**5. Tools for the evaluation of impact assessment**

farmers' association fostering their market position. A stronger policy impact could be achieved by promoting bio-diesel crops that have alternative markets and fit more easily into the current farming system, reducing trade-offs with current crop activities and allowing synergies between fuel and feed production. Better enforcement of resource providing contracts is critical to avoid default and to alleviate labour and land constraints, thereby

In this context, another example endemic to the neotropics - the oil-producing macaw palm (*Acrocomia aculeata*) - has to be mentioned. Macaw palm recently gained economic importance in Paraguay and Brazil. In contrast to the African oil palm (*Elaeis guineensis*), it is adapted to a much wider range of environmental conditions which allows its production outside of the humid tropical zone, reducing negative impact on tropical rain forests. Another advantage of macaw palm is that it does not contain toxic compounds. The palm is a non-domesticated species with a high yield potential of an estimated 2.5 to 10.9 tons oil per hectare and year [43-45] and a life time of 70 years [46]. It grows well under various soil and weather conditions, naturally occurring in tropical and subtropical environments from southern Mexico to northern Paraguay and Argentina [46, 47]. It is often found on degraded grasslands as single trees, providing some extra feed to cattle which eat both, the fruits and leaves. The palms sustain longer periods without rain, and dry periods may last up to several months. Macaw palm fruits have a wide range of market opportunities with local and international perspec‐ tives as they are able to provide food, feed, fibre, and fuel (Fig. 2) [45]. The production and use of macaw palm can, therefore, provide a good example for a bio-economy crop that can fulfil food and fuel demands at the same time. Macaw palms growing in the Brazilian cerrados show a huge variability in biomass production and oil yield within and across various sites which highlights the importance of protecting biodiversity hotspots as source of future crops and in

improving farmers' ability to engage in bio-diesel crop production [42].

view of their domestication potential [48].

132 Agroecology

**Figure 2.** Processing, dry matter yield fractions and uses of macaw palm products [45]

The negative experiences of the large scale and partially forced introduction of the new crop jatropha [43] have shown that the introduction of a novel crop shall be guided by an *ex-ante* assessment of its ecological, economic and social impacts, including questions of local likelihood of acceptance.

Newcomers such as jatropha often lack proper long-term research on feasibility, trade-offs and environmental consequences, contributing to a better acceptance as well as public and private sector commitments for understanding needs of rural communities [39]. Macaw palms adapted to a wide range of environments are naturally occurring from Central America down to the north of Argentina and Paraguay [45], hence having potential for being cropped in many areas of South America or even outside of this continent, e.g. in Africa or Asia. Out-scaling of promising novel plants, however, also bears risks and requires an approach looking at all aspects of production from selection of genetic material and propagation, testing of cropping systems and crop management options to harvesting, transport, storage, and processing as well as considering development of new products. This needs an analysis of the entire value chain.

An *ex-ante* look at ecological and socio-economic aspects of macaw palm production or other novel crops allows identification of potential benefits, constraints and risks. Modelling is one approach in the portfolio of tools and techniques available to unravel dynamics of land use and their impact on the ecosystem associated with introduction of novel or alien species. Land use systems research addresses issues such as agricultural policy making, land use planning and integrated water management and involves for this purpose multiple stakeholders with various potential roles. Models are appreciated for both, their characteristic system research features and their integrative capacity [49]. Land use change models are tools for understand‐ ing and explaining causes and consequences of land use dynamics, The term land cover refers to the attributes of a part of the earth's land surface and immediate subsurface, including biota, soil, topography surface, groundwater and human structures. In that sense, modelling of land use changes provides insights into the extent and location of land use changes and its effects [50].

Especially important are arrangements regarding participation of stakeholders, and account‐ ability in governance. Improving the ability of research programs to produce useful knowledge for sustainable development will require both greater and differentiated support for multiple forms of boundary work. Key issues are the use of knowledge for enlightenment, decision support, and negotiation support associated with boundaries between scientists and farmers, scientists and local policy-makers, and multiple knowledge sources and multiple users. Important determinants for their success are participation, accountability, and boundary objects to foster credibility and long-term success. Boundary objects are benchmark sites which allow studying human use of and impact on forest margins to gain knowledge for viable ways for a sustainable use of these sites. For decision support, joint creation of tangible products by scientists and farmers linking research with action, collaborative field trials, on-farm nurseries, and the production of training materials on effective land use practices are important. In terms of policy-makers, essential means are synoptic country reports, particularly when prepared as "policy briefs" on key issues and models focusing at regional scales. Moreover, it is important to note that context matters and challenges of boundary work need differentiation leading to strategies that follows context; however this is not possible without participation of all stakeholders [51].

Appropriate tools for that are:


PaLA was developed by the World Agroforestry Centre for agro-ecological analysis [52]. It captures local knowledge at relevant temporal and spatial scales, and provides insights on farmers' perception on the relationship between land use and landscape functioning: farmer's management options and the actual choices made, flows of water, sediment, nutrients and organisms, and internal filter functions that determine landscape functioning based on land use practices and interactions between landscape units. PaLA consists of the following eight steps [52]:

**Step 1.** Identification of ecological and administrative domains with clear boundaries.

**Step 2.** Sampling of representative stakeholders to be interviewed, using questionnaire and/or ranking methods. Criteria of representativeness are selected on the basis of specific project purposes.

**Step 3.** Formulation of the survey interdisciplinary group, planning and designing checklist and matching PRA tools.

**Step 4.** Making of a village sketch/model in order to identify the land use patterns and focus points in the landscape by using semi-structured interviews with male and female groups. The village sketch/model provides local names of area, distribution of land use, and main land‐ scape features such as rivers, streams, mountains, roads.

**Step 5.** Transect walk are necessary to obtain an understanding of the soil-plant-water interactions along the landscape. Transects need to represent most of the land use types of the study area. The methods used are simultaneous transect walks and semi-structured inter‐ views; delivered outputs are representative transects and sketches of the areas.

**Step 6.** A timeline for each land use type along transects or/and the fields located in the representative areas of the study catchment or village, is made to study land use changes over time, based on semi-structured interviews and timeline drawing.

**Step 7.** Feedback meeting in order to report findings to the farmers/stakeholders involved to get their feedback. The methods used are posters using visualising tools and group meetings.

**Step 8.** Data analysis: Qualitative data of each PRA tool, i.e. sketch transect, timeline, and secondary data is analyzed separately by the team. Thereafter, results are evaluated to identify landscape patterns and issues.

RaCSA also developed by the World Agroforestry Centre is a negotiation support tool that aims at providing reliable data on above and below ground carbon stocks in a defined landscape, its historical changes, the impact of on-going land-use change on projected emissions, and a framework for data generation on land-use options and their changes over time [52]. This approach assesses local ecological knowledge, explores its economic potential and uses carbon stocks as an indicator for the health or fertility status of soils. Furthermore, drivers of land use change and impact on environmental services such as biodiversity can be assessed. Simultaneously it provides knowledge on alternative land use options and mitiga‐ tion strategies by means of mid- to long-term scenarios at landscape level using the Forest, Agroforest, Low value Landscape Or Wasteland (FALLOW) model. Finally, RaCSA was developed as negotiation support tool providing a basis for stakeholder discussions.

RaCSA consists of the following six steps [52]:

of policy-makers, essential means are synoptic country reports, particularly when prepared as "policy briefs" on key issues and models focusing at regional scales. Moreover, it is important to note that context matters and challenges of boundary work need differentiation leading to strategies that follows context; however this is not possible without participation of all

PaLA was developed by the World Agroforestry Centre for agro-ecological analysis [52]. It captures local knowledge at relevant temporal and spatial scales, and provides insights on farmers' perception on the relationship between land use and landscape functioning: farmer's management options and the actual choices made, flows of water, sediment, nutrients and organisms, and internal filter functions that determine landscape functioning based on land use practices and interactions between landscape units. PaLA consists of the following eight

**Step 1.** Identification of ecological and administrative domains with clear boundaries.

**Step 2.** Sampling of representative stakeholders to be interviewed, using questionnaire and/or ranking methods. Criteria of representativeness are selected on the basis of specific

**Step 3.** Formulation of the survey interdisciplinary group, planning and designing checklist

**Step 4.** Making of a village sketch/model in order to identify the land use patterns and focus points in the landscape by using semi-structured interviews with male and female groups. The village sketch/model provides local names of area, distribution of land use, and main land‐

**Step 5.** Transect walk are necessary to obtain an understanding of the soil-plant-water interactions along the landscape. Transects need to represent most of the land use types of the study area. The methods used are simultaneous transect walks and semi-structured inter‐

**Step 6.** A timeline for each land use type along transects or/and the fields located in the representative areas of the study catchment or village, is made to study land use changes over

**Step 7.** Feedback meeting in order to report findings to the farmers/stakeholders involved to get their feedback. The methods used are posters using visualising tools and group meetings. **Step 8.** Data analysis: Qualitative data of each PRA tool, i.e. sketch transect, timeline, and secondary data is analyzed separately by the team. Thereafter, results are evaluated to identify

views; delivered outputs are representative transects and sketches of the areas.

time, based on semi-structured interviews and timeline drawing.

stakeholders [51].

134 Agroecology

steps [52]:

project purposes.

and matching PRA tools.

landscape patterns and issues.

Appropriate tools for that are:

**•** Participatory Landscape Analysis (PaLA), **•** Rapid Carbon Stock Appraisal (RaCSA), and

**•** Integrated Renewable Energy Potential Assessment (IREPA).

scape features such as rivers, streams, mountains, roads.

**Step 1.** Initial appraisal of landscape (see PaLA), focused on dynamics of tree cover.

**Step 2.** Explore Local Ecological Knowledge (LEK) and economics of local tree/forest man‐ agement combined with a rapid household socio-economic survey.

**Step 3.** Plot-level C data of representative land cover units using an updated version of the ASB Cstock protocol provides time-averaged carbon stock data for above-ground vegetation and soils.

**Step 4.** Remote sensing and ground-truthing are used to provide spatial analysis of land cover change, based on a sufficiently sensitive 'legend'.

**Step 5.** The Public/Policy Ecological Knowledge (PEK) kit is used to obtain information on tree/forest management and existing spatial planning rules.

**Step 6.** Scenario studies of changes in C stocks and welfare through modelling land use and carbon stock dynamics in the landscape by using FALLOW.

FALLOW was developed by the World Agroforestry Centre for trade-off analysis [53, 54]. It allows a spatially explicit and dynamic modelling of land-use cover change (LUCC) in datapoor regions and merges bio-physical and socio-economic information to evaluate impacts of LUCC on food security, watershed functions, biodiversity, and carbon stocks. FALLOW was successfully applied in South-East Asia to assess local land use change dynamics without the need for long-term and data-intensive studies without the need for long-term and dataintensive studies based on farmers' knowledge [55], to explore livestock fodder options and their consequences for carbon stocks [56] and stakeholders' perceptions [57].

The introduction of new technologies, such as renewable energy technologies (RET) is comparable to the introduction of new plant species into local agricultural systems: new plant species as well as new technologies imply a change in the daily routine of local livelihoods. Traditionally, the planning of rural energy development projects took place in central gov‐ ernment offices far away from rural communities [58]. The applied decision support tools aimed at the identification of the most efficient technology with the lowest costs [59]. The technologies selected in such 'top-down' approaches, were afterwards (involuntarily) imposed into rural communities [58]. Amigun *et al*. explored the community perspectives on the introduction of large-scale biodiesel production from canola (*Brassica napus*) and soybean (*Glycine max*) in South Africa [60]. The local population was overwhelmingly against the proposed biodiesel production. Their reasons for the rejection included a variety of especially social and environmental factors: land regarded as identity; competition with food security; distortion of the social community fabrics; doubts about the credibility of the developers and possible air and water pollution with respect to health risks of local population.

Several studies pointed out that the acceptance of RET depends on the complex interaction of social, institutional, environmental and techno-economic factors on a very local level [61, 62]. Smallholder agricultural systems are very diverse and therefore an assessment of these factors on individual basis is required that is based on public participation and pooled learning among the relevant stakeholders [60].

IREPA provides a people-centred, bottom-up approach for the assessment of the implemen‐ tation potential of renewable energy technologies into smallholder agricultural systems [63]. This participatory approach explores the renewable resource base and the livelihoods of smallholder farmers to characterize the role of energy in the daily routine (social, institutional, environmental, technical, and economic factors) to select appropriate RET. The researcher acts as facilitator to guide the assessment while the local stakeholders become researchers who contribute knowledge and expertise. For that the IREPA approach comprises the following steps [63]:


This participatory, bottom-up research structure provides a shift from the traditional top-down approaches to a holistic consideration of the local diversity. It aims to successfully induce changes into prevailing structures and behaviour patterns of smallholder agricultural systems in order to make sustainable use of the local natural resource base.

An important impact to the ecological and economic performance of land use systems is the productivity of these systems. To address this, the selected land-use systems and their performance can be modelled by using the **Wa**ter, **Nu**trient, **L**ight **C**apture in **A**groforestry **S**ystems model. This model deals with a wide range of agroforestry systems and annual single cropping systems with minimum parameter adjustments [67]. Hence, it is possible to explore the performance of various land-use systems under a wide range of management options and changing environmental conditions. In Brazil this model has already been used to assess the performance of sugar cane in agroforestry systems [68]. The model can be applied and adapted to various climate, soil, and cropping conditions [69, 70].

Agriculture is a major water consumer [71]. Furthermore, the semi-arid and arid areas of Brazil will suffer from a decrease of water resources due to climate change [72]. Brazil is a water rich country, for which the Amazon region is an example. But inappropriate land uses in Atlantic forest and cerrado ecoregions have led to a degradation of the soils which makes these resources scarcer in terms of biodiversity [73]. Different land use systems, small family farmers and industrial agriculture properties require different water quantities for their production processes. Therefore, water use efficiency will become an increasingly important sustainability indicator for land use systems. Water use in land use systems can be measured by evapo‐ transpiration of crops. A water balance of the systems is an approach to quantify the water cycle. The water footprint is an indicator of water use, e.g. during agricultural processes, differentiating water use into three water categories: green, blue, and grey [74]. Green represents the volume of rainwater consumed during the production process and blue the natural run-off through groundwater and rivers minus environmental flow requirements, while grey is an indicator of freshwater pollution that can be associated with the production of a product over its full supply chain. The global water footprint in the period 1996-2005 was 9087 Gm³ per year (74% green, 11% blue, and 15% grey) in which agricultural production contributed 92% [75].

We, therefore, propose linking and integrating the above mentioned methods once promising novel species or crops are identified to (i) develop and test land use options and model their agronomic performance; (ii) analyse, quantify and compare the ecological and socio-economic performance of land-use systems; and (iii) test options for developing new bioproducts. This procedure will help identifying optimised value chains, addresses trade-offs and consequences for environmental services, and looks at development of proper production systems for smallholder farmers. Using participatory approaches will foster farmers' participation and help identify options meeting their needs. It will also allow an exchange of ideas and infor‐ mation among all stakeholders.
