1

POLICY NOTE: LANDSCAPES AND RESILIENCY IN RURAL YEMEN1

This policy note describes the analysis and offers policy recommendations of the findings of the

Rural Landscapes and Resiliency study, a component of the Yemen Rural Development Agenda

Economic and Sector Work (ESW). The objective of this study was to refocus policy makers and

other stakeholders on the importance, needs and potential for the development of Yemen’s

neglected, rain-fed rural areas which account for approximately 70 percent of Yemen’s

agriculture. Rain-fed areas include rural populations with deep poverty pockets, are the most

vulnerable to climate change, and are critical to social and political stability particularly

among the chronic poor, the disenfranchised and unemployed youth. A particular feature of

rain-fed areas is the increased marginalization of certain groups such as women, despite being

the main actors in rain-fed production activities. The rain-fed agricultural areas of Yemen have

experienced dramatic social and bio-physical changes over time, adversely impacting

livelihood and increasing vulnerability under limited coping mechanisms. As a result, rain-fed

communities lack the necessary resilience for withstanding bio-physical shocks (such as reduced

rain water) from climate change. Suggested policy recommendations include strengthening

formal and informal rural institutions for natural resource management, increasing support for

the production of cash crops as alternatives to, and supporting off-farm income generating

activities in rain-fed areas.

1 This policy note was prepared as part of analytical work on “Scaling Up Rural Development in Yemen,” an

Economic and Sector Work (ESW) study carried out by the World Bank, along with other background papers on

rural institutions, rural investment, and rural public expenditure. The team members who contributed to this policy

note are: Hanane Ahmed (hahmed2@worldbank.org) and Erick C.M. Fernandes (efernandes@worldbank.org), both

from the World Bank, Jeffrey Richie from the University of Washington (jrichey@u.washington.edu), and Ahmed

Al-Wadaey from the Sana’a University. In the World Bank office in Sana’a, Naif Abu-Lohom provided technical

guidance, Fowzia Yahya Musleh Al-Qobi, and Nabila Al-Mutawakel provided support to the team during their

mission to Sana’a. Kwaw Andam (kandam@worldbank.org) was the Task Team Leader for the ESW study. The

ESW was carried out under the guidance of Steven N. Schonberger, sector manager for water and agriculture in the

World Bank’s Middle East and North Africa Region. The study has benefited from information and ideas from

government officials, academia, and NGOs in Yemen, and from other members of the study team. Ideas in the paper

are offered as information and analysis for consideration, and do not represent the official policy or

recommendations of the World Bank or any other organization.

2

A. THE APPROACH

Since rain-fed agriculture is a farming system

reliant on rainfall for water, understanding the

causes of vulnerability and sources of

resilience in rain-fed rural areas requires

linking the understanding of the socio-

economic context to the “landscape” which

determines livelihood opportunities and

vulnerability to rainfall shocks. This approach

is key since resilience in the rain-fed areas is

determined by how households and

communities engage with their biophysical

environment to support their livelihoods, and

how they are able to adjust to changes which

come from either the social (population growth,

intra and inter-migration) or the biophysical

(land degradation, climate change) contexts.

Therefore, the challenge was to develop and

examine information that would summarize

individual community responses to the

environment across the rain-fed agriculture

regions of Yemen. How does the physical

environment set the stage for the communities?

The next section describes the tasks undertaken

to answer this question, followed by a section

assessing the socio-economic context of the

study area.

B. RAIN-FED LANDSCAPES ANALYSIS

Descriptions of the environment can be

thought of as a “sandwich,” of multiple layers

of information, each representing a key part of

the landscape. Layers include the physical relief

(the topography), soils (their types and

ultimately properties), vegetation (land-use and

land-use change), and finally water resources.

Superimposed on this physical world, then, are

the communities that live there, and their

economic and social attributes. There were 7

tasks involved in describing the bio-physical

environment of rain-fed areas.

Task 1: Delineation of the Study Area

Six sites within the Sana’a basin were selected

for this study2. The Sana’a basin (the basin is

defined as the drainage basin, meaning that rain

falling in the basin flows to a common drainage

point and to, for example, the next basin over)

is a region already suffering from extreme

water stress both in terms of rain water and

groundwater reserves. Climate models predict

that this scenario is going to further deteriorate

under the impacts of climate change

(decreasing rainfall and increasing

temperature). With a population growth rate of

seven percent, and extensive pumping for

irrigation, Sana’a’s water sources are being

rapidly depleted. This creates important

implications for water resource demand and

management.

Figures 1 and 2 delineate the focus areas of the

study.

Figure 1: Delineation of the regional scale (western

Yemen), in red line. The Sana’a basin is shown, in

black outline

The focus area of the study was the Sana’a basin: Bani Matar,

Bani Al-Hareth, Amaran, Bani Hushaish, Manakhah, Hamdan.

3

Figure 2: Image of the 6 study sites

Task 2: Digital Elevation Model

A Digital Elevation Model (DEM) establishes

the topography of the Sana’a basin. This

describes the “physical structure” of a region

(Sana’a basin) i.e. what the elevation and slope

gradients are, where drainage patterns

culminating in river networks are, etc. When

more local data of higher resolution becomes

available, the topography of the Sana’a basin

will be accurately established.

Figure 3: Topography of Sana’a basin

Task 3: Develop a Soils Map

Soil maps include not only the “names” of

soils, but, more importantly, the properties of

the soil (depth, texture, chemical composition). At the regional scale, global data sources for

soils showed little differentiation across the

region. At the scale of the Sana’a basin, two

different classifications (Russian and American

classification schemes) presented a more refined

view of the basin. Interestingly, there are

considerable differences between the two

classifications depending on the interpretation

of the soil names. Further conclusions can be

drawn when more data in digital format with

higher resolution becomes available.

Task 4: Develop maps of Land-use and Land-

Use Change (LU/LUC)

How land is used (LU) relative to what is

available is a key determinant of rural

livelihoods. Evaluating what current land use is

provides the baseline of what current conditions

are. How that use changes over time (LUC)

establishes how the landscape has evolved, with

insight for the future. This was done for years

2000, 2005, 2010, and 2013.

Regional vegetation patterns (from the MODIS

satellite) clearly show how focused vegetation is

on the highlands. The patterns suggest that most

of the eastern and northern parts of the

highlands are open shrub-lands and

cropland/natural vegetation mosaics. More

purely croplands are present on the western

slope. The Sana’a basin is at the transition,

sloping from the more vegetated highlands

down to the more barren lowlands (see Figure

4).

4

Figure 4: Land cover classes for Western region and

Sana’a basin

In addition, the more western part of the area

has considerable rain-fed crops/natural

vegetation. To the north/northwest is vegetation

that is mostly qat combined with irrigated mixed

crops (local knowledge is important to

distinguish between the two). To the

central/east is more grape production, mixed

further east with qat/mixed (see Figure 5).

Figure 5: Preliminary vegetation classification in the 6

study sites

The next question is how has this mix of

vegetation evolved over time? To examine this

question, a Landsat satellite image from March

3, 2003 was classified in the same way, and the

difference examined (“Land-use change”).

Results, while highly preliminary, are

intriguing. Overall, there is a quite substantial

change in the land-use patterns. The extent of

grapes appears to be less. Rain-fed crops may

have increased, relative to grapes. The most

pronounced change is an increase in qat/mixed

irrigated agriculture (see Figure 6).

Figure 6: Land-use change in the 6 study sites

More refined image analysis is needed.

However, the available data suggests that there

has been a significant evolution in the landscape

of the Sana’a basin over the last 11 years.

Task 5: Regional Climatology (1948-2010)

The next issue was to examine patterns in the

climatology of the region; that is, the rainfall

(P), daily maximum temperature (Tmax), and

daily minimum temperature (Tmin), covering

the period 1948-2010. Over the period of

record, precipitation showed pronounced

phases. From 1948 to about 1970, annual

rainfall was highly variable, with an average of

471 mm/year, with standard deviation of 247

mm/yr (see Figure 7). During high rainfall

years, rainfall was in the 600-800 mm/yr range.

From 1970 on, average rainfall decreased to

5

311 mm/yr, with a standard deviation of 98

mm/yr. High rainfall years would serve to

replenish ground water reserves.

Figure 7: Mean yearly records (1945-2015)

Figure 8: Mean yearly minimum and maximum

temperatures

Both minimum and maximum temperature

showed a statistically significant increase over

the period of record (see Figure 8).

The combination of decreased rainfall and

increased temperature adds further stress to the

water cycle of the basin.

Task 6: Changes in Total Water Storage

Closely related to rainfall patterns is the

“change in total water storage” across the

region. With information from the GRACE

(Gravity Recovery and Climate Experiment)

satellite, 2003-2013, clear seasonal patterns in

storage were seen, corresponding to the

precipitation regime. There has possibly been

a gradual reduction in overall storage over the

period of record (see Figure 9). Given the

large errors associated with a very small signal,

it is too early to make significant statements

about these trends, but results are suggestive.

Figure 9: Terrestrial Water Storage

Task 7: Dynamic Information Framework

(DIF) web-site

The Dynamic Information Framework (DIF)

was developed as a decision support system

using available datasets. This a user-friendly

web-site which: i) represents and provides

ready access to the data gathered and results to

a broad audience, and ii) to provide relevant

information for future rural and urban planning

by considering the entire landscape.

Information from landscapes analysis rapidly

accumulates. The information generated from

the six tasks was uploaded into the DIF web-

site, and can be updated as further data

becomes available. This kind of information

6

allows planners to consider the entire landscape

rather than a segment of it when undertaking

planning/development activities. Having a top-

down view of the landscape under analysis

provides important insights about interactions

among different elements of the natural

resource base which is useful for their long-

term sustainability.

The Center for Environmental Visualization

(CEV) at the University of Washington has set

up an interactive Yemen Dynamic Information

(DIF) web-site. Ultimately, the objective is for

the DIF web-site to be hosted by the Ministry

of Planning and International Cooperation.

Subsequently, the MOPIC will maintain, and

update the site as yearly data on the different

biophysical layers become available.

The web-site can be viewed at:

http://test3.ocean.washington.edu/ Figure 10: Yemen DIF Website

C. SOCIO-ECONOMIC ASSESSMENT

This section of the policy note describes how

rural rain-fed communities in the Sana’a

basin have coped with the bio-physical changes

of their environment (described in the

preceding section), in addition to the social

changes (population growth, and migration) in

pursuit of their livelihoods. The coping

mechanisms as reflected in strategies

determine how resilient these communities are

to the changes.

Understanding how the rural communities

adapted to the bio-physical and socio-economic

changes over time to reduce their vulnerability,

thereby, enhancing their resilience requires

descriptive panel information for rural

households both at the household and farm

levels captured through household surveys.

While this type of information is not readily

available for Yemen, this policy note offers

brief static descriptive background on rain-fed

communities in the rural Sana’a region3. This

was done using the 2005-2006 Yemen

Household Budget Survey (HBS)4. As a second

best approach and complementing this effort,

qualitative description is provided of the

communities in the rural Sana’a region in

terms of how they adapted to the bio-physical

and socio-economic pressure. The qualitative

description is followed by the main findings

and policy recommendations.

The aforementioned data constraints on lack

of available up-to-date household survey

information limits the establishment of

linkages between changes in the biophysical

environment and how the communities until

the recent years have managed to adopt to

3 The 2005-2006 HBS does not capture the Sana’a

basin area. Therefore, the descriptive statistics are

based on information on the Sana’a region (rural).

4 Ideally, information from a follow-up survey would

have been useful to analyze changes in household

behavior from bio-physical and social changes. The

2013-2014 Yemen Household Budget Survey (HBS)

is currently ongoing, and is expected to be available

by the end of 2013/early 2014. While this HBS

captures a variable which identifies whether

households rely on rain for water as a source of

irrigation, concerns exist about the quality of data

entry and consistency of questions captured in the

survey questionnaire and in the actual database.

7

these changes. As a result of this limitation,

qualitative information has been used to

provide an overall picture of the socio-

economic changes over time. Currently a

survey is being undertaken (Yemen Household

Budget Survey 2013/2014) which is an update

of the 2005/2006 HBS. There is potential to

update the socio-economic assessment once the

survey data becomes available.

Additional constraints which have limited

establishment of linkages between the

biophysical and socio-economic sections

involved the lack of opportunity to undertake

field work due to security concerns in Yemen.

i) Descriptive Background

According to the 2005-2006 Yemen Household

Budget Survey, the poverty incidence in the

overall Sana’a region was 28% and nearly 10%

were food insecure.

Out of the 4,527 Yemeni households captured in

the agriculture section of the HBS, nearly 68%

rely on rain water as a source of irrigation, and

25% rely on wells for water. In the rural Sana’a

region, 56% and 39% of the households in rural

Sana’a region rely on rain and wells

respectively for irrigating their lands. In

addition, 91% of households in rain-fed rural

Sana’a region engaged in livestock activities,

and 15% have enterprises. As for the type of

land ownership, 48% of rain-fed lands are

owned, and 39% are share-cropped in.

The majority of the household heads in rain-fed

rural Sana’a region are illiterate (84%) and only

1% are female headed households, while the

percentage of female headed households in the

rain-fed areas in rural Yemen is 6%. Even

though women are central agents in the rain-fed

areas, estimates on percentage of female headed

households may not reveal this fact, due to for

example the definition that may have been used

in the survey to classify a household member as

a household head.

As for source of income during the past 12

months (of the survey year) in the rain-fed rural

Sana’a region, qat production and sales is a

primary mean of deriving agriculture income.

According to the 2005-2006 HBS, income

proceeds from the sale of qat averaged 193,503

Yemeni Riyals relative to sales from non-qat

crops amounting to 10,366 Yemeni Riyals.

While income from qat sales contributed

substantially to the overall household income,

the households seemed to have also relied on

alternative source of income for diversifying

their livelihood strategies. According to the

2005-2006 HBS, on average households in the

rural Sana’a region received 3,932 Yemeni

riyals from their retirement payment income,

2160 riyals from the Social Welfare Fund, and

2,628 riyals from remittances. Remittance

income is a vital source of livelihood support to

households in rain-fed areas but the question

remains about its sustainability with increased

limitations of migration opportunities.

According to the 2005-2006 HBS, households

in the rural rain-fed Sana’a region have obtained

loans for several reasons. The most cited reason

is borrowing to meet household’s consumption

needs and this was the response of 57% of the

households, followed by borrowing for

ceremonials such as weddings, funerals, etc…

(28%) and 27% of the households cited

borrowing to meet emergency needs. The

sources of the loans are diverse but the majority

of the rural rain-fed households in the Sana’a

region (73%) borrowed from friends/neighbors.

Overall, while income from the sales of qat is

the primary source of livelihood in the rain-fed

communities of the rural Sana’a region,

alternative sources of income as indicated are

also important. However, given constraints in

water availability for crop production including

qat fiscal constraints, and limits on migration,

8

the coping mechanisms for increasing resilience

against bio-physical and socio-economic

changes may not be sustainable potentially

leaving rain-fed communities vulnerable.

ii) Study Site Characteristics

The study site characteristics relate to sites in

the Sana’a basin.

Specifically, in the Bani Matar site, almond is

recognized as a strategic crop. It is considered a

drought tolerant cop and grows in areas low

rainfall amounts (150-400 mm). However,

during the flowering and fruiting stage, water

availability becomes crucial. Communities view

this crop as a substitute for the qat production

which is a promising source of income for

farmers in Bani-Matar.

In the Bani Hushaish site, 35% of the

cultivated area is used for grapes production,

30% for vegetables, and 35% for qat production.

In this area, poor living standards of local

people led many owners and users of wells to

look for other income sources through selling

water to others. This was the situation before the

initiative to prevent selling water was

introduced.

In the Bani Al-Hareth site, key crops produced

include coffee, almond, and qat. The challenges

facing this area include the wastewater and

drainage system from the capital being directed

to Bani Al-Hareth. Therefore, the community

continues to face challenges from biological and

chemical contamination, which can significantly

affect rural landscape development.

In the Manakhah site, this area is famous for

planting coffee as a valuable source of income.

However, recently qat is grown as much as

coffee is, and competes with vegetable crops as

well. Land degradation is present in the terraced

farms especially during the rainy season.

Helping the farmers in adopting optimal land

and water management practices will

significantly raise income and reduce poverty.

In the Amran site, this area has the highest level

of poverty in both rural and urban Amran. In

Amran governorate, farming is the main source

of income after livestock production. The

governorate which is considered agriculturally

fertile, meets a major share of the country’s

agricultural needs. Most farmers in the

governorate practice subsistence farming of

maize, wheat, millet and vegetables. However,

the series of conflicts in the area in recent

previous years had a severe impact on the

agricultural sector. Farmers were hard hit as the

agricultural cycle was interrupted, inflation

soared, and public services deteriorated and

became unreliable. With the aim of improving

farming practices, the Yemeni government took

the step some time ago of setting up agricultural

research farms in different parts of the country.

In the Amran governorate, for example, an

agriculture research station was set up to

introduce improved varieties of grains and

fruits, and to help local farmers adopt new

technologies.

iii) Main Findings

(a)With regard to the social changes of the

Sana’a basin over time, the abandonment of

agricultural land is a phenomenon mostly

driven by socio-economic factors such as

immigration into areas where new economic

opportunities are offered to rural people. Ecological drivers such as elevation and land

mismanagement (biophysical factors) leading to

soil erosion and reduced soil fertility are of

secondary importance. However, water remains

the central constraint for the agriculture of the

basin.

(b) Poverty is relevant in the Sana’a basin. The

poverty rate is considered relatively high

9

according to the poverty map prepared by the

World Food Program (WFP) and the World

Bank. The poverty rate ranges between 25.8%-

35%. However, this percentage is expected to

increase due to increased limitation over time in

opportunities for on-farm and off-farm income

generating activities.

(c)Poverty is not directly related to rain-fed

areas alone. However, rain-fed agriculture that

takes place in most parts of Yemen needs to be

supported more strongly, as the option is either

supporting low levels of production or

desertification. Also the continuous

demographic pressures in terms of population

growth will further exacerbate the situation and

undermine most of the efforts exerted to

alleviate poverty. This will be reflected in the

high rates of poverty and widespread

malnutrition, especially with the serious

challenges the government is facing in covering

educational, health and infrastructure services

provision.

(d)Livelihood strategies are diverse but

vulnerable leading to weaker resilience in the

context of limited coping strategies. Rain-fed

households typically own a few small ruminants

(usually from 5-10), in addition to shared

animals (cattle and small ruminants mainly).

(e)Casual seasonal labor exists in agriculture

locally, mostly in harvesting and in weeding. Usually women’s daily pay rates are one-third

to one-half those of men. In some cases

women’s only payment for weeding is the

weeds/grasses that they collect. These are used

for their own livestock activities or even sold.

Collection and selling of firewood for fuel is a

common mean of obtaining energy resources.

(f)Women are occasionally involved in off-

farm activities such as sewing, henna

decoration, bead work etc. and even for a poor

household these do not generate more than

25% of their income. Financial and material

support from others include widow’s pensions

payments from the Social Welfare Fund and

contributions from charitable associations

(mostly at the time of Ramadan and Eid) but

also from wealthier families in the community,

who help them, whether they are related or not.

(g)The lack of income opportunities steadily

forces people to apply negative coping

mechanisms, including child labor and the sale

of assets (including livestock) and relief items.

(h)Women are central agents in the

management of water and land resources in

rain-fed agriculture. Male migration is a key

factor for the increased involvement of women

in rain-fed agriculture. More men in rain-fed

rural communities migrate to the cities or to the

Gulf countries to find jobs. As a result, women

become more and more involved in water

resource management in rain-fed agriculture

activities i.e. production and livestock. Here,

traditional agriculture relies mainly on dry

terraced farming which are maintained and

managed by women, while young girls take on

the responsibility of fetching drinking water, in

addition to other water activities related to

household responsibilities. Adult women are

responsible largely for daily agricultural

activities, while men are in charge of seasonal

activities such as plowing and harvesting. Rain-

fed regions today suffer from terrace

degradation and land erosion, as rain‐fed

systems are neglected for unsustainable irrigated

agriculture practices. While devoting most of

their daily activities in rain‐fed agriculture

practices and livestock rearing, many women

have also developed knowledge in water and

soil management.

D. POLICY RECOMMENDATIONS

The dominant land-use/water linkage is clearly

the expansion of qat and other irrigated crops,

with the associated impacts, both social and

environmental. The shift from more traditional

10

crops to qat, coupled with changes in climate

(deceasing rainfall, increasing temperature) adds

additional layers of water stress to Yemen.

The second type of constraint facing the rain-

fed rural communities of the Sana’a basin is

the abandonment of agricultural land driven

by socio-economic factors i.e. immigration of

males into areas with better income

opportunities. This socio-economic change has

implications for the management of both land

and water practices in terms of for example

building new and maintaining existing soil and

water conservation structures.

So far the communities have adopted several

unsustainable coping mechanisms to increase

their resilience to both sets of challenges

through contributions from the Social Welfare

Fund, migration, child labor, sales of assets

(including livestock), and relief items.

Given the aforementioned challenges and the

unsustainable coping strategies, the following

are three suggested policy recommendations to

enhance the resilience of rain-fed communities

against bio-physical and social changes:

First, strengthen the existing traditional land

and water resource management practices that

have been adopted for many generations in

rural Yemen to undertake integrated

landscapes management. It is important for

policy makers to support existing formal and

informal institutions which already have the vast

local knowledge for supporting their natural

resources, to develop local adaptive capacity

and resilience5.

Existing global experiences exhibit the positive

outcome that can be realized if natural

resources are managed efficiently through

integrated landscapes management. For

example, in the lower Amu Darya river basin in

Uzbekistan the approach has included

5 See associated policy note on rural institutions.

improving water management for drainage and

salinity control and wetland restoration,

increasing productivity of irrigated agriculture

and restoring grazing lands. In China, the Loess

Plateau Watershed rehabilitation project has

returned the devastated Loess Plateau to

sustainable agricultural production, improving

the livelihoods of 2.5 million people and

securing food supplies in an area where food

was scarce in the past. In Colombian hillsides,

the landscape approach has been integrating

livestock, trees and a range of crops, depending

on the slope of the land and the direction of the

streams, to increase incomes while conserving

the landscape.

Second, it is important to re-think the

implications of an expanded production of qat

as part of an effort to slow down the depletion

of water resources. Qat is a water intensive crop

and crowds out the production of food crops and

export crops. First and foremost, this requires a

concerted political and social will from the

policy makers. Second, increase the

attractiveness of producing other cash crops such

as coffee and almond so farmers have the

incentive to alter their marginal rate of

substitution between qat and alternative

profitable cash crops. Doing so will require

building a strong and organized value chain for

alternative cash crops production which already

exists in the case of qat production6.

Third, suggested policy recommendation to

address the socio-economic challenges driven

by migration from rain-fed areas would be to

provide support to increase off-farm income

generating activities in the rain-fed areas. Essentially this means promoting a dynamic

rural non-farm sector with linkages to both the

agriculture and the urban economy (World

Development Report, 2008). This was the case

in China in which industry was brought to rural

towns allowing for diversification of rural

6 See associated policy note on the rural investment

climate.

11

incomes. While this approach is not readily

available in the short-run, policy makers need to

re-think this as a vital measure for the

transformation of the rural economy, and

strengthening resilience in the medium and

long-run.