developed to deal with these constraints. For example, the niches created by the topography and biophysical processes have been further enforced by site-specific selective management practices of crop choice, intensity of cultivation, and fertilization, see Figure 3. Soil organic matter levels in fields of the study area are high (NSS 1994 and unpublished field data) in spite of continuous cultivation, which can be seen as an indicator that ecological capital is generated by soil conservation practices.

15

Khatsa Genei Dindirmo Intsi Description Flat field at base of

slope or valley floor

Sloping field facing west (in front of the house)

Flat field near or on ridge

Sloping field facing east (behind the house)

Land use Crops: Maize and beans

Less common: Sorghum, finger millet, Irish potatoes, sweet potatoes

Maize and beans, sweet potatoes, Irish potatoes, wheat, sorghum, finger millet, coffee

Tobacco, pumpkins, vegetables, maize, wheat

Maize and beans, sweet potatoes, finger millet, sorghum, coffee, Irish potatoes

Other: Grazing, fodder, wild vegetables, fodder, water for home consumption and cattle

Housing Timber trees

Fruit trees, timber trees, small grazing area

Grazing

Cropping seasons

Jul-Feb Oct-Mar, Apr-Sep Sep-Mar Sep-Mar

Management Occasional adding of manure Drainage

Adding manure Mulching Ridging

Adding manure Adding manure and crop residues Ridging

Soils Sandy; brownish or whitish

Brownish, whitish or, reddish

Reddish Brownish, whitish or, reddish

Problems Flooding, water logging, heavy soils (wet and silted), prosperous weeds

Strong sun dries soil quickly, soil erosion, fertility loss

Strong sun, fertility loss

Windy (cold), soil erosion, fertility loss

Soil fertility Moisture content

Percentage of households with access

89% 89% 29 % 65%

Figure 3. A transect view of the agricultural landscape of Iraqw'ar Da/aw based on a workshop with farmers (men and women) in the ridge community Hhay Geay. The farmers classify fields according to slope and which direction they face, which has implications for e.g. soil and surface temperature, wind exposure, soil depth and nutrient status. The rankings of soil fertility and moisture content are farmers’ perceptions.

Land rights in the Mbulu highlands are related to clan affiliation, traditionally distributed by the elders, who also are responsible for communal lands. A farmer generally owns his fields, but long or short term land loans is also common (Börjeson 2004). Figure 4 summarizes management practices and institutional arrangement at different scales within the study area, ranging from family-based management of private fields, over common

16

property resources to networks that extend outside the Mbulu highlands. For details see Paper I.

Börjeson (2004) reports an 80% increase of cultivated land from 1958- 1988 in the study area. Internal drivers of land changes are population increase, declining yields, tenure insecurity and thus competition for land (Snyder 1996, Lawi 1999). Climate change, towards drier and less predictable conditions as perceived by the local farmers, is another external factor influencing land use decisions (Source: interviews). External drivers such as government campaigns of e.g. tree planting, valley cultivation and coffee growing, also has affected land use dynamics within the case study (Lawi 1999). Conversion of pasture to fields as well as tree planting implies a conversion of communal lands to private, which shifts the focus of landscape management from the elders to individual farmers (interviews, Paper I). The authority of the elders has also been challenged by the formal institutional arrangements imposed by colonial and post-independence governments (Snyder 1993, Lawi 1999).

Summary of results of Papers I and II

The persistence of the agricultural system in Iraqw’ar Da/aw based on permanent cropping indicates that for a certain time period, the local farmers have been able to sustain the capacity to produce food without severely deteriorating important assets such as soil

Outside Iraqw’ar Da/aw Food exchange networks, markets

Ridge community Neighborhood assistance, work parties, management and monitoring of communal pasture and wetlands

Wider village community: Long- and short-term cattle loans, land loans, pasture management

Private management: Micro-scale adaptations and optimization (fertilization, crop choice, etc.)

Figure 4. Management practices and institutions at different scales in Iraqw’ar Da/aw, Mbulu highlands, ranging from household based management of private fields, over different spheres of communal control and collaboration, to networks that allows for access to resources outside the area.

17

fertility, or passing thresholds of the regenerating capacity of the agroecosystem.

In Paper 1, a multitude of management practices reflecting traditional ecological knowledge in Iraqw’ar Da/aw was found to be directed towards functioning of the whole ecosystem rather than merely resource output. In particular, a number of practices were found that improve the nutrient re-circulation and erosion control of an intensive land use system in a hilly landscape. The practices are embedded in a decentralized but nested system of institutions, such as common property pastures and networks for food exchange, that can buffer for localized disturbances and enables response to feedback signals such as declining production on several levels. These features allow for adaptive response to ecosystem dynamics and improve the resilience of the linked social-ecological system, in terms of increased capacity to deal with recurrent disturbance such as temporary droughts and pest outbreaks.

When comparing the practices identified in the Mbulu highlands with management practices found among farmers in southeastern Sweden, we found similar approaches for dealing with uncertainty and change in the two cases, in spite of different biophysical and socio-economical conditions. Both cases were small-scale in terms of monetary flows, have integrated production of crops and livestock, and use limited amounts of agro-chemicals. Using the model of front and backloop practices (see Figure 1), we found that farmers in both cases applied practices for buffering crop production through diversification, sustaining processes such as pest control and pollination, and using qualitative ecosystem indicators for circumscribing uncertainty. Further, a number of practices identified in both cases, such as crop diversification, were also found to dampen the effects of disturbance. Practices are adjusted and developed in response to ecosystem dynamics, and we found examples of revitalization of old practices for example for disease control following recurrent outbreaks associated with climatic change. We emphasize the role of continuous learning for developing sustainable agriculture in complex agroecosystems.

Local taboos and tropical dry forest dynamics: Androy, southern Madagascar

Deforestation of Madagascar has previously been estimated to be very high, where a growing population practicing slash-and-burn agriculture and pasture burning has been seen as the main driver of land cover change (see McConnell 2002). Recent studies however paint a more nuanced view on the deforestation rates and reveal a more complex set of drivers for land use change (e.g. McConnell 2002, Richard et al. 2002, Casse et al. 2004, Kull 2004). Two recent policies are part of the countries strategy to counteracting the

18

degenerative processes of forest loss. The first, announced at the World Park Congress in Durban, South Africa, in November 2003, prescribes an extension of the county’s protected area from 1.7 million ha to 6 million ha until 2008 (Anonymous 2003). The second approach is a decentralization of forest management, in line with an international focus on the role of local communities in conservation and natural resource management (Kull 2004).

The dry forest of south and southwest Madagascar has been reported to decline since the early 1970’s, principally due to clearing for agriculture, cattle herding, timber harvest and charcoal production (Sussman and Rakotozafy 1994, Sussman et al. 2003). Despite the global recognition of the value of the dry southern forest, there have been surprisingly few studies on the ecology of the dry spiny forest, including studies on the impact on human activities (Ganzhorn et al. 2003, Elmqvist 2004).

In Androy, located in the extreme south of Madagascar, the Tandroy that are the dominating ethnic group are recorded in the area since the 17th century, when they were primarily settled in the areas in between the rivers in the south (Heurtebize 1986). Over time, seasonal settlement in the sparsely settled areas further north became permanent. However, population density in northern part of Androy of 10-30 persons/km2 is still significantly lower than in the south, where density ranges from 100-350 persons/km2 (SAP 2002. The zebu, the local cattle, has a central role for the Tandroy, economically as well as culturally (Heurtebize 1986).

The southern plains are sandy with occasional patches of dense forest, whereas the northern part is a hilly upland on crystalline Precambrian bedrock with more extensive forest cover (Battistini and Richard-Vindard 1972). On the sandy areas of the south, cultivations of primarily maize, cassava, beans, and sweet potato, dominate the landscape. Characteristic of this landscape are also the distinct patches of dense forest (Clark et al. 1998, Paper III), of which most are shown in Paper IV to be subject to protection by local taboos. The pattern of forest patches has existed at least since the 1950’s. The history of these patches, i.e. to what degree they are remnants of an earlier more extensive forest cover, or have developed through human interventions (cf. Fairhead and Leach 1996a), is not known.

The forested areas further north are used extensively for grazing, but there are also cultivations around the scattered settlements. The forest further local communities with a number of resources, such as construction wood, fuel, medicines, game, off-season or emergency food, see Table 1 for examples of species utilized (source: interviews). The forest of southern Madagascar is a mixture of spiny bush and dry deciduous forest

19

Table 1. Examples of forest species that provides important goods for local livelihoods.

Vernacular name Species Family

Timber Fantiolitse Alluaudia procera Didieraceae Mendorave Albizia greveana Fabaceae Handy Neobeguea mahafaliensis Meliaceae Fuelwood Katrafay Cedrelopsis grevei Burseraceae Hazolava Neobeguea mahafalensis Meliaceae Livestock fodder Try Cynanchum marmerianum Asclepiadaceae Mozotse Euphorbia stenoclada Euphorbiaceae Supplementary food

Fangitse Fandra

Dolichos fangits Dioscorea fandra

Diascoreaceae Diascoreaceae

Fio Asparagus sp. Liliaceae Kily/ Tamarind Tamarindus indica Fabaceae Medicinal plants Katrafay Cedrelopsis grevei Burseraceae Vahombey Aloe vahombe Liliaceae Famata Euphorbia decorsei Euphorbiaceae Game Lambro/ Wild pig Potamochoerus larvatus Akanga/ Helmeted guinea fowl Numida meleagris mitrata Tivoke/ Blue coua Coua caerulea

and is dominated by species of the family Didieraceae and woody species of Euphorbiaceae (Koechlin 1972).

Androy region one of the poorest in Madagascar and most often struck by drought and seasonal famine (Dostie et al. 2002, see also Paper III). The region is characterized by semi-arid climatic conditions with very irregular rainfall averaging less than 500 mm per year (Battistini and Richard-Vindard 1972). Precipitation can vary greatly within short distances, which sets the preconditions for seasonal migration with cattle herds. In response to the temporal and spatial vagaries in precipitation, the Tandroy move with their herds 3 up to 8 months a year. Since the 1980’s the region has experienced an increased aridity expressed through shorter wet seasons, more unpredictable rainfall patterns, and recurrent drought conditions that almost has become chronic (Gould et al. 1999, Casse et al. 2004). Figure 5 in Paper III shows an increasing trend in number of days per months with rainfall < 1 mm.

The dry spiny forest of southern Madagascar has been classified as one of 200 ecoregions in the world to be prioritised for conservation (Olson and Dinerstein 2002). As very few formal reserves exist in this region, it is a priority area for establishing new protected areas. An initial proposal for potential areas to protect includes the northern part of the study area of Papers III and IV (Fenn 2003). As the proposal is very recent, it remains to be seen what types of protected areas that will be applied in the dry spiny forest.

20

Access to land resources in Androy is traditionally governed through clan affiliation and ancestors land (tannin-drazana), and agreements between clans or individuals. This is part of the fokonolona, traditions and customs that guide human behavior. There is also the fanjakana, i.e. the formal institutions of Malagasy society, including village presidents, the regional government, and the Forest Service, who grant permits for harvest of forest resources such as timber and wood for charcoal production. Part of the fokonolona is the taboos that are given by the ancestors. Taboos, or fady, meaning forbidden, or ‘you shall not’ are important components in social interactions and creation of identity in Androy and all of Madagascar (Gennep 1904, Ruud 1960, Lambek 1992, Walsh 2002). Further, many of these taboos are related to interactions with the environment, such as habitat taboos restricting human use of certain areas (Ruud 1960, Paper IV).

Summary and conclusions of Papers III, IV and V

Taken together the three papers address land cover changes during the last decades, present social and ecological processes behind the landscape of today and some ecosystem services it generates, and glimpse into the future by assessing the consequences of a scenario of loss of forest habitat.

In Paper III we identified large-scale forest regeneration in Androy during the last 15 years, comparing Landsat images from 1986 and 2000. Forest regeneration is an aspect of forest dynamics that often has been overlooked, in southern Madagascar as well as in tropical dry forests in general. In Androy, we found that spontaneous regeneration covering approximately equal amounts as forest being degraded during the period 1986-2000 (18 300 ha), giving an overall marginal decrease of forest cover of 2%. The study reveals locally different outcomes of social and ecological drivers, with patterns of regeneration, deforestation, or stable forest cover over a short distance. An analysis of local drivers and dynamics of social institutions revealed that degraded forest was found in areas characterized by insecure property rights, while areas with well-defined property rights and well-enforced local rules for forest protection has either regenerating or stable forest cover.

In Paper IV, further investigation of the local practice of reserving certain areas from human resource extraction showed that numerous patches or parcels of forest in the landscape are protected by local taboos. In southern Androy all forest patches larger than 5 ha and many smaller ones were found to be taboo forests. Eight different types of taboo areas were differentiated. The protected forest patches did not differ in species richness compared to non-protected ones, but in species composition. Taboo areas were also found in northern and western Androy. The forest patches are suggested to contribute to regional biodiversity conservation and the generation of a number of ecosystem services, including

21

cultural. The system of taboos assigned to certain portions of forest patches represent a well-functioning sanctioning system that secure the protection this patches.

In Paper V we investigated generation of two supporting ecosystem services important for maintaining crop production and forest habitats in the fragmented landscape of southern Androy: pollination and seed dispersal. Based on two spatially explicit models, the present landscape was found to be adequate in generating the investigated services. However, when testing the effects on generation of the services against successive removal of forest patches, the services appeared as highly vulnerable, displaying drastic decline and thresholds. Our results suggest that for maintaining connectivity and crop pollination cover in southern Androy, the geographical locations of the remaining forest patches are more crucial than their size. We argue that small forest patches should increasingly be viewed as essential components in a production landscape.

The papers illustrate that landscape structure and ecosystem resilience in southern Madagascar are related to social and ecological processes.

Discussion Securing the capacity of an ecosystem to generate services is dependent on the functioning of internal processes, such as nutrient recycling and succession, and the presence of habitat that can act as source areas for ecosystem renewal following a disturbance (Nyström and Folke 2001, Bengtsson et al. 2003). First, I will discuss how certain management practices affect the structure and function of ecosystems and their capacity to sustain a flow of ecosystem services, addressing in turn the functioning of internal process and linkages between habitats. Figure 4 presents a simplified view of these connections, and summarizes local management that can build resilience of these components. Secondly, I will discuss the context of management practices at different scales.

Maintaining capacity for internal renewal

A test of how certain management practices can contribute to the capacity of agroecosystems to absorb and recover from disturbance was provided by the devastations of the hurricane Mitch in Costa Rica (Holt-Gimenez 2002). An extensive monitoring of indicators such as amount of topsoil, field moisture, erosion levels and economic income revealed that small scale farms applying a bundle of practices referred to as sustainable, suffered less damage and recovered faster, as compared to similar farms applying conventional practices such as mechanical cultivation and application of chemical fertilizers and pesticides. Sustainable practices included planting along contour bounds, creation of terraces and ditches, intercropping including legumes, reduced use of chemical

22

Mobile links: lemurs, birds,

insects

Source ecosystems

Managed production ecosystem

Distur-bance

Examples of ecosystem services

Goods: Food production Pasture production Fuelwood, timber Supporting processes: Nutrient recycling Soil formation Local climate stabilization (Paper I and II)

Pollination Seed dispersal Biological control Ecosystem renewal (Paper I, II, IV, V)

Sources of genetic material for renewal Sources of mobile links (Paper III, IV, V)

Examples of management practices

Crop and landscape diversification Soil and water conservation Pest control management

Protection of agents of pest control, pollination, and seed dispersal (Paper II, IV)

Protection and creation of source areas/habitats (Paper II, IV)

Underlying institutions

Property rights regimes Collaboration Exchange networks

Social sanctions (Paper II), or taboos (Paper IV)

Taboos system with rules and sanctions (Paper IV)

Figure 4. Managed ecosystem and linkages to other ecosystems illustrating components in sustaining a flow of ecosystem services. The table lists examples of management practices identified in this thesis that contributes to maintain ecosystem processes for buffering change and allow for ecosystem renewal following change.

inputs, use of compost and animal manure, integrated pest management, mixing fields with woodlots and vegetative strips. Similar practices were found in the Mbulu highlands in Tanzania (Paper I), to some extent also in the study in Roslagen, Sweden (with the exception of the terracing and contour plowing)(Paper II), as well as in other studies from different parts of the world (e.g. Reijntjes et al. 1992, Netting 1993). Here, I emphasize their role in maintaining ecosystem function during periods of disturbance. Practices directed toward e.g. soil and water conservation, biological control, and the diversification of crops and fields can frame the impact of disturbances such as temporary droughts, pest outbreaks, and storms (Papers I and II).

The findings of Holt-Gimenez (2002) also indicate that there are certain management practices that can improve the capacity of the agroecosystem to return to a desired state

23

and prevent irreversible degradation of the agroecosystem. The developing science of agroecology emphasize the role of functioning ecosystem processes such as biological pest control, nutrient recycling, local climate stabilization, in sustaining long term agricultural production and rural food security (Carroll et al. 1990, Reijntjes et al. 1992, Altieri 1995). Management practices that for example stimulate soil biodiversity, increase soil organic matter, protect topsoil, improve local climate regulation, provide habitats for enemies of pests, and prevent weed infestation, can contribute to the capacity of an agroecosystem to return to a productive state after a disturbance (Reijntjes et al. 1992, Altieri 1994, Swift 1997, Janzen 1999).

Although general similarities in practices can be found across cases, a key role of local practices is the adaptation to site-specific conditions, e.g. taking advantage of small scale heterogeneity, or in responses to feedback such as temporary declines in production, or alterations in species composition (Scoones 2001, Papers I and II, see also Figure 3). The qualitative indicators identified in Paper II, developed to circumscribe uncertainty and avoid crop damage due to e.g. temporary drought or frost, provide additional examples. Paper I argues that local ecological understanding can function as a filter for adapting innovations and improvements, such as new crops or management techniques, to the local conditions.

Practices found in Papers I and II aim to spread risk, based on an awareness of recurrent but unpredictable disturbances. Risk spreading as a greater imperative than maximization of production has been identified in smallholder farming systems all over the world (e.g. Niemeijer 1996, Brookfield 2001). Diversification as a strategy to buffer crop production for disturbances such as temporary drought, pest outbreak or frost, was identified at several levels in Papers I and II. This includes total crop diversity on a farm, diversification within a field, and the utilization of small-scale heterogeneity across the landscape. For example, in the Mbulu highlands they use maize varieties that have different maturing time, intercrop maize, beans and pumpkins, and commonly plant a couple of rows of cassava somewhere on the farm. Cassava has low production in the cool climate but is very drought resistant and can remain in the soil for a long time until need arises. The cropping pattern is further diversified along niches in the landscape with different microclimatic conditions (see Figure 3).

There is mounting scientific evidence of the role of diversity in buffering change and providing insurance capital for disturbance. For example, within field crop diversity have been shown to increase efficiency of production during variable climatic conditions and pest outbreaks (Drinkwater et al. 1998, Wolfe 2000, Zhu et al. 2000). Further, the portfolio of species and landscape niches utilized has similarities with the recent proposal that

24

resilience in ecosystem function can be derived from a variety of responses to disturbance within a functional group, such as primary producers (see Elmqvist et al. 2003).

Project interventions to e.g. improve agricultural production have often failed to see the management strategies existing in local agroecosystems, and also failed to see the ecological processes underlying production, which in many cases has led to environmental degradation (Matson et al. 1997, Scott 1998, Sayer and Campbell 2002). Interventions with a narrow focus aiming to stabilize production through eradication of variability and disturbance have been found to decrease resilience, often through the alteration of slow ecological variables such as soil organic matter, landscape heterogeneity, and species diversity (Holling and Meffe 1996, Sayer and Campbell 2002). Further, external interventions can disrupt the underlying structure for management and thus remove mechanisms for creative, adaptive response by people (Scott 1998, van der Leeuw and team 2000).

Sustaining landscape processes

Three factors have been pointed out as crucial for ecosystem regeneration following disturbance: a) remaining legacies in the landscape struck by disturbance, such as seed banks, stumps, or large trees, b) the existence of nearby areas that can supply propagules and organisms to re-colonize the disturbed area, c) the presence of organisms that can transport new seeds and propagules to the disturbed area (i.e. mobile links, Lundberg and Moberg 2003), see Figure 4 (Nyström and Folke 2001, Bengtsson et al. 2003).

The section above addresses processes relevant within factor a). In the following I will also address factors b) and c). The practice of protecting areas in space and time is a component of local systems for ecosystem management that is found all over the world, especially in tropical areas (see e.g. Ramakrishnan 1998). In Africa, sacred forests have been identified in most countries of Sub-Saharan Africa (see for example Schoffeleers 1978, Wilson 1993, McClanahan et al. 1997, Kokou et al. 1999, Byers et al. 2001, Virtanen 2002, Mgumia and Oba 2003), including Madagascar (e.g. Horning 2003b). The practices can be seen as nurturing source of renewal of local ecosystems (Berkes and Folke 2002, Colding et al. 2003b).

The dynamics of the dry forests of southern Madagascar is not well understood (Ganzhorn et al. 2003, Elmqvist 2004). Paper III reports of large-scale regeneration of the forest in Androy, which contrasts to previous assumptions that the potential for regeneration in these dry ecosystems is low. In Androy, the practices of setting aside certain areas from human use has created a network of forest patches that can be seen as “islands” located in a landscape intensively managed for crop and livestock production (Paper IV). Such a

25

network of forest patches can have important functions as source areas of propagules for forest regeneration (Janzen 1988, Chazdon 2003), and also for maintenance of regional species diversity and ecological functioning (Khan et al. 1997, Mgumia and Oba 2003).

For the taboo forests to act as source areas, it requires functioning mechanisms for seed dispersal, i.e. through wind or animal vectors. Vertebrates that eat fruits, thus can play a key functional role in the regeneration of tropical forests hit by disturbance by bringing in seeds from surrounding ecosystems (Cox et al. 1991, Ganzhorn et al. 1999, Elmqvist et al. 2002). The taboo patches can serve as important habitats for mobile link species, such as birds and lemurs (Paper IV). In Paper V, we investigate Ring-tailed lemur, Lemur catta, as seed disperser in southern Androy. Provided that L. catta can move more than 1000 m in the fragmented landscape of southern Androy, we find that landscape connectivity and thus the potential for seed dispersal is fairly good. Paper IV suggests and modeling exercises in Paper V strengthen that the forest patches in southern Androy has a significant value in sustaining the generation of ecosystem services in the landscape. Paper V further illustrates that the spatial configuration of the forest patches has important implications for the potential for seed dispersal as well as for crop pollination in the agricultural landscape surrounding the forest patches.

The landscape pattern of forest patches in Androy is strongly associated with the practice of assigning protection to certain forest patches and clusters of trees (Paper IV), with implications both for where the patches are located in the landscape and for their continued existence. The taboos have preserved a number of patches for at least 50 years and created legacies in the landscape with high cultural value and unique species composition in the highly managed production landscape. Furthermore, practices of tree planting and reserving new areas as taboo may add forested patches in the landscape (Paper IV). In northern Androy mobility of people is high. As taboo forests often are located close to human settlements, they may play an important role for forest regeneration of abandoned fields and settlements.

In the case of habitat taboos or forest protection by local belief systems, there seems to be significant overlap between local people and governments and external actors with an interest in preventing resource degradation and protect biodiversity (Berkes 2004). In spite of this, there have been few attempts to integrate such local management systems into formal natural resource management (Gadgil et al. 1998, Mgumia and Oba 2003). In the debate on policies for protecting biodiversity, the current emphasize on focusing protection of biodiversity on setting aside “hot spot” areas of high species richness does not take the contribution of natural habitats to sustain functions in productive landscapes into account. Although of great importance for biodiversity maintenance and generation of ecosystem

26

services at a regional level, such a strategy for biodiversity protection will in itself not be a viable option in human-dominated environments (Folke et al. 1996, Bawa et al. 2004, Palmer et al. 2004). Thus, in productive landscapes, additional approaches are needed to sustain biodiversity necessary for the generation of ecosystem services. Bengtsson et al (2003) argue that present reserves, which often are aimed at preserving a static ecosystem, should be complemented with dynamic reserves, such as ecological fallows and dynamic successional reserves, to sustain species and processes required for renewing and reorganizing desired ecosystem states after disturbance. A network of small high-value protected areas within productive landscape as in southern Androy can be a valuable tool for sustaining landscape ecosystem services (Paper V, De Marco and Coelho 2004).

Practices, knowledge, and governance systems

Shifting focus from managing harvest levels of natural resources to managing the capacity of the ecosystem that generates them requires an understanding of the underlying ecological processes. Scott (1998) argues that experience based practical knowledge, generated through trial-and-error learning, is the most appropriate way to address complex problems and conditions. Such experience based learning can be generated through management practices that are responsive to ecosystem feedback. The practices identified inPaper II, such as the qualitative indicators related to ecosystem succession and protection of species contributing to pest control and pollination (Table 5), indicate that the local users of ecosystem services have an understanding of ecosystem processes and their dynamics. Similarly, the listed benefits generated by taboo forests that were appreciated by local informants (Table 4 Paper IV) indicated awareness of the ecological roles of the protected forests.

To maintain, re-evaluate, and transfer such knowledge, it needs to be embedded in institutions, such as customary rules and practices. Berkes and Folke (2002) emphasize the importance of responding to change or crisis with experience, which can lead to adequate responses and prevent the build up of a larger scale crisis. Management practices can function as repositories of knowledge that can be reactivated following disturbance and change, as in the case of the Swedish farmers in Paper II who are returning to the old practice of multi-species leys that is locally known to produce more stable harvest during variable climatic conditions that the farmers have experienced during the last decade. Experience of earlier disturbances that can be drawn upon during times of crisis can reside in the ecological system, e.g. in species composition and landscape structures such as the age classes of trees regenerated after a forest fire, as well as within the social system, carried in practices such as described in this thesis, in institutions such as the taboos of Androy, in narratives, or rituals (Berkes 1999, McIntosh et al. 2000). Traditional

27

management of natural resources is often embedded in ritual and local religious practices (Gadgil et al. 1993, Berkes 1999), and cultural values may thus be closely related to ecological values and local ecological understanding.

For example, the practice of protecting certain habitats is often regulated in local institutions such as taboos (Colding and Folke 2001, Paper IV). The taboos in Androy have secured the protection of numerous forest patches in spite of strong pressures on forest resources (Paper III and IV). Generally, the taboos that are associated with belief in the power of ancestral or other spirits ban harvest of all living resources from the forests. In southern Androy, practically all the remaining forest habitats are protected by taboos, and this is the area with the most well enforced taboos (Paper IV). Paper V showed the generation of some ecosystem services in this area to be vulnerable to even small changes in forest cover. Gadgil and Vartak (1976) argue that given the complexity of ecosystems and the difficulty in regulating the quantity of resource harvesting, a total ban of resource extraction or access may be the most easily perceived and most efficient way of guarding against resource depletion. Ascribing cultural or spiritual values to a certain resource has been described as a way of securing resources for the benefit of a wider community (Schoffeleers 1978, Barrow 1996).

The landscape of taboo forests in Androy is not static. Continued protection of forest patches in Androy was found to be dependent on active enforcement of sanctions, and new areas are established, e.g. to form new burial grounds or ceremonial places. Thus, we see a potential to build on such social mechanisms to protect vulnerable areas or even for restoration of forest habitats. The tools used in Paper V can be used to identify areas that are particularly strategic for the generation of ecosystem services. In India, there are examples of forests protection based on religious beliefs that had been abandoned but were revived again when their value for generating ecosystem services was realized by the local community (Gadgil et al. 1998). In Oceania, local practices for marine resource management, including species and habitat taboos, have been revived and found to be successful in managing coastal ecosystems (Johannes 2002).

Paper II concludes that similar approaches to dealing with uncertainty and change can be found in different parts of the world. However, it does not claim that all local communities possess adequate local ecological knowledge or that the existing institutional framework is relevant for ecosystem management. The capacity of local communities to adaptively govern ecosystems cannot be taken for granted (Barrett et al. 2001). Whereas some communities have a rich and longstanding knowledge about their local ecosystem that is actively used in management, others communities knowledge base may be malfunctioning, lost or have never developed, depending on internal and external drivers. Paper III

28

suggests that the dynamics of local institutions have played a key role in the locally contradictory patterns of forest regeneration, degradation, and protection. For example, lack of enforced property rights has contributed to a situation of land clearings to secure entitlements to the land. In the Mbulu highlands, a number of social mechanisms acting at different scales affect land-use practices, see Figure 2 and Paper I. For example, access to pasture resources was accessed via common property institutions, which included monitoring and regulation of use. Further, the risk of crop failures are shared across scales through social networks of food exchange, where harvests are shared across areas of different biophysical conditions in case of localized or temporary crop failures (Paper I, see also Loiske 2003). Mazzucato and Niemeijer (2000) showed that farmers’ land use planning and application of soil conservation methods was strongly associated with local networks for exchange of goods and labor.

Thus, knowledge may exist without being due to drivers such as economic opportunities, insecure property rights, and societal turbulence. In the Mbulu highlands, a shift in land rights from communal pastures to private fields affects the capacity of the farmers to monitor and respond to feedback on the state of pasture resources (Paper I). In Paper II, farmers in Roslagen, Sweden, were aware of multiple benefits of intercropping, but were constrained in applying this knowledge by retailers that did not accept mixed grains. Kepe and Scoones (1999) showed that farmers have knowledge of the disturbance regimes of fire and grazing pulses that create grasslands with different species community and different value for pasture production. However, factors such as limited access to pasture lead to failure to regulate stocking densities and maintain the desired states.

Ekblom (2004) argues that from a local perspective, forest loss and degeneration is associated with loss of local control. Similar findings are reported in, e.g., Gibson et al. (2000) and by Horning (2003a). In many areas of the world local protection of e.g. forest areas has limited capacity to deal with strong external drivers and rapid change and erosion of local institutions due to cross-scale factors such as cultural changes, policy interventions, tenure instability, migration, and education (e.g. Sheperd 1992, Wilson 1993, Cox and Elmqvist 1997, Horning 2003b, Kajembe et al. 2003, Gokhale 2004). Cross-scale linkages between local and higher level institutions and organizations, such government authorities or NGOs, that can secure local entitlements to land and resources and provide some insulation from the pressures of global markets, are important conditions for maintaining the local capacity to adaptively manage ecosystem resilience (Alcorn et al. 2003, Brown 2003, Pretty 2003).

29

Conclusions People are part of and not apart from ecosystems. This is particularly true for production landscapes where the landscape is more or less transformed to generate a set of desired ecosystem services. The capacity of the landscape to sustain a flow of ecosystem services is strongly connected to human management practices and sustainable ecosystem management needs to include the local users and inhabitants.

This thesis identifies management practices that make ecological sense in complex adaptive systems and contribute to resilience in social-ecological systems. Specifically, we have identified practices that a) sustain internal processes that buffer for disturbance and enhance capacity to re-organize following disruptive change, and b) protect source ecosystems in the landscape that can contribute to regeneration and ecosystem renewal. Further, these practices can contribute to the learning capacity in the social-ecological system, provided a governance system that can store, re-evaluate, and transmit the experience generated. The thesis demonstrated how local management practices are dependent on and embedded in a governance system and how social dynamics have implication for ecosystem management and ecological dynamics. The thesis emphasizes the importance of sustaining governance systems that generates, re-evaluates, and maintains local management practices for ecosystem resilience. It illustrates the substantial role of social capital for development of sustainable ecosystem management (Pretty and Ward 2001). Thus, identifying and mapping the governance system that underlies the desired landscape and the generation of the desired ecosystem services – be it food production or biodiversity conservation - is an essential part in developing schemes for resource and ecosystem management.

Landscape and systems ecologists are increasingly seeing the need to manage landscapes, not only parks and their buffer zones, and the structures and ecological processes that maintain their values in the long run (Bengtsson et al 2003). Increased understanding of the dynamics and regenerative capacity of e.g. forest ecosystems and of how management practices in the landscape can strengthen or disrupt landscape processes and connections between formally or informally protected areas is essential in this context. Such understanding may contribute to maintaining biological diversity and foster resilience both within and outside natural reserves.

The thesis suggests that a sustained local capacity to self-organize around and in response to local ecosystem dynamics as a prerequisite for resilient management practices to emerge and evolve. Pretty (2003) shows that new local groups for ecosystem management can and are emerging in different parts of the world, given an appropriate institutional context. Similar findings are reported by Olsson and colleagues (Olsson et al. 2004a, Olsson et al.

30

2004b). However, as was shown by Huitric (2003), a good governance system does not necessarily imply a resilient social-ecological system. More studies are needed on the role of locally devised management institutions in crafting schemes for sustainable management. The rationale for doing this stems from evidence of failures of community based management in Africa and other areas (du Toit et al. 2004), which has been attributed to lack of appreciation of local conditions and informal institutions (Songorwa 1999, Murphree 2002, Brown 2003), as well as from the growing literature of case studies describing the capacity of self-organized management systems to adaptively govern ecosystem dynamics in developing as well and developed countries (Papers I, II and IV)(Fairhead and Leach 1996b, Gadgil et al. 1998, Niamir-Fuller 1998, Johannes 2002, Olsson et al. 2004a).

Expansion of the network of protected areas in Madagascar needs to address the key challenge of constructing reserves that do not challenge local livelihoods. Protected areas that are not anchored locally will remain vulnerable and costly to enforce. In Androy, there is a strong mistrust of top-down intervention. Yet, there exists a local practice assigning protected areas through taboos that could be built upon when designing protected areas in the unique dry spiny forest of southern Madagascar. The present IUCN categories for protected areas are focused on ecological parameters and do not take governance characteristics into account (IUCN 1994). I agree with the recommendation from the Vth IUCN World Parks Congress where it is argued that alternative governance systems, such as community conserved areas, should be included in the system of protected area categories (Recommendation 5.26. IUCN 2003).

To conclude, failure to recognize social-ecological connections in management can lead to unintended consequences for the capacity for ecosystem management resulting from changes in parameters that seem unrelated to ecosystem interactions, such as belief systems or social networks. The work in this thesis demonstrates the need for the development of an interdisciplinary conservation science that includes an expanded understanding of the coupled dynamics of social-ecological interactions. Increased understanding of locally devised practices and rules as linkages between social and ecological systems can provide both site-specific and generally relevant insights for management of resilient ecosystems for human well-being.

NOTE 1 Local ecological knowledge (LEK) has been defined as the knowledge held by a specific group of people about their local ecosystems (Olsson and Folke 2001). This definition recognizes that ecological knowledge used in local resource management can also be generated by and reside in communities that lack historical and cultural continuity. Local ecological knowledge is a blend of knowledge generated locally through practice and experience plus knowledge incorporated from other sources, such as scientific knowledge (se Figure 1 in Paper II). Another definition of local understanding for ecosystem management is Traditional

31

ecological knowledge (TEK): as a cumulative body of knowledge, practices, and beliefs about the relationships of living beings, including humans, to one another and to the environment. This body of knowledge evolves by adaptive processes and is handed down through generations by cultural transmission (Gadgil et al. 1993).

Acknowledgements Many people have contributed to this thesis. Thanks to all of you! In particular I would like to mention:

Thomas Elmqvist, my supervisor, for inviting me to the Madagascar project and for your reliable support and supervision since day one. It is deeply appreciated!

Carl Folke, for generously spreading enlightenment and for sharing your deep engagement and visions for a better world. You have been there for me at some crucial times. I am also grateful to you for providing opportunities to participate in Resilience Alliance and Millennium Ecosystem Assessment contexts, which has meant a lot to me.

Monica Hammer, for making my PhD-project possible, supervision especially during the first half of my thesis, and for your support and inspiration.

My co-authors. Particular thanks goes to Örjan B who stood up to my deadlines. Kristina, I enjoyed our work but even more the friendship that developed.

Markku Pyykönen for the maps.

Lisa, Thomas H, Miriam, and Erik, for last minute help with the papers and the kappa. Without that I wouldn’t have made it!

All my colleagues at Natural Resource Management, CTM, and Beijer. You are such a wonderful, supportive, and warm group of people. Each of you would deserve a paragraph of your own, for inspiring discussions, pep talks, shared lunches, and smiles, but that would take too much of my last precious hours to write…. Some of you have to be mentioned in particular though. Line, for instant help, numerous discussions (including therapy), great hospitality, and for arranging fun things, such as skiing trips and parties! Miriam, for providing lattes and chocolate, virtual as well as real ones, at the right times. Lisa, for sharing your energy, especially during these last months. My roommates, Åsa S, Cecilia, Rebecka, Erik, and Elin, for sharing aha’s and frustrations, fruit bowls and whey cheese (in strange combinations), fika’s and recipes, and laughter.

Lasse, for invaluable computer support. It is so comforting just to know that you are there! Barbara and Birgitta, for help always served with a smile, and Ragnar and Siw for support. Bosse for interesting talks and smiles in the corridor.

The PLATINA network, based the Geography departments at SU, for providing an inspiring context.

Gia Kjellén, my mentor, for sharing your experience and for important pep-talks!

All my informants and other people in Tanzania and Madagascar that have shared their knowledge and contributed to the success of fieldwork, and, not least, great personal experiences. In the Mbulu highlands, I am particularly grateful for the help and hospitality

32

of Mehhi Amsi and Dominik Habiye with families in Kwermusl, and for the assistance of Basili Awett, Arkadi Akonaay, Ero Boay, and Yusefu, Forest officer in Murray. Deogratius Hillu for advice, assistance, and translations. Rose Basili, Willibrord Maqway, and Flora Madamgi also acted as translators during interviews. Ally Msuya for logistical help. Thanks also to Vesa-Matti Loiske, who first took me to the Mbulu highlands, for establishing the conditions to do research in and around the area.

In Androy, I am grateful for the hospitality in the villages of Manave, Mitsoriake, Mareñy, Lahabe, and Ambonaivo. In particular I would like to mention the help and friendship of Manandongo, Safimana, Nicola, and Fingana, who opened doors for us and stood by us.

Jean Aimé Rakotoarisoa, who can solve any problem, Elisabeth Rabakonandrianina, and Chantal Radimilahy, for your support in Madagascar and throughout the project. I am also grateful to George Heurtebize, for his great hospitality and friendly sharing of his extensive knowledge, Ramilisonina, Retsihisatse, Mike Parker-Pearson, and Karen Godden.

All my fieldwork companions. Doing fieldwork in remote places creates strong bonds between people. Anna-Carin, you shared my first Africa experience and were my partner during my first fieldwork in Tanzania, which has played a very important role in this thesis. Thank you for good collaboration and friendship! Together with Marie and Lowe, we had a great time in the Mbulu highlands, of discoveries and giggle. Lowe Börjeson, our discussions and your help throughout this time have been very important to me!

Further, Anna, Markus, Herilala and Fafa, we went through good and bad adventures in the spiny bush together. I am happy to have shared it with you! Anna, for also being my GIS support. Andriamaherilala Jean Aimé for being my window to people and understanding of life in Androy and Madagascar. I learned many things with you. Victor Razanatovy – simply the best driver with many talents!

All my friends, I am privileged to have you! You know who you are. Lotta, for always knowing who I am. Linn, for important pepping at the right time.

My family: My parents, Jan and Lisa, for your always-present support, love and good advices, and for bringing on your respect for nature and joy of dwelling there! My great brothers Pelle and Oskar, and the most recent members, Vivi and Julian. Lottie and Calle, for your generosity and for stimulating discussions! And Simon, you have contributed to this thesis in so many ways, as field partner, cover designer, illustrator, computer support, discussion partner and more. But on top of all that is the invaluable everyday support and love which I can’t thank you enough for. You are the greatest! and I am so happy to be with you.

Finally, Sida/Sarec for funding my thesis work.

33

References

Anonymous. 2003. Good news at the Parks Congress. Plant Talk 34. Alcorn, J., J. Bamba, S. Masiun, I. Natalia, and A. G. Royo. 2003. Keeping ecological

resilience afloat in cross-scale turbulence: an indigenous social movement navigates change in Indonesia. Pages 299-327 in F. Berkes, C. Folke, and J. Colding, editors. Navigating Social-Ecological Systems. Building resilience for complexity and change. Cambridge University Press, Cambridge.

Altieri, M. A. 1994. Biodiversity and pest management in agroecosystems. Hayworth Press, New York.

Altieri, M. A. 1995. Agroecology. The science of sustainable agriculture. Westview Press, Boulder. Colorado.

Arrow, K., B. Bolin, R. Costanza, P. Dasgupta, C. Folke, C. S. Holling, B. O. Jansson, S. Levin, K. G. Maler, C. Perrings, and D. Pimentel. 1995. Economic growth, carrying capacity, and the environment. Ecological Economics 15:91-95.

Arthur, W. B. 1999. Complexity and the economy. Science 284:107-109. Barrett, C. B., K. Brandon, C. Gibson, and H. Gjertsen. 2001. Conserving tropical

biodiversity amid weak institutions. Bioscience 51:497-502. Barrow, E. G. C. 1996. The drylands of Africa. Local participation of tree management.

Inititives Publishers, Nairobi. Battistini, R., and G. Richard-Vindard. 1972. Biogeography and ecology in Madagascar.

Dr. W Junk B.V. Publishers, Hague. Bawa, K. S., W. J. Kress, N. M. Nadkarni, S. Lele, P. H. Raven, D. H. Janzen, A. E. Lugo,

P. S. Ashton, and T. E. Lovejoy. 2004. Tropical ecosystems into the 21st century. Science 306:227-228.

Bengtsson, J., P. Angelstam, T. Elmqvist, U. Emanuelsson, C. Folke, M. Ihse, F. Moberg, and M. Nyström. 2003. Reserves, Resilience, and Dynamic Landscapes. Ambio 32:389-396.

Berkes, F., and C. Folke, editors. 1998. Linking Social and Ecological Systems. Management Practices and Social Mechanisms for Building Resilience. Cambridge University Press, Cambridge.

Berkes, F. 1999. Sacred ecology: traditional ecological knowledge and resource management. Taylor & Francis, London.

Berkes, F., J. Colding, and C. Folke. 2000. Rediscovery of traditional ecological knowledge as adaptive management. Ecological Applications 10:1251-1262.

Berkes, F., and C. Folke. 2002. Back to the future: Ecosystem change, institutions and local knowledge. in L. H. Gunderson and C. S. Holling, editors. Panarchy : Understanding Transformations in Human and Natural Systems. Island Press.

Berkes, F., J. Colding, and C. Folke, editors. 2003. Navigating Social-Ecological Systems: Building Resilience for Complexity and Change. Cambridge University Press, Cambridge.

Berkes, F. 2004. Rethinking community-based conservation. Conservation Biology 18:621-630.

Börjeson, L. 2003. The history of Iraqw intensive agriculture. in M. Widgren and J. Sutton, editors. Islands of intensive agriculture in Eastern Africa. James Currey Publishers, Oxford.

Börjeson, L. 2004. A history under siege. Intensive Agriculture in the Mbulu highlands, Tanzania, 19th century to the present. PhD. Stockholm University, Stockholm.

Brookfield, H. 2001. Exploring agrodiversity. Columbia University Press, New York.

34

Brown, K. 2003. Integrating conservation and development: a case of institutional misfit. Frontiers in Ecology and the Environment 1:479-487.

Byers, B. A., R. N. Cunliffe, and A. T. Hudak. 2001. Linking the conservation of culture and nature: A case study of sacred forests in Zimbabwe. Human Ecology 29:187-218.

Carpenter, S., B. Walker, J. M. Anderies, and M. Abel. 2001. From Metaphor to Measurement: Resilience of What to What? Ecosystems 4:765-781.

Carroll, C. R., J. H. Vandermeer, and P. M. Rosset. 1990. Agroecology. McGraw-Hill, New York.

Casse, T., A. Milhoj, S. Ranaivoson, and J. R. Randriamanarivo. 2004. Causes of deforestation in southwestern Madagascar: what do we know? Forest Policy and Economics 6:33-48.

Chazdon, R. L. 2003. Tropical forest recovery: legacies of human impact and natural disturbances. Perspectives in Plant Ecology Evolution and Systematics 6:51-71.

Clark, C. D., S. M. Garrod, and M. P. Pearson. 1998. Landscape archaeology and remote sensing in southern Madagascar. International Journal of Remote Sensing 19:1461-1477.

Colding, J., and C. Folke. 2001. Social taboos: "Invisible" systems of local resource management and biological conservation. Ecological Applications 11:584-600.

Colding, J., T. Elmqvist, and P. Olsson. 2003a. Living with disturbance: building resilience in social-ecological systems. in F. Berkes, J. Colding, and C. Folke, editors. Navigating Social-Ecological Systems: Building Resilience for Complexity and Change. Cambridge University Press, Cambridge.

Colding, J., C. Folke, and T. Elmqvist. 2003b. Social institutions in ecosystem management and biodiversity conservation. Tropical Ecology 44:25-41.

Costanza, R., L. Wainger, C. Folke, and K. G. Maler. 1993. Modeling Complex Ecological Economic-Systems - toward an Evolutionary, Dynamic Understanding of People and Nature. Bioscience 43:545-555.

Cox, P. A., T. Elmqvist, E. E. Rainey, and E. D. Pierson. 1991. Flying foxes as strong interactors in South Pacific Island Ecosystems: A conservation hypothesis. Conservation Biology 5:448-454.

Cox, P. A., and T. Elmqvist. 1997. Ecocolonialism and indigenous-controlled rainforest preserves in Samoa. Ambio 26:84-89.

Daily, G., editor. 1997. Nature's services. Societal dependence on natural ecosystems. Island Press, Washington DC.

Davis, A., and J. R. Wagner. 2003. Who knows? On the importance of identifying "Experts" when researching local ecological knowledge. Human Ecology 31:463-489.

De Marco, P., and F. M. Coelho. 2004. Services performed by the ecosystem: forest remnants influence agricultural cultures' pollination and production. Biodiversity and Conservation 13:1245-1255.

Diaz, S., and M. Cabido. 2001. Vive la differance: plant functional diversity matters to ecosystem processes. Trends in Ecology & Evolution 16:646-655.

Dietz, T., E. Ostrom, and P. C. Stern. 2003. The struggle to govern the commons. Science 302:1907-1912.

Dostie, B., S. Haggblade, and J. Randriamamonjy. 2002. Seasonal poverty in Madagascar: magnitude and solutions. Food Policy 27:493-518.

Drinkwater, L. E., P. Wagoner, and M. Sarrantonio. 1998. Legume-based cropping systems have reduced carbon and nitrogen losses. Nature 396:262-265.

35

du Toit, J. T., B. H. Walker, and B. M. Campbell. 2004. Conserving tropical nature: current challenges for ecologists. Trends in Ecology & Evolution 19:12-17.

Ekblom, A. 2004. Changing landscapes. An environmental history of Chibuene, Southern Mozambique. Uppsala University, Uppsala.

Elmqvist, T., M. Wall, A. L. Berggren, L. Blix, A. Fritioff, and U. Rinman. 2002. Tropical forest reorganization after cyclone and fire disturbance in Samoa: Remnant trees as biological legacies. Conservation Ecology 5.

Elmqvist, T., C. Folke, M. Nyström, G. Peterson, J. Bengtsson, B. Walker, and J. Norberg. 2003. Response diversity, ecosystem change, and resilience. Frontiers in Ecology 1.

Elmqvist, T. 2004. The forgotten dry forest of southern Madagascar. Plant Talk 35:29-31. Fabricius, C., and E. Koch. 2004. Rights, resources, and rural development. Earthscan,

London. Fairhead, J., and M. Leach. 1996a. Misreading the African landscape; Society and ecology

in a forest-savanna mosaic. Cambridge University Press, Cambridge. Fairhead, J., and M. Leach. 1996b. Enriching the landscape: Social history and the

management of transition ecology in the forest-savanna mosaic of the Republic of Guinea. Africa 66:14-36.

Fenn, M. 2003. The spiny forest ecoregion. in S. M. Goodman and J. P. Benstead, editors. The natural history of Madagascar. University of Chicago Press, Chicago.

Folke, C., C. S. Holling, and C. Perrings. 1996. Biological diversity, ecosystems, and the human scale. Ecological Applications 6:1018-1024.

Folke, C., J. Colding, and F. Berkes. 2003. Building resilience and adaptive capacity in social-ecological systems. in F. Berkes, J. Colding, and C. Folke, editors. Navigating Social-Ecological Systems: Building Resilience for Complexity and Change. Cambridge University Press, Cambridge.

Folke, C., S. Carpenter, B. Walker, M. Scheffer, T. Elmqvist, L. Gunderson, and C. S. Holling. 2004. Regime shifts, resilience, and biodiversity in ecosystem management. Annual Review of Ecology, Evolution, and Systematics 35:557-581.

Gadgil, M., F. Berkes, and C. Folke. 1993. Indigenous Knowledge for Biodiversity Conservation. Ambio 22:151-156.

Gadgil, M., N. S. Hemam, and B. M. Reddy. 1998. People, refugia and resilience. Pages 30-47 in F. Berkes and C. Folke, editors. Linking Social and Ecological Systems. Management Practices and Social Mechanisms for Building Resilience. Cambridge University Press, Cambridge.

Ganzhorn, J. U., J. Fietz, E. Rakotovao, D. Schwab, and D. Zinner. 1999. Lemurs and the regeneration of dry deciduous forest in Madagascar. Conservation Biology 13:794-804.

Ganzhorn, J. U., S. Goodman, and A. Dehgan. 2003. Effects of forest fragmentation on small mammals and lemurs. Pages 1228-1234 in S. M. Goodman and J. P. Benstead, editors. The natural history of Madagascar. University of Chicago Press, Chicago.

Gennep, A. v. 1904. Tabou et totémisme à Madagascar: Étude descriptive et théorique. Bibliothèque de l'École des hautes études, Paris.

Gibson, C., M. A. McKean, and E. Ostrom, editors. 2000. People and Forests. Communitites, institutions and governance. The MIT Press, Cambridge, Massachusetts.

Gokhale, Y. 2004. Reviving traditional forest management in Western Ghats. Economic and Political Weekly:3556-3559.

36

Gould, L., R. W. Sussman, and M. L. Sauther. 1999. Natural disasters and primate populations: The effects of a 2-year drought on a naturally occurring population of ring-tailed lemurs (Lemur catta) in southwestern Madagascar. International Journal of Primatology 20:69-84.

Gunderson, L., and C. S. Holling, editors. 2003. Panarchy: understanding transformations in human and natural systems. Island Press, Washington (DC).

Hanna, S., C. Folke, and K.-G. Mäler, editors. 1996. Rights to nature. Island Press, Washinton DC.

Heurtebize, G. 1986. Histoire des Afomarolahy (Extrême-Sud de Madagascar). CNRS, Paris.

Holling, C. S. 1973. Resilience and stability of ecological systems. Annual Review of Ecology and Systematics 4:1-23.

Holling, C. S. 1986. The resilience of terrestrial ecosystems: local surprise and global change. Pages 292-317 in W. C. Clark and R. E. Munn, editors. Sustainable development of the biosphere. Cambridge University Press, Cambridge.

Holling, C. S., and G. K. Meffe. 1996. Command and control and the pathology of natural resource management. Conservation Biology 10:328-337.

Holling, C. S., F. Berkes, and C. Folke. 1998. Science, sustainability and resource management. Pages 342-362 in F. Berkes and C. Folke, editors. Linking Social and Ecological Systems. Management Practices and Social Mechanisms for Building Resilience. Cambridge University Press, Cambridge.

Holling, C. S. 2001. Understanding the complexity of economic, ecological, and social systems. Ecosystems 4:390-405.

Holt-Gimenez, E. 2002. Measuring farmers' agroecological resistance after Hurricane Mitch in Nicaragua: a case study in participatory, sustainable land management impact monitoring. Agriculture Ecosystems & Environment 93:87-105.

Horning, N. R. 2003a. How rules affect conservation outcomes. Pages 146-153 in S. M. Goodman and J. P. Benstead, editors. The natural history of Madagascar. Chicago University Press, Chicago.

Horning, N. R. 2003b. The cost of ignoring rules: How Madagascar’s biodiversity and rural livelihoods have suffered from institutional shortcomings. in The International Conference on Rural Livelihoods, Forests and Biodiversity 19-23 May 2003, Bonn, Germany.

IUCN. 1994. Guidelines for Protected Areas Management Categories. The World Conservation Union (IUCN), Cambridge, UK and Gland, Switzerland.

IUCN. 2003. Recommendations: Vth IUCN World Parks Congress, Durban, South Africa. IUCN, Gland, Switzerland.

Janzen, D. 1988. Management of habitat fragments in a tropical dry forest: growth. Annals of the Missouri Botanical Garden 75:105-116.

Janzen, D. 1999. Gardenification of tropical conserved wildlands: Multitasking, multicropping, and multiusers. Proceedings of the National Academy of Sciences of the United States of America 96:5987-5994.

Johannes, R. E. 2002. The renaissance of community-based marine resource management in Oceania. Annual Review of Ecology and Systematics 33:317-340.

Kajembe, G. C., E. J. Luoga, M. S. Kijazi, and C. S. Mwaipopo. 2003. The role of traditional institutions in the conservation of forest resources in East Usambara, Tanzania. International Journal of Sustainable Development and World Ecology 10:101-107.

Kasperson, R. E., J. X. Kasperson, and B. L. Turner. 1995. Regions at risk: comparisons of threatened environments. United Nations University Press, New York.

37

Kates, R. W., W. C. Clark, R. Corell, J. M. Hall, C. C. Jaeger, I. Lowe, J. J. McCarthy, H. J. Schellnhuber, B. Bolin, N. M. Dickson, S. Faucheux, G. C. Gallopin, A. Grubler, B. Huntley, J. Jager, N. S. Jodha, R. E. Kasperson, A. Mabogunje, P. Matson, H. Mooney, B. Moore, T. O'Riordan, and U. Svedin. 2001. Environment and development - Sustainability science. Science 292:641-642.

Kepe, T., and I. Scoones. 1999. Creating grasslands: Social institutions and environmental change in Mkambati area, South Africa. Human Ecology 27:29-53.

Khan, M. L., S. Menon, and K. S. Bawa. 1997. Effectiveness of the protected area network in biodiversity conservation: A case-study of Meghalaya state. Biodiversity and Conservation 6:853-868.

Kinzig, A., and e. al. 2000. Nature and Society; an imperativ for intergrated environmental research. Workshop report National Science Foundation, Tempe, Arizona.

Kinzig, A., S. W. Pacala, and D. Tilman, editors. 2002. The functional consequences of biodiversity. Princeton University Press, Princeton, N.J.

Koechlin, J. 1972. Flora and vegetation of Madagascar. in R. Battistini and G. Richard-Vindard, editors. Biogeography and ecology in Madagascar. Dr. W Junk B.V. Publishers, Hague.

Kokou, K., G. Caballe, K. Akpagana, and K. Batawila. 1999. Forest islands of southern Togo: dynamics and relationship with surrounding vegetations. Revue D Ecologie-La Terre Et La Vie 54:301-314.

Kull, C. A. 2004. Isle of fire. The political ecology of landscape burning in Madagascar. The University of Chicago Press, Chicago.

Lambek, M. 1992. Taboo as Cultural-Practice among Malagasy Speakers. Man 27:245-266.

Lawi, Y. Q. 1999. Where physical and ideological landscapes meet: Landscape use and ecological knowledge in Iraqw, northern Tanzania, 1920s- 1950s. International Journal of African Historical Studies 32:281-310.

Levin, S. A. 1999. Fragile dominion. Perseus Books, Cambridge, Mass. Loiske, V.-M. 2003. Institutionalised exchange as a driving force in intensive agriculture:

an Iraqw case study. in M. Widgren and J. Sutton, editors. Islands of intensive agriculture in eastern Africa. James Currey Publishers, Oxford.

Loreau, M., S. Naeem, P. Inchausti, J. Bengtsson, J. P. Grime, A. Hector, D. U. Hooper, M. A. Huston, D. Raffaelli, B. Schmid, D. Tilman, and D. A. Wardle. 2001. Ecology - Biodiversity and ecosystem functioning: Current knowledge and future challenges. Science 294:804-808.

Manson, S. M. 2001. Simplifying complexity: a review of complexity theory. Geoforum 32:405-414.

Matson, P. A., W. J. Parton, A. G. Power, and M. J. Swift. 1997. Agricultural intensification and ecosystem properties. Science 277:504-509.

Mazzucato, V., and D. Niemeijer. 2000. The cultural economy of soil and water conservation: Market principles and social networks in eastern Burkina Faso. Development and Change 31:831-855.

McClanahan, T. R., H. Glaesel, J. Rubens, and R. Kiambo. 1997. The effects of traditional fisheries management on fisheries yields and the coral-reef ecosystems of southern Kenya. Environmental Conservation 24:105-120.

McConnell, W. 2002. Madagascar: Emerald Isle or Paradise lost? Environment 44:10-22. McIntosh, R. J., J. A. Tainter, and S. K. McIntosh, editors. 2000. The way the wind blows.

Climate, history, and human action. Columbia University Press, New York.

38

Meindertsma, J. D., and J. J. Kessler. 1997. Towards a better use of environmental resources. A planning document of Mbulu and Karatu Districts, Tanzania. Mbulu District Council, Mbulu, Tanzania.

Mgumia, F., and G. Oba. 2003. Potential role of sacred groves in biodiversity conservation in Tanzania. Environmental Conservation 30:259-265.

Millennium Ecosystem Assessment. 2003. Ecosystems and Human Well-being. Island Press, Washinton DC.

Murphree, M. W. 2002. Protected areas and the commons. The Common Property Resource Digest 60:1-4.

Netting, R. M. 1993. Smallholders, householders. Farm families and the ecology of intensive, sustainable agriculture. Stanford University Press, Stanford.

Niamir-Fuller, M. 1998. The resilience of pastoral herding in Sahelian Africa. Pages 250-284 in F. Berkes and C. Folke, editors. Linking Social and Ecological Systems. Management Practices and Social Mechanisms for Building Resilience. Cambridge University Press, Cambridge.

Niemeijer, D. 1996. The dynamics of African agricultural history: Is it time for a new development paradigm? Development and Change 27:87-110.

NSS. 1994. Land resources inventory and land suitability assessment of Mbulu District, Arusha Region, Tanzania. Ministry of Agriculture, Tanzania, National Soil Service, Tanga.

Nyström, M., and C. Folke. 2001. Spatial resilience of coral reefs. Ecosystems 4:406-417. Olson, D. M., and E. Dinerstein. 2002. The Global 200: Priority ecoregions for global

conservation. Annals of the Missouri Botanical Garden 89:199-224. Olsson, P., and C. Folke. 2001. Local ecological knowledge and institutional dynamics for

ecosystem management: A study of Lake Racken Watershed, Sweden. Ecosystems 4:85-104.

Olsson, P., C. Folke, and F. Berkes. 2004a. Adaptive comanagement for building resilience in social-ecological systems. Environmental Management 34:75-90.

Olsson, P., C. Folke, and T. Hahn. 2004b. Social-ecological transformation for ecosystem management: the development of adaptive co-management of a wetland landscape in southern Sweden. Ecology and Society 9.

Ostrom, E. 1990. Governing the Commons. The evolution of institutions for collective action. Cambridge University Press, Cambridge.

Palmer, M., E. Bernhardt, E. Chornesky, S. Collins, A. Dobson, C. Duke, B. Gold, R. Jacobson, S. Kingsland, R. Kranz, M. Mappin, M. L. Martinez, F. Micheli, J. Morse, M. Pace, M. Pascual, S. Palumbi, O. J. Reichman, A. Simons, A. Townsend, and M. Turner. 2004. Ecology for a crowded planet. Science 304:1251-1252.

Peterson, G., C. R. Allen, and C. S. Holling. 1998. Ecological resilience, biodiversity, and scale. Ecosystems 1:6-18.

Pretty, J., and H. Ward. 2001. Social capital and the environment. World Development 29:209-227.

Pretty, J. 2003. Social capital and the collective management of resources. Science 302:1912-1914.

Ramakrishnan, P. S., editor. 1998. Conserving the sacred for biodiversity management. Oxford adn IBH publishing, New Dehli.

Reijntjes, C., B. Haverkort, and A. Waters-Bayer. 1992. Farming for the future: an introduction to low-external-input and sustainable agriculture. ILEIA, Leusden.

39

Richard, A. F., R. E. Dewar, M. Schwartz, and J. Ratsirarson. 2002. Life in the slow lane? Demography and life histories of male and female sifaka (Propithecus verreauxi verreauxi). Journal of Zoology 256:421-436.

Rockström, J., and L.-O. Jonsson. 1999. Conservation tillage system for dryland farming: on farm research and extension experiences. E. Afr. agric. For. J. 65:101-114.

Ruud, J. 1960. Taboo: A study of Malagasy customs and beliefs. Oslo University Press, Oslo.

SAP. 2002. Données structurelles concernant la sécurité alimnetaire. Systèm d'Alerte Précoce, Union Européene, Ambovombe.

Sayer, J. A., and B. Campbell. 2002. Research to integrate productivity enhancement, environmental protection, and human development. Conservation Ecology 5.

Scheffer, M., S. Carpenter, J. A. Foley, C. Folke, and B. Walker. 2001. Catastrophic shifts in ecosystems. Nature 413:591-596.

Schoffeleers, J. M., editor. 1978. Guardians of the land : essays on Central African territorial cults. Mambo Press, Gwelo.

Scoones, I. 1999. New ecology and the social sciences: What prospects for a fruitful engagement? Annual Review of Anthropology 28:479-507.

Scoones, I. 2001. Dynamics and diversity: soil fertility and farming livelihoods in Africa. Earthscan Publications Ltd., London.

Scott, J. C. 1998. Seeing like a state: How certain schemes to improve the human condition have failed. Yale University Press, New Haven.

Sheperd, G. 1992. Managing Africs's tropical dry forests. A review of indigenous methods. Overseas Development Institute, London.

Snyder, K. A. 1993. Like water and honey. Moral ideology and the construction of community among the Iraqw of Northern Tanzania. PhD-dissertation. Yale University.

Snyder, K. A. 1996. Agrarian change and land-use strategies among Iraqw farmers in northern Tanzania. Human Ecology 24:315-340.

Songorwa, A. N. 1999. Community-based wildlife management (CWM) in Tanzania: Are the communities interested? World Development 27:2061-2079.

Spear, T. 1997. Mountain farmers : moral economies of land and development in Arusha and Meru. James Currey Publishers, Oxford.

Steffen, W., A. Sanderson, P. Tyson, J. Jäger, P. Matson, I. B. Moore, F. Oldfield, K. Richardson, H.-J. Schnellhuber, I. B. L. Turner, and R. Wasson. 2004. Global change and the earth system: A planet under pressure. Springer Verlag, Berlin.

Sussman, R. W., and A. Rakotozafy. 1994. Plant Diversity and Structural-Analysis of a Tropical Dry Forest in Southwestern Madagascar. Biotropica 26:241-254.

Sussman, R. W., G. M. Green, I. Porton, O. L. Andrianasolondraibe, and J. Ratsirarson. 2003. A Survey of the Habitat of Lemur catta in Southwestern and Southern Madagascar. Primate conservation 19:32-57.

Swift, M. 1997. Agricultural intensification, soil biodiversity and agroecosystem function in the tropics. Applied soil ecology 6:1-2.

Tagseth, M. 2000. Knowledge and development in mifongo irrigation systems : three case studies from Mt. Kilimanjaro, Tanzania. Norwegian univ. of science and technology, Trondheim.

Thornton, R. 1980. Space, time and culture among the Iraqw of Tanzania. Academic Press, New York.

Tiffen, M., M. Mortimore, and F. Gichuki. 1994. More people, less erosion : environmental recovery in Kenya. Wiley, Chichester.

40

van der Leeuw, S. E., and t. A. r. team. 2000. Land degradation as a socio-natural process. in R. J. McIntosh, J. A. Tainter, and S. K. McIntosh, editors. The way the wind blows: climate, history, and human action. Columbia University Press, New York.

Virtanen, P. 2002. The role of customary institutions in the conservation of biodiversity: Sacred forests in Mozambique. Environmental Values 11:227-241.

Walker, B. H. 1993. Rangeland Ecology - Understanding and Managing Change. Ambio 22:80-87.

Walsh, A. 2002. Responsibility, taboos and 'the freedom to do otherwise' in Ankarana, northern Madagascar. Journal of the Royal Anthropological Institute 8:451-468.

Wilbanks, T. J., and R. W. Kates. 1999. Global change in local places: How scale matters. Climatic Change 43:601-628.

Wilson, A. 1993. Sacred forests and the elders. in E. Kemf, editor. Indigenous peoples and protected areas : the law of Mother Earth. Earthscan, London.

Wolfe, M. S. 2000. Crop strength through diversity. Nature 406:681-682. Zhu, Y. Y., H. R. Chen, J. H. Fan, Y. Y. Wang, Y. Li, J. B. Chen, J. X. Fan, S. S. Yang, L.

P. Hu, H. Leung, T. W. Mew, P. S. Teng, Z. H. Wang, and C. C. Mundt. 2000. Genetic diversity and disease control in rice. Nature 406:718-722.

41