Tree Planting Zones in Nepal - an ecological approach based on vegetation types

Full text

DFSC Case Study No. 1
TISC Technical Paper No. 103
December 2001
Tree Planting Zones in Nepal
- an ecological approach based on vegetation types
by
Jens-Peter Barnekow Lillesø, Lokendra P. Dhakal, Tirtha B. Shrestha, Rathna P. Nayaju, Rabin Shrestha,
and Erik Dahl Kjaer
Tree Improvement and Silviculture Component (TISC) of Department of Forest, HMG/Nepal Ministry
of Forest and Soil Conservation/Natural Resource Management Sector Assistance Programme
(NARMSAP) supported by Danida
and
Danida Forest Seed Centre

Citation:
Lillesø, J.P.B., Dhakal, L.P., Shrestha, T.B., Nayaju, R.P., Shrestha, R. and Kjaer, E.D. 2001. Tree Planting Zones in
Nepal - an ecological approach based on vegetation types. DFSC Case Study No.1. TISC Technical Paper No. 103.
Danida Forest Centre, Humlebæk. Tree Improvement and Silviculture Component, Kathmandu.
The publication can be requested from:
Danida Forest Seed Centre
Krogerupvej 21
DK-3050 Humlebaek
Denmark
Tlf: +4549190500
Fax: +4549160258
Email: dfsc@dfsc.dk
Web Site: www.dfsc.dk
Tree Improvement and Silviculture Component
Hatisar, Naxal
Kathmandu
Nepal
P O Box 6055
Tlf: +977 01 434504/437784
Fax: +977 01 434546
E-mail: tisc@mos.com.np
Cover photo: Landscape with terraces in Kathmandu highland, Middle Hills, Nepal. Agriculture on terraces makes
effective use of the steep mountain sides. H. Keiding phot. 1992.
HMG/Danida Tree Improvement and Silviculture Component (TISC) is one of the Components of His Majesty’s
Government of Nepal and Danida Natural Ressource Management Sector Assistance Programme (NARMSAP). The
TISC main focus is on: 1) basic and advanced tree domestication; and 2) participatory silviculture. Community/ far-
mers’ organisations are encouraged in the protection and management of important farmland and forest tree species
where TISC facilitate mostly in organisation building and technical backstopping.
Danida Forest Seed Centre is a Danish non-prot institute, which has been working with development and transfer of
know-how in management of tree genetic resources since 1969. The development objective of DFSC is to contribute
to improve the benets of growing trees for the well-being of people in developing countries. The programme of
DFSC is nanced by the Danish International Development Assistance (Danida).
ii

Trees are important to people in Nepal; for timber,
rewood, household articles etc. and not least for
food security, because livestock in Nepal to a
large extent feed on tree-fodder, and because the
major source of fertiliser in agriculture comes from
animal dung and branches cut from trees.
Many tree species, and in particular fodder trees,
are planted by farmers in Nepal. It is also expected
that the demand for many species will increase
both among the Forest User Groups for enrich-
ment plantings and for reforestation by the poorest
segments of society e.g. through the Hills Lease-
hold Programme. Priority species for different uses
vary substantially depending i.a. on ecological fac-
tors.
Choice of species should rst of all reect the
tree planters’ priorities, but a good match of plan-
ting material to planting site will improve the bene-
t that the farmers obtain from growing the trees.
The correct choice of seed source may constitute
the difference between success and failure when
trees are planted. The use of the most appropriate
seed source compared with a more or less random
source may easily increase the value production
with 20-50 % or more.
Successful tree planting requires a certain degree
of ecological and technical skills: the species should
be planted at a site where it is able to survive and
grow, and it should be planted and nursed in an
appropriate way depending on the species’ require-
ments and the use of the trees.
Given the extraordinarily large ecological varia-
tion in Nepal, special attention needs to be given
to ecological considerations. This goes for choice
of species for a given site, but also for choice of
seed sources for a given species. Different species
and seed sources will be best at different planting
sites and the challenge is therefore to match species
and seed sources to planting site.
A zonation system, where zones of similar eco-
logical conditions are identied, can be a power-
ful tool to assist in providing a good match of
species and provenances to planting site, but also
to enable the planning of seed collection, seed
source development and genetic resource conser-
vation.
The present document describes development of
such a zonation system, here called a ’tree planting
zone system’ for Nepal. The principal purpose of
the system is to support provision of better plan-
ting material to tree planting farmers in Nepal, and
thereby improve their benets from domestication
of tree species.
The document provides documentation and
explanation of the system for professional staff
involved in tree domestication and tree seed pro-
curement in Nepal, but may also serve as inspira-
tion for professionals involved in management of
genetic resources in other countries.
The tree planting zone system is together with
the GIS system of the vegetation of Nepal mainly
intended as a planning tool and as a starting point
for preparation of extension material for eld use.
The planting zones may form a framework that
NARMSAP, and hopefully all other organisations
involved in tree seed supply and tree planting, can
use for planning and implementation of supply
and utilization of planting material in Nepal.
The tree planting zones in Nepal has been develo-
ped as a joint effort between TISC and Danida
Forest Seed Centre within the overall framework
of the collaborative natural resource management
programme between HMG/Nepal and Danida.
K. R. Shrestha
Chief TISC
Preface
iii
L. Graudal
Director DFSC

Executive Summary
A substantial number of species (especially fodder
trees) are planted by farmers and it is likely that
the demand for many species will increase in the
future. Given the extraordinarily large ecological
variation in Nepal, special attention should be
given to ecological considerations. Different spe-
cies and seed sources will be best at different plan-
ting sites, and the challenge is therefore to match
species and seed sources to planting site. Well-
adapted seed sources may ensure a reliable yield for
farmers, while maladapted seed sources may result
in loss - or even total failure.
In the present ’Tree Planting Zones’ planting sites
with similar environmental conditions are grouped
together into zones for which specic seed sources
can be developed and thereby increase farmers’
planting success. The ’Tree Planting Zones’ can be
recognised in the eld by farmers and will be uti-
lised where there is the greatest potential for seed
demand.
The distribution of vegetation types is the best
indicator of growing conditions for trees and
bushes. ’Tree Planting Zones’ have been develo-
ped in the warmer ecological zones where most
planting by farmers take place and consequently
where most seed transfers are required. In the
colder ecological zones transfer of seeds for the
relatively limited planting should be avoided and
seed should be collected at the planting site.
Planting Zones in Nepal
Lower Tropical Ecological Zone
Sal Forest: semi-evergreen forest type, monsoon forest type
Upper Tropical Ecological Zone
Valley Systems: wet central valley, eastern tropical valleys, Sun Koshi tropical valley, western valleys
Siwaliks: eastern wet Siwaliks, western dry Siwaliks
Sub-Tropical Ecological Zone
Upper and Lower Schima-Castanopsis Forest: eastern S-C; central S-C; western S- C
Upper and Lower Chir Pine-Broadleaved Forest: eastern C-P-B; western C-P-B
Chir Pine Forest: Chir Pine Forest
Not all zones will be relevant for all species, because
only very few species (such as e.g Alnus nepalensis)
are relevant for planting across all zones (cf. Sec.
3.3, below).
The system of ’Tree Planting Zones’ will be part
of a tree seed distribution system that will contri-
bute towards the improvement of the living condi-
tions for the poor farmers (smallholders) in Nepal.
As increased understanding is gained, it will
become relevant to develop more species specic
recommendation for establishment and use of seed
sources within Nepal. However, as long as a decen-
tralised seed distribution system is developing, the
planting zones based on vegetation types will pro-
bably remain the backbone for the tree seed sector
in Nepal.
iv

1. Ecological Zonation of Planting Sites for Improved Benet 1
1.1 It begins with genotype by environment interaction 1
1.2 The question of local optimality 4
1.2.1 When is a seed source local? 4
1.2.2 Are local seed sources the best choice? 4
1.3 Seed zones: a concept with several meanings 5
1.3.1 Seed procurement zones: geographic areas with uniform seed sources 5
1.3.2 Seed application zones: areas with fairly uniform planting sites 6
1.3.3 Seed application zones (tree planting zones) common to all untested species 7
1.4 Elaboration of tree planting zones in Nepal - environmental factors to be considered 8
2. Climates and Vegetation in Nepal 10
2.1 Climates in Nepal 11
2.1.1 Precipitation, potential evapotranspiration and dry/wet seasons 12
2.2 Distribution of vegetation in Nepal 14
2.2.1 Alpine, Sub-Alpine and Trans-Himalayan Ecological Zones 16
2.2.2 Temperate Ecological Zone 18
2.2.3 Subtropical and Tropical Ecological Zones 19
3. Delimiting Planting Zones in Nepal 25
3.1 Tropical planting zones 25
3.1.1 Lower Tropical Ecological Zone (Sal and Mixed Broad Leaved Forest) 26
3.1.2 Upper Tropical Ecological Zone (Hill Sal Forest) 27
3.2 Sub-tropical planting zones 28
3.2.1 Schima-Castanopsis Forest 28
3.2.2 Chir Pine-Broad Leaved Forest 30
3.2.3 Chir Pine Forest 31
3.3 Examples by species 32
4. Application of the Planting Zones in Nepal 35
4.1 Guidelines for seed transfer within and between seed zones 35
4.2 Seed supply, seed source development and seed source description 35
4.3 Indigenous knowledge and revision of planting zones as increased
experience is gained 36
5. Conclusions and Recommendations 36
References 37
Contents
v

Figures
Figure 1. Genotype x environment interaction - Growth in three environments.
Figure 2. The performance of a seed source across site type values is described by three parameters, which characterise
the seed source.
Figure 3. Distribution of settlements in relation to Ecological Zones and Physiographic Regions
Figure 4. Floristic provinces of Asia
Figure 5. Moisture patterns in Nepal: described by moisture regions
Figure 6. Distribution of Alpine Ecological Zone in Nepal
Figure 7. Distribution of Sub-Alpine Ecological Zone in Nepal
Figure 8. Distribution of Trans-Himalayan Ecological Zone in Nepal
Figure 9. Distribution of Temperate Ecological Zone in Nepal
Figure 10. Distribution of Sub-Tropical Ecological Zone in Nepal
Figure 11. Distribution of vegetation types in Subtropical and Tropical Ecological Zones and length of dry season
Figure 12. Distribution of Upper Tropical Ecological Zone in Nepal
Figure 13. Distribution of Lower Tropical Ecological Zone in Nepal
Figure 14. Relation between forest vegetation and total rainfall (ordinate) as well as length of dry period in months
(abscissa) in India.
Figure 15. Assumed relation between forest vegetation and rainfall (ordinate) as well as length of dry
period in months (abscissa) in Upper and Lower Tropical Ecological Zones in Nepal.
Figure 16. Planting zones of Lower Tropical Sal and Mixed Broad Leaved Forest type in Nepal
Figure 17. Planting zones of Hill Sal Forest type in Nepal
Figure 18. Planting zones of Schima-Castanopsis Forest type in Nepal
Figure 19. Planting zones of Chir Pine-Broadleaved Forest type in Nepal
Figure 20. Planting zones of Chir Pine Forest type in Nepal
Figure 21. Bauhinia purpurea distribution in vegetation types in Nepal (for detailed zonation, see table 7)
Figure 22. Brassaiopsis glomerulata distribution in vegetation types in Nepal (for detailed zonation, see table 7)
Figure 23. Ficus neriifolia distribution in vegetation types in Nepal (for detailed zonation, see table 7)
Figure 24. Garuga pinnata distribution in vegetation types in Nepal (for detailed zonation, see table 7)
Table 1. Guidelines for maximum transfer of seed sources of interior natural stands of some North American conifers
Table 2. Classication of patterns of seed source variation in eucalyptus species
Table 3. Thermal and moisture proles of the ecological zones
Table 4. Temperature extremes (°C ) by ecological zones
Table 5. Number of vegetation types per ecological zone
Table 6. Mapped vegetation types in Nepal; number of species in tentative lists; % of Nepal’s area
Table 7. Distribution of some priority species in planting zones
vi
Tables

1
ECOLOGICAL ZONATION OF PLANTING SITES FOR IMPROVED BENEFITS
1. Ecological Zonation of Planting Sites
for Improved Benets
Trees are important to people in Nepal, both for
timber, rewood and household articles. However,
trees are also of major importance for human food
security, because the livestock to a large extent feed
on tree-fodder, and because the major source of fer-
tiliser in agriculture comes from bedding based on
manure and branches cut from trees (LSMP, 1993).
Tree planting programmes have been carried out on
thousands of hectares over the past decades based
on fairly few species (Jackson, 1994). In recent years
large-scale plantation establishment has been largely
abandoned and instead the community forestry
programmes are increasingly focused on manage-
ment of natural forest. Nevertheless a substantial
number of species (especially fodder trees) are still
planted by farmers and it is likely that the demand
for many species will increase in the future for enri-
chment planting in the natural forest held by Forest
User Groups and for reforestation by the poorest
segments of society through the Hills Leasehold
Programme (Yadav and Dhakal, 2000). Unlike indu-
strial trees where a few species dominate plantings,
farmers utilise a large number of species, and prio-
rity species for different uses vary substantially bet-
ween villages depending i.a. on ecological factors
(Lillesø et al. 2000).
Successful tree planting requires a certain degree
of ecological and technical skills: the species should
be planted at a site where it is able to survive and
grow, and it should be planted and nursed in an
appropriate way depending on the species require-
ments and the use of the trees. Given the extra-
ordinarily large ecological variation in Nepal, spe-
cial attention should be given to ecological consi-
dera- tions. This goes for choice of species for a
given site, but also for choice of seed sources for
a given species. Different species and seed sources
will be best at different planting sites, and the chal-
lenge is therefore to match species and seed sources
to planting site. Choice of species will (and should)
rst and foremost reect the farmers’ priorities, but
a systematic approach that supports good match of
planting material to planting site will improve the
benet that the farmers obtain from growing the
trees.
The objective of the present document is to
develop a tree planting zoning system that in gene-
ral can support choice of species and seed sources
in Nepal. Such a system is often referred to as seed
zones or breeding zones (cf. below), but we prefer
to use the term ‘tree planting zones’. This is because
the purpose of the system is to support provision
of better planting material to tree planting farmers
in Nepal, and thereby improve their benets from
domestication of tree species. The planting zones
shall form a framework that NARMSAP, and in
principle all other organisations involved in the
tree seed sector in Nepal, can use for planning and
implementation of supply with planting material.
An important aspect of the system is to nd the
level of complexity that on the one hand eases ef-
cient use, but on the other hand integrates and
handles the extreme ecological complexity that
characterises Nepal. As discussed below, a too com-
plex system will not work and will therefore have
no impact, whereas a too simple system may actu-
ally counteract good match between seed sources
and sites rather than supporting it. Obviously,
this ‘best trade off’ in complexity depends on the
future use of the zonation system, and the applica-
tion of the system must therefore be kept in mind
in the development phase. Also, it is important to
outline how, and for which purposes, the suggested
planting zone system can be used, and how and for
which purposes it cannot be used.
For Nepal, the objective of the present planting
zone system is to form a framework for:
Development of guidelines for procurement
of seed, including site specic seed source
recommendations
Evaluation of seed source availability in relation
to tree planting activities
Support to domestication of priority species
Support to choice of species for planting out-
side their natural range
These objectives must therefore be kept in mind in
the development phase (cf. chapters 2 - 4 below).
In chapter 1, a few general aspects are rst discus-
sed in relation to these objectives.
1.1 It Begins with genotype by environment
interaction
Many of the tree species planted in Nepal have
large natural distribution areas, where they grow
across strong environmental gradients. Only little
is known about the genetic variation between such
tree populations within Nepal. However, the gene-
ral experience is that important genetic variation
is most likely to exist between populations for
plant species that cover large and/or environmen-
tally heterogeneous distribution areas. Studies from
other parts of the World - tropical, subtropical and
temperate - have conrmed this picture for a large
number of tree species, and revealed genetic vari-
ation in key quantitative traits. Both for survival
and growth rate in plantings, but also quantity and
quality of many types of end products (Mouna,
1990; Mandal and Gibson, 1998; Zobel and Jett,

2
Figure 1. Genotype x environment interaction -
Growth in three environments.
A: Interaction of practical importance.
B: Interactions not of practical importance (from
Matheson and Cotterill, 1990).
1995). In other words, growth and development of
planted trees depend on the origin of the seed used
for raising the seedling.
Selection of the most appropriate seed source
can therefore signicantly increase tree production
as compared to random seed sources, while a bad
seed source may result in loss - or even total failure
if the seed source is so maladapted to the given site
that the tree does not survive.
Identication of the best seed source is unfor-
tunately not an easy matter, because the ‘optimal
seed source’ depends on the environment in which
it shall grow. This is referred to as genotype by en-
vironment interaction (in short GXE), and from a
practical point of view it means that the seed sour-
ces change rank when compared in different en-
vironments (A in Fig. 1). What is best in a given
test environment may be inferior at another, and
seed sources therefore have to be selected speci-
cally depending on the environmental features of
the planting site.
From an application point of view, it is the pre-
sence – or potential presence – of GXE that leads
to a need for zonation of planting sites for efcient
seed procurement and domestication. In the pre-
sence of GXE, one cannot identify and use a uni-
versal, overall best, seed source for planting, but
one has to identify ‘the best seed source for a given
type of sites’. A seed source (or a group of similar
seed sources) will have a range of environments
within which it is possible to realise a high and
reliable yield, but use of the seed source beyond
these environments will in principle result in
lower yield due to poor adaptation. Tree planting
zoning therefore attempts to separate the planting
sites into a number of zones, which can form the
basis for operational procurement of good (or even
‘best’) seed sources for all potential planting sites
within the commercial planting areas for the spe-
cies. Without GXE there is no need for a zoning
system from a seed procurement and domestica-
tion point of view, the same seed sources can in
principle be procured and used everywhere (see B
in Fig. 1).
GXE reects that genotypes have different
‘response functions’, i.e. they respond differently
to different environments. One seed source may
for example grow relatively best on semi-dry sites,
whereas another seed source has it optimal perfor-
mance on wetter sites. The‘response functions’ of
a seed source is not only a question of where it
has its optimum, but certainly also how much is
lost when grown at other sites (cf. Fig. 2). In gene-
ral, the atter the seed source curves are (larger C
cf. Fig 2); the larger (fewer) planting zones can be
developed.

3
ECOLOGICAL ZONATION OF PLANTING SITES FOR IMPROVED BENEFITS
Figure 2. The performance of a seed source across site type values is described by three parameters, which characte-
rise the seed source. (A) = optimum performance of a seed source at site requirement value (B); (C) = exibility of
a seed source, equivalent to extent of the zone within which the seed source achieves above 80% of its optimum
performance. The exibility is thus the range of sites where a seed source has an acceptable performance (in this case
80% of the optimum) (from Lindgren and Ying, 2000).
Response functions and deployment strategies
The response functions (how individual seed sour-
ces perform in different environments) may be quite
differently shaped for different seed sources. Some
seed sources may be quite exible (at curves, large
C), while others may be much more selective in
environmental preferences (steep curves, small C).
Traditionally, a seed source is therefore character-
ised by two parameters (i) its average performance
over a number of sites, and (ii) how stable the per-
formance is from site to site (cf. e.g. Finley and
Wilkinson, 1963; Matheson and Raymond, 1984).
Based on this, it is common to separate seed sources
into three groups:
Seed sources that on average perform well, and
in general does so everywhere (exible, only
little GXE).
Seed sources that on average are of inferior per-
formance, and in general perform poorly
everywhere (exible, only little GXE).
Seed sources that perform well or very well
on quite specic sites, but are quite poor on
others (high GXE).
‘Type II’ seed sources do exist, and will in general
be of limited interest for planting programmes. The
choice between‘type I’ and ’type III’ can be more
complicated, and reects two different ‘deploy-
ment strategies’. Application of ‘type III’-seed sour-
ces will require relatively smaller seed zones and
involve higher risks for ‘mis-match’, but they may
be able to use the full potential of different plan-
ting sites.‘Type I’ seed sources are easier (and safer)
to apply, but may not utilise the full potential of
the planting sites.
In the context of GXE-analysis, the term ‘geno-
type’ is used in a very broad sense to cover both
clones (‘true’ genotypes), families or even seed
sources. Although generally accepted, it is a pro-
blematic generalisation because the degree of GXE
may indeed be different for genotypes, families
and seed sources. In principle, one would expect
more GXE for clones than for families, and more
GXE for families than for seed sources. This is
because family performance is an average of seve-
ral genotypes, and seed source performance is ave-
rage of several families. One can decide to select
and use a mixture of improved clones with large
exibility (large C, cf. Fig. 2), either each clone by
itself or in mixed combination, and such a strategy
can lead to a low number of required seed zones.
The importance of GXE for a given species and
set of sites will therefore depend on the type of
applied planting material. GXE is obviously not
only a function of the diversity of planting sites
(environments), but also the applied planting mate-
rial. The size (and number) of planting zones for a
given species should therefore also reect the exi-
bility of the applied seed sources (their C-factor, cf.
Fig. 2).
I.
II.
III.

4
1Actually, elevation is all
taken into account by cal-
culating a so-called equiva-
lent latitude = latitude*
+ (elevation-300m)/300m
(Lindgren & Ying, 2000,
p.96).
1.2 The Question of Local Optimality
1.2.1 When is a seed source local?
Given the huge ecological diversity of Nepal with
corresponding risk of using maladapted seed sour-
ces, one must consider the option of using local
seed sources, i.e. seed sources that have grown close
to the planting site for many generations. They are
therefore expected to be adapted to the site or at
least the long-term performance can be predicted.
In practice the rst problem is to dene ‘local’ in
a practical context. When is a seed source close to
the planting site in terms of ecological similarity?
Of course, a seed source next to the planting site
will be ‘local’, but what if the nearest seed source
is 50 kms away, or 150 kms? In general, this will
depend on the species distribution at the landscape
level (scattered or continuous, single trees, clusters
or large stands), its degree of genetic isolation
(pollen and seed ow) and environmental hetero-
geneity in the region. Species can be placed in
three broad groups (partly based on Eriksson et al.
1993):
A single population: all individuals of a species
jointly form one single large, random mating,
population growing in a range where ecological
variation is small and pollen ow is substantial.
Here, one would expect only limited genetic
differentiation between populations within the
natural range (there is basically only one popu-
lation). In this simple situation, we suggest all
seed sources from the natural range of the spe-
cies to be ‘local’ to all planting sites within the
natural range of the species.
Continuous distribution: trees are distributed
more or less continuously over a large area that
covers substantial ecological variation. Seed and
pollen ow takes place and homogenises gene-
tic variation over small to intermediate distances,
but natural selection has generated a large scale
genetic structure that forms continua (clinal vari-
ation) in which genetic differentiation among
individuals is inuenced by separation distance.
There may be more than one cline, in which case
the variation becomes more multidimensional.
Looking at more than one trait can complicate
things further. Here, we suggest a seed source as
‘local’ to the planting site, if it originates from
seed sources that is not too far from the planting
site – not in terms of geographic distance, but
in terms of difference in the clinal gradient(s).
Further, we consider a seed source to be pseudo-
local to a planting site, if it originates from a site
that is similar to the planting site in terms of the
ecological gradient, even if it is (very) far away
from the planting site. This may even be the case
if the planting site is outside the natural distri-
bution range of the species, however, this will
require that the gradients can be identied and
quantied for each seed source.
More complicated patterns consist of ‘island
type’ of separated populations. These may differ
genetically from each other both because of
genetic drift (random processes combined with
genetic isolation) and selection. In some spe-
cies, island populations exchange genes mostly
with adjacent colonies rather than equally with
all colonies. Exchange of genes may follow a
‘stepping-stone model’ where gene movement
occurs in steps from population to population,
but often gene ow follows a more complex
web structure often referred to as meta-popula-
tions. Such structure may exist in one, two, or
possibly three spatial dimensions, but the pat-
terns are much more difcult to predict. Here,
it is probably only seed sources geographically
close to the planting site that can be considered
to be ‘local’, i.e. seed sources that are part of the
sub-population that covers the area in which the
planting site is located.
Even for species with a continuous distribu-
tion (‘type 2’), one still needs to ask what is
‘ecologically close enough’ in terms of both
local and pseudo-local? This will depend on the
genetic exibility of a species (discussed above,
‘C-factor’ cf. Fig. 2), and therefore also be spe-
cies specic. The broader the exibility, the
more seed sources can be considered ‘local’. So,
what is ‘local’ will depend on the species, even
in a given environmental range.
1.2.2 Are local seed sources the best choice?
Local seed sources have grown in environments
that are similar to the planting site. In principle,
the local seed source can therefore through natu-
ral selection have developed into a genetic resource
particularly suited for the planting site. Haman et
al. (2000) present an example of local adaptation.
They demonstrate through analysis of three prove-
nance tests within the area of natural distribution
(British Columbia) for red alder (Alnus rubra) that
the best performing families almost exclusively ori-
ginate from the same region where planting sites
were located. However, the performance cannot be
described by altitude. Haman et al. (2000) see this
as adaptation to some kind of regional (geographic)
climate where gene ow among local populations
(may) prevent local adaptation on a ne scale as
red alder is wind pollinated and has minute seeds
disseminated by wind.
There are several other examples from the
Northern Hemisphere showing local adaptation in
the sense that the seed sources perform poorly
if planted far from their places of origin. Langlet
(1936) pioneered in Sweden by showing a close
correlation between latitude of origin, and the per-
formance of Pinus sylvestris. Since then, the relati-
onship has been quantied further, and Lindgren
and Ying (2000) nd that the use of a common
seed source of Pinus sylvestris should not be used
1.
2.
3.

5
ECOLOGICAL ZONATION OF PLANTING SITES FOR IMPROVED BENEFITS
over more than 2 degrees of latitude in Sweden1.
Similar studies of Fraxinus americana (Roberds et al.
1990) and Pseudotsuga heterophylla (Campbell, 1991;
Rehfelt, 1983) in North America also showed geo-
graphical patterns that set limits to transfer of seed
sources.
A local seed source may not always be the best
performer at a given site (Namkoong, 1969; Eriks-
son et al. 1993) because local seed sources may
actually not be the best adapted (natural selection
is subject to many constraints and does not neces-
sarily assure ‘optimality’, cf. Lewontin, 1984). Espe-
cially when economically important traits are taken
into consideration, non-local seed sources may be
able to produce more valuable products. However,
nding an alternative, non-local, better seed source
requires substantial resources and time (range-wide
seed collection, establishment of (maybe repeated)
trials in the specic environments, evaluation and
analysis). The long term performance of local
seed sources is known and non-local seed sources
should therefore in general only be used as alter-
native to viable local seed sources if the choice is
based on solid testing (Kjaer, 1997).
Use of ‘local’ seed sources will therefore not
ensure that the best possible seed sources are used,
but will signicantly reduce the risk of using mala-
dapted planting material, and thereby reduce the
risk that plantings fail to grow and fail to provide
benets to planters.
The majority of tree species used in Nepal have
not been subject to seed source testing. Therefore,
seed procurement should aim at using ‘local’ seed
sources. These ‘local’ seed sources should consist
of a reasonable number of trees, and should not
have been degraded due to human selection or use.
The ‘local’ seed source can be improved through
selection based on genetically sound principles.
It is therefore essential for practical purposes
to be able to decide what is ‘local’ in a given con-
text and this is the major objective of the present
system for zonation of planting sites.
1.3 Seed zones: a Concept with Several
Meanings
So-called seed zones have been proposed in many
countries in order to assist the choice of seed sour-
ces to specic sites. Examples from tropical coun-
tries are from Thailand (Eis, 1986), Uganda (Lillesø
and Kaumi, 1993), Sudan (Aalbæk and Kananjii,
1995), Indonesia (Pilegaard Hansen et al. 2000)
and Burkina Faso (Diallo et al. in prep.). These
examples all represent so-called general seed zone
systems, which means that the seed zones are not
developed for any specic species, but should in
principle apply to all species. Many species-specic
seed zone systems have been developed in tem-
perate areas (see e.g. Campbell, 1991 and referen-
ces herein), but fewer examples are available from
the tropics. Often, but not always, it is recommen-
ded that seeds are not transferred between seed
zones, and often (but not always) movement of
seeds within zones are subject to some restrictions,
mainly in terms of elevation.
The seed zone concept is obviously related to the
planting zones that we seek to establish in the pre-
sent document. However, no common denition
of seed zones is used and various authors apply
different denitions, often without stating what is
actually meant by a ‘seed zone’ in their specic
context. This seems especially to be the case for
seed zones of the ‘general type’ (i.e. non-species-
specic). Lack of clear denitions, and lack of focus
on application of the proposed seed zones reduce
their practical applicability, especially when com-
bined with poor documentation of the parameters
that dene the borders of the seed zones.
In order to reduce confusion and misuse of seed
zoning systems, it is especially valuable to distin-
guish between ‘seed procurement zones’, and ‘seed
application zones’ (cf. e.g. Buijtenen, 1992; West-
fall, 1992), and these concepts will therefore be
briey presented below.
1.3.1 Seed procurement zones: geographic
areas with uniform seed sources
Seed procurement zones can be applied with the
sole purpose of providing a framework for identi-
fying the origin of collected seed. The seed procu-
rement zone thus becomes the basic unit for seed
collection.
In natural forest it may serve little purpose to
delineate small seed sources in large areas of con-
tinuous forest and it will make more sense for the
seed collector to set up a common system of seed
collection zones, and identify the origin of the seed
according to these. It may also be easier for the user
of the planting material, if the origin refers to a well-
established relatively simple seed zone system. The
question is ‘how simple?’ i.e. how many zones?
The guiding principle behind seed procurement
zones is that there will be no signicant genetic
differences within small distances because the eco-
logical conditions are expected to be fairly uniform
within small distances. Therefore there is no reason
to assume that strong divergent natural selection
has been taking place within the seed procurement
zone. Furthermore, any genetic differences due to
past genetic processes are likely to be homogeni-
sed by pollen and seed ow. This requires that
the seed procurement zones are continuous areas.
They should not be too large, because they may
then include genetically different parts, and seed
from a given seed zone would then show varying
performance depending on collection site.
Therefore establishment of seed procurement
zones would entail the establishment of a network
of relatively small seed collection zones. Corre-
spondingly it will require a relatively long list of
seed zones and many seed lots to handle if seeds
are distributed in a centralised way.
Seed collection zones as described above are
only relevant for indigenous species; but ‘natura-
lised species’, i.e. species that have been grown

6
in the country for more generations, may also be
relevant. This requires that it is fair to assume
that the observable genetic patterns are caused by
recent adaptation to local condition and exchange
of pollen, rather than inherited differences that go
back to the introduction (founder effects). In Nepal
an example of a widely used introduced species is
Melia azedarach (Bakaino).
1.3.2 Seed application zones: areas with fairly
uniform planting sites
A different use of seed zones is to categorise plan-
ting site into larger groups. Focus is put on plan-
ting sites rather than native seed sources. This is
conceptually different from the seed procurement
zones discussed above, because the emphasis is less
on genetic similarity of the seed sources and more
on the grouping of planting sites with similar env-
ironmental conditions.
The guiding principle behind seed application
zones is that different seed sources should be used
at different sites (due to GXE, cf. Section 1.1). The
challenge is therefore to group planting sites (env-
ironments) into zones, where there is only little
GXE between sites within zones, and where sub-
stantial GXE therefore only occurs between the
zones.
Seed application zones may cover fairly large
areas. Often a zone will consist of several smaller
areas that all have similar ecological conditions,
but are not necessarily located next to each other.
Seed application zones are often dened by their
geographic limits, but may as well be dened by
ecological or topographic limits (i.e. rainfall, tem-
perature and/or elevation range). One can say that
a seed application zone basically refers to a set of
uniform planting sites in the sense that the same
seed source will be suitable for planting in that
zone. The planting zones established in this docu-
ment is of the ‘seed application zone’ type, but we
still prefer simply to call them ’tree planting zones’
to avoid any confusion.
Species may react quite differently along ecolo-
gical gradients, and it may therefore be wise to
develop specic seed application zones for each of
the most important species, or for groups of spe-
cies that in general react in the same way. Unfortu-
nately, even closely related species may react quite
differently. Again practical and experimental expe-
rience mainly exists from the Northern Hemisphere
and Australia. In Western Canada it is for example
recommended to transfer seed of Pinus contorta no
more than 100-150 meters (up/down), while Pinus
monticola can be transferred 700 meters (up/down),
cf. Table 1.
The GXE patterns may thus vary between species
in an unpredictable way. Intense studies of euca-
lyptus species have for example shown that species
of this genus can show quite different GXE pat-
terns. Based on Eldridge et al. (1994) and Raymond
(1989), the species can roughly be classied into
four groups (cf. Table 2).

7
ECOLOGICAL ZONATION OF PLANTING SITES FOR IMPROVED BENEFITS
The rst group shows large seed source diffe-
rences, and the performance of the seed sources
reects geographic areas. GXE at the seed source
level are generally small, and fairly large seed appli-
cation zones may therefore be applied. Actually,
for this group it is possible to have only one seed
application zone covering all the tested locations,
but it is important to identify and use the best
seed source. The second group also shows large
seed source differences reecting geographic origin.
However, strong seed source by environment int-
eractions (GXE) exists for these species, and seve-
ral breeding populations are therefore required in
order to obtain good match of seed sources to plan-
ting site. A third group has few differences between
seed sources, relatively small interactions between
seed sources and environments and no single seed
sources has been shown to be outstanding. One
seed zone would be sufcient, as for group 1, but it
seems less important to identify the best seed sour-
ces, as they are not so different. A fourth group of
eucalypt species is characterised by signicant - but
quite complex - differences between seed sources,
and apparently only little seed source by environ-
ment (GXE) interactions. To conclude, it is dif-
cult a priori to know how many seed application
zones are required, because the GXE patterns may
differ signicantly between species.
1.3.3 Seed application zones (tree planting
zones) common to all untested species
One cannot expect that all species have the same
GXE pattern, and as such it is in principle not pos-
sible to construct tree planting zones that cover all
species. Nevertheless, there is a denite need for
tree planting zones to be used for all untested spe-
cies in order to avoid use of maladapted seed sour-
ces and planting failures. This is certainly also the
case in Nepal, where at present data is only availa-
ble from Dalbergia sissoo tested at two locations in
the Tarai (Hansen et al. 2001).
As one cannot predict the likely GXE in a general
way, the only way to construct a general seed appli-
cation zoning (tree planting zoning) will therefore
be to construct a system of homogeneous environ-
mental zones (ecologically similar areas), and not
move the seed sources across to different zones.
This will reduce the risk of failures caused by plan-
ting material that is not well adapted to an area.
It will not assure that the ‘best’ seed sources are
used at a given area (cf. discussion above), or that
the relatively large number of tree planting zones
actually is required. This will depend on the GXE
pattern that remains unknown for the majority of
species.
The cost of simply using the local seed source
can be illustrated by generalising the example from
the eucalypts (see table 2), where the GXE pattern
has been estimated for different species.
For group 1 species (large genetic differences,
but little GXE), general seed zoning will pro-
bably impose unnecessary constraints on use
of seed sources over a wide area. Much can
be gained from identifying and using the best
seed sources, but an overly restrictive zonation
system will decrease the possibilities for doing
so. Also, for domestication, an unnecessary large
number of seed sources will be established for
this type of species. The size of these extra costs
will mainly depend on the type of established
seed sources and domestication activities, but
also the type of seed distribution system. In
a decentralised system, the extra costs will be
small, whereas the extra costs can be substantial
in a centralised seed distribution system.
For group 2 species (large genetic differences
and also large GXE), some kind of zonation will
be necessary to capture the GXE, and for these
species a general zonation system is likely to
protect against use of maladapted seed sources.
For group 3 species (small genetic differences
and little GXE) and group 4 species (some gene-
tic differences, but irregular) a general zonation
system may not be required. The extra costs will
be as for group 1.
In Nepal, a very decentralised seed distribution
is considered the most appropriate and correspon-
dingly a relative large number of local seed sour-

8
ces of priority species are envisaged (Dhakal et al.,
2001). In practical terms the extra ‘cost’ for having
a fairly large number of zones will therefore be
small, as the main factor determining the number
of seed sources will be the decentralised distribu-
tion system, while the seed zoning system will
determine the location of the seed source. For prio-
rity species, it may be decided to engage in more
intense breeding in Nepal. A breeding programme
for Dalbergia sissoo is well into its 1st generation,
programmes for a number of MPT species have
also been initiated and additional species are plan-
ned to follow (Lillesø et al., 2000). For these spe-
cies, specic breeding and planting zones will be
developed based on results gained from eld trials.
Still, the general tree planting zone system will be
used as the starting point from which species-speci-
c zones will be developed.
1.4 Elaboration of the Tree Planting Zones
in Nepal - Environment Factors to be
Considered
Basically no experimental data is available on
‘response functions’ of tree species in Nepal i.e.
how they react when seed sources from one env-
ironment is used in another environment. There-
fore, the planting zone system must be based on
establishment of zones with fairly uniform ecolo-
gical conditions and must be restrictive in transfer
between these zones without prior testing. Nepal
covers areas with extreme ecological variation -
from alpine to tropical vegetation - often with dra-
matic clines and variation over small geographic
distances. This feature of Nepal complicates estab-
lishment of planting zones, but also increases the
importance of doing so.
The most extreme variation over small geogra-
phic distances in Nepal is probably due to effects of
elevation, where the temperature is highly depen-
dent on altitude. Many of the tree species planted
in Nepal have a natural distribution that covers
large gradients both in terms of altitude and mois-
ture. Some examples are listed in Table 7, and clinal
variation in altitude is thus a parameter that will be
important to include in zonation.
Rainfall, and more general availability of mois-
ture over the year, also varies substantially within
relatively small geographic distances (Nayaju and
Lillesø, 2001), to some extent as co-variation with
altitude, but also due to a much more complex
combination of physiographic aspects and pre-
vailing wind-directions during the monsoon season.
Furthermore the aspect of planting site on the
slopes (directed North, South, East or West) will
inuence the microclimatic conditions on the plan-
ting site, and soils also vary substantially (LRMP,
1986). Moisture is therefore also a parameter to be
considered for zonation, especially as a combined
effect of total annual rainfall and the length of the
dry season (discussed below).
As may be expected, Nepal has a large corre-
sponding variation in vegetation types. This varia-
tion has been reviewed recently by Shrestha et al.
(2001), who elaborated a vegetation map with 38
vegetation types in 7 major ecological zones. Ele-
vation and moisture cannot by themselves describe
this variation, although the two factors play a clear
role. Many other factors determine the natural
vegetation, and these factors are most likely impor-
tant for the zonation. For the purpose of zoning
planting sites, we therefore consider the distribu-
tion of natural vegetation (or in reality the poten-
tial vegetation) to be the best starting point for
the zonation. Some of these vegetation types cover
large variation in altitude in which case we suggest
subdivision. This will result in a fairly detailed
zonation that can then be simplied for specic
purposes as will be discussed in details in chapter 3.
The rst step in elaboration of the planting zones
is however to look more carefully at the variation
in climate and vegetation (see chapter 2).
The majority of settlements in Nepal are located
in the Tropical and Sub-Tropical Ecological Zones
as can be seen in g. 3 and the main prospects for
planting by local communities and individual far-
mers are in these two zones. The zones are heavily
utilised for agriculture, and farmers plant a large
number of trees, especially for fodder. Historically
the most wide-ranging (and probably mismatched)
transfer of seed has taken place in these two major
ecological zones. At present, limited planting acti-
vities take place in the colder parts of Nepal and
for most species seed use can be covered by local
collection of seed. This may change in the future if
cultivation of medicinal plants increase2.
Detailed planting zones will therefore be ela-
borated in more detail in the Tropical Ecological
Zone and in the Sub-Tropical Ecological Zone,
while in the ecological zones at higher elevations
it is recommended that collection should be from
the same vegetation type in the immediate sur-
roundings of the planting site, taking aspect into
account.
2
For example for Seabuck-
thorn (Hippophae spp.) and
some medicinal plants
considerable demand may
arise and would require
elaboration of seed trans-
fer guidelines. Such guide-
lines could be established
following the distribution
of potential vegetation
types.

9
ECOLOGICAL ZONATION OF PLANTING SITES FOR IMPROVED BENEFITS
Figure 3. Distribution of settlements in relation to Ecological Zones and Physiographic Regions.

10
The climatic criteria that have been found to be
most correlated with vegetation from empirical stu-
dies around the world are mean temperature of
the coldest month, annual rainfall, and length and
number of dry seasons (Ashton, 1995).
The ecology of Nepal is dominated by the
extreme topography of the country coupled with
the impact of the monsoon, and furthermore the
country is biogeographically a meeting point of two
oristic realms. The extreme topography means
that soil factors are less important in the overall
natural distribution of vegetation types and species
(Stainton, 1972; Dobremez, 1976; Shrestha et al.,
2001).
The extreme topography has the effect that
moving a few kilometres in Nepal corresponds to
moving between tropical, sub-tropical, temperate
or even arctic zones because the abrupt changes in
elevation closely corresponds to a change in tempe-
rature regime (Nayaju and Lillesø, 2001). The distri-
bution of vegetation types very strongly reects
the vertical change in climate and correspondingly
most species have distributions that cover only part
of the vertical variation in the country (but almost
always several vegetation types).
The change in vegetation types from east to west
is less easily interpreted as the distribution of vege-
tation types probably is a mixture of general and
2. Climates and Vegetation in Nepal
local variation in precipitation patterns as well as
the biogeography of individual species. Nepal is
situated at the interface between the Indo-Malayan
Realm (an ecological region with tropical dry or
deciduous forests) and the Paleartic Realm (an eco-
logical region with temperate and sub-polar vegeta-
tion types, and is furthermore a meeting point of
several oristic provinces, see g. 4 and Shrestha et
al. (2001).
In most of Nepal, altitude and aspect are of
paramount importance in determining the type
of forest found at any particular place (Stainton,
1972). In contrast, the landscape in the Lower Tro-
pical Ecological Zone is more or less at, and
soil type and water-holding capacity of the subsoil
become relatively more important in determining
the distribution of species and vegetation types.
The Land Resources Mapping Project (LRMP,
1986) mapped soil distributions in large parts of
Nepal, while several other projects have mapped
parts of the Lower Tropical Ecological Zone. The
maps are, however, of such small-scale that they
cannot be used in a large-scale mapping. TISC
intends to compile completed soil explorations
into regional and country-wise maps for further
elaboration of planting zones.
Figure 4. Floristic provinces of Asia (from Shrestha et al., 2001).

11
2.1 Climates in Nepal
The key climatic factors are rainfall and tempera-
ture as well as potential evapotranspiration. When
considering rainfall it is not only the total annual
rainfall that is of interest, but also the annual distri-
bution of rainfall and the variation between years.
From table 4 with temperature extremes by eco-
logical zone it can be seen that there is an increase
in number of cold days (and a corresponding decre-
ase in number of hot days) from Lower Tropical to
Sub-Alpine ecological zones.
The occurrence of frost is likely to have consi-
derable adaptive value for tree species and an ave-
rage frost line for Nepal has been estimated to
be at around 1000m elevation (Nayaju and Lil-
lesø, 2001), which is along the border between the
Tropical and Sub-Tropical ecological zones. Occur-
rence of frost in Nepal is, however, quite a complex
phenomenon. It depends upon many factors such
as geographical position, ridge, valley, slope aspect,
elevation, moisture availability, and soil characteri-
stics. In particular the frequency and duration of
frost on northern slopes is higher than on southern
slopes. Since air drainage is better on ridges than in
valleys, frequency of frost on ridges is lower than
in the valley.
The ecological zones also describe differences in
temperature regimes that are strong indicators of
the growth conditions and will therefore be uti-
lised as one of the parameters in the planting zone
system. The distribution of precipitation across
Nepal is, however, more complex.
From table 3 with average gures for tempera-
ture and precipitation by ecological zones it can
be seen that temperature falls with increasing ele-
vation of the zones, while there is overlap of preci-
pitation extremes within the zones.
CLIMATES AND VEGETATION IN NEPAL

12
2.1.1 Precipitation, potential evapotranspira-
tion and dry/wet seasons
The number of climatological stations is insuf-
cient to provide a complete picture of precipita-
tion variation in Nepal, as only about 270 stations
have been installed (Nayaju and Lillesø, 2001). It
is possible, however, to nd some major patterns,
across the country. Within these major patterns a
repeated phenomenon, which is not described by
the stations, is caused by the different aspects ohill-
sides such that north and west faces tend to be
shadier and retain moisture longer and correspon-
dingly have a more varied ora (Stainton, 1972,
Shrestha et al., 2001).
In Nepal most of the precipitation falls within
the monsoon period from June to September and
the corresponding potential ‘monthly evapotran-
spiration balance’ is positive for these months in
most parts of Nepal (Nayaju and Lillesø, 2001).
In the following the climate has been described
in two different ways (i) using Thorntwaithe’s mois-
ture index (see Nayaju and Lillesø, 2001) and (ii)
using the number of dry/wet months3 as the clas-
sication criteria. In both cases the meteorological
stations have been grouped on the digital map and
climate lines have been drawn following the natu-
ral contours (main ridges) of the map.
Map using Thorntwaithe’s moisture index
Moisture Regions in accordance with Thorntwai-
the’s moisture index were identied and mapped
(see g. 5, Ecological Zones with Moisture Regi-
ons). The map (red lines) shows that 4 types of
moisture regions can be described (1: Arid; 2:
Semi-Arid; 3: Humid; and 4: Per-Humid).
The Arid moisture region is a large rain-shadow
area on the high altitude leeward side of the Hima-
layas. The Semi-Arid moisture region is mainly in
the high altitude north and north-western part of
the country, but also in the eastern and western
upper tropical valleys and southern parts of the
Lower Tropical Ecological Zone. The Humid mois-
ture region is prevalent in the Subtropical, Upper
Tropical and Lower Tropical ecological zones. The
Per-Humid moisture region is relatively small and
mainly occurs on the slopes of Himalayas in cen-
tral and eastern parts of Nepal, only reaching the
Lower Tropical Ecological Zone in the extreme east
of the country.
The moisture regions only very broadly corre-
spond to the variation in distribution of vegetation
types and contribute relatively little to an explana-
tion for the detailed occurrence of perennial tree
species.
Map based on length of the wet/dry season
Trees are perennial crops and it is likely that the
length of the growing season has adaptive signi-
cance (inuencing leaf ush, leaf fall, owering,
fruiting, and seedling survival). In g. 5 the Tro-
pical and Sub-Tropical vegetation types have been
overlaid by a description of the climate based on
the length of the wet/dry season (yellow lines -
calculated by the monthly potential evapotranspi-
ration minus monthly precipitation). The wet/dry
season map shows that the western part of Nepal
generally has a longer dry season than the eastern
part. The longest dry season is experienced in the
Trans-Himalayan Ecological Zone. The shortest
dry season is experienced in a belt stretching from
the Lower Tropical Ecological Zone in the east
across the eastern Upper Tropical Ecological Zone
and into the slopes of Himalayas in the Central
Nepal as well as along the eastern slopes of the
Himalayas. Parts of the Upper Tropical, Sub-Tropi-
cal and Temperate ecological zones in the western
part Nepal have a shorter dry season due to the
occurrence of winter rains in December to Febru-
ary.
The wet/dry season map exhibits a much better
correspondence with the distribution of vegetation
than the moisture map in the Sub-Tropical Eco-
logical Zone as discussed below and illustrated in
gure 11. The correspondence is not complete, but
by and large climate probably has a large part of
the explanation for the distribution of the three
major forest types in the Sub-Tropical Ecological
Zone.
Moisture index or dry/wet months or vegetation
types?
The distribution and number of meteorological
stations is inadequate to fully explain the distri-
bution of rainfall regimes in Nepal, although the
information can be used to describe some overall
patterns in the country. The moisture index map
and the wet/dry zone map do correspond in the
major patterns, while the wet/dry zone map appe-
ars to have a better correspondence with the distri-
bution of vegetation in the Sub-Tropical Ecological
Zone. The distribution of vegetation types is there-
fore probably a better predictor of growth conditi-
ons for tree species than the available climatological
information.
3Dry months are dened
as having monthly precipi-
tation smaller than poten-
tial evapotranspiration

13
CLIMATES AND VEGETATION IN NEPAL
Figure 5. Moisture patterns in Nepal: described by moisture regions - Thorntwaite’s moisture index (red lines) and length of dry season (yellow lines).

14
2.2 Distribution og Vegetation in Nepal
The Potential Vegetation Map of Nepal developed
by TISC provides details of the 38 potential vegeta-
tion types covering Nepal. The vegetation types are
grouped into ecological (elevation) zones as shown
in table 5. Descriptions and maps of individual
vegetation types and a list of species can be found
in Shrestha et al. (2001).
The Potential Vegetation Map depicts the natural
environment of any particular place as indicated by
the vegetation type and the database of plant spe-
cies occurring in each vegetation type. Although
at present the species database is not complete for
all vegetation types and all species, it can be used
as a basis for estimating the potential growth area
for many indigenous species. See tables 5 and 6. A
detailed inventory of species in the potential vege-
tation types is given in Shrestha et al. (2001).
The vegetation types are most often spread as
relatively narrow bands along the elevation zones
and many vegetation types especially in the Tem-
perate, Sub-Alpine and Alpine ecological zones
consist of widely separated (disjunct) areas. The
disjunct areas have many species in common, but
a substantial number of species most probably do
not occur in all disjunct areas of a vegetation type,
and furthermore the climatic conditions may vary
considerably from one area to another.
The environment within the vegetation types is
not entirely homogeneous, especially due to the
variation in aspect. Northern aspects are generally
moister, shadier, and cooler than southern aspects
with western and eastern in between. It is likely
that the occurrence of species within each vegeta-
tion type will vary with the ‘micro’ climates of the
aspects, but this has not (so far) been investigated
in great detail.

15
CLIMATES AND VEGETATION IN NEPAL

16
2.2.1 Alpine, Sub-alpine and Trans-Himalayan
Ecological Zones
The Alpine and Sub-Alpine ecological zones (g.
6 and 7) are relatively narrow bands spanning the
width of Nepal with shrubs and herbs as domi-
nant life forms and the Trans-Himalayan Ecologi-
cal Zone (g. 8) mainly covers the high altitude
Figure 6. Distribution of Alpine Ecological Zone in Nepal.
rainshadow area in Central-Western part of the
country also with shrubs and herbs as the domi-
nant life forms. Relatively few people occupy these
zones permanently, andthese zones have very few
planting activities.

17
CLIMATES AND VEGETATION IN NEPAL
Figure 7. Distribution of Sub-Alpine Ecological Zone in Nepal.
Figure 8. Distribution of Trans-Himalayan Ecological Zone in Nepal.

18
2.2.2 Temperate Ecological Zone
The Temperate Ecological Zone (g. 9) has nearly
50% of the vegetation types of Nepal, but only
around 12% of the area. The zone is extremely
heterogeneous due to the steep terrain. Compa-
ratively few people permanently live in the zone
and relatively few planting activities of tree species
take place, although the vegetation is most often
heavily utilised by people living in the neighbou-
ring Sub-Tropical Ecological Zone.
Figure 9. Distribution of Temperate Ecological Zone in Nepal.
ring Sub-Tropical Ecological Zone.

19
CLIMATES AND VEGETATION IN NEPAL
2.2.3 Subtropical and Tropical Tcological Zones
The Subtropical Ecological Zone (g. 10) spans
Nepal, but the forest types are different in the
Eastern and Western part of the country, most pro-
bably reecting the variation in precipitation regi-
mes.
Figure 10. Distribution of SubTropical Ecological Zone in Nepal.

20
In gure 11 the distribution of the three major
forest types in the Subtropical Ecological Zone and
the two main forest types in the Tropical Zone is
overlaid by the wet/dry season (yellow lines). The
Schima-Castanopsis forest occurs in the wettest parts
(shortest dry season) of the zone in Eastern and
Central Nepal interrupted by a relatively drier Chir
Figure 11. Distribution of vegetation types in Subtropical and Tropical Ecological Zones and major
precipitation zones (dry/wet seasons).
Pine-Broadleaved forest (longer dry season). The
driest Chir Pine forest (shortest dry season) occurs
on the lower subtropical slopes of western Nepal
with Chir Pine-Broadleaved on the upper slopes.
The lack of complete correspondence is probably
to a large extent due to incomplete coverage of
meteorological stations.

21
CLIMATES AND VEGETATION IN NEPAL
The Upper Tropical Ecological Zone (g. 12) and
Lower Tropical Ecological Zone (g. 13) contain
vegetation types as relatively broad bands spanning
the width of Nepal. The zones have not been the
subjects of systematic investigations by botanists.
At present each zone is categorised by one forest
type (Shrestha et al., 2001).
Figure 12. Distribution of Upper Tropical Ecological Zone in Nepal.

22
4Siwaliks or Churia hills
are the rst low hills in the
upper tropical belt run-
ning along the length of
Nepal. The Siwaliks have
soils of low water-holding
capacity (Ives and Mes-
serli, 1990).
Figure 13. Distribution of Lower Tropical Ecological Zone in Nepal.
It is likely that rainfall patterns and soil factors
have had some inuence on the natural distribu-
tion of species, but the details of the natural distri-
bution of species will never be known with great
accuracy as large areas of forests have been cleared
for agriculture and the remaining forests have been
profoundly inuenced by human activities.
The precipitation patterns do, however, indicate
that there may be substantial differences. In the
Upper Tropical Ecological Zone the large valley
system in the central part of Nepal is relatively wet,
while the valley systems to the east and west are
drier. In the Lower Tropical Ecological Zone the
most eastern part is substantially wetter than the
rest. TISC will carry out rapid vegetation surveys
to accumulate species lists and to the extent pos-
sible map distributions of the most important spe-
cies.
The Lower Tropical Ecological Zone is densely
populated, especially in the eastern part. The
Upper Tropical Ecological Zone has a considerable
human population in the valleys, while the Siwa-
liks4 has only dense populations at a few locations
(see g. 3).

23
CLIMATES AND VEGETATION IN NEPAL
Precipitation and vegetation types in the Tropi-
cal Ecological Zone
An indication of the likely subdivision of the Tro-
pical Ecological Zone can be found from Walter
(1971), who plotted the tropical forest types of
India with respect to annual rainfall and length of
the dry season. Walter (1971) found that neither
the amount of precipitation nor the length of the
dry period alone was decisive for delimiting diffe-
rent forest types. A longer dry period can be com-
Figure 14. Relation between forest vegetation and total
rainfall (ordinate) as well as length of dry period in
months (abscissa) in India. I: Evergreen; II: Semi-ever-
green tropical rain forest, III Monsoon forest (A: Humid,
B: Dry), IV: Savannah (thornscrub forest) (from Walter,
1971, gure 123 page 212).
pensated by higher precipitation. By entering the
rainfall on the ordinate and the length of the dry
period on the abscissa, the points representing a
particular vegetation type fell into a well-dened
pattern (Fig. 14). Vertical separation would be
expected if the length of the dry season were the
only factor. In fact, however, the mean lines run at
steeper angle in the moist area and more horizon-
tally in the dry area.
Walter (1971) concluded that the change from
tropical rainforest to semi-evergreen rainforest and
from this to the rain-green dry forest is determi-
ned more by length of the dry season than by the
annual rainfall. The annual rainfall is more critical
than the length of the dry period for the change of
rain-green dry forest into savannah and desert.

24
Figure 15. Assumed relation between forest vegetation and rainfall (ordinate) as well as length of dry period in
months (abscissa) in Upper and Lower Tropical Ecological Zones in Nepal. I: Evergreen and II: Semi-evergreen tro-
pical rain forest, III Monsoon forest (A: Humid, B: Dry), IV: Savannah (thornscrub forest). Lines are approximate
and taken from Walter (1971), see gure 14.
Using the same criteria as Walter (1971) the meteo-
rological stations can be plotted as shown in Fig. 15.
The diagram in g 15 indicates that some of the
meteorological stations fall in the semi-evergreen
tropical rain forest and in fact most of these sta-
tions fall in a contiguous area in the eastern wet
zone, which should therefore probably be consi-
dered separate from the remainder of the Tropical
Ecological Zones.
The geographical separation of the remaining sta-
tions is less straightforward, because there are no
clear physical barriers nor can stations of Humid
Monsoon Forest be separated from Dry Monsoon
Forest, except parts of the Dun valley in central
Nepal. The Dry and Wet Monsoon stations appear
to be a patchwork across Nepal, perhaps with a ten-
dency for the Siwaliks in the Upper Tropical Eco-
logical Zone to be more humid than the Lower
Tropical Ecological Zone. There are insufcient
stations in the tropical valleys to evaluate how
humid they are; we therefore suggest (judging from
the surrounding vegetation) tentatively to separate
the tropical valley systems as presented in the fol-
lowing chapter.
lowing chapter.

25
DELIMITING PLANTING ZONES IN NEPAL
A tree planting zoning system should be a compro-
mise between the recognised ecological variation
(and the assumed variation in ‘species’ popula-
tions), the cost of establishing seed sources and
the assumed benet of establishing more seed sour-
ces. Through use of the system i.e. by proper
seed source registration and establishment of BSOs
(Breeding Seed Orchards) and adherence to the seed
transfer guidelines, more information on species
and demand of species will accumulate and chan-
ges in detailed zoning will probably be required;
such changes can easily be elaborated.
In chapter 1 we deliberate on the importance of
taking potential Genotype by Environment inter-
action (GXE) into consideration to avoid maladap-
tion of planted trees and bushes and we recommend
that Tree Planting Zones with fairly uniform ecolo-
gical conditions are elaborated. Local seed sources
would then be sources recommended for particular
planting zones.
A seed distribution system must be as simple
as possible to be followed in practice. Two impor-
tant requirements in Nepal would be (i) that the
tree planting zones can be recognised in the eld,
and (ii) that seed sources are established for tree
planting zones where there is in fact a demand for
seed.
The potential vegetation has the advantage that
farmers can easily identify the seed zones by
the use of indicator species. The seed zoning
system which is based on vegetation types has
no distinct boundaries because the vegetation
types continuously intergrade, but with the help
of indicator species it will be possible to deter-
mine each zone in the eld, Shrestha (in prep.).
We recommend to elaborate more detailed plan-
ting zones in the subtropical and tropical ecolo-
gical zones, because most trees are planted by
farmers in these zones. In the colder ecological
zones much less planting takes place and we
recommend that planting should utilise sources
from similar vegetation type in the immediate
surroundings of the planting site, but if species
are to be utilised for more intensive planting in
the colder ecological zones, then more detailed
planting zones should be elaborated.
In chapter 2 we deliberate on the variation in
climate and ecology in Nepal and recommend
that the potential distribution of natural vegetation
should be used as the best indicator for the growth
conditions for trees and bushes.
As mentioned in chapter 2 and discussed in detail
in Lillesø et al. (2000) and Dhakal et al. (2001),
a decentralised tree seed distribution system with
3. Delimiting Planting Zones in Nepal
emphasis on fodder species could help improve
living conditions for the poor farmers in Nepal
and in particular would help improve the plight
of poor women. A decentralised seed distribution
system would require many local seed sources,
which corresponds well with the requirements of
the suggested tree planting zoning system. The tree
planting zones are primarily seed application zones
(see chapter 1.3.3) and seed should be procured
from well-dened seed sources within the zones.
3.1 Tropical Planting Zones
The elevation spans of the ecological zones are
300m for Lower Tropical, 700m for Upper Tro-
pical, while Sub-Tropical Ecological Zone spans
1000m. Experience from tested species suggests
that there is most often a change in the genetic
constitution of populations along elevation gra-
dients but the degree of change is different for
different species (see table 1). It is not known
how species in Nepal change with elevation gradi-
ents, except that they most likely will be sensitive
to transfer across elevation gradients smaller than
1000 meters.
The subtropical vegetation types will therefore
have a lower and upper elevation zone to avoid
moving populations more than 500 meters up or
down the elevation gradient. Most planting activi-
ties in the Upper Tropical Ecological Zone will be
in the valley systems which are above 500 m, the-
refore only one elevation zone will be dened. The
Lower Tropical Ecological Zone has a span of 300
m and will likewise have only one elevation zone
dened.
3.1.1 Lower Tropical Ecological Zone (Sal and
Mixed Broad Leaved Forest)
The Lower Tropical Ecological Zone has one eleva-
tion zone from 0 to 300 meters. The zone spans
the entire width of Nepal, but as described in
details above, the climate in the easternmost part
of the zone corresponds to a semi-evergreen Forest
type (east of Koshi River), while west of this area
the climate corresponds to a monsoon forest type
(mostly dry monsoon forest type). The borders
between the potential vegetation types cannot be
observed in the eld, as the variation is clinal.
The degree of variation in the monsoon forest
type is suggested by the analysis of two Breeding
Seedling Orchards of Dalbergia sissoo. It was found
that some genotype x environment interactions
were present in vigour (but not in stem form)
and indicate that Dalbergia sissoo may have geno-
type x environment interactions also within the
drier planting zone (Hansen et al., 2001). TISC will
further investigate this.
(i)
(ii)

26
This planting zone is the wettest and easternmost part of the Lower
Tropical Ecological Zone (east of Koshi River). Most of the area has
been converted to agricultural land, but pockets of natural vegetation
remain.
This planting zone (west of Koshi River) mainly consists of dry
monsoon forest interspersed with localised areas with humid monsoon
forest. At present, this area is considered as one zone, but further work
with Dalbergia sissoo may show the need for a further subdivision, at
least for this species
Figure 16. Planting zones of Lower Tropical Sal and Mixed Broad Leaved Forest type in Nepal.
Semi-Evergreen Forest type
Monsoon Forest type
Lower Tropical Sal and Mixed Broad Leaved Forest planting zones

27
DELIMITING PLANTING ZONES IN NEPAL
3.1.2 Upper Tropical Ecological Zone (Hill Sal
Forest)
The Upper Tropical Ecological Zone consists of the
Siwalik range with soils with low water holding capa-
city and of tropical valley systems. The upper limit of
1000 meters corresponds to the lower limit for occur-
rence of frost.
The Siwalik spans the entire width of Nepal but
the climate in the easternmost part of the zone cor-
responds to a Semi-Evergreen Forest type, while west
of this area the climate corresponds to a monsoon
forest type (mostly dry monsoon forest type). The bor-
ders between the potential vegetation types cannot be
observed in the eld, as the variation is clinal.
Most of the valley systems penetrating far into the
hills are relatively dry, while the large Central Valley
is relatively moister. The distribution of meteoro-
logical stations in the tropical valley systems does
not provide for a very detailed analysis. The valley
system has therefore been split into several sub-
zones partly based on data from meteorological
stations and partly on the type of adjacent forest
types. Not all borders can be observed in the eld,
as the variation is clinal.
Figure 17. Planting zones of Hill Sal Forest type in Nepal.

28
Hill Sal planting zones
Wet Central Valley
Eastern Tropical Valleys
Sun Koshi Tropical Valley
Western Valleys
Eastern Wet Siwaliks
Western Dry Siwaliks
This planting zone is the largest tropical valley system in the central
Nepal. It is bordered to the north and west by Schima-Castanopsis
forest, to the west by Chir Pine-Broadleaved forest and to the south by
Lower Tropical Forest in the Dun valley.
This planting zone consists of the tropical valleys through which the
Tamur and Sapt Koshi rivers run surrounded by Schima-Castanopsis
forest at higher elevations. The southern parts of the valleys are drier
than the northern parts.
This planting zone consists of the long valley system through which
the Sun Koshi river run surrounded by Chir Pine-Broad Leafed at
higher altitudes. The whole length of the valley is relative dry.
The western valleys through which the Mahakali, Seti, Karnali and
Bheri rivers run are relatively drier than the eastern and central valleys
and are surrounded by dry Chir Pine Forest type.
Corresponding to the Lower Tropical Planting zone to the south the
Eastern Wet Siwaliks are much wetter than the rest of Upper Tropical
Ecological Zone, but the wet zone extends as far as west of the central
valley.
This planting zone starts west of the Central Valley and extends to the
western-most part of Nepal.
3.2 Sub-tropical planting zones
The Sub-Tropical Ecological Zone reaches from
1000 to 2000 meter elevation. There are three major
forest types of increasing aridity: Schima-Castanop-
sis; Chir Pine-Broadleaved and Chir Pine. The Chir
Pine-Broadleaved Forest type has many species in
common with Schima-Castanopsis Forest type but
the ecological conditions are such that the species
composition and abundance of species are very dif-
ferent.
The three forest types should be considered sepa-
rate planting zones.
3.2.1 Schima-Castanopsis Forest
The Schima-Castanopsis Forest type consists of
three major disjunct areas. Due to the decentra-
lised seed distribution network, it is recommen-
ded that seed sources (farmland seed sources and
Forest User Group seed sources in natural forest)
be initially established for each of the major dis-
junct areas. Further domestication of selected spe-
cies may breed for the whole forest type.
The forest type covers an elevation span of 1000
meters and seed sources should be established
for lower (1000-1500m) and upper elevation levels
(1500-2000m).

29
DELIMITING PLANTING ZONES IN NEPAL
Figure 18. Planting zones of Schima-Castanopsis Forest type in Nepal.
Schima-Castanopsis planting zones
Eastern Schima-Castanopsis upper and
lower sub-zones
Central Schima-Castanopsis upper and
lower sub-zones
Western Schima-Castanopsis upperand
lower sub-zones
This planting zone occurs in the eastern part of Nepal sur-
rounding the Arun and Tamur valleys.
This planting zone occurs in the Kathmandu valley and neigh-
bouring subtropical valley systems.
This planting zone occurs in the central part of Nepal sur-
rounding the Central Valley to the north and east.

30
3.2.2 Chir Pine-Broad Leaved Forest
The Chir Pine Broad Leafed Forest type occurs in
two widely separated areas. Due to the decentra-
lised seed distribution network, it is recommended
that seed sources (farmland seed sources and FUG
seed sources in natural forest) be initially estab-
lished for each of the two disjunct areas. Further
domestication of selected species may breed for the
whole forest type.
Figure 19. Planting zones of Chir Pine-Broadleaved Forest type in Nepal.
Chir Pine-Broadleaved Planting Zones
Eastern Chir Pine-Broadleaved upper and lower sub-
zones
Western Chir Pine-Broadleaved upper and lower sub-
zones
This planting zone occurs in a large relatively con-
tiguous area between the eastern Schima-Castanop-
sis and the Upper Tropical Central Valley
This planting zone occurs as relatively narrow
bands above Chir Pine forest in the western part of
Nepal.
the whole forest type.

31
DELIMITING PLANTING ZONES IN NEPAL
3.2.3 Chir Pine Forest
The Chir Pine Forest type occurs in one conti-
guous area in western Nepal and will be considered
as one zone with no sub zones.
Figure 20. Planting zones of Chir Pine Forest type in Nepal.
Chir Pine Planting Zones
Chir Pine Forest upper and lower sub-zones This planting zone occurs in the western part of
Nepal and covers the lower ridges surrounded by
Dry Siwaliks below and Chir Pine-Broad Leafed
above.
as one zone with no sub zones.

32
3.3 Examples by species
Most species occur in more than one forest type
and correspondingly they will occur in several
zones. For many species, planting only occurs in
some of the zones within their distribution. For
each species a prioritisation of seed zones will take
place and seed sources will be developed in the
zones with the highest planting activities. To illu-
strate the rst step in planning establishment of
seed sources for a species, the distributions of four
species are presented g. 21, 22, 23, and 24 and the
corresponding potential planting zones are listed
in table 7.
Bauhinia purpurea is one of the more popular
fodder species and it is considered to consist of two
types of trees by farmers (but not by botanists),
one of which provides better quality fodder. The
two types can be distinguished by leaf characters,
but the genetic background for the separation into
types is not known. The species is distributed in
central to eastern parts of Nepal in Hill Sal forest
and in Schima-Castanopsis forest. Brassaiopsis glo-
merulata is an important fodder species for far-
mers. It is distributed from east to west in Nepal in
Hill Sal Forest. Ficus neriifolia is one of a series of
Ficus species that are some of the most important
fodder species for farmers in Nepal. Garuga pinnata
is very useful timber species for both commercial
and domestic use.
Figure 21. Bauhinia purpurea distribution in vegetation types in Nepal (for detailed zonation, see table 7).

33
DELIMITING PLANTING ZONES IN NEPAL
Figure 22. Brassaiopsis glomerulata distribution in vegetation types in Nepal (for detailed zonation, see table 7).
Figure 23. Ficus neriifolia distribution in vegetation types in Nepal (for detailed zonation, see table 7).

34
Figure 24. Garuga pinnata distribution in vegetation types in Nepal (for detailed zonation, see table 7).

35
APPLICATION OF THE PLANTING ZONES IN NEPAL
4. Application of the Planting Zones in
Nepal
4.1 Guidelines for Seed Transfer Within and
Between Seed Zones
The planting zones follow the forest types, but
have further been split into sub zones based on the
ecological criteria of elevation. Zones should thus
be ecologically uniform within which a seed source
can be considered ‘local’ to a given planting site.
Use of ‘local’ seed sources would thus imply that
seed should originate from the same planting zone.
Still, elevation cannot explain all gradients, and use
of seed sources from far away should be avoided
if good seed sources are available closer by (cf. the
example of Alnus rubra in North America descri-
bed in sec.1.2). This will especially be the case for
the sub-zones, such as e.g. Schima-Castanopsis that
consist of disjunct areas. If more intensive dome-
stication starts through establishment of Breeding
Seedling Orchards, the sub-zoning by disjunct areas
should be tested and families from each disjunct
area should be included in the Breeding Seed
Orchards. If based on results of testing, it can also
prove advisable to use selected seed source in seve-
ral planting zones, but such guidelines for transfer
across planting zones should be species specic
as they rely on results from specic eld trials or
otherwise gained experience.
The subdivision of the Sub-Tropical forest types
into elevation zones is based on the experience
from many species around the world that shows
that most species respond quite strongly to move-
ment up and down elevation gradients. The upper
and lower sub-zones should therefore be conside-
red very important and practical guidelines should
be developed on how to implement establishment
of farmland seed sources in particular and how to
distribute seed in each of the elevation sub-zones.
For species that have never been tried in Nepal,
an introduction of a species can be guided by cli-
matic matching of the seed zones with climates of
the foreign seed sources. Naturally occurring spe-
cies should in general only be planted within their
natural area of distribution. Planting outside the
natural distribution may be successful, but the risk
of mal-adaption may be considerable. For species
that may be in great demand outside their natural
range (for example subtropical fodder species in
demand by hill people that have migrated to the
Lower Tropical Ecological Zone) it is therefore
recommended that the Department of Forests
establishes trials to evaluate their growth and survi-
val before supporting introduction on a large
scale.
4.2 Seed Supply, Seed Source Development
and Seed Source Description
TISC has a central role in supporting development
of a decentralised seed distribution system, and
in the establishment of seed sources, especially of
farmland fodder species. The main objective of the
seed distribution system is to contribute towards
the improvement of the living conditions for the
poor farmers (smallholders) in Nepal. The key is
to increase the availability of kinds and quality
of fodder seedlings throughout the country and
this will be done by (i) making better seeds availa-
ble and (ii) making the actors in the seed and nur-
sery business aware of the value of using improved
seed. Detailed planning and implementation pro-
posals have been reported elsewhere (Lillesø et al.,
2001; Dhakal et al., 2001), but are based on the pre-
sent planting zones.
The main actors in distribution will be decen-
tralised entities such as farmers’ associations, seed
cooperatives and Forest User Groups. The objec-
tives for the seed cooperatives will be (i) to collect
and supply seeds of known origin and broad gene-
tic base to a transparent market; and (ii) conserva-
tion of the gene resource base of the economically
important farmland tree species and improvement
of desired traits.
The planting zones will be the natural units for
the operation of seed co-operatives. If so, seed
distribution will support use of local seed sources
and thereby reduce risk of poor match between
species/provenance and planting site. This will be
opposed to more centralised seed distributors that
tend to move seed across zones as part of commer-
cial seed trade.
Priority species will eventually be established in
Breeding Seed Orchards (BSOs). These will pro-
vide the possibility of selection of the best mate-
rial, but will also provide the opportunity to test
how the species react to the seed zoning system,
i.e. how sensitive the species are to the environ-
mental variables (a few families will be exchanged
between seed zones in order to be tested). The
established planting zones will thus be the star-
ting point and reference for development of spe-
cies specic programmes.
Registered seed sources will be geo-referenced to
the seed zone map and included in a seed source
database. The database will be used for planning
and for extension/marketing. This information on
planting zone can be used for seed documentation,
and also for specic recommendations regarding
choice of seed sources for given species. If included
as part of seed documentation, it will allow the tree
planters to realise when they use seed of non-local
origin.

36
5. Conclusions and Recommendations
Nepal has one of the most complex natural env-
ironments in the world, but at the same time this
variation is quite well documented compared to
many other developing countries.
The tree planting zone system proposed in this
document is based on the distribution of the Poten-
tial Vegetation Types in Nepal; the mapped distri-
bution of the vegetation is the result of more than
30 years of work by botanists and foresters and is as
such a remarkably detailed work.
The tree planting zone system is the practical
application of knowledge accumulated on the
distribution of vegetation types and plant species.
Through application of the system an understan-
ding of the distribution of populations of the spe-
cies will be accumulated, and this understanding
will be of direct usefulness to farmers in Nepal and
will help towards improving the productivity of
the farmland and plantations in Nepal.
We would recommend that the tree planting
zone system should be regarded as a tool that can
be improved through sensible use. In the short
term it can help a decentralised seed distribution
system become more efcient and enable provi-
sion of reliable information to farmers; and in the
longer term it can provide better information on
how to increase the productivity of species useful
to farmers.
4.3 Indigenous Knowledge and Revision of
Planting Zones as Increased Experience
is Gained
The seed supply will initially be from selected sour-
ces in farmland (and for some species also in natu-
ral forest). Many farmers will have experience with
use of these sources including nding appropriate
planting sites for the species. This is relevant for
understanding the microsite matching that the seed
zoning system does not describe, e.g. if species pre-
ferentially grow on either south (drier and sunnier)
or north (moister and shadier) aspects. The seed
source descriptions will compile information from
the farmers that constitute each seed source.
It has already for some species been reported that
farmers recognise several varieties, sometimes with
different elevation distribution (Ficus semicordata -
Khasro Khanyu and Rai Khanyu (Amatya, 1989),
and Bauhinia purpurea, two types based on leaf cha-
racters (Thapa, 1994; TISC staff, personal commu-
nication). This information should be compiled
as part of the preparation of seed source descripti-
ons and should be taken into account when seed
zoning for individual species is prepared.
As increased understanding is gained, it will
become relevant to develop more species-specic
recommendation for establishment and use of seed
sources within Nepal. However, as long as a decen-
tralised seed distribution system is being develo-
ped, the planting zones based on vegetation types
will probably remain the backbone of the tree seed
sector in Nepal.

37
REFERENCES
Aalbæk, A. and Kananjii, B. 1995. Tree seed zones for Sudan. Sudan Tree Seed Project and Danida
Forest Seed Centre. Humlebaek, Denmark.
Amatya, S.M. 1989. Variation of Ficus semicordata Buch. Ham. Ex Smith Sensu Loto, its Taxonomy,
Distribution and Use as a Fodder Tree in Nepal. M.Sc. Oxford University, UK
Ashton, P.S. 1995. Towards a regional forest classication for the humid tropics of Asia. In Box, E.O
et al. (eds). Vegetation Science in Forestry, pages 453-464. Kluwer Academic Publishers. Netherlands.
Buijtenen, van J.P. 1992. Fundamental Genetic Principles. Pages 29-68 in Fins, L., Friedmand, S.T. and
Brotschol, J.V. Handbook of quantitative forest genetics. Kluwer Academic Publishers, Boston, USA.
Campbell, R.K. 1986. Mapped genetic variation of Douglas-r to guide seed transfer in southwest
Oregon. Silvae Genetica 35, 2-3, 85-96.
Campbell, R.K. 1991. Soils, Seed-zone Maps, and Physiography: guidelines for seed transfer in
Southwestern Oregon. Forest Science 37: 973-986.
Dhakal, L.P, Lillesø, J.P.B., Jha, P.K., Aryal, H.L. and Kjaer, E.D. 2001. Addressing small holders demand
for propagation material of woody species. Part. II: Guidelines and workplan. HMG/Danida
NARMSAP/TISC, Kathmandu, Nepal.
Diallo, R. Balima, Sina, S., Thomsen, A. and Graudal, L. in prep. Seed Zones for Burkina Faso. Centre
National de Semences Forestieres (CNSF), Burkina Faso and Danida Forest Seed Centre. Denmark.
Dobremez, J.F. 1976. Le Nepal ecologie et biogeography. Editions du Centre National de la Researche
Scientique, Paris, France.
Eis, S. 1986. Seed zones in Thailand, an environmental interpretation. The Embryon 2: 40-48. Muak
Lek, Thailand.
Eldridge, K., Davidson, J., Harwood, C. and van Wyk, G. 1994. Eucalypt domestication and breeding.
Clarendon Press, Oxford, UK.
Eriksson, G., Namkoong, G. and Roberds. J.H. 1993. Dynamic gene conservation for uncertain futures.
Forest Ecology and Management 62: 15-37.
Finley, K.W. and Wilkinson, G.N. 1963. The analysis of adaptation in a plant breeding programme.
Australian Journal of Agric. Research 14: 742-754
Haman, A., Koshy, M.P., Namkoong, G. and Ying, C.C. 2000. Genotype X environment interactions
in Alnus rubra: developing seed zones and seed transfer guidelines with spatial statistics and GIS.
Forest Ecology and Management 136: 107-119.
Hansen, C. Pilegaard, Therkildsen, P. and Joy, B. 2000. Tree seed zones of Kalimantan. Indonesia Forest
Seed Project and Danida Forest Seed Centre. Humlebaek, Denmark.
Hansen, L.N., Kjaer, E.D. and Dhakal, L.P. 2001. Genetic variance in Dalbergia sissoo in Nepal.
Manuscript.
Ives, J.D. and Messerli, B. 1990. The Himalayan dilemma (Second edition). Reconciling development
and conservation. Routledge, New York, USA.
Jackson, J.K. 1994. Manual of Afforestation in Nepal. FORESC, MOFSC, Kathmandu, Nepal. Second
Edition.
Kjaer, E.D. and Foster, G.S. 1996. The economics of tree improvement of Teak. Technical Note 43.
Danida Forest Seed Centre, Humlebaek, Denmark.
Kjaer, E.D. 1997: Sustainable use of forest genetic resources. The Arboretum. Horsholm. Denmark.
Langlet, O. 1936. Studier over tallens fysiologiska variabilitet och des samband med klimatet. Medl.
Statens Skogsforsogsanstalt 29: 1-188. Sweden.
Lewontin, J.W. 1984. Adaptation. In: Sober (editor): Conceptual Issues in Evolutionary Biology,
p.235-251. MIT Press.
Lillesø, J.P.B., Dhakal, L.P., Jha, P.K., Aryal, H.L and Kjaer, E.D. 2001. Addressing small holders demand
for propagation material of woody species. Part I: Analysis and Strategy Proposal. HMG/Danida
NARMSAP/TISC, Kathmandu, Nepal.
Lillesø, J.P.B. and Kaumi, S.Y. 1993. Baseline survey for tree seed project Uganda. Danida Forest Seed
Centre. Humlebaek, Denmark.
Lindgren, D. and Ying, C.C. 2000. A model integrating seed source adaption and seed use. New Forests
20: 87-104.
Linhart, Y.B. and Grant, M.C. 1996. Evolutionary signicance of local genetic differentiation in plants.
Annual Review of Ecology and Systematics 27:237-277.
LRMP. 1986. Land Systems, Land Utilization and Agriculture Forestry Reports. Land Resources Map
ping Project Kenting Earth Sciences Ltd. Ottawa, Canada.
LSMP. 1993. Livestock Sector Master Plan Volume I-IV, 1993. HMG Department of Livestock/Asian
Development Bank/APROSC. Kathmandu. Nepal.
Mandal, A.K. and Gibson, G.L. (editors) 1998. Forest Genetics and Tree Breeding. CBS Publishers.
New Delhli. India.

38
Matheson, A.C. and Cotterill, P.P. 1990. Utility of genotype x environment interactions. Forest Ecology
and Management 30: 159-174.
Matheson, A.C. and Raymond, C.A. 1984. Provenance x environment interaction: its detection,
practical importance anduse with particular reference to tropical forestry. In Barnes, R.D. and Gibson,
G.L.J. (eds.). Proceedings of a joint work conference on provenance and genetic improvement
strategies in tropical forest trees held in Harare, Zimbabwe, April 1984. Commonwealth Forestry
Institute, Oxford, UK.
Mouna, O. 1990. Population Genetics in Forest Tree Improvement. In: Plant population genetics,
Breeding and Genetic Resources (Edited by Brown, A.H.D., Clegg, M.T., Kahler, A.L. and Weir,
B.S.), p 282-298. Sinauer Associates Inc. Massachusetts.
Namkoong, G. 1969. Non-optimality of local races. I: Proceedings of the 10th Southern Forestry Tree
Improvement Conference. Springeld, VA. US National Tree Improvement Society.
Nayaju, R.P and Lillesø, J.P.B. 2001. Nation-wide Climatic Tables. HMG/Danida NARMSAP/TISC,
Kathmandu, Nepal.
Raymond, C.A. 1989. Developing breeding populations for temperate eucalypts in relation to natural
population structure. Pages 423-424 in Gibson, G.L., Grifn, A.R. and Matheson, A.C. (eds) Breeding
Tropical Trees: population structure and genetic improvement strategies in clonal and seedling
forestry. Proc. IUFRO Conference, Pattaya, Thailand 1988. Oxford Forestry Institute, UK.
Rehfelt, G.E. 1983. Ecological adaptations in Douglas r populations. Canadian Journal For. Research:
13: 626-632.
Roberds, J.H., Hyun, J.O., Namkoong , G. and G. Rink. 1990. Height response function for white ash
provenances grown at different latitudes. Silvae Genetica 39: 121-129.
Shrestha, T.B., Shrestha, R., Lillesø, J.P.B., Dhakal, L.P. and Tamrakar, P.R. 2001. Overview of the
Elaboration of Iso-potential Forest Type/Ecological Zone Map and Species Database. HMG/Danida
NARMSAP/TISC, Kathmandu, Nepal.
Shrestha, T.B. (in prep.) Guide to eld identication of vegetation types in Nepal. HMG/Danida
NARMSAP/TISC, Kathmandu, Nepal.
Stainton, J.D.A. 1972. Forests of Nepal. John Murray Ltd. London, UK.
Thapa, B. 1994. Farmers’ ecological knowledge about management and use of farmland tree fodder
resources in the Mid-hills of Eastern Nepal. PhD. Thesis. University of Wales.
Thapa, H. B., Lillesø, J.P.B. and Poulstrup, E. 2000. Species leaets for 131 woody species. Danida Tree
Improvement and Silviculture Component. NARMSAP. Kathmandu, Nepal.
Walter, H. 1971. Ecology of tropical and subtropical vegetation. Oliver and Boyd. Edinburgh. UK.
Westfall, R.D. 1992. Developing seed transfer zones. Pages 313-398 in Fins, L., Friedmand, S.T. and
Brotschol, J.V. Handbook of quantitative forest genetics. Kluwer Academic Publishers, Boston, USA.
Yadav, R.P. and Dhakal, A. 2000. Leasehold forestry for poor. An innovative pro-poor programme in
the hills of Nepal. Policy Analysis in Agriculture and related resource management. Series No. 6.
HMG Ministry of Agriculture/Winrock International. Kathmandu, Nepal.
Zobel, B.J. and Jett, J.B. 1995. Genetics of wood production. Springer Verlag. Berlin.

39