0

FURTHER SUPPORT TO LAND USE

EULUP

Project Number 2010/230-489

Rural Land Management Planning:

Step 1 and 2:
Optimised Land Use Plan
and
Land Use Intensity Plan

TECHNICAL REPORT No. 9

Neda Nordin
KE2 on Environment and Spatial Planning

Robert Palmer
SSTE on Land Use Zoning

2011

An EU funded project managed by Implemented by:

European Commission Liaison Office
 1

Contents

1 EXECUTIVE SUMMARY.....................................................................................................................4
2 INTRODUCTION...............................................................................................................................7
3 OUTLINE OF RLMP FRAMEWORK....................................................................................................8
3.1 Structure of RLMP...................................................................................................................8
3.2 Decision Tree technique..........................................................................................................8
4 OPTIMISED LAND USE PLAN (oLUP)................................................................................................9
4.1 Structure, rules and data.........................................................................................................9
4.2 oLUP Decision Trees..............................................................................................................11
4.2.1 Arable land (CLC Class 2.1).............................................................................................11
4.2.2 Permanent crops (CLC Class 2.2)...................................................................................12
4.2.3 Pastures (CLC Class 2.3).................................................................................................12
4.2.4 Heterogeneous agriculture (CLC Class 2.4)....................................................................12
4.2.5 Forests and semi-natural areas (CLC Classes 3.1, 3.2, 3.3).............................................13
4.3 oLUP results and recommendations......................................................................................13
5 LAND USE INTENSITY PLAN (LUIP).................................................................................................15
5.1 Structure, rules and data.......................................................................................................15
5.2 LUIP Decision Tree.................................................................................................................16
5.2.1 Agriculture (CLC Class 2)................................................................................................16
5.2.2 Forest and semi-natural vegetation (CLC Class 3)..........................................................17
5.3 LUIP results and recommendations.......................................................................................17
6 LIMITATIONS OF oLUP AND LUIP...................................................................................................18
7 CONCLUSIONS...............................................................................................................................19
8 REFERENCES..................................................................................................................................20

ANNEX Figures
Fig. 1 Basis for Rural Spatial Planning
Fig. 2 oLUP Decision Tree for arable land use and permanent crops
Fig. 3 oLUP Decision Tree for pastures
Fig. 4 oLUP Decision Tree for heterogeneous agriculture
Fig. 5 oLUP Decision Tree for forests, natural grasslands, transitional
woodland/shrub and sparsely vegetated areas
Fig. 6 LUIP Decision Tree

An EU funded project managed by Implemented by:

European Commission Liaison Office
 2

Abbreviations and Terms

ALS Agricultural Land Suitability

BRSP Basis for Rural Spatial Planning
CAP Common Agricultural Policy (EU)
CLC CORINE Land Cover
CLC2006 CORINE Land cover data for 2006
CORINE Co-ordinate Information on the Environment
CZP Construction Zoning Plan
DT Decision Tree
EBV Estimated Biodiversity Value
EEA European Environmental Agency
ELS Environmental Land Sensitivity
EU European Union
EULUP Acronym for project “Further Support to Land Use”
FAO Food and Agriculture Organisation
GIS Geographic Information System
GWS Groundwater Sensitivity
GWR Groundwater Risk
ICMM Independent Commission for Mines and Minerals
IDWR Institute for Development of Water Resources
ILA Integrated Land Suitability
JRC Joint Research Centre (of EU)
KE2 Key Expert for Environment
LUIP Land Use Intensity Plan
MAFRD Ministry of Agriculture, Forestry and Rural development
MESP Ministry of Environment and Spatial Planning
oLUP Optimised Land Use Plan
RLMP Rural Land Management Planning
SER Soil Erosion Risk
SES Soil Erosion Sensitivity
SSTE Senior Short-term Expert
USDA United States Department of Agriculture

An EU funded project managed by Implemented by:

European Commission Liaison Office
 3

1 EXECUTIVE SUMMARY

Robust rural land management planning is critical if Kosovo is to achieve a sustainable and
competitive agricultural sector. Fertile agricultural land is too precious to be lost under
widespread un-planned development. Rural Land Management Planning (RLMP), at
municipality level, is a vital part of future spatial land use planning in Kosovo.

Use of these plans will help to protect the best and most flexible land for agriculture and
horticulture. At the same time they will help to protect the environment by restricting the
potentially damaging effects of soil erosion and water pollution from agriculture. Furthermore,
their use will also promote care for biodiversity both in the wild parts of Kosovo and in many
agricultural areas.

RLMP provides a practical framework that fosters improved land use structure that
encourages land to be managed in an integrated and sustainable way. RLMP has been
divided into four stages (Fig. 1 Annex 1) all of which rely on land assessments that have been
incorporated into Agricultural Land Suitability (ALS) and Environmental Land Sensitivity (ELS)
values.

 Stage 1 of RLMP identifies the optimum land use across Kosovo through the
Optimised Land Use Plan (oLUP).
 Stage 2 builds on the oLUP by recommending the intensity of land use within a Land
Use Intensity Plan (LUIP).
 Stage 3 provides recommended guidelines for soil/land management to safeguard the
environment through the Rural Environmental Action Plan (rEAP).
 Stage 4 brings together all previous data to develop a Construction Zoning Plan (CZP)
at Municipality and Village level.
Technical guidance for the preparation of oLUP and LUIP are provided in this report.
Separate Technical Reports have been prepared for ALS, ELS, ILS, rEAP, and CZP. These
technical reports will be combined into one user-manual ‘Guidelines for Rural Spatial
Planning’.
The conceptual models for oLUP and LUIP were created by using a statistical technique
called ‘Decision Tree’. This allows the establishment of logical rules by asking questions and
analysing their answers that are later grouped into the most appropriate decisions.

The oLUP map identifies the optimal land cover for Kosovo in terms of ensuring the most
efficient use of land resources for economic and environmental purposes. oLUP is based on
land cover classes (CLC 2006), but integrates agricultural land suitability (ALS),
environmental land sensitivity (ELS), flood risk and irrigation infrastructure. The oLUP
Decision Tree assessment resulted in 5 recommendations, where 4 of them suggest land
use changes:

1. Current land use

An EU funded project managed by Implemented by:

European Commission Liaison Office
 4

2. Heterogeneous areas or permanent crops

3. Pasture
4. Arable land
5. Non-agriculture
The proposed land use changes (Recommendations 2 to 5) concerns only 12% of Kosovo
territory while the rest is recommended to be used as currently is (Recommendation 1). The
naming of recommended land use changes were kept consistent with CLC system which is
still new for Kosovo.

The oLUP has been integrated with CLC 2006 data to derive a national land cover map
identifying the optimum land cover across the whole of Kosovo (oLUP+CLC).

oLUP+CLC provides the input data for developing the Decision Tree for LUIP – the second
plan within the RLMP suite. LUIP assigns land management proposals and intensity of use to
help decision makers to strategically guide agricultural and forestry inputs and businesses.

The LUIP conceptual model identifies 5 distinct types of land management (Fig. 1 Annex 1) -
3 for agricultural land and 2 for forest/semi-natural areas:

1. Intensive Agriculture or horticulture

2. Moderate agriculture
3. Extensive agriculture or livestock
4. Commercial or bio-energy forestry
5. Sustainable forestry or nature management
The most important agricultural land management type defines Intensive agriculture and
horticulture (Zone 1). This is the area that should be preserved for agriculture with a strict
presumption against development. This is both the highest quality agricultural land and also
the land where environmental limitations on agriculture are either small or adverse effects
from agriculture can be adequately mitigated by good soil/land management techniques.

The second agricultural land management type – Moderate agriculture (Zone 2) also contains
some good agricultural land but here environmental limitations are greater and cannot always
be adequately mitigated. Although there will be pockets of high quality land, generally the
intensity of agricultural use will be lower than Zone 1. Nevertheless, much of this zone is
worth preserving for agricultural use but inevitably some development will have to be allowed
according to local pressures and intensities identified by Municipality Planners.

The third agricultural land management type is Extensive agriculture and livestock (Zone 3).
This represents the poorest agricultural land, some of which is not suited to arable cultivation
and remains under pasture, grasslands or mixed use. Environmental limitations, such as soil
erosion risk, groundwater sensitivity or high biodiversity value may locally preclude
agricultural development. The cost of development in Zone 3 will often be considerably
higher than in the other agricultural zones, but nevertheless this is where urban development
in agricultural areas should be encouraged as long effective and relevant environmental
protection measures or technologies are applied.

The LUIP divides areas of forests and semi-natural vegetation into two broad management
zones. The important zone for economic development of Kosovo is Commercial or Bio-

An EU funded project managed by Implemented by:

European Commission Liaison Office
 5

energy forests (Zone 4). These are the areas where commercial forests for industrial or
energy use could be located as these areas have few environmental limitations relatively
high ALS (class 1-5). Implementation of Bio-energy land management would help Kosovo to
better implement climate change mitigation measures.

The remaining zone Sustainable forestry or nature conservation (Zone 5) contains the most
valuable areas of Kosovo in terms of biodiversity resources or the land that is intrinsically is
not suitable for commercial forestry (i.e. land cover type 3.3.3 – Sparsely vegetated areas).,
Municipality Planners must exercise caution and in relevant areas begin to identify potential
sites for NATURA 2000 to ensure that the EU Habitats Directive requirement for Special
Areas of Conservation and the EU Birds Directive call for Special Protection Areas for birds
can be implemented in future.

The quality of proposed recommendations for land use changes and land management types
greatly depends on the quality of the input data, such as CLC, DTM, and on all other data
used to compile Land Assessment models (ALS, SES, GWS, EBV), which is sufficient for
national and municipal level planning.

The proposed land management zones represent the first attempt to consistently implement
an integrated land management approach throughout the whole Kosovo. Using this
methodology all municipalities will be able to apply a national dataset for coherent planning of
urban and agricultural developments Furthermore relevant ministries will have this same
dataset to assist and enhance policy making, which will be especially useful during the period
of EU accession.

An EU funded project managed by Implemented by:

European Commission Liaison Office
 6

2 INTRODUCTION

Rural Land Management Planning (RLMP), at Municipality level, is a vital part of future spatial
planning in Kosovo. Widespread use of these plans will help to protect the best land for
agriculture and horticulture and manage it in a sustainable way. At the same time they will
help to protect the environment by restricting potentially damaging effects of soil erosion and
water pollution from agriculture. Their use will also promote an increase in biodiversity both
in the forested parts of Kosovo and in agricultural areas.

The aim of this Technical Report is to help Municipality Planners to understand the system of
planning models developed by this project. It will also inform Government when formulating
agricultural policies.

The current technical guidance relates to Stages 1 and 2 of RLMP (Fig. 1 Annex). The
methodology is fully described, step by step, from producing an Optimised Land Use Plan
(oLUP) through to defining land management zones within a Land Use Intensity Plan (LUIP)
that can be used directly or enhanced further by Municipality Planners as part of their
Municipality Development Plans.

One of the key elements of this RLMP framework is the insight it gives to inherent land quality
coupled with the environmental consequences of intensively using land. Land quality in
Kosovo has been assessed by developing a national system of Agricultural Land Suitability
(ALS) (Palmer and Nordin 2011a). The detrimental effects of mechanised arable cultivation
on the environment have been considered by implementing an Environmental Land
Sensitivity (ELS) assessment (Nordin  and Palmer 2011b). Linking ALS and ELS to derive an
Integrated Land Suitability (ILS) provides a powerful tool for land use planning and zoning in
rural areas. ILS is a basic input to RLMP in agricultural areas and its use will help target rural
development away from the best land and hence contribute to the competitiveness of the
agricultural sector.

The integration of nature conservation and environmental protection into the management of
agricultural land follows the principles of CAP reform as laid out in 1992. EU agricultural
subsidies no longer depend on production but are used to support some of the non-
productive functions of rural land, such as; environmental interactions, ecology, cultural
heritage and social and employment issues, all linked to greater efficiency.

Robust rural land management planning is critical if Kosovo is to achieve a sustainable and
competitive agricultural sector. Fertile agricultural land is too precious to be lost under
widespread un-planned development.

An EU funded project managed by Implemented by:

European Commission Liaison Office
 7

3 OUTLINE OF RLMP FRAMEWORK

3.1 Structure of RLMP

The RLMP framework is made up of four stages (Fig. 1 Annex) all of which rely on land
assessments that have been incorporated into Agricultural Land Suitability (ALS) and
Environmental Land Sensitivity (ELS) values.

 Stage 1 of RLMP identifies the optimum land use across Kosovo through the
Optimised Land Use Plan (oLUP).

 Stage 2 builds on the oLUP by recommending the intensity of land use within a Land
Use Intensity Plan (LUIP).

 Stage 3 provides recommended guidelines for soil/land management to safeguard the

environment through the Rural Environmental Action Plan (rEAP).

 Stage 4 brings together previous stages at Municipality and Village level to provide the
Construction Zoning Plan (CZP).

Technical guidance for the preparation of oLUP and LUIP are provided in this report.
Separate Technical Reports have been prepared for ALS, ELS, ILS, rEAP, and CZP. These
technical reports will be combined into one user-manual ‘Guidelines for Rural Spatial
Planning’.

3.2 Decision Tree technique

Decision tree (DT) analysis is a powerful and popular tool for classification and prediction. A
DT uses a tree-like graph or model of decisions and their possible consequences to analyse
certain patterns – rules – that are easily understandable. This way the competing alternatives
can be compared by posing a series of questions (DT nodes) and analysing their answers
(DT decisions).

The decision tree frameworks used in oLUP and LUIP are described below and graphically
presented in the Annex.

An EU funded project managed by Implemented by:

European Commission Liaison Office
 8

4 OPTIMISED LAND USE PLAN (oLUP)

The first task within the RLMP framework is to derive a map identifying the optimum land user
across the varying landscapes of Kosovo. Its implementation will ensure a more efficient use
of land resources in terms of economics and environment protection.

An assessment of the most favourable sustainable land use for an area takes account of the
inherent quality of the land. The considered land qualities are related to land capacity to
support a particular land use together with the environmental implications of carrying out that
land use on the site. A land use that degrades the environment is unlikely to be the best
possible use of that land. If, however soil erosion, water pollution or degrading biodiversity
caused by using land in a particular way can be mitigated by good land management
techniques, then the land use is sustainable in the long term. Inherent land quality is
identified by ALS and environmental implications of arable cultivation are assessed by ELS.

oLUP is based on the most up to date available CORINE Land Cover (CLC) data. The most
recent CORINE (COoRdinate INformation on the Environment) land cover map (CLC2006)
was made available on the EEA website in 2010 and was produced from 2006 satellite and
orthophoto data. The CLC2006 GIS database was produced jointly by the European
Commission and37 European countries. For further information on CLC please refer to ELS
Technical Report, Section 5.2.1 and Annex (Palmer and Nordin 2011) and its Annexes or the
web site at http://www.eea.europa.eu/publications/COR0-landcover/page001.html.. In Kosovo
there were identified 28 CLC classes (see Annex in ELS Report).

4.1 Structure, rules and data

The structure of oLUP conceptual model consists of five decision trees (DTs) which are
based on CLC2006 classes integrated with the following data:

 Agricultural Land Suitability (ALS),

 Environmental Land Sensitivity (ELS),
 Estimated (modelled) flood risk (FLR) areas,
 Irrigated areas (IRA).
The main function of each DT in the oLUP model is to review data from the interaction of
ALS, ELS and as the extent of flood risk and irrigation areas. Using basic GIS data analysis,
data classes were compiled for each oLUP DT. The main rules applied in the DT’s structure
are as follow:

 Where ALS class is High (Class 1-3) land must be kept under agriculture, wherever
possible, because its economic value exceeds any environmental value.

 A High ESL score (> 40) indicates where land has sufficient environmental sensitivity
to be an unsustainable impediment to intensive arable cultivation. This level of ELS

An EU funded project managed by Implemented by:

European Commission Liaison Office
 9

indicates that one or more of erosion, groundwater or biodiversity limitations cannot

be mitigated by simple land management techniques. Land with an ELS score > 40
should be left for environmental protection purposes or used for extensive agriculture.

 Areas with a flood risk must not be used for arable cultivation and natural vegetation
or mixed use should be maintained.

 Irrigation infrastructure is an important agricultural investment, and all areas currently

under irrigation should remain in agricultural use.

The assessment of optimum land use has been based on the CLC 2nd level land cover
classes that relate to agriculture and forests/semi-natural areas in Kosovo. . A separate DT
has been devised to identify the optimum land use for individual CLC 2nd level classes or
groups of classes (Table 1). . CLC land cover 1st level category 1- Artificial surfaces and
category 5 – Water Bodies have been excluded from this analysis.

Include these
DT CLC 2nd level classes
CLC 3rd level classes
1 2.1 Arable land 2.1.1
2 2.2 Permanent crops 2.2.1, 2.2.2
3 2.4 Heterogeneous agriculture 2.4.2, 2.4.3
4 2.3 Pastures 2.3.1
3.1 Forests 3.1.1, 3.1.2, 3.1.3
5 3.2 Scrub and/or herbaceous vegetation 3.2.1, 3.2.2, 3.2.3, 3.2.4
3.3 Open spaces with little or no vegetation 3.3.3, 3.3.2, 3.3.3, 3.3.4
Table 1 Groups of DTs as related to CLC 2nd level classes for assessing optimum land use in Kosovo

The ALS conceptual model and GIS data has been produced by EULUP and described in
Technical Report No.3 (Palmer and Nordin 2011). The ELS model along with three sub-
models of GWS, EBV and SES and GIS data has been produced by EULUP and described in
Technical Report No.8 (Nordin and Palmer 2011).The flood risk (FLR) assessment has been
modeled by EULUP. A simple model estimating the likely extent of regular flooding in Kosovo
has been devised for use in oLUP, EBV and rEAP (Table 2). FLR is based on the river
network and is defined where slopes, adjacent to the river, are 0-10, and the elevation of the
land parcel (i.e. grid cell in GIS data) is the same as the river bed. Such areas must also
coincide with those CLC classes (2.3.1, 2.4.3, 3.1.1, 3.2.1 and 3.2.4), which indicate the
possibility of frequent flooding. . If the CLC mapped polygon is larger than the area delimited
by slopes 0-10 then the extent of slopes 0-1 is taken as the outer extent of land under flood
risk.

FLR Factors Criteria

Slope 0-10
Elevation Same as river bed
CLC classes: 2.3.1, 2.4.3, Polygons must cross or lie
3.1.1, 3.2.1, 3.2.4 immediately adjacent to rivers
Table 2 Factors used in Flood Risk (FLR) assessment

An EU funded project managed by Implemented by:

European Commission Liaison Office
 10

In some cases it may be possible to augment these simplistic nationally-derived FLR data by
more detailed records held by individual Municipalities.

Land irrigation (IRA) data were derived from the project “KIRP – Kosovo Irrigation and
Rehabilitation Project”, implemented in 2005. The beneficiaries of this project were MAFRD
and private water companies. This project, from field investigation, collected the extent of
irrigation infrastructure together with a range of other attribute data in a GIS database. For
modeling purposes, irrigation channels were buffered in order to identify single and unified
areas (polygons) under irrigation, irrespective of whether water is supplied from a channel or
pipe.

4.2 oLUP Decision Trees

A separate DT has been developed for each of the five CLC 2nd level classes given in Table1
and are discussed below.

4.2.1 Arable land (CLC Class 2.1)

This group of CLC classes should include both non-irrigated (2.1.1) and permanently irrigated
arable land (2.1.2), however CLC2006 in Kosovo identified only the 2.1.1 class. Undoubtedly
there are areas of 2.1.2 class, and hopefully they will be identified in the future CLC update.

These DT rules for arable land (Fig. 2: Annex) should be applied in strict chronological order
and the GIS methodology is provided by Ducke (2011). First ALS is reviewed. Where it is
high (ALS Class 1, 2 and 3) the optimum land use shall always be intensive arable use
irrespective of the presence of any environmental sensitivity. This is likely to be some of the
most productive land in Kosovo and any environmental sensitivities, if and when they occur,
can be readily mitigated by a range of good land management techniques.

Where arable use occurs on lower quality land (ALS Class 4 or 5) the optimum land use
again shall remain as arable unless ELS exceeds a score of 40. . Optimum land use in
areas where ELS exceeds score 40 shall not be intensive arable but heterogeneous
agriculture or permanent crops. This is a mixture of annual crops, permanent crops, and
natural vegetation often in complex patterns with small areas of arable cultivation.

On poor quality land (ALS Class 6, 7 and 8) arable use is not the best economic use of land
and, therefore, pasture is recommended for these areas.

Flood risk assessment was not included into this DT as there are no areas of flood risk in this
CLC class. The presence or absence of an irrigation network was not considered as in areas
of ALS class1-3 it is not a relevant criteria, while for ALS classes 4-5 GIS analysis showed
that there were no irrigated areas where ELS score exceeded 40.

An EU funded project managed by Implemented by:

European Commission Liaison Office
 11

4.2.2 Permanent crops (CLC Class 2.2)

This group of land cover classes in CLC includes vineyards, orchards and berry plantations.
All crops require expensive infrastructure with an on-going commitment to high maintenance
costs. It is impracticable, therefore, to expect landowners to change land use irrespective of
land quality and thus the DT recommends “keep the current land use”. A very simple DT is
provided for these crops (Fig. 2: Annex). Informal field observations suggest that the area of
Permanent Crops has increased recently because of the expansion of vineyards. However,
this increase is not yet reflected in CLC2006 data.

4.2.3 Pastures (CLC Class 2.3)

Where the FLR model identifies a flood risk in these areas, land use should not be changed
as pasture remains the optimum use (Fig. 3: Annex). Locally, pasture occurs on good quality
land (ALS Class 1, 2 or 3) and here there is the potential for more intensive land use. This
potential can be realized wherever irrigation is available but areas outside the reach of an
irrigation system shall be assessed for environmental limitations. Wherever ELS >40 pasture
remains the best land use option. Where the irrigation network is in place and ELS is < 40, it
is recommended to turn pastures into arable land.

Where Pasture is found on average or poor quality land (ALS Classes 4 to 8) pasture shall
remain as the optimum land use.

4.2.4 Heterogeneous agriculture (CLC Class 2.4)

The CLC 2nd level class 2.4 includes the following 3rd level classes that are identified in
Kosovo: 2.4.2-Complex cultivation patterns and 2.4.3-Land principally occupied by agriculture
with significant areas of natural vegetation.

Where a flood risk is present current heterogeneous land use is considered as suitable land
use (Fig. 4 Annex). Outside the flood risk areas land quality (ALS Class) is the main factor
determining optimum land use. Areas of high quality land (ALS Class 1 and 2) shall be used
for intensive arable cultivation as long as environmental sensitivities are low (ELS score <40).
If ELS is >40 then the current land use should remain.

In areas of ALS Class 3 or 4, where irrigation is available, and where environmental
sensitivity is low (ELS score <40) an intensive arable use would be economically beneficial.
All other areas of ALS 3 or 4 should remain with current land use.

In areas of ALS Class 5 or 6 current land use is always considered as optimal.

On poor quality land (ALS 7 or 8) optimal land use is defined by the scale of environmental
limitations. Where there are low environmental sensitivities, then the current land use will be
remain the optimum use. However, where environmental sensitivities are higher (ELS>40) on
this poor quality land then non-agricultural use is recommended.

An EU funded project managed by Implemented by:

European Commission Liaison Office
 12

4.2.5 Forests and semi-natural areas (CLC Classes 3.1, 3.2, 3.3)

These CLC 2nd level classes include the following 3rd level classes in Kosovo: 3.1.1- Broad–
leaved forests, 3.1.2- Coniferous forest, 3.1.3- Mixed forests, 3.2.1- Natural grasslands, 3.2.2-
Moors and heathland, 3.2.3- Sclerophyllous vegetation, 3.2.4- Transitional woodland/shrub,
3.3.1- Beaches, dunes, sands, 3.3.2-Bare rocs, 3.3.3- Sparsely vegetated areas, 3.3.4- Burn
areas.

Optimum land use for these land cover categories depends on land quality (Fig 6: Annex).
Locally, where ALS is Class 1, 2 or 3 there may be the possibility for arable cultivation.
However, in most cases environmental limitations can be high and where ELS >40 current
land use remains the optimum use. Where environmental sensitivities are low (ELS <40)
arable should be recommended wherever there is no risk of flooding, the blocks of land
exceed 25 ha and irrigation is available. This land size stipulation is used because
mechanized arable cultivation is only viable and sustainable on blocks of land above this size.

Elsewhere for high quality land (ALS 1, 2 or 3) and for all land of only average or low quality
(ALS Class 4-8) current land use remains the optimum use.

4.3 oLUP results and recommendations

A comparison of the oLUP and CLC indicates that in order to attain a sustainable and more
effective land use structure the land use of approximately 12 per cent of Kosovo should be
changed (Table 3). The largest change recommended by the oLUP assessment is the
identification that 30% of the area of Heterogeneous agriculture (CLC 2.4) is best suited to
arable use and nationally this accounts for almost 10 per cent of Kosovo. A further seven per
cent of Heterogeneous crops (2 per cent of Kosovo) is identified by oLUP as to being better
suited to non-agricultural use.

Several other small changes in land use are recommended by the oLUP. Nationally, within
Kosovo, just over two per cent of arable land as indicated by the current CLC2006 data is
better suited to Heterogeneous or permanent crops or to pasture. Over six per cent of
Pastures should be converted to arable cultivation. A very small area of CLC land class 3 –
Forests, scrub, open spaces etc should be turned into arable cultivation.

Recommended changes in CLC classes by oLUP assessment (%)

Arable Pastures Heterogenous Forests and Kosovo
land agriculture semi-natural
(CLC 2.1) (CLC 2.3) (CLC 2.4) areas (CLC 3)
1. Keep current x x x x x
Optimised land use
land use 2. Heterogeneous 1.86 0.18
categorie or perm. crops
s 3. Pasture 0.68 0.07
proposed
4. Arable Land 6.48 34.2 0.05 9.88
by oLUP
5. Non- 7.39 2.1

An EU funded project managed by Implemented by:

European Commission Liaison Office
 13

Agriculture
Total x x x x x

2.1
5 - Non-agriculture
9.2 4 - Arable use
0.1 3 - Pasture

0.2 2 - Heterog/perm.crops
1 -Keep current use
88.4

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0

Table 3 Recommended land use changes (%) as related to to CLC classes

If the oLUP recommendations for land use change were implemented throughout Kosovo by
municipality planning provisions , the country’s agricultural sector would be both more
prosperous and more environmentally sustainable. The implementation of proposed land use
changes should be encouraged by new or adapted MAFRD policies that shall be in line with
EU CAP and environmental policies. The practical implementation of oLUP should be the
responsibility of local planners, municipal agricultural specialists, Rural Advisory Service, and
ultimately – farmers and agricultural businesses.

An EU funded project managed by Implemented by:

European Commission Liaison Office
 14

5 LAND USE INTENSITY PLAN (LUIP)

The second stage of RLMP framework is the LUIP, which aims to identify the most
appropriate methods and intensities of land management to help decision makers to
strategically guide agricultural and forestry inputs and businesses. The LUIP conceptual
model identifies 5 Zones of land management (Fig. 1 Annex 1) - 3 for agricultural land and 2
for forest/semi-natural areas.

In principle, these are very broadly defined land management/intensity types, as the currently
available data for all Kosovo on soil, climate and agronomic parameters did not allow the
identification of more precise or better defined land management classes. Ideally, agriculture
practices or crop types (i.e. wheat, vines, orchards, vegetables, etc.) would be recommended.

The same arguments stand for forestry/nature management types. Only two broad
management types were identified due to the lack of specific forestry data covering all
Kosovo. On the other hand, it was not the task of the project to consider in great detail the
forestry sector. The most significant achievement of LUIP is that areas suitable for
commercial and bio-energy forests were identified, which is the first such assessment for
Kosovo.

5.1 Structure, rules and data

For further modelling purposes the proposed land use changes identified by oLUP are joined
with the current CLC2006 dataset to produce an optimised land use map for Kosovo
(oLUP+CLC). The land use changes recommended by oLUP are made at all relevant levels
in the CLC classification. Wherever oLUP suggests that Heterogeneous or permanent crops,
pasture or arable land as the optimised land use, these classes are included in the CLC
Class 2. The area recommended for change to non-agriculture, however, is included in the
CLC Class 3. The new colours are assigned to the recommended land use changes that are
consistent with the CLC2006 colour pattern. The optimised land cover map is used as the
basis for the production of LUIP.

The structure of the LUIP conceptual model is much simpler than the corresponding model
for oLUP, and comprises one decision tree (DT), which has two branches relating to the
relevant CLC 1st level classes:
 Class 2 – Agricultural areas,
 Class 3 – Forest and semi-natural areas.
The model has been built from data derived from the Land Assessment models. The
Agriculture class and Forest/nature classes of integrated oLUP+CLC map have different
modelling structure and different data as noted below:
1. Agriculture classes
o Integrated Land Suitability (ILS),
o Estimated (modelled) flood risk (FLR),

An EU funded project managed by Implemented by:

European Commission Liaison Office
 15

o Irrigated areas (IRA),

o Slope.
2. Forest/nature classes
o Grouping of CLC 3rd level classes,
o Elevation,
o Agricultural Land Suitability (ALS),
o Soil Erosion Sensitivity (SES),
o Estimated Biodiversity Value (EBV).
CLC Class 1 (Artificial surfaces and Urban) and Class 5 (Water) receive no further
assessment in LUIP and are unchanged from the oLUP+CLC map and used directly in the
LUIP.

Following extensive GIS data analysis, data were divided into suitable classes for use in the
LUIP DT. The rules for the DT structure are as follows:

 ILS score is used as the main criteria defining land management intensity in
agricultural areas. The classes used to define ILS were identified after thorough GIS
analysis of ILS dataand its combination and interaction with CLC, ELS and ALS data.

 Slope and irrigation are supplementary criteria used to define land management
intensity,

 In forests/natural areas the principal criteria used to identify of forests/nature land

management types are those relating to favourable economic conditions for industrial
forests management (ie land quality, size of plot etc).

5.2 LUIP Decision Tree

The LUIP’s conceptual model is shown graphically in Figure 6 (Annex) and the two major DT
branches for Agriculture and Forestry/nature are discussed below.

5.2.1 Agriculture (CLC Class 2)

The starting point for this DT branch is Class 2 land derived from the ‘oLUP+CLC’ map (Fig. 6
Annex). ILS is the combined assessment that incorporates land quality (ALS) and
environmental sensitivities (ELS). ILS is calculated by summing the scores for ALS and ELS.
This provides ILS scores ranging from 0 to 200, though very high and low values do not
occur. The highest ILS scores are given to good quality agricultural land where
environmental limitations are small, whereas low ILS scores are given to poor quality land
with many environmental limitations. Land with an ILS score >145 is interpreted as the best
land for agriculture. Where this land occurs on slopes <3 0 it is assumed to be suited to
‘Intensive agriculture or horticulture’. However, where slopes are >30 the land is suited to
‘Moderate agriculture’.

Agricultural land with intermediate ALS score (120-145) is identified as land for ‘Moderate
agriculture’ unless it has access to irrigation, in which case it is identified as suited to

An EU funded project managed by Implemented by:

European Commission Liaison Office
 16

‘Intensive agriculture or horticulture’. The recommendation termed ‘Moderate Agriculture’

contains a broad range of land quality, locally affected by marked environmental limitations.
These limitations may be soil erosion, vulnerable groundwater or restricted biodiversity. It
can include pockets of intensive arable cultivation but environmental limitations restrict its
flexibility. Much of this land is under annual and permanent crops and there can be small
areas of natural vegetation.

Land with and ALS score <120 is suitable for ‘Extensive agriculture or livestock’. This
includes annual and permanent crops grown with low inputs of fertilisers and probably
achieving low yields in association with pasture supporting livestock and maybe interspersed
with areas of natural vegetation.

5.2.2 Forest and semi-natural vegetation (CLC Class 3)

The starting point for this DT branch is Class 3 derived from the ‘oLUP+CLC’ map). The main
purpose of this DT branch is to identify areas within these land cover types that are suitable
for ‘Commercial or Bio-energy Forests’. As Kosovo citizens are heavily dependant on wood
fuel, an increased bio-energy production capability would economically benefit local people,
as well as helping Kosovo comply with EU energy policies and the UNCCCF Kyoto
agreement in due course.

If forestry enterprises are to be economically successful for commercial and bio-energy

forests several land criteria must be met. The land should be of average or better quality
(ALS Class 1-5), climate should not restrict tree growth and forests must be close to
populated areas (elevation below 800 m), soil erosion should not be a major issue during the
establishment and harvesting of forests (SES score <60). Biodiversity value should be low
(<65) so that disturbance during harvests is not damaging valuable nature resources. When
all these stipulations are met the land is suitable for Commercial exploitation or the planting
and maintenance of fast growing tree species.. The LUIP then ascribes all remaining areas
of Class 3 land to ‘Sustainable forestry or nature management type’.

5.3 LUIP results and recommendations

Seven distinct land types are identified by the LUIP model. They all have a different
significance to planning decisions. The land types and their significance are as follows.

Land type 1 – Intensive agriculture or horticulture; probably the most important separation
within the LUIP framework. This is intensive arable and horticultural land that should be
strictly protected for agricultural use. It represents the best quality land within Kosovo that
has manageable environmental limitations. As long as a high standard of land management
is carried out there is likely to be minimal environmental detriment from this land use. These
areas are suitable for vegetable production and other intensive types of agricultural
industries, depending on local conditions, accessibility and soil parameters.

Land type 2 – Moderate Agriculture; this contains a broad range of land quality. Generally
where land quality is high there are unsustainable impediments to arable cultivation and one
or more of erosion, groundwater or biodiversity limitations cannot be mitigated by simple land

An EU funded project managed by Implemented by:

European Commission Liaison Office
 17

management techniques. In most cases land quality is average and there is often some form
of environmental sensitivity. This land supports an often complex mixture of annual and
permanent crops, natural vegetation and localised areas of intensive arable cultivation. This
represents valuable agricultural land which should be protected where possible but in some
municipalities development will be unavoidable. In these cases the Rural Environment Action
Plan (Palmer and Nordin 2011d) should be consulted to identify areas of Land type 4 with the
greatest environmental limitations for agricultural as the best areas to be sacrificed to
development. Depending on local conditions and soil quality, this land type is mostly suitable
for mixed agricultural use, wheat and other crops.

Land type 3 – Extensive agriculture or livestock; these are agricultural areas often
interspersed with areas of natural vegetation. They mainly consist of annual and permanent
crops generally achieving low yields, together with pasture supporting livestock and natural
vegetation. Some of these areas will be steeply sloping, which is likely to increase the costs
involved in development, but nevertheless this is the agricultural land that should be
developed if better quality land is to be preserved. These areas are mostly suitable for
producing hay, livestock, honey cultivation and other alternative low-input agriculture.

Land type 4 – Commercial or energy forestry; illegal felling of trees for fuel must be
controlled. If this is to happen sustainable forests must be developed from which future
needs can be resourced. This land represents areas where trees will grow well and there are
no environmental risks associated with growing and subsequently felling trees. There should
be a presumption against development or mineral extraction in these areas. Some State
forests already have good quality forest management plans, but an indication of this Land
Type should be integrated into existing forest plans. Existing trees and shrubs could be also
used in special bio-energy plantations depending on local climate, accessibility and soil
conditions.

Land type 5 – Sustainable forestry or nature management; these are the areas that have
one or more environmental constraints require the current land use (forests or nature) to be
maintained. ,It is imperative that Municipality Planners are vigilant in preserving potential
sites for NATURA 2000 to ensure that the EU Habitats Directive requirement for Special
Areas of Conservation and the EU Birds Directive can be implemented in future in suitable
parts of Kosovo.

The LUIP plan is used as the basis for developing the Rural Environmental Action Plan
(rEAP), which should be read jointly with the LUIP plan. rEAP provides more detail
information on the necessary environmental actions and measures that ensure sustainable
agriculture and forestry is possible without damaging the environment.

6 LIMITATIONS OF oLUP AND LUIP

Inevitably there are limitations to these Plans but it is important to stress that they are
designed as simple broad frameworks to help Municipality Planners integrate land physical
properties into rural spatial planning decisions and, in general, into the municipal

An EU funded project managed by Implemented by:

European Commission Liaison Office
 18

development plan. Each of the RLMP models can be further updated and enhanced
depending on the resources of individual municipalities, but they could be also used as they
are. However,in the absence of research and more precise data, these plans, as developed
by EULUP, still provide a big step forward in the sustainable planning of rural developments.

The outputs from oLUP remain unvalidated as it would require a concerted programme of
field investigation to review the validity of a reasonable proportion of the land use changes
proposed by oLUP. The system suggests that <12 per cent of Kosovo requires change to
achieve optimized land use, which appears realistic.

The land use management types identified by LUIP have been to some extent reviewed in
the field by N.Nordin, mainly in Southern-Western municipalities. In fact LUIP model outputs
were tweaked to better fit real conditions on the ground following field observations.

All the recommendations made within the oLUP and LUIP plans are based on the spatial data
that are available nationally for Kosovo and that was used to develop the models and
datasets that form the Basis for Rural Spatial Planning (BRSP). These data are often
inadequate but will remain the only data sources for the foreseeable future. Their
deficiencies can only be overcome by long-term (5-10-year) investment. The main limitations
are:

 The poor quality of most of the existing data severely restricts the accuracy and
flexibility of any results: hence there is a strict limit of 1:50,000-scale that has to be
imposed on the plotting of any results.

 The results will not be valid for investigations of environmental conditions in areas of
less than about 25 ha. Where greater detail is required there is no alternative but to
carry out a site investigation. However, if better soil data becomes available in future
the models can be readily updated to provide more detailed and reliable assessments.

 Water quality monitoring is urgently needed to check the state of Kosovo’s

groundwater reserves. Once these data are available they can be fed into the
Groundwater risk assessment to improve the quality of this model.

 The estimated biodiversity value model is provisional and based on very little
measured data. If further relevant information becomes available describing Kosovo’s
natural environment it may be necessary to review the EBV model in the light of the
scope of new data.

7 CONCLUSIONS

This pragmatic framework is provided as a basis for compiling RLMPs at municipality scale.
The LUIP provides a simple 5-category classification of land use, which can be used directly
by Municipality Planners. Agricultural land is divided into 3 categories of differing intensity
based on land quality, environmental sensitivity and landscape. Non-agricultural land is given

An EU funded project managed by Implemented by:

European Commission Liaison Office
 19

just two categories with land suitable for Commercial or bio-energy forestry separated from
those only suitable for – sustainably managed forests or nature. Water bodies and artificial
and urban areas were excluded from analysis and mapped directly from CLC2006 data.

RLMP is only designed as a framework, which can be built on by local Planners with local
knowledge and understanding. The principal advantages of implementing this framework in
all Kosovo are that this framework is:

 an Integration of land physical properties in the form of land assessments into rural
spatial planning and decision making.

based on existing data that are available at the same scale of resolution for the whole of
Kosovo;

usable at the scale required for Municipality planning (1:50,000 scale);

implemented in GIS, therefore easily applied and the simple clear methodology can be readily
updated as new information becomes available, either nationally or from the Municipality;

ready for use by Municipality Planners now;

a consistent and reliable system that will allow comparisons between Municipalities;

clearly described in explanatory Technical Reports and GIS User Manuals

capable of wide dissemination and implementation throughout Kosovo and requires no further
input from soil scientists or geologists.

RLMP data and maps could be very beneficial to forming national agriculture and
environment protection policies, as they cover all Kosovo. Coupled with statistical and
economic data the RLMP plans could serve as initial data for correctly adopting EU legislation
in Kosovo.

Kosovo is changing incredibly rapidly with examples of non-coordinated development

occurring on good quality agricultural land. Agriculture is and will remain a vital part of the
Kosovo economy and the protection of good quality land, which is a major objective of this
RLMP framework, is required now, not in 2 or 5 or 10 years time. The implementation of this
type of RLMP (oLUP, LUIP, rEAP and CZP) into the Kosovo spatial planning system should
provide a quick, robust, but reliable, solution to regulate irreversible changes in land use that
can potentially threaten Kosovo’s ability to function as a sovereign state.

8 REFERENCES

CORINE Land Cover (2006). Joint Research Center, European Environmental Agency, EU.

Ducke, B (2011) GIS Technical Documentation, EULUP Technical Report 16, Nov. 2011

An EU funded project managed by Implemented by:

European Commission Liaison Office
 20

Palmer, R.C. and Nordin, N (2011) Agricultural Land Suitability Classification, final version.
EULUP Technical Report 3.

Nordin, N, and Palmer, R.C. (2011) Environmental Land Sensitivity Assessment. EULUP
Technical Report 8.

Palmer, R.C. and Nordin, N (2011) Rural Environmental Action Plan. EULUP Technical
Report 10.

An EU funded project managed by Implemented by:

European Commission Liaison Office
 21

ANNEX

Figures

Fig. 1 RLMP Decision Tree

Fig. 2 oLUP Decision Tree for Arable and Permanent Crops

Fig. 3 oLUP Decision Tree for Pastures

Fig. 4 oLUP Decision Tree for Heterogeneous agriculture

Fig. 5 oLUP Decision Tree for Forests, natural grassland etc

Fig. 6 LUIP Decision Tree

An EU funded project managed by Implemented by:

European Commission Liaison Office