Sustainable Landscape Management Project (P154698/P157909)

Project Appraisal Document

Supplemental Technical Notes

Contents Sustainable Landscape Management Project (P154698/P157909)................................................. 1

A World Bank Group operations from which lessons learned for this Project have been

drawn........................................................................................................................................... 2

B Detailed analysis of Bealanana Landscape .......................................................................... 6

1. Description Bealanana landscape ..................................................................................... 6

2. SWOT Synthesis ............................................................................................................ 10

3. Main Issues and Intervention strategy of the Project ..................................................... 12

C Dynamic Information Frameworks (DIF) for Integrated, Local to Landscape Scales

Decision Support to Resource Managers and Policy Makers ................................................... 16

1. Introduction .................................................................................................................... 16

2. Definition of a landscape: .............................................................................................. 16

3. What is a Dynamic Information Framework (DIF) and why is it needed? .................... 17

4. Methodology & Components of a DIF for Agriculture, Environment, and Natural

Resources .............................................................................................................................. 18

D Dynamic Information Framework ..................................................................................... 20

1. DIFs for Decision Support from local to landscape scales ............................................ 21

2. Steps for implementing the development of a PADAP-DIF .......................................... 23

E Payments for Environmental Services ............................................................................... 25

1. Rationale for the use of PES in PADAP ........................................................................ 25

2. PES experiences in Madagascar ..................................................................................... 27

3. Challenges to implementing PES in Madagascar .......................................................... 30

A WORLD BANK GROUP OPERATIONS FROM WHICH LESSONS LEARNED FOR

THIS PROJECT HAVE BEEN DRAWN

RWANDA: LAND HUSBANDRY, WATER HARVESTING AND HILLSIDE IRRIGATION PROJECT

UTTAR PRADESH SODIC LANDS RECLAMATION PROJECT

TURKEY - EASTERN ANATOLIA WATERSHED REHABILITATION PROJECT

At the global level, the WBG is supporting clients to enhance and sustain ecosystem and

agroecosystem productivity. Based on lessons learned from WBG operations in other regions,

common priorities and considerations for climate resilient and productive landscapes include:

a. Over time, projected agroecosystem shifts will offer opportunities to re-zone food &

forest production priority areas both nationally and regionally.

b. New crop varieties that are both high temperature and drought tolerant and animal

breeds that are heat and disease tolerant must be urgently developed and deployed

globally.

c. High efficiency irrigation could buy one or two decades of time against the

devastating impacts of frequent and prolonged droughts and rising temperatures.

d. Ground water aquifers need to be urgently mapped and assessed for subsurface water

storage potential and surface flood control via infiltration capture zones to collect and

hold water flows (floods) from extreme events from local to national and sometimes

regional, transboundary scales.

e. Climate smart landscapes and agroecosystem design to reduce surface water flows

and promote greater infiltration will significantly add resilience against high intensity

rainfall events (especially important where the models project increased hurricane &

rainfall intensity).

f. Previous and evolving El Niño & La Niña and other extreme events (e.g. cyclones)

should be analyzed for reference baselines to assess adaptation needs to make human

settlements, agriculture, and infrastructure more resilient to climate shocks.

g. Insurance options can help buffer against short to medium term economic losses but

will be increasingly dependent on data systems.

h. Terrain mapping and distributed hydrological modeling can identify hotspots for

flooding and actions including relocation of highly vulnerable populations or

appropriate protective infrastructure put in place to protect critical assets.

i. Functional decision support systems (DSS) require high quality, long term and

interlinked climate, terrain, land cover & land use, infrastructure, population &

settlement data sets.

j. Re-Zoning and/or reinforcement to existing infrastructure: Based on outputs of

above DSS’s, cost-benefit analyses and prioritization of the significant reinforcement

to existing infrastructure (roads, bridges, canals, dam spillways, sea walls, high

capacity pumps).

B DETAILED ANALYSIS OF BEALANANA LANDSCAPE

2. As part of the preparation of a new interministerial program (agriculture, environment,

water) for natural resource management, the Ministry at the Presidency in charge of Agriculture

and Livestock (MPAE) of Madagascar carried out a preparatory study financed by the French

Development Agency (AFD).

3. This preparatory study was entrusted to consultants BRLI / BRLMadagascar and concerns

the Project for Sustainable Agriculture Landscape Approach (PADAP) aimed at increasing

agricultural productivity in close connection with the sustainable management of natural resources

in 5 selected landscapes ( Iazafo (Region Analanjirofo) Soanierana Ivongo (Region Analanjirofo)

Andapa (Region Sava) Bealanana (Sofia District) Marovoay (Boeny Region). This project has

been formulated in the following of the Projet Bassins Versants et Périmètres Irrigués (BVPI) and

the third phase of the Environmental Program (PE3).

4. To achieve the objectives set for it, the PADAP project will receive funding from IDA,

GEF and AFD which could amount to US $ 65 million for IDA, US $ 13.6 million for the GEF

(grant) and 25 million euros for AFD (loan).

5. The team of BRLI / BRLMadagascar experts focused in one of the 5 selected landscapes,

the Bealanana in the Sofia region. The expected results were:

a. The production of a thorough diagnosis of the different practices and land use in the

landscape;

b. The identification and proposal of optimized alternative land uses and practices

developed in consultation with local and central stakeholders;

c. The development of a strategy and an action plan to implement.

6. The methodology was developed and implemented based on an landscape approach. This

approach represents a versatile tool, operational and interdisciplinary serving environmental

issues. By its work in situ, this approach constitutes a lever to find analytical and participatory

solutions for a development adapted to socio-economic and environmental issues of the territories

while integrating a truly ecosystemic dimension.

7. By providing inputs from the field on one of the landscape, this study will contribute to the

formulation of the whole PADAP program.

1. Description Bealanana landscape

8. The site Bealanana is located in the Sofia region Northwest of Madagascar, specifically

within the Bealanana district. This district consists of 20 municipalities among which 13 are part

of the study area. The study area corresponds to the watershed of Maevarano and Bealanana rivers

and mainly between 1000 and 2000m altitude. The annual average temperatures range between 20

° C and 25 ° C. The average annual rainfall in the city Bealanana is 1333mm.

9. The Maevarano river is the main focus of this study and the outer limits of watershed of

this river is the boundary of the landscape. In terms of habitat, field observations allow to classify

the environments into 4 main categories:

a. forestry formations which are as follows:

i. large forests North of the area consist of protected areas Bemanevika

managed by The Peregrine Fund, Mahimborondro and COMATSA,

ii. Natural forests isolated massifs, degraded

iii. The gallery forests and riparian present in the valleys

iv. Planting forests dominated Eucalyptus

b. grassland dominated with Aristida, savannas dominated with Philippia on the high

river basins and grassy savannas with pioneer tree species;

c. wetlands consist of marshes and swamps. Marshes are deep water reserves that are

located upstream watersheds. They are a real water tower in the area. These media also

play a crucial ecological and agro-ecological role for the region where birds are housed

like Busard de Madagascar Circus macrosceles. They are a source of income for the

people (fishing, basketry).

d. plains, some of them with rice fields. These plains are predominantly cultivated by the

people and provide the staple food (rice).

FIGURE 1 : TOPOSEQUENCE OF THE LANDSCAPE OF THE STUDIED AREA

PHOTO 1 : REPRESENTATION OF THE LANSCAPE OF THE STUDIED AREA

10. Despite consistency issues across the study area, the diagnosis made from the available

bibliographic data and their intersection with visits leads us to distinguish three sub-landscapes in

the area: the landscape of Ambatoriha, landscape Bealanana, and the Ambatosia one. These three

sub-landscapes specificities are described below :

11. • Ambatoriha landscape corresponds to the upstream watershed Maevarano,

encompassing the vast plain located East of the study area. This landscape is characterized by

agricultural activities almost exclusively rice and large forest areas covering the North and East

part of the watershed. Rice plain is marked by a kind of water control gradient from upstream to

downstream with:

a. a rather good control of the water upstream thanks to belt channel systems and allow

irrigation of the valleys.

b. A water control which gradually degrades closer to the minor riverbed, not allowing

the inter-season crops. Only broadcast sowing is possible for rice which is inundated

in the major riverbed.

c. Some forest areas (especially near Anjorzoromadosy) being degraded because of wild

exploitation within the forest.

12. • Bealanana landscape match watersheds located in the Northwest of our study area. It

includes plain located near the city of Bealanana and goes up to the confluence of the river with

the Bealanana Maevarano. This area is especially characterized by rice production and inter-season

crops: mainly rice, or beans, and more marginally some vegetable crops. These are usually

produced at plain limit and in the tanety in Agroforestry (eg edge of town Bealanana) or terrace

irrigated areas with good water control as is the case in the periphery of Antsamaka. This area is

also characterized by highly fragmented and degraded forests in the North, and wetlands and

lowland marshes sometimes competing with rice lands.

13. • Ambatosia landscape is the most downstream part of our study area, located around the

plain of Ambatosia. This landscape is characterized by significant wetland peripheries where rice

is grown with very low water control, broadcast, at an early stage in order to limit the impact of

floods on crops (rice already high). Besides rice, garlic and onions are also important production

and also other vegetable crops but on more marginal areas: bean, tomato, carrot, cucumber, etc.

These vegetable crops are an important source of revenue for local people and supply Bealanana

market, but also regional and national markets in terms of garlic and onion via Antsohihy,

Mahajanga, Diego Suarez and Antananarivo.

14. On the overall, each landscape are tightly interconnected to the 2 other, in terms of natural

resources, hydraulic flows with major impacts (erosion, silting,…), and human activities.

FIGURE 2 : EXAMPLES OF INTERACTION BETWEEN THE 3 SUB-LANDSCAPES

2. SWOT Synthesis

15. The main diagnosis results are summarised below through a SWOT analysis broken down

according to the 3 + 1 pillars of sustainable development through the matrices below.

Economic Dimension

Strength

• soil and climatic conditions

• Pre-existing of social and associative network favourable for technical leadership and changing practices

Weakness

• Accessibility conditions

• Variable water control

• No large local industry (rice mill)

• Lack of collective organization to market agricultural products (cooperative)

• land status still not formalized and regularized

Opportunity

• Develop rice in inter season (better valued on the domestic market)

• Develop interseason diversification of crops (market gardening)

• Develop agroforestry in the shallows and upstream of the confluence

• Establish private industries (mills)

Threat

• Lower value of agricultural products if their productions are increasing but that the access conditions remain difficult or worsen

Social Dimension

Strength

• Pre-existing of social and associative network

• collective enthusiasm for the technical development since it is water control

• attractive area for new generations and migrants

Weakness

• Lack of collective action in the context of economic activities (marketing cooperatives) or environmental (protection of collective projects)

• Land access difficulty in some areas

• Lack of social infrastructure

Opportunity

• Mobilization of NGOs, associations and local leaders to introduce and facilitate the adoption of technical and organizational innovations

• Increased productive areas

• Valuation of multifunctionality of new infrastructure

Threat

• preponderance of private interest over the public interest even when the sum of individual benefits is less than the collective benefits (logic of collective action)

• Increasing inequalities within the local population

• Deterioration of infrastructure existing and future

• Increased health pressures on human health and the environment

Environmental Dimension

Strength

• various physical and climatic conditions

• Existence of endemic species

• Diversity of habitats (forests, savannahs, plains, wetlands, etc.)

• Existence of protected areas and national park

• Many ecosystem services provided by natural environments (especially in forest and wetland: providing clean water, wood, fruit, climate regulation, carbon storage, etc.)

Weakness

• Lack of control over the use of environment and natural resources (forestry, forest fires, etc.)

• natural conditions and human activities increasing the risk of erosion and creating serious problems downstream (silting, floods, landslides, etc.)

Opportunity

• Extension of protected areas and national park and development of tourist activity in connection with the landscape, environment and agricultural activities in the area

• Development of agricultural products to protect and enhance the forest areas through agroforestry (vanilla, coffee ...)

• Establishment of Payment for Ecosystem Services (PES) or environmental offsets

• REDD + Extension zoning neighbour

Threat

• Continued degradation of forests by over-exploitation of forest resources

• Worsening of the consequences of erosion (soil fertility loss, threats to existing or future infrastructures and facilities)

• Reduced or loss of ecosystem services previously providing free benefits to people

• Potential impact of mining on the area

Institutional and Governance Dimension

Strength

• Existence of local dynamics aimed at taking responsibility

Weakness

• Lack of resources and needs for technical frameworks to establish routines

Opportunity

• Interministerial Initiative PADAP Top-Down and Bottom-Up local expectations

Threat

• Misuse or inefficiency of the actions and measures implemented for the community for the benefit of private interests if the checks and / or supervision are inadequate

3. Main Issues and Intervention strategy of the Project

16. The diagnosis made on the ground reveals many issues that can be classified into three

themes:

a. the water control,

i. The absence, failure or lack of irrigation infrastructures. Traditional or

funded infrastructure exists in some areas but they are not always functional

or adapted: some of them do not meet the needs of users. This is aggravated

by the lack of management structure, maintenance and infrastructure

protection. A damaged part of the infrastructure usually leads to progressive

and total destruction.

ii. The presence of hydraulic locks and choke points of the river Maevarano

due to erosion, causing flooding upstream, and consequently reducing

available low-land for agriculture. It is the main source of flooding in the

three studied plains. Obstructions are formed by natural shrinkage or

created by the riverbed. Users have already tried several times to solve the

problem at their level but they can do little compared to the scope of work.

iii. Degradation of riverbanks and rambling river water. This problem is due to

the presence of hydraulic lock and sometimes river silting. Indeed, lack of

water drainage during the rainy season, the river rambles and form

temporary beds destroying the rice fields and farms. Farmers are sometimes

forced to organize almost every year to stem the breaches until the water

creates a new passage the following year. The problem of silting is more

important at the confluences.

b. reducing pressure on the environment

i. Control and reduction of bushfires. Bushfires are widely generalized over

the entire landscape of the study area. During the dry period, these fires

appear everywhere and even in the large forests: forest managers found

forest fires inside protected areas.

ii. Better control the use and exploitation of forest resources. Wood theft, both

plantation forests on national forests around the villages, is increasing.

According to some farmers met, even seedlings of Eucalyptus transplants

are sometimes stolen because everyone wants to have its own plantations.

Thus, these conditions increase the pressure on forests and riparian gallery

in valleys and near villages. These types of forests are not provided with

specific management arrangements as they play important roles in the

protection of irrigated areas downstream.

iii. Dike erosion problems and their consequences. The different forms of

erosion observed in the field and especially of lavaka can be linked to

environmental degradation by bush fires and forest exploitation.

iv. Preserving wetlands ecosystem with high interest. Although in connection

with the recurrent flooding in the area and sometimes competing with

agricultural activities (loss area with invasion of reeds or papyrus), some

wetlands also provide many ecosystem services such as: supplying fish and

waterfowl (duck), fodder for the cattle in the dry season and goose farms,

plant fibers used for artisanal basketry, phyto water purification, water

storage, favorable habitats for endemic biodiversity, etc. Accordingly, these

wetlands must draw special attention to preservation, including the

development of ecotourism complementary of the one from upstream

protected areas.

v. Avoid, reduce or offset the potential impacts of mining. The mineral

resources present in the area, including bauxite attract the interest of foreign

investors. Surveys campaigns have already been conducted in the

municipality of Ambatosia and trails have been created over hundreds or

thousands of hectares by a Chinese company. This mining activity will

impact water quality among the watershed, and will cause additional

erosion, threatening the hydraulic infrastructure investment sustainability.

Facing such challenges, solutions to prevent, reduce and compensate the

social and environmental impacts should be found, in line with sustainable

development actions developed by the PADAP.

c. improving living conditions

i. Ensuring food security of the population by intensifying production through

inter season cropping. There are not self-sufficient households and found

that the self-sufficient households have low diversification of production

(or even a monoculture strategy for some) to ensure a good supply in

quantity and quality. Many farmers already practicing successfully a large

number of other crops other than rice. The conditions for extension,

intensification and diversification are limited so that the landscape has great

agricultural potential, in terms of area and soil quality, with opportunities to

improve productivity.

ii. Improving access conditions to open up the area. The conditions of access

to the area is particularly difficult and sometimes even tracks become

impassable in wet season. This isolation does not allow a good development

of products by farmers who are sometimes forced to sell off their harvest to

ensure runoff before the tracks are no longer practicable or before the next

harvest.

iii. Diversify productions. Such diversification nevertheless requires technical

support in terms of training, and the prior identification of lead markets in

the long term and the ability to develop new industry or to structure new

sectors (vanilla, coffee, etc.).

17. The project implementation strategy suggested for the Bealanana landscape has been

developed according to the following overall objective: "to increase agricultural productivity in

close connection with the sustainable management of natural resources in selected landscapes".

18. Main elements of the strategy are:

a. To improve the economic life conditions by improving market access, improving the

development of production and secure incomes.

b. Better valuing of the existing production by sustainable rehabilitation of the main

access roads to production areas and the creation of group of farmers, with technical

assistance to farmers' organization to help them organize themselves and support in

marketing their productions.

c. Increase production by (i) improving water management, (ii) enlarge rice area in

plains by better water control (reducing the seasonal floods), (iii) the creation and

establishment of institutions and modes of governance (related to water management

(WUA), the land regulation (land office); the regulation of sectors (farmers

'organizations or interprofessional union)) and (iv) technical supervision of farmers'

organizations to disseminate improved technologies, agroecology (climate smart and

sustainable agriculture, agroforestry, etc.) and encourage diversification in interseason.

d. Diversifying income-generating activities by interseason cultures, crops grown in

agroforestry, non-agricultural activities like artisanal transformation units (small IAA,

weaving raffia mats, etc.) and / or service delivery (eco-tourism).

e. Contractualising with downstream actors of the value chain through support for

business partnership (local, national, international) and legal technical assistance in the

establishment of contracts and possible PPP.

f. improving environmental life conditions in order to have a rational use and a

sustainable management of natural resources (water, soil, forest, etc.), preserve

biodiversity in the most sensitive areas (wetlands and forests), reduce erosion and

downstream impacts (siltation, flooding, etc.)

g. improve institutional life conditions and governance through the establishment of

local consultation groups on the scale of Municipalities to participate in the

development of strategic municipal documents, support for natural resource users

(WUA for water, VOI for the forest, etc.), and the establishment of an observatory at

the district level in order to have economic, social and environmental indicators and

reliable data to ensure the monitoring and evaluation.

19. The project could be structured into three components: a "soft" component on capacity

building and planning, an "investment" component, and a project management component. The

total budget of the project has been estimated at 25 million euros on a total of 5 years.

Total cost (euros)

Component 1. Definition and setting up a landscape approach 1 770 000

7%

160 000

160 000

50 000

40 000

10 000

1 500 000

300 000

650 000

150 000

300 000

100 000

60 000

30 000

30 000

18 412 000

76%

7 660 000

7 350 000

130 000

180 000

2 420 000

2 210 000

60 000

150 000

1 747 000

520 000

350 000

150 000

120 000

517 000

30 000

60 000

2 765 000

1 700 000

885 000

180 000

3 820 000

30 000

3 650 000

140 000

3 906 000

16%

750 000

1 073 000

1 343 000

75 000

665 000

Total excluding various unforeseen & diverse costs 24 088 000

Unforseen & Diverse costs 5% 1 204 400

TOTAL 25 292 400

Activity 1.3 - Capacity building for better resources management, application and enforcement of the law, etc.

Activity 1.1 - Development of information and management tools such as hydrographic maps, land use plans, maps of habitats and

1.1.1 Establishment of an observatory of the local rural development

Activity 1.2 - Development of tools for planning and decision making on the use of natural resources (land, water, forests, etc.)

1.2.1 Update of PCD and SAC

1.2.2 Support for the development / revision of a Forest Management scheme at the scale of landscape

2.1.2 Study, control and supervision of works

1.3.1 Support and extension service

1.3.2 Setting up Water User Associations

1.3.3 Establishment of community associations for the management of natural resources

1.3.4 Support for monitoring and control of managers actors of natural resources

1.3.5 Capacity building for forest officers

Activity 1.4 - Pilot system of payment for environmental services

1.4.1 Design and implementation of compensation mechanisms for VOI

1.4.2 Design and implementation of private reforestation compensation mechanisms

Component 2. Adoption of integrated approach in the targeted landscapes

Activity 2.1 - Rehabilitation of roads

2.1.1 Rehabilitation of roads

2.3.6 Support for marketing and contracting with the downstream value chains

2.1.3 Implementation, training and supervision of user associations

Activity 2.2 - Development of irrigation for better water control

2.2.1 Hydraulic Infrastructures

2.2.2 Meteorological and hydrological equipment

2.2.3 Pilot project of sprinkler irrigation

Activity 2.3 - Structuring, organization and development of agricultural value chains

2.3.1 Sizing and creation of 10 input supply stores

2.3.2 Rehabilitation of research station

2.3.3 Support to Farmers Organisations

2.3.4 Provision of small equipment / Farm equipment

2.3.5 Sizing and creation of 9 warehouses

Activity 3.2 Local team of theproject

2.3.7 Support for the installation of private operators for transformation

Activity 2.4 - Protection of degraded lands upstream

2.4.1 Prevention and control of erosion

2.4.2 Promotion and development of agro-forestry chains

2.4.3 Securing land

Activité 2.5 - Protection and restauration of forests

2.5.1 Forest inventory work

2.5.2 Conservation activities, support for the restoration of degraded natural forests and reforestation

2.5.3 Promotion of technology "clean cooking" for households

Component 3. Project coordination and management

Activity 3.1 Project coordination and management

Activity 3.3 Technical Assistance

Activity 3.4 Accountability

Activité 3.5 Specific studies, Monitoring&Evaluation, audits

C DYNAMIC INFORMATION FRAMEWORKS (DIF) FOR INTEGRATED, LOCAL TO

LANDSCAPE SCALES DECISION SUPPORT TO RESOURCE MANAGERS AND

POLICY MAKERS

1. Introduction

20. Globally, as population increases, urban areas expand, agriculture intensifies, and climate

changes, both natural and human dominated landscapes are under increasing pressure. For

example, land use actions (deforestation, dam construction) at a discrete spatial location can have

hydrological flow impacts hundreds to thousands of miles away. Floods and droughts are

increasingly impacting rural and urban settlements, biodiversity, freshwater availability,

agriculture and livelihoods. Climate variability and change is forcing changes in temperature and

rainfall regimes, reduction of mountain glaciers, rising sea levels, and the frequency and intensity

of extreme events. Planners, resource managers, and policy makers are increasingly faced with

having to make decisions with limited information in the face of dynamic and interacting forces

linked to population growth, changing climate and food, energy, and water demands from local to

landscape scales. There is an urgent need for data systems and frameworks that can be used to

assemble and link multisector geospatial data in a way that can be used by models to simulate

outcomes across a range of investment and action scenarios. Policy makers can then use these

scenarios to comparatively assess investment costs and priorities that optimize tradeoffs and

synergies.

2. Definition of a landscape:

21. Section Error! Reference source not found. defined a landscape for the purpose of the

proposed Project. The definition for the Project is meant to provide clear and yet operationally

meaningful boundaries for project activities. More in general:

“a ‘landscape’ is a socio-ecological arrangement that consists of a mosaic of

natural and/or human-modified ecosystems, with a characteristic configuration of

topography, vegetation, land use, and settlements that is influenced by the

ecological, historical, economic and cultural processes and activities of the area.

The mix of land cover and use types (landscape composition) usually includes

agricultural lands, native vegetation, and human dwellings, villages and/or urban

areas. The spatial arrangement of different land uses and cover types (landscape

structure) and the norms and modalities of its governance contribute to the

character of a landscape.

Depending on the management objectives of the stakeholders, land-scape

boundaries may be discrete or fuzzy, and may correspond to watershed

boundaries, distinct land features, and/or jurisdictional boundaries, or cross-cut

such demarcations. Because of this broad range of factors a landscape may

encompass areas from hundreds to tens of thousands of square kilometers.”

(EcoAgriculture, Policy Note 10, 2013)

22. Examples of some the questions faced by planners, resource managers and policy makers

include:

a. What effects would changing climate (temperature, rainfall) have on agriculture, water

resources and biodiversity?

b. What are the impacts of changes in agriculture and forestry practices on local to

regional water balances?

c. How would changes in land use practices affect water supply, water quality, and

biodiversity?

d. What are the linkages between biodiversity and agricultural productivity?

e. Can climate data over a growing season be used to improve crop selection (and fire

management)?

f. Can floods or drought hotspots be projected, plausible livelihood impacts simulated,

and mitigation measures be proactively designed and implemented?

3. What is a Dynamic Information Framework (DIF) and why is it needed?

23. To respond objectively and quantitatively to the questions posed above, it is essential to

have (a) actionable information, (b) synthesis of the information, and (c) "bringing to life"

(simulation) of the key information to provide integrated and local to landscape scale impact and

outcome scenarios that can be compared and assessed by decision makers.

24. The objective of a Dynamic Information Framework (DIF) is to provide a geospatial

gateway and simulation platform for:

a. Multisector data repositories and geospatial data organization – including institutional

roles & governance (Figure 1 and Figure 2);

b. Coupling the data to new generation and ‘first principles’ Earth System Science models

(Figure 3); and

c. Using the models to produce dynamic simulations of integrated, nested local to

landscape scale impact scenarios (Figure 4).

25. The scenarios can be modified to suit a variety and levels of natural resource (topography,

soils, vegetation, biodiversity) and other change drivers in both space and time so that decision

makers can evaluate best case-worst case outcomes as a basis for assigning priorities, timing, and

levels of needed investments. The DIF can also serve as an efficient Monitoring, Reporting, and

Verification (MRV) tool for governments and donors.

FIGURE 1. GEOSPATIAL, LOCAL TO LANDSCAPE DATA

GRID

FIGURE 2. DATA INFRASTRUCTURE & INSTITUTIONAL

LINKAGES

4. Methodology & Components of a DIF for Agriculture, Environment, and

Natural Resources

26. A central component in the DIF-based, local to landscape approach of natural resource

planning and management is based on how water and the landscape converge, in space and time.

Water provides spatial, time-based, and operational connectivity among the multitude of DIF

layers and everyone understands water (one has it or not, it is of adequate quality or not, it is

available in the right place at the right time or not). Most important, water is observable,

measureable, and subject to being modeled, as a function of known drivers and spatial-temporal

relationships.

27. To meet these challenging criteria, the DIF approach uses a new class of open and

publically accessible hydrology models, which also serve as overall landscape models, because of

the processes (and data layers) they represent. The requirements of the model dictate what data

modules must be assembled and the output variables for the Decision Support System (DSS).

28. The Earth System model (Figure 3), the core of the computation engine, is a geospatial

hydrology model that explicitly represents the effects of vegetation, topography, and soils on the

exchange of moisture and solar energy between land and atmosphere. The core model can then be

coupled to other crop, nutrient, soil, infrastructure, and economic models, and compared to

independent data sources, to ultimately provide the basis for management-focused applications in

the DSS. The results of model runs are complex, multi-layer, 4-dimensional (including space, time)

analyses of landscapes and their resources that require visualization (graphs and video) to make

complex technical outputs understandable to policy makers (Figure 4).

FIGURE 3. EARTH SYSTEM MODEL WITH DATA

LAYERS VISUALIZATION

FIGURE 4. COMPLEX DATA TO EASY 4-D

29. The information required to support modeling and decision support is derived from

multiple sources. Even in very remote, data-sparse regions, global coverages can provide at least

first-order estimates (e.g. Google Earth). There are three types of data that are needed:

30. Static data such as the basic structure of the river basin (topography, river networks), soil

properties (how deep are the soils, what is their texture), vegetation properties (rooting depth,

height, leaf area index).

31. Climate forcing data, which includes the daily average precipitation, minimum and

maximum temperature, and winds. These dynamic data "drive" the model and can be derived from

meteorological observation networks, climate weather models, or directly from satellite

observations. Changing the climate forcing data, allows testing of different climate change

scenarios at scales that are much more relevant for policy makers than the scale of global or

regional climate models.

32. Model calibration and validation data e.g. river/stream flow as measured by gauges. These

data are used for model calibration and to test the calibrated model, against observed data from a

different time period than used for calibration.

FIGURE 5. EXAMPLES: PROTOTYPE DIFS UNDER DEVELOPMENT IN CENTRAL ASIA & LATIN AMERICA

http://goo.gl/XobnCI http://goo.gl/cVIICW

D DYNAMIC INFORMATION FRAMEWORK

33. Madagascar has a large amount of natural resource, social, political, and economic data

that are needed for effective planning and decision making in the face of a growing population,

changing land use, and climate change drivers. A tested approach in many regions to organize,

enhance access by authorized users, and to use the data for planning and decisionmaking is via the

development of Dynamic Information Frameworks (DIF).

34. The objective of a DIF is to provide (i) an integrated geospatial repository for existing but

disaggregated and diffusely distributed data from different Ministries and government agencies,

and (ii) a gateway for dynamic understanding, planning and management of any landscape (Figure

6). The goal is to deliver actionable information, as the foundation for a high-level decision support

system by providing (quantitative) analyses of complex, systemic interdependent environmental

problems.

FIGURE 6. SCHEMATIC OF THE BIOPHYSICAL COMPONENTS OF ANY LANDSCAPE INCLUDING THE CARBON,

ENERGY, AND WATER CYCLES THAT CONTRIBUTE TO THE PRODUCTION AND ENVIRONMENTAL SERVICES

ESSENTIAL FOR SUSTAINABLE DEVELOPMENT

Source: Adapted from Fernandes (2006), “Sustainable Land Management: Challenges, Opportunities & Tradeoffs”

35. The DIF approach draws on the emergence of Earth Systems Sciences based on the rapidly

evolving capabilities for addressing global change issues through the use of satellites, new

generations of dynamic computer models, and field measurements/observations.

36. The architecture of the DIF computing framework consists of streaming information from

multiple sources (e.g. satellites, weather records and operational climate models, soil profiles,

stream gauges, species lists) and rendering them into data layers identified as the required inputs

for the geospatial hydrology and landscape models (Figure 7). The core of the information

framework is built by targeting the outputs required for decision support; for example, targets of

soil moisture, floods and drought, agricultural production, or hydropower. These geospatial targets

can be derived from computational tools, such as land surface models. These models in turn require

information from multiple sectors, which in turn can serve multiple purposes.

FIGURE 7. FRAMEWORK FOR MULTISECTOR DECISION SUPPORT

1. DIFs for Decision Support from local to landscape scales

37. The overarching “framework” component of the DIF is to place the data layers and

computational tools within the context of decision-making requirements and visualization of

results in a format useful to resource managers. The approach is to coordinate the ensemble of data

products that span a wide spatial and temporal range in a single, centralized location.

38. There are three main data modules categories based on their function in relation to the

computational model: static input data, dynamic input data, and observed data. Even in very

remote, data-sparse regions, global coverages can provide at least first-order estimates (viz. Google

Earth). These datasets include: digital elevation maps from Shuttle Radar Topography Mission

(SRTM) to delineate the river network and watershed boundaries, remotely sensed land cover

maps from MODIS, soil distributions and parameters from the Food and Agricultural

Organization, observed streamflow from the Global Runoff Data Centre, and terrestrial water

storage observations from the Gravity Recovery and Climate Experiment satellite mission to

validate against the water balance.

39. The “dynamic” component of the DIF is based on two concepts. First, it is a decision

platform that can analyze historical, current, and future patterns so that outputs are dynamic in

time. Second, the components of the DIF platform are dynamic in that the key pieces used for

assessment might change as a project and stakeholder needs evolve. The initial DIF applications

have emphasized hydrologic modeling (envisaged in Error! Reference source not found.) as the

omputational engine that integrates the climate and biophysical data layers DIF to generate the

dynamic outputs and scenarios shown in Figure 8.

FIGURE 8. FRAMEWORK FOR MULTISECTOR DECISION SUPPORT

40. The existing core computational model(s) used in DIFs are open source (code freely

available) “land-surface models,” representing the flow of water across the landscape. Ideally, for

broader scales (15kmx15 km) a semi-distributed, grid-based macroscale model that explicitly

represents the effects of vegetation, topography, and soils on the exchange of moisture and solar

energy between land and atmosphere is needed. For higher resolution applications (5kmx5km or

less), fully-distributed models that recognize the spatial heterogeneity of the watershed are

recommended if adequate local data can be accessed. Currently available DIF model components

are powerful enough and of sufficient resolution to provide simulations and scenario options for

the PADAP Landscape profiles presented in Error! Reference source not found.. Figure 8 shows

one example of landscape scale land cover and land use scenarios and associated surface flows

generated for landscapes in EAP.

41. Because the DIF models integrate rainfall, temperature, evapotranspiration variables they

can also be parameterized to incorporate projected climate change and thereby can provide

outcome scenarios that are essential for objectively designing climate smart land management

plans. For Payments for Environmental Services (PES), the DIF could be used to select several

candidate pairs of catchments at each of the sites. Monitoring paired catchments (one treated, one

not) can help assess the effect of treatments and to verify that the land use interventions and

hydrological models (Error! Reference source not found.) make reliable predictions of these

effects. Some of these candidate catchment pairs will be selected (based on planned interventions)

for close monitoring and re-calibration of the DIF modules.

42. The DIF could be coupled with additional data simulation and/or analytical modules. For

example:

a. Collect Earth, an open source analytical platform developed by FAO that geo-

synchronizes the visualization and use of imagery of varying spatial and temporal

resolutions, including DigitalGlobe, SPOT, Sentinel 2, Landsat and MODIS imagery

within Google Earth, Bing Maps and Google Earth Engine. Images from multiple years

are supplemented by seasonal and multi-year graphs of several indices (e.g., Landsat 8

32-day Normalized Difference Vegetation Index (NDVI), Normalized Difference

Water Index (NDWI), Enhanced Vegetation Index (EVI), Moderate Resolution

Imaging Spectroradiometer (MODIS) 16-DAY NDVI and Landsat 7 Monthly NDVI

Composite).

b. NDVI and Land Degradation: Use of satellite data for assessment of land degradation

is a big improvement from past land degradation assessment, which heavily relied on

expert opinion. A global study led by IFPRI created a new product derived from NDVI,

which captures intensity of greenness of land cover, inferring net primary productivity

(NPP), i.e. the net biomass produced by the soil and other natural resources – after

controlling for soil fertility amendment (e.g. application of fertilizers and manures) and

rainfall variability. The corrected NDVI is used as a quantitative indicator for land

degradation. Household surveys are used to ground truth the NDVI-derived outputs.

c. Dinamica EGO, an open source simulation platform developed by researchers at the

University of Minas Gerais, Brazil. One module, OTIMIZAGRO, is a spatially explicit

model that simulates land-use, land-use change, forestry, deforestation, regrowth, and

associated CO2 emissions, under various scenarios of agricultural land demand and

deforestation/forest restoration policies. OTIMIZAGRO models nine annual crops,

including single and double cropping (soy, sugarcane, corn, cotton, wheat, bean, rice,

cassava, tobacco) and five perennial crops (arabica coffee, Robusta coffee, oranges,

cocoa, and banana), and forest plantations. Such a model could and will need to be

parameterized for Madagascar based on local policies and biophysical conditions.

2. Steps for implementing the development of a PADAP-DIF

43. Through previous DIF development processes in WBG operations in ECA, LAC, and SAR,

feedback from local stakeholders has pointed to three main components that are necessary to make

the DIF a viable decision support tool:

1. Assembling a local, cross sector expert team that will be engaged in assembling the DIF is

critical for empowering national agencies, enhancing local capacity, and ensuring the short

to long-term national commitment to a sustainable and functional DIF. The local agency

buy-in via their experts improves both the access to and the transparency of data use as

well as encouraging reproducibility of modeled outcomes.

2. Complex modeled outcomes and scenarios need to be presented in compelling and easy to

understand formats via user interfaces that are comparable with local needs and culture.

Increasingly, web-based and ICT-based visualization tools are being used by a range of

stakeholders (including farmers in many African countries) to access information in near

real-time.

3. The lack of computing and multi-sector data infrastructure can be a barrier to utilizing the

tools within a DIF. To address this challenge, the DIF modules can be deployed using cloud

services in order to reduce the need for each Ministry requiring expensive and not easy to

maintain computing infrastructure, especially in countries where computing resources are

limited.

44. The development of an effective PADAP-DIF will require investments that address the

constraints identified above. A key starting point is the assembly of a multi-sector DIF team

(Ministry, local government agencies, and local academia) that forms the core for an initial

capacity enhancement workshop. The outcome of such a workshop is an empowered national and

multisector DIF team with specific responsibilities for data access and acquisition and institutional

participation in the DIF building process. Once the DIF data layers are assembles, the DIF becomes

a valuable tool for the multitude of capacity enhancement activities identified in the components

of the project.

45. Based on work in other WBG regions developing a functional PADAP-DIF with strong

local participation is expected to cost about US$400,000-500,000 depending on (i) the extent and

quality of local data, (ii) the quality of available local data, (iii) the need for new data acquisitions,

and (iv) the adequacy of existing data and cyber infrastructure and potential to transform local

server-based resources to more secure cloud-based platforms.

E PAYMENTS FOR ENVIRONMENTAL SERVICES

46. Payments for Environmental Services (PES) will be an important tool for the PADAP

project, since it will allow the conservation or restoration of landscape elements that are important

for hydrological services, biodiversity, and/or carbon but which are not profitable for the local

population. This annex explains the rationale for the use of PES in PADAP, and the challenges it

will face in doing so. It also identifies the ways in which the project’s activities will lead to the

design and implementation of PES pilots, as these activities are part of several sub-components

rather than being grouped together.

1. Rationale for the use of PES in PADAP

47. Depending on how they are used, landscapes can provide a wide variety of benefits. In

addition to producing agricultural products, they can produce hydrological services, conserve

biodiversity, and sequester carbon. However, the landholders who actually manage the landscape

generally only receive only a small part of these benefits while others benefit downstream water

users such as irrigation systems, domestic water supply systems, and hydroelectric power

producers, in the case of hydrological services, and the global community in the case of

biodiversity conservation and carbon sequestration. As a result, landholders don’t take these

broader benefits into account in their land use decisions, and many valuable parts of the landscape

have been lost, replaced by much less environmentally friendly land use practices, while others

are under severe threat of conversion.

48. The improvements to be undertaken at the project sites will often be ‘win-win’ in the sense

that they will benefit both the landholders themselves and environmental service users. Replacing

unsustainable upland rice (tavy) production with a sustainable agroforestry system, for example,

would increase income for landholders while reducing erosion, increasing carbon sequestration,

and providing a more biodiversity-friendly habitat. In cases in which the new practices to be

adopted are more profitable to landholders than their current practices, short-term support should

be sufficient to stimulate their adoption sustainably. This support might take many forms, such as

the financial support to the necessary investment, technical assistance (TA), and/or provision of

required inputs, depending on the nature of the obstacles to adoption of the practice.

49. However, all elements of the desired landscape are not necessarily ‘win-win’. Some of the

practices that generate environmental services such as watershed protection, biodiversity

conservation, and/or carbon sequestration are less profitable to individual landholders than

alternative, less environmentally friendly practices. Landholders are unlikely to be willing to adopt

these practices voluntarily. Short-term support might persuade them to do so, but they are then

likely to abandon them once the short-term support ends. Such practices are only likely to be

adopted sustainably if landholders are offered long-term support that compensates them for the

opportunity cost of foregoing more profitable (but less environmentally friendly) practices.

50. The various possible cases are shown in Figure 9. In each case, the benefits shown are the

net benefits to landholders of undertaking a given activity. The benefits to downstream users or

the global community of undertaking activities such as forests or agroforestry are not shown, and

neither are the costs imposed on others by environmentally harmful activities such as tavy.

FIGURE 9. TYPES OF LAND USE CHANGES IMPLEMENTED IN A LANDSCAPE

51. In Panel A, tavy is replaced by productive practices such as agroforestry or silvopastoral

practices, which are more profitable for landholders than tavy once established. Their profitability,

mean that landholders are likely to adopt them readily, once obstacles to their adoption have been

removed, and then retain them even after the project ends. This is the ‘win-win’ case. Most of the

land use practices supported under component 2 are likely to be of this kind. Because short-term

support is sufficient, efforts to induce adoption of these practices can easily be financed by the

project itself.

52. Panel B shows a case in which tavy is replaced by a conservation practices such as forest,

which generates limited or no returns to landholders, either because of its nature or because of

restrictions on its use (for example, if the land is located inside a protected area). In this case, net

returns to landholders may well be lower than those of tavy (if that were not the case, the forest

would not have been cleared for tavy). On addition to the initial investment cost, there would thus

be an opportunity cost for landholders from foregoing income from tavy. A sufficiently large short-

term subsidy might induce landholders to adopt such practices, but they will abandon them in favor

of the more profitable tavy once the subsidies end. Landholders are only likely to retain such

practices if they receive a payment sufficient to offset the opportunity cost of foregoing tavy.

Crucially, this payment must be made annually, and indefinitely. As such, it cannot be financed

by the project itself beyond the first few years (although the project can certainly finance the initial

cost of reforestation).

53. Panel C shows a case in which a forest remnant is being conserved to avoid it being cut

down for tavy. Maintaining the forest would impose an opportunity cost to landholders in the form

of foregoing income from tavy. Landholders are thus unlikely to retain forest unless they receive

a payment sufficient to offset this opportunity. Here, too, this payment must be made annually,

indefinitely, and so cannot be financed by the project itself beyond the first few years.

54. There are thus likely to be at least some elements of the landscape that would be desirable

because of their hydrological, carbon sequestration, or biodiversity conservation benefits that will

not voluntarily be adopted by landholders in the absence of long-term payments. Because such

payments must last far beyond the end of the project, however, they cannot be financed by the

project itself.1

55. One approach to avoiding this problem is to try to increase the returns that landholders

receive from forests. Developing ecotourism, for example, could generate an income stream for

landholders. Developing new value chains for forest products could have the same effect. The

project will support such efforts, wherever possible, under Error! Reference source not found..

In some cases, these efforts will be sufficient to make forests more valuable than the alternatives,

thus converting a case such as that shown in Panel B of Figure 9 to one similar to that shown in

Panel A. In some cases, however, such efforts will either be impractical (for example, areas where

access is too limited for commercialization of forest products to be viable), disallowed (for

example, forests within protected areas), or insufficient (ecotourism and/or forest products do not

enough generate additional income to compensate for the opportunity cost of foregoing tavy).

There will thus remain some practices whose presence in the landscape would be desirable that

would not be adopted (or retained) by landholders.

56. Providing long-term compensation to landholders who adopt environmentally friendly land

use practices is precisely the objective of Payments for Environmental Services (PES), which are

contingent payment to those who manage natural resources so as to benefit others. The World

Bank has supported PES programs in a large number of projects.2

2. PES experiences in Madagascar

57. Although use of PES has been concentrated in Latin America, there is growing interest in

and use of the approach in other regions, including Africa. Several PES pilots are being

implemented in Madagascar (Figure 10).

1 The project could finance long-term payments if it established a trust fund, as was done in the Costa Rica

Mainstreaming Market-based Instruments for Environmental Management Project (P093384/P098838) and the

Mexico Environmental Services Project (P087038/P089171), for example. This approach is not broadly

applicable, however, as it requires a very large capital investment to generate a sufficient payment stream,

particularly with low interest rates. 2 There are currently ten projects under implementation with PES components (in Brazil, Colombia, Mexico,

Nicaragua, Ghana, Kenya, Albania, and Bhutan) and one under preparation (in Kenya), in addition to this one.

FIGURE 10. PES PILOTS IN MADAGASCAR

58. Andapa.3 A PES program to protect Andapa town’s drinking water supplies was proposed

in 2009 by a local NGO (Association des Paysans de Montagne du Monde, APMM) with the help

of WWF. It planned to pay upstream farmers with land in the Sahamazava watershed near the

springs that supply Andapa’s water to switch from tavy to perennial crops. A proposed fee on water

users of MGA1000 (US$0.31) per month per household would have financed the program.

However, this financing was not forthcoming, because of opposition from water company

JIRAMA, which administers the town’s water supply system.4 Eventually, an alternative funding

source was developed with the assistance of a local community association, Plateforme Toham-

pontsy, formed specifically for this purpose. Plateforme Toham-pontsy collects MGA2000

(US$0.62) per month per standpipe from the users of the town’s 86 standpipes and channels them

to the PES program, which now also covers a second watershed south of Andapa where another

3 This pilot is being implemented in one of the PADAP project’s sites. 4 JIRAMA offered a variety of arguments for refusing to finance the PES project, including that: the company

could not afford it, as it makes heavy losses; landholders would probably not comply; small-scale interventions

such as local reforestation are sufficient to protect the watershed; (any benefits the company receives from the

‘cheap’ gravity-fed water system at Andapa are already used in cross-subsidies to more expensive systems;

other service users should also contribute; and JIRAMA was unable legally to add extra fees to its water bills

(Andriamahefazafy, 2016).

water intake was recently built with AfDB support. However, financing from Plateforme Toham-

pontsy (US$620/year) only covers a small fraction of the cost of efforts to change land uses in the

watersheds; Belgian NGO Graine de Vie provides the balance through in-kind contributions.

Farmers in upstream areas are provided with seedlings for a mix of tree crops (cloves, cacao,

coffee) and native trees to plant in areas currently under tavy. The program thus focuses on

activities such as those in Panel B of Figure 9. Support is provided in consultation with associations

of local farmers. It is important to note that this support is provided through ex ante support rather

than as conditional payments. Another important aspect of the situation in this case is that most of

the people who farm the target watersheds live in Andapa itself, and so benefit themselves from

its water system.

59. Tolongoina. A PES program was proposed to protect a run-of-the-river micro-hydrolectric

plant built on the river running through Tolongoina to supply local communities with support of

the EU-financed RHYviere Program. The plant is operated by a private company, SM3E. Part of

the plant’s watershed has considerable forest cover and few threats, but the northern part of its

watershed, where there is a small storage reservoir, is threatened by erosion. The PES program

aims to protect and restore this area. Electricity users at first resisted the idea of compensating

upstream landholders, who are migrants with no downstream family ties, but eventually agreed

after traditional authorities confirmed the upstream landholders’ right to land. Likewise, local

authorities also resisted the idea of paying for conservation as they considered compliance with

conservation rules as a duty. A 3-year agreement was finally reached in 2013, soon after electricity

started to flow from the plant. Electricity users pay 2.5 percent of their bill (a total of ca

US$31/month), the Commune pays US$33/month, and plant operator SM3E pays US$36/month,

for a total of about US$1200/year. Payments are not made in cash, but by financing agreed sub-

projects in the watershed (compiled by a farmer association, TAMIS, and implemented by the

watershed committee, KOMHASA). Some of these sub-projects involve changing land uses, but

others focus on alternative income sources (such as raising ducks), in the hope that this will result

in reduce pressure on forests (a strategy commonly followed by Integrated Conservation and

Development Projects, ICDPs, with very mixed results). Watershed households were convinced

to accept a payment considerably lower than their opportunity costs (estimated at about

US$4500/year, or US$151/household), based on arguments that future degradation would reduce

their income and that the payment, though small, constituted an acceptance by downstream users

of the important role played by upstream households. Whether this arrangement will result in

effective and lasting protection of the watershed remains to be seen.

60. Makira. A number of projects in Madagascar are using sales of carbon credits to help

finance conservation activities. The Wildlife Conservation Society’s (WCS) Makira REDD+

Project5, for example, uses revenue from carbon sales to finance the creation of the Makira Natural

Park and to support sub-projects that contribute to the wellbeing of communities and households

living around the Park, in part by helping them adopt sustainable land use practices and/or develop

alternative income sources. The Makira REDD project has been certified under the Climate,

Community and Biodiversity Standards (CCBS) and Verified Carbon Standards (VCS). Two sales

of pre-certified carbon credits were made before 2011, and then a first sale of 72,369 Verified

Carbon Units (VCUs) was made in December 2013. Although transfers to communities are based

on sales of carbon credits (communities receive 50 percent of the revenue from carbon sales) and

so are conditional on results, they are not conditional on the land use decisions of individual

5 The Makira Natural Park is located on the southern edge of the PADAP Project’s Andapa site.

communities or households. Moreover, transfers to local communities are made through in-kind

support of investment projects rather than as cash payments.

61. Corridor Ankeniheny-Zahamena (CAZ). Conservation International (CI) developed the

Corridor Ankeniheny-Zahamena (CAZ) Project to generate financing for conservation from

carbon sales, leading to the signature of an emissions reduction purchase agreement (ERPA)

between the Malagasy government and the BioCarbon Fund in 2008.6 As in the case of Makira,

50 percent of the revenues from carbon sales will be distributed to local communities. Separately,

CI itself made grants to local community organizations in Maroseranana and Didy, two towns near

the CAZ, to patrol the PA. The grants were meant in part to pay for the patrols themselves, and in

part to stimulate local development. The payments were not explicitly conditional, but did result

in patrolling being undertaken satisfactorily in the first two years. Patrolling fell off in the third

year, however, when the looming end of the agreement meant that further payments were not at

risk. Most local development funds were spent on activities that did not involve people actually

living in and using the forest areas, as they were not part of the traditional local communities.

62. Menabe. Although a great many actors are seeking to preserve Madagascar’s rich

biodiversity, the paucity of funding for biodiversity conservation means that most are trying to

leverage other funding, such as carbon finance (as in the case of the WCS’s Makira REDD+

Project). Nevetherless, there is one project that is explicitly tying payments to local communities

to biodiversity conservation. A PES program implemented by the Durrell Wildlife Conservation

Trust in the Menabe region in western Madagascar offers participating communities annual

payments, based on the state of the ecosystem on their land (as measured in an annual transect,

which records species of interest and threats). About US$8500 is paid annually, in-kind (with

communities deciding what they wish to purchase). Payments are for results relative to those of

others, so the distribution of payments varies from year to year.

63. Overall, the track record of PES in Madagascar to date has been limited. The few efforts

that have been made to convince service users to pay to conserve the ecosystems that benefit them

(at Andapa and Tolongoina) have only yielded nominal funding, more important for their symbolic

value than for their monetary value. On the supplier side, there have been few efforts to implement

truly conditional payments, with the Durrell Trust’s project in Menabe being the only exception.

Neither has there been any significant effort to make actual payments: most ‘PES’ projects have

followed the standard approach of most rural development projects in Madagascar of financing

specific sub-projects. Although some

3. Challenges to implementing PES in Madagascar

64. Implementing PES generally requires four parallel sets of activities: (1) understanding the

linkages between land use and the desired environmental services; (2) putting in place financing

arrangements; (3) putting in place field arrangements to contract providers, monitor compliance,

and make payments; and (4) putting in place the institutional framework. This section reviews the

challenges faced by these activities in Madagascar, in light of the experience of the few existing

PES pilots in the country, experience with PES implementation worldwide, and country

characteristics. It also indicates the activities that the PADAP project will undertake to meet these

challenges.

6 Although the project is widely known as the ‘CAZ Project’, the BioCarbon Fund refers to it as the

‘Ankeniheny-Zahamena-Mantadia Biodiversity Conservation Corridor (REDD) Project.’

65. Links between land use and service generation. As in many countries, the understanding

of how different land uses affect the various environmental services of interest is imperfect.

However, advances in the use of remote sensing imagery and hydrological modeling make this

much less of an obstacle than it once was. Moreover, the P4GES initiative has been carrying out

research on impacts of land uses change on hydrological ecosystem services in Madagascar.

The dynamic information framework developed under Error! Reference source not found.

will provide the basis for the hydrological modeling necessary for the development and

implementation of PES pilots. These hydrological models will be developed under Error!

Reference source not found., as part of the preparation of river basin plans for the PADAP

project sites. They will identify critical hydrological areas in the target watersheds, the land

uses that could contribute to protecting downstream water users, and estimate the potential

impacts of their adoption on downstream water services.

66. Working with service users. As noted, conserving or restoring landscape elements such

as forests that are important for hydrological services, biodiversity, and/or carbon but which are

not profitable for local landholders, requires long-term compensation to landholders and, therefore,

long-term financing sources. In some countries, such as Costa Rica and Mexico, the government

provides such financing (usually through dedicated income sources to avoid year-to-year

variations in funding levels). Given its fiscal constraints and other pressing demands, however, the

Government of Madagascar is unlikely to be able to provide such financing. In the case of PADAP,

the obvious source of financing required for PES would be those who benefit from these activities:

farmers in irrigated areas, whose productivity is substantially increased through upstream

conservation activities.7 It is in their interest to ensure that these improvements are adopted and

maintained. Their contribution to the compensation of those who adopt conservation practices

upstream would be an investment in their own future productivity, similar to expenditures on O&M

of the irrigation system itself.

67. Although there is a strong logic to having service users pay for conservation, securing the

cooperation of farmers in irrigated areas will not be easy. Despite the benefits they derive from

irrigation, such farmers are often accustomed to paying little or nothing for irrigation water. In

Madagascar, this problem is exacerbated by the fact that farmers in irrigated areas are poor. While

it is true that farmers in irrigated areas are often poor, however, they tend to be relatively better off

than farmers in upstream areas, and the improvements offered by the PADAP project will increase

their income even further. [As shown in Error! Reference source not found., yields/returns in

irrigated areas are expected to increase by xx percent thanks to PADAP.] More to the point, the

losses farmers in irrigated areas would sustain in terms of reduced productivity if watershed

degradation deteriorates the efficiency of irrigation would be far greater than the payments needed

to avert this threat.8

Efforts to obtain financing from producers in irrigated areas would be incorporated into efforts

under Error! Reference source not found. to ensure that they pay for the O&M of the

irrigation system itself—indeed, watershed conservation is simply another form of

maintenance. Under this sub-component, the project will help design rules for the operation of

7 Proper water management is critical for improved productivity in irrigated areas. Moreover, erosion resulting

from upstream degradation would substantially increase O&M costs in irrigated areas. 8 Solonitompoarinony (2001), for example, found that rice yields in areas that were moderately affected by

upstream erosion were only 60 percent as high as yields in areas that were not affected, while those in areas

heavily affected by erosion were only 50 percent as high.

the irrigation systems that it will help improve under which farmers using irrigation water will

be responsible for paying for all operations and maintenance (O&M) costs of the system,

including (a) the costs of providing compensation to upstream landholders who maintain

conservation practices that protect those systems, (b) the costs of operating the PES program

itself.

To ease the burden on producers of making payments, the cost of payments for conservation

and of the administrative costs of the PES program, would initially be borne by the project,

with the producer contribution being gradually phased in until producers bear 100 percent of

the costs at end of project.

To emphasize the links between the productivity improvements resulting from PADAP’s

activities, payments for O&M (including conservation) could be made proportional to

increases in yield above the baseline level.

68. In some cases, there will also be other water users (drinking water systems, hydroelectric,

etc.) that will benefit from improvements in upstream landscapes, and so could also contribute to

financing them.

Where opportunities to include other water users are encountered, the projects will seek to

incorporate them into plans for the pilot.

A study undertaken under Error! Reference source not found. will help identify

administrative, bureaucratic, and legal obstacles that might constrain the participation in PES

of such water users.9

69. Carbon payments from the REDD+ program could also contribute to payments for forest

conservation. Madagascar has recently signed an Emission Reduction Purchase Agreement

(ERPA) with the Carbon Fund for the purchase of up to US$50 million of carbon credits.

The project will work with the national REDD+ coordination office (BNC-REDD) to examine

options for channeling part of REDD revenues to landholders who conserve or restore forests

on their lands.

70. Working with landholders. To achieve the desired conservation outcomes, a PES

program must make payments to the actors that manage areas at risk of degradation or degraded

areas that need restoring in ways that induce them to adopt the desired land uses. An immediate

challenge in the PADAP sites, as in much of Madagascar, is that most of the affected areas are not

individually owned. Moreover, the threats come primarily from shifting cultivation, making it

difficult to identify who is actually managing any given piece of land at a given time. Under these

conditions, there is a danger offering payments for conservation of particular areas may,

perversely, attract more people to vulnerable areas, thus increasing rather than decreasing pressure.

A study undertaken under Error! Reference source not found. will evaluate the lessons

of other PES programs that have worked in areas where tenure is weak or uncertain, such

as the Bolsa Floresta program in the Brazilian state of Amazonas.

71. On the other hand, many of the areas affected are managed by local communities. There is

a long experience with community forest management (CFM) in Madagascar. Under the GELOSE

9 Such as the inability to add extra fees to its water bills that JIRAMA cited as a reason not to participate in

Andapa PES project.

and GCF laws, management of many areas has been devolved to local organizations known as

Communautés de Base (COBA, or VOI in Malagasy).10 In the existing PES pilots at Andapa and

Tolongoina, associations of upstream landholders were established specifically to act as

interlocutor with the downstream service users. These local groups, whether COBAs or ad hoc

associations, could act as service providers, managing the entire area under their purview, and then

distributing the payments received among its members as necessary to compensate them for their

opportunity costs. There is considerable experience with making PES payments to communities

rather than individual households (in Brazil’s Bolsa Floresta program, Costa Rica’s PSA program,

and Mexico’s PSAB program, for example, in addition to some of the Malagasy PES pilots), that

the project will be able to draw on.

The efforts undertaken under Error! Reference source not found. to establish and strengthen

COBAs will contribute to efforts to develop PES pilots in their areas. In turn, payments

received under a PES program could help further strengthen the COBAs by providing them

with additional resources and incentives to work together.

The project will work with existing and new COBAs, or with ad hoc organization of

landholders, to develop appropriate rules for land management and the corresponding

compensation.11

72. Institutional arrangements. The institutional arrangements required for a PES program

include the logistical arrangements to contract participants, monitor compliance, and make

payments; the rules under which the program works; and the broader legal and policy framework.12

If PES programs work through local community organizations, such as COBAs, many of the

logistical problems would be greatly simplified, as there would a single contract and a single

payment. The biggest challenge is likely to be that of monitoring the areas receiving payments, to

ensure that they are under the agreed land uses. The community would carry out its own

monitoring, but this would have to be verified by a third party. Advances in the use of remote

sensing imagery and/or drones for such monitoring are likely to make this task much less onerous

than it would have been in the past.

The project will evaluate various options to conduct monitoring of land use in the areas covered

by the PES pilots, with emphasis on methods that could be easily replicated or scaled up.

73. The challenges to developing PES pilots in Madagascar are thus significant, but far from

insurmountable.

10 In some cases, COBAs will include both farmers from irrigated areas and farmers from upland areas—indeed,

they are sometimes the same people. The development of PES programs may prove easier in such cases. In

other cases, downstream and upstream farmers will belong to different communities. The Social Safeguards

study will provide an initial indication of community organizations found at the project sites. More in-depth

surveys will carried out during the first year of implementation. 11 The Bolsa Floresta program in the Brazilian state of Amazonas (Viana and others, 2013) offers one possible

model. Rather than making payments per hectare conserved, as most PES programs do, it makes payments to

individual households who agree not to deforest, as well as separate payments to finance the adoption of

sustainable land use practices on areas that have already been cleared. In addition, it makes payments to local

communities (based on the number of participating households within the community) to finance social

investments and to support community associations. As monitoring the actions of individual households would

be impractical, penalties for non-compliance are based on reducing or withholding the community payments. 12 The Ministry of Environment is coordinating a committee to assess the need for, and possible approaches to, a

national PES policy.