-reviewed open-ac

BioRisk 3: 137-160 (2009) Wace A es ee
doi: 10.3897/biorisk.3.5 RESEARCH ARTICLE BioRisk

www.pensoftonline.net/biorisk Biodiversity & Ecosystem Risk Assessment

Complex ex situ - in situ approach for conservation of
endangered plant species and its application to
lris atrofusca of the Northern Negev

Sergei Volis', Michael Blecher’, Yuval Sapir’

I Life Sciences Department, Ben Gurion University of the Negev, Israel 2. Ein Gedi Nature Reserve, Israel Na-
ture and Parks Authority, Israel 3 Porter School for Environmental Studies and Department of Plant Sciences,
Tel Aviv University, Israel

Corresponding author: Sergei Volis (volis@bgu.ac.il)

Academic editors: LJ. Musselman, E Krupp | Received 4 February 2009 | Accepted 14 December 2009 | Published 28 December 2009

Citation: Volis S, Blecher M, Sapir Y (2009) Complex ex situ - in situ approach for conservation of endangered plant
species and its application to Jris atrofusca of the Northern Negev. In: Krupp E Musselman LJ, Kotb MMA, Weidig I (Eds)
Environment, Biodiversity and Conservation in the Middle East. Proceedings of the First Middle Eastern Biodiversity
Congress, Aqaba, Jordan, 20-23 October 2008. BioRisk 3: 137-160. doi: 10.3897/biorisk.3.5

Abstract

We introduce a novel approach for conservation of endangered plant species in which ex situ collections
maintained in natural or semi-natural environment are a part of a complementary ex situ — in situ con-
servation strategy. We provide detailed guidelines for 1) representative sampling of the populations; 2)
collection maintenance; and 3) utilization for im situ actions. Our approach is the first that explicitly takes
into account ecologically significant (i.e. adaptive) variation of plants in both ex situ and in situ conserva-
tion actions. We propose that an important part of the conservation strategy is preserving both neutral
and adaptive genetic diversity through a quasi im situ conservation approach. Finally, we demonstrate this

approach using a critically endangered plant species, [ris atrofusca from the northern Negev, Israel.

Keywords

In situ, ex situ, conservation strategy, relocation, translocation, local adaptation

Introduction

A large body of theoretical and empirical work has been devoted to particular questions
about optimal conservation, such as the importance of inbreeding depression (Hedrick
and Kalinowski 2000, Keller and Waller 2002), population size (Barrett and Kohn

Copyright S.Volis, M. Blecher, Y. Sapir. This is an open access article distributed under the terms of the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

138 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

1991, Ellstrand and Elam 1993), isolation (Newman and Tallmon 2001), genetic di-
versity (Lande and Barrowclough 1987; Newman and Pilson 1997) and outbreeding
depression (Hufford and Mazer 2003, Tallmon et al. 2004). But, to-date there were
limited attempts to conceptually unite different aspects of population viability as part
of a conservation methodology. Such unification is especially lacking for ex situ conser-
vation. In this paper we review the state-of-art in ex situ conservation with an emphasis
on its utilization within a more general strategy having a final in situ output. Then we
introduce a detailed approach for conservation of endangered species that integrates
ex situ and in situ conservation as complementary and that could be used as a tool for
finding an efficient solution to a particular conservation task. Finally, we present a
study where this approach is applied for an endangered plant species.

Ex situ conservation

Ex situ conservation methods samples genetic diversity of species using certain criteria
and store/propagate the collected material outside the natural environments in which
the species grows (Heywood and Iriondo 2003). Importance of ex situ collections for
conservation in situ was realized when collections in botanical gardens and arboreta
helped implementation of population management and recreation (Falk 1987; Given
1987; Millar and Libby 1991). At the same time, limitations of their usefulness be-
came evident. ‘The latter include poor genetic or demographic management almost
inevitably resulting in genetic erosion, artificial selection and spontaneous hybridiza-
tion. To prevent/reduce negative effects of genetic drift, inbreeding depression and
mutational meltdown, that all happen as a result of small (effective) population size of
a collection, sampled individuals must be maintained separately or through controlled
breeding and pedigree design. This introduces other limitations of ex situ collections,
such as space limitations and high cost of maintenance.

A need of a conceptually sound link between conservation-oriented ecological
and genetic research, and its routine application to ex situ management has been
recognized (Maunder et al. 2004a). Ex situ conservation needs biologically effective,
financially realistic and easy-to-use guidelines that can be applied to a wide range of
situations. The development of such guidelines must take into consideration basic
issues of conservation biology. Traditionally, the germplasm sampled for ex situ col-
lection is supposed to represent potential adaptive variation within a species (Brown
and Briggs 1991, Brown and Marshall 1995, von Bothmer and Selberg 1995). In
case of limited resources for collecting and maintaining plants, minimal sampling
can precede additional sampling, which will be performed upon availability of more
resources (Brown and Briggs 1991). The key issue is choosing a limited, but repre-
sentative, number of populations, using the correct criterion. This criterion, in our
view, should be ecologically significant (i.e. not the potentially adaptive, but the cur-
rently adaptive) variation. Therefore, research that allows detection of spatial pattern
of morphological, life history and fitness traits, should be the first priority tool for

Complex ex situ - in situ approach for conservation of endangered plant species... 139

providing sampling guidelines (Husband and Campbell 2004). Although variation
revealed with molecular markers can provide valuable insights into the importance
of different non-selective processes in species evolution, this information is secondary
for making conservation decisions.

An endangered species is usually represented by small and isolated populations that
already underwent strong effects of genetic drift and/or inbreeding, i.e. comprise a lim-
ited number of genetically different individuals (Aguilar et al. 2008). Open pollination
in a sample from such a population may result in inbreeding depression due to high
probability of mating between genetically identical or closely related genotypes. On
the other hand, interbreeding of individuals from environmentally dissimilar habitats
planted in close proximity often leads to outbreeding depression. ‘These two risks rarely
apply to obligate or predominant selfers, but can be extremely important for outcross-
ing, and especially for self-incompatible species. Outbreeding depression is an opposite,
as compared with beneficial gene flow between genetically differentiated (isolated) pop-
ulations, hybridization process. The parents do not necessarily have to be taxonomically
distinct, viz. be recognized as different subspecies. Therefore, for an endangered spe-
cies, creation of ex situ collections and decisions about suitable material for relocation/
reintroduction should take into account the potential risks of inbreeding/outbreeding
depression, in addition to local adaptation and spatial structuring of adaptive variation.

The above issues should be considered as the basic principles in developing an ex
situ conservation approach that would be an integral part of a more general strategy
with an ultimate final zz situ output. Several approaches combining ex- and in situ
conservation were proposed in the past, but none is satisfying as a general conservation
strategy. For instance, the inter-situ approach proposes an off-site collection maintained
within the natural habitat. This approach was considered potentially promising, but
was not tested and no detailed methodology for practical use was developed (Husband
and Campbell 2004). A slightly different approach are the “forest gene banks” (Uma
Shaanker and Ganeshaiah 1997, Uma Shaanker et al. 2001, 2002). In this concept, a
particular existing population acts as an in situ sink, into which genetic material from
several source sites is introduced and maintained. ‘Thus the genetically diverse sink
population serves as a repository of the species gene pool and at the same time allows
for random interbreeding. This approach might be useful in certain cases (lack of local
adaptation, low genetic diversity, self-incompatability), but may lead to outbreeding
depression when locally adapted genotypes are brought together. Therefore this con-
cept cannot be used in a general application.

“Quasi in situ” conservation

Here, we propose the use of ex situ collections in natural or semi-natural environments
as a part of a complex ex situ — in situ conservation strategy. Here below we provide
detailed guidelines for 1) representative sampling; 2) collection maintenance; and 3)
utilization for in situ conservation actions. The novelty of our approach is in that it ex-

140 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

plicitly takes into account potential local adaptation of plants in both ex situ and in situ
conservation actions. An integral part of this strategy is preserving the species genetic
diversity (both neutral and adaptive) through “quasi in situ” conservation.

The proposed strategy starts with an analysis of the species distribution to identify
potential locally adapted populations or population groups. This analysis is crucial for
understanding the extent of local adaptation and its spatial pattern. As intensity of lo-
cal selection varies, either gradually with increase in distance or abruptly with change
in a habitat, only knowledge of the local selection regimes can tell us whether material
is adapted or not. Two main procedures exist to identify local adaptation: transplanta-
tion experiments (e.g. Joshi et al. 2001, Volis et al. 2002), and tests for outbreeding
depression, when a locally adapted genotype is crossed with plants originating from
elsewhere (Hufford and Mazer 2003). If local adaptation is important, the introduced
genotypes must fit into the local biotic and abiotic conditions, i.e., it should come
from within the area defined as that of intensive local selection, or from a habitat with
closely similar local selection effects.

Experimental determination of a scale of local adaptation (Fig. 1) is a highly desired
second step of the procedure. A potential for local adaptation and its spatial scale can be
roughly predicted from knowledge of a species’ breeding structure and life history. Self-
pollination and short seed or pollen dispersal distance are known to be associated with
a smaller scale of local adaptation than outcrossing and long-distance seed or pollen
dispersal (Linhart and Grant 1996). However, these considerations are too general to

Breeding, life history Scale of local adaptation Conservation implications

sampling introduction
inbreeding
seed/potlen | MALML|| IN YT
a =
dispersal Small-scale adaptation, discrete pattern 3 Ae
distance 2g re es
ov o| >
; at ee [aes
Small-scale adaptation, clinal pattern Sts BS) eae] Mee
ec eB] Yc
ce yu =| 3s] ¢
Oo 0 y 5 oO
MMs 8) 2 eS
: F 9s > Oo} of U
Large-scale adaptation, discrete pattern ago ool (©)
es 3
Rn) |}

outcrossing Small-scale adaptation, clinal pattern

Figure |. A scheme of relationship between (i) species properties (breeding structure, life history) and (ii)

scale of local adaptation, and implications of (i) and (ii) for sampling and introduction. The scheme shows
that predominant outcrossing and large dispersal distance are associated with large scale local adaptation,
necessitating increased size, genetic diversity and connectivity of introduced populations.

Complex ex situ - in situ approach for conservation of endangered plant species... 141

be used as guidelines for conservation of a particular species because of numerous other
factors, such as patterns of environmental variation (e.g. discrete or clinal), number
of available habitats, a species’ evolutionary potential, time since the last colonization,
etc. Therefore, an experimental assessment of the scale of local adaptation is crucial.
Such a test should include all habitats where the species is found (discrete variation)
or locations along an environmental gradient (clinal variation). Of course, logistical
considerations may limit a range of habitats or locations to be tested, or prevent testing
for local adaptation at all. In this case variability in morphology, phenology and life his-
tory traits across a species’ range must be known. This variability matching important
environmental parameters (e.g. temperature, soil type, rainfall) or being associated with
distinct habitats or vegetation communities are indications of local adaptation.

If phenotypic or genetic variation is spatially structured, even though it is not a
result of natural selection, it may represent distinct evolutionary lineages within spe-
cies, being an important part of the species’ diversity. This variation, although neutral,
is important for preserving species evolutionary potential, i.e. ability to adapt to future
climate or habitat changes. As neutral genetic variation in many cases is not reflected
in phenotypic or in molecular markers variation, its existence must be presumed when
a species consists of populations not connected through gene flow.

With knowledge about spatially structured adaptive and neutral variation, a sampling
design is worked out. The optimal design for ex situ collection is a stratified one, with a
lower level (of neutral variation) nested within a higher one (of adaptive variation). Prac-
tically, this means that for a species present in several habitats (e.g. soil types, regions of
different aridity, vegetation communities), sampling in each habitat must include several
geographically isolated populations. A representative number of spatially separated indi-
viduals must be collected in each population, to ensure sampling of different genotypes.

The next step is planting the germplasm sampled as living collection. In the quasi
in situ approach, a choice of a site must take into account local adaptation (tested
or presumed). This means that there must be a close environmental match of ex situ
location with locations of sampled natural populations. A close match can be biologi-
cally meaningful only if the ex situ location is in a natural (or at least semi-natural)
environment. In addition, this site should be protected by national law and practically
(regularly) inspected by rangers. In our view, different classes of strictly protected ter-
ritories are the areas that fully satisfy these requirements. An optimal design would
be planting several populations representing a particular eco-geographical region or
habitat in a protected area in the same eco-geographical region or habitat. A repre-
sentative number of genetically different individuals per population should be planted
separately at distances, allowing subsequent identification of planted genotypes. ‘This
is important for both estimating rate of survival and enabling controlled pollination, if
necessary. Following the recommendations of the Center for Plant Conservation in the
United States (Maunder et al. 2004a), number of plants per population should be 10
to 50, and five populations per habitat or eco-region should be sampled.

A comparison between quasi in situ and traditional ex situ conservation in botani-
cal gardens is summarized in Table 1. We argue that the quasi in situ approach provides

142 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

better representation of genetic diversity, increases chances of germplasm survival, and
better suits the purpose of propagating material for im situ conservation actions.

In situ conservation aims at either enhancement of existing populations or crea-
tion of self-supported new populations via reintroductions and translocations, using
sampled or propagated material (reviewed in Bottin et al. 2007, Menges 2008). When
a natural population exists, or existed in the recent past, a choice of material is quite
straightforward. A large population from the same or the geographically closest popu-
lation within the same habitat would be the best choice. If, however, relocation is
planned, viz. introduction of material into locations where a population never existed,
a decision is more problematic. There are many examples of unsuccessful relocation
into sites that seemed highly suitable and were located in close proximity to a location
where natural populations had been extirpated (e.g. Holland 1980, Morton 1982,
Cranston and Valentine 1983). This implies that the recommendation by Schaal and
Leverich (2004) to use for relocation a large sample from the closest population repre-
senting the same habitat, should be treated with caution and only applied when data
on a species’ environmental requirements are very limited. A much better option is a
limited relocation within an experimental framework, and a full relocation in those
sites where survival and reproduction are high.

As soon as the relocation sites are chosen, material for propagation may be taken
from natural or ex situ collections. The major issues are required origin, genetic diversity
and quantity. Acquiring sufficient quantity of propagation material (seeds, bulbs, root
cuttings, saplings) from the closest natural population can negatively affect the latter's
growth rate and threaten its viability. In addition, single source material may lead to
inbreeding depression and high susceptibility to diseases. Quasi in situ collection may
effectively solve both problems. Plants, grown in the collection, are (presumably) lo-
cally adapted and genetically different as they originate from several populations. Ad-
ditionally, naturally occurring cross-pollination of plants in a collection should neither
lead to breakdown of co-adapted gene complexes, nor to dilution of local adaptation
because all plants in the collection originated from the same environment and no
maladapted genes will participate in recombination and segregation. Therefore, the
offspring of cross-pollination in a collection should well suit the relocation purpose,
and can be collected in large quantities to meet the needs for successful relocation.

The last step in the quasi in situ strategy is determination of spatial parameters of
introduced populations (size, distance from the nearest population) and monitoring of
relocation success. Again, as with choosing a relocation site, experimentation should

Table |. A comparison between quasi in situ and ex situ conservation.

Parameter Ex situ Quasi in situ

Space for maintaining the collection _| Limited Less limited

Suitability of environment Usually un-suitable Suitable

Maintenance and renewal of material | Artificial Natural

Cost High Very low and only at the initial stage

Complex ex situ - in situ approach for conservation of endangered plant species... 143

be a common practice when several populations of different size are planted and moni-
tored over a number of years.

Application: Iris atrofusca of the northern Negev as a case study

Tris atrofusca Baker (Fig. 2) belongs to the section Oncocyclus (Siems.) Baker (Iridaceae)
that are characterized by dense clonal growth and conspicuous large, mostly dark flow-
ers that grow individually on a stem (Avishai and Zohary 1980, Sapir et al. 2002).
Eight species of Oncocyclus that grow in Israel (Feinbrun-Dothan 1986, Danin 2004)
have high conservation priority (Sapir et al. 2003) and are included in the Red Data
Book of the country (Shmida and Pollak 2007). Jris atrofusca is one of the most threat-
ened species of Oncocyclus irises in Israel these days (Shmida and Pollak 2007).

Tris atrofusca is relatively widely distributed. It occurs from the northern Negev
in the south to the Golan heights in the north. This distribution is the widest of all
Oncocyclus species in Israel. Identification is not always easy. Sapir et al. (2002) showed
that morphologically it does not differ from its closest relatives, J. haynei and I. petrana.
Morphological traits are also associated with the aridity gradient (Arafeh et al. 2002, Sa-
pir et al. 2002). However, morphological and genetic analyses indicated that /. atrofusca

Figure 2. Leaf fans, flowers and rhizomes of Jris atrofusca from the Goral Hills, northern Negev (water-
color by Irene Blecher © 2006).

144 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

populations of the northern Negev form a cluster within the general pattern (Arafeh et
al. 2002, Sapir et al. 2002), and might even represent a separate taxon (Kushnir 1949).

The habitats of J. atrofusca in the northern Negev are the most vulnerable through-
out its distribution. In the last decade, /. atrofusca populations of the northern Negev
have been suffering mainly from anthropogenic disturbance that decreased population
sizes, with some populations becoming extinct. These disturbances include urbaniza-
tion, infrastructure works, intensive and extensive agriculture, overgrazing, forestry
works, and illegal Bedouin settlements. Recently, a plan for expanding the area of Beer
Sheva, the main town of the northern Negev, is threatening the largest and the densest
populations of 1. atrofusca, which grow in Goral Hills, north of the town. These issues
lead to urgent research of /. atrofusca populations in the Negev. Here we present studies
we did under the guidelines we drew above for the quasi in situ conservation approach.

Methods
Research area

Tris atrofusca grows in the northern Negev in two main groups of populations: in Goral
Hills (central coordinates: 31°18'N 34°48'E), north to Beer Sheva, and Arad Valley
(central coordinates: 31°16'N 35°07'E) (Shimshi 1979/8).

The two regions differ in climate. While the Goral Hills area is above the 200 mm
isohyet (semi-arid conditions), the Arad Valley is close or below to the 150 mm iso-
hyet, which indicates arid conditions (Atlas of Israel 1985, Jaffe 1988).

The topography of the Goral Hills area is mostly slopes of shallow hills. The angles
of the slopes are up to 20%. ‘The soil is a shallow calcareous lithosol overlying frac-
tured Eocene limestone (Shimshi 1979/80). Depressions between the hills are filled
with shallow loessial soil. Arad Valley, on the other hand, is a relatively flat plain (with
wadis and gullies), covered with Quaternary aeolian loess of considerable thickness
(> 2 m), with some isolated outcrops of calcareous lithosols (Shimshi 1979/80), which
are mostly the heads of insulated hills.

In hard and fissured limestone and dolomite with calcareous lithosol some of the
rain water penetrates the soil and is accumulated in the fissures and crevices, where it is
protected from direct evaporation. Loessial soils have a different moisture regime. Due
to the high moisture holding capacity of the fine-grained substratum, some of therain
water is absorbed by the upper soil layers, but most of this water is consequently lost
by direct evaporation from the soil surface (Danin 1988).

Inventory and demographic observations

A field survey based on previous knowledge on the distribution of the species in the
northern Negev was conducted in 2006. The survey aimed at documenting the pre-

Complex ex situ - in situ approach for conservation of endangered plant species... 145

cise locations of populations, the distributing area of each population, and to record
ecological conditions and anthropogenic impacts. In each population, clumps of /ris
atrofusca were recorded for their size, estimated by the diameter of the clump.

To assess population growth rates, we started, in 2006, a detailed census of two
populations representative of two regions, the Goral Hills and Arad Valley (Gvaot
Goral, G-G, and Tel Arad, T-A, respectively) (Fig. 3). Two permanent observation
plots were established: 120 m x 6 m (G-G) and 60 m x 6 m (T-A), respectively).
In March 2006 we counted all individuals within the plots and classified them
as either immatures (1* or 2"! year juveniles with a single fan), vegetative (non-
flowering, but with > 1 fan) adults or reproductive adults (Fig. 4). We marked
each established clump (=genet) of /. atrofusca individually, measured its diameter
and counted the number of leaf-fans (=ramets). Also in April 2007 and 2008, we
recorded number, size, and reproductive status of adults in the plots. Each season
(2006, 2007 and 2008) we calculated the average number of fruits per repro-
ducing plant (i.e. a clump comprising >1 ramets), average number of seeds per
fruit, and resulting fecundity. Measurements and counting were done when plants
started to senesce.

Since the actual age of individual plants of 1 atrofusca in the field can not be de-
termined, the population structure analysis was based on the number of individuals

31°25'0"N

31 20'0"N

37°15'0"N

eer Sheva

Sh ae Sa

e| as ere, no Aris atrofusca ca surveyed p population Bi : | SS a, sf V3 : ne ek
Maia Loh = ~ Demographic observations plots Ve BS (eee Pood |?
j ig ; | i” L ge scale experiment sites Bah é A 2 at Be Z|

eE ig 1:130,000 all Seis: experiment site Wwgies sae ad

J ay
<— i eK =o 4 £ a ao 5 aby: i)
34°40°0"E 34°45'0"E 34°60 '0"E 34°55'0"E 35°0'0"E 35°5'0"E

Figure 3. Map of populations surveyed, experimental relocation sites and populations in which perma-

nent demographic observations plots were established.

146

Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

seedling |-year juvenile 2-year juvenile 3-year juvenile

Figure 4. Plants of J. atrofusca followed from seed germination on and representing different age-stages

(seed, seedling, juvenile and vegetative adult).

in the different ontogenetic stages of the life cycle. During the years 2005 to 2008 we
followed the germinated seeds to distinguish age-stages in J. atrofusca. We identified

the
BD
os

following age-stage categories (Fig. 4):

Seeds;

Seedlings (individuals developed shortly after germination of seeds, with cotyle-
dons and often with one pair of leaves);

Juveniles (individuals with a single leaf fan comprising more than two leaves and
having a poorly developed root system); a major difference between 1-year and
2-year juveniles is in the development of the root system, where the latter starts to
develop a rhizome;

Vegetative adults or immatures (non-flowering individuals with more than two leaf
fans and fully developed root system, which has a developed rhizome);
Generative adults (individuals bearing flowers).

Oncocyclus plants are dormant between the end of the flowering season (end of

spring) and the start of the next winter. Our observations indicate that plants can stay
dormant during not only summer, but also during the next growing season, from fall
to spring (Volis & Blecher, pers. obs.). This adds another ontogenetic stage for adults —
dormants, which will be verified as more demographic data become available.

Complex ex situ - in situ approach for conservation of endangered plant species... 147

Adult plants can be distinguished from juveniles in the field by the compact assem-
blage of leaf fans, which emerge close to each other from the below-ground rhizome
(Fig. 4). All these leaf fans are genetically identical individuals (ramets) that may be-
come independent plants after fragmentation of the mother rhizome (genet).

Plant sizes and number of fans per clump were log-transformed and analyzed across
years and habitats using repeated-measures ANOVA. Only alive and non-dormant
adults were analyzed for these two traits.

Soil seed-bank and germination trials

We created three experimental permanent soil seed banks and started monitoring the
fate of the seeds in fall 2005 at Tel Arad National Park (Fig. 3). Seeds of 1. atrofusca
were collected by MB from plants in proximity to the plots under observation in 2005.
The seeds were buried at three sites along the slope in: (1) plastic trays filled with soil
of the site of transplanting and containing one spikelet per cell (221 seeds per tray, one
tray per site), placed about 2 cm below ground level, and (2) in furrows (100 seeds per
furrow, two furrows per site).

Similar soil seed banks were established in fall 2006 at Gvaot Goral site (Fig. 3).
Two trays, each containing 221 locally collected seeds, were buried at the top and the
bottom of the hill. The experimental soil seed banks were monitored for seed germina-
tion during 2006 to 2008.

Effect of rhizome initial size on growth and flowering

This experiment was conducted during the growing seasons of 2005 and 2006. We
used only rhizomes with one distinct bud to test the effect of initial rhizome weight
on probability of flowering. We also measured several morphological traits to iden-
tify potential morphological indicators for probability of flowering. Rhizomes were
planted in 3-liter pots filled with loess soil, one weighted rhizome per pot. Pots were
placed 25 cm apart in a nethouse, and watered regularly with 2-liter/hour drippers,
one dripper in each pot. During the experiment (November — April) plants were get-
ting natural rainfall (208 mm) plus supplementary watering (equivalent to 95 mm of
rainfall), to compensate for the higher evaporation rates from the pots. At first sign of
leaves senescence the following measures were taken: length and width of the longest
leaf, diameter at the base of the leaf fan, number of ramets, and number of flowers.
After complete drying out of above-ground biomass the rhizomes were dug out and
weighted. The sample size was 204 plants. Large rhizomes rescued by MB in 2005
from the Goral Hills (road-building strip for new railroad tracks) that represented a
group of (potentially) independent ramets were cut into pieces and used in this and
the following experiments.

148 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

Creation of quasi in situ gene banks

Between 20 to 50 large genets of 1. atrofusca, comprising many ramets, were sam-
pled from four populations from Goral Hills and Arad Valley regions. Populations
were chosen based on: (1) the threat of habitat destruction — the populations chosen
were critically endangered (construction, agriculture, etc.) and required immediate
relocation; and (2) their representation for the distributional range of /. atrofusca in
the northern Negev. The plants were planted in two replicates at both Tel Beer Sheva
and Tel Arad National Parks (Fig. 3) that represent ecological conditions like those of
Goral Hills and Arad Valley areas, respectively. However, as the two population groups
(Goral and Arad areas) were found to differ in habitat, demography and morphology
(Shimshi 1979/80; Blecher 2007 and this study), we decided that Tel Beer Sheva and
Tel Arad National Parks will harbor populations from their respective regions only,
and the populations planted outside their region of origin will be relocated to the ref-
uge in their respective region at the next stage of the project. Meanwhile, we are moni-
toring the transplanting success of plants of different origin across the two regions.

Relocation experiments

Rapid disappearance of J. atrofusca populations in the Negev necessitates measures
of species conservation, such as relocation to safe areas, protected by law. In order
to determine species’ habitat preferences we set relocation experiments at two scales,
large (tens of kilometers) and small (hundreds of meters), respectively, using rhizomes
rescued from sites of habitat destruction and immediate threat for the plants, i.e. from
populations that required relocation.

Large scale relocation experiment

Two sets of five large (> 20 g) and 15 small (5-10 g) rhizomes of 1. atrofusca of
Arad Valley and Goral Hills origin were planted in six locations that embraced
the whole species range in the Negev and beyond it. The locations were: KKL
experimental site near Ofakim, Tel Beer Sheva National Park, Nevatim Basis, La-
hav North Nature Reserve, Tel Arad National Park, Har Amasa Nature Reserve
(Fig. 3). In spring 2007 and 2008 we recorded the numbers of surviving plants,
flowers and fruits per plant.

Small scale relocation experiment

Rhizomes rescued in spring 2006 in the Goral Hills area (construction of new railroad
tracks) were planted in fall 2006 in sets of 62 rhizomes at 22 microhabitats in Lahav

Complex ex situ - in situ approach for conservation of endangered plant species... 149

North Reserve (Fig. 3, 9). Each set comprised the following size classes: <5 g (14),
5-10 g (10), 10-20 g (23), 20-30 g (10), 30-40 g (3) and >40 g (2). In spring 2007

and 2008 we recorded the numbers of surviving plants, flowers and fruits per plant.

Results
Distribution and habitats

The results of this survey (Table 2) clearly show that in the Arad Valley about a third
of the plants of /. atrofusca grow in wadis. In the Goral Hills area, no population was
found in wadis. Detailed geographical interpretation of data on /. atrofusca survey in
the northern Negev, including categorization of the populations for protection pur-
poses, is presented in Blecher (2007) with proposals for new protected areas and en-
largement of the existent Parks.

The two populations studied (Gvaot Goral and Tel Arad) differed in average plant
density, estimated in spring 2006 (0.88 vs. 0.30 plants/m? in G-G and T-A, respective-
ly). There was a marginally significant difference between two populations in clump
size, with clumps at G-G being consistently larger during three years than at T-A. The
number of ramets per genet exhibited significant season/population interaction. ‘This
trait was more constant over the years at G-G than at T-A (Table 3).

The two populations differed in the stage structure, with juveniles comprising
86%, 70% and 46% vs. 23%, 24% and 20% of established plants (over three years)
in G-G and T-A populations, respectively (Table 4). Percentage of flowering adults
and average fruit-set per reproducing plant over three years were higher in G-G than
in T-A population (40413% vs. 67+10% and 1.6140.38% vs. 0.52+0.30%, respec-
tively). The two populations also dramatically differed in fecundity (Table 4).

Germination

Low germination rates were observed in the soil seed banks established in 2005 at Tel
Arad National Park. During three seasons, 2005-6, 2006-7 and 2007-8, no germina-
tion event was recorded in any of the buried trays with seeds. In furrows, no germina-
tion was observed in 2005-6 and 2006-7, but in 2007-8 germination fraction was

Table 2. Number of plants and distribution of Jris atrofusca in two geographic regions.

Total distributing area | Total cover of | Total number | Hills and slopes | Wadis
(Hectare) plants* (m7) of clumps

Goral Hills | 19.3 186 1968 0
Arad Valley | 34.8 439 4716 1785

* Total cover of plants is a summary of all clumps.

150 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

4%, 15% and 12% (hill top, middle and foot, respectively) In the Gvaot Goral site,
where 2 trays (experimental soil seed banks) were buried in fall 2006, one seed germi-
nated in the following winter (season 2007—2008). These results suggest strong seed
innate dormancy in the first year after dispersal with increase in germination fraction

in following years.

Table 3. Repeated measures ANOVA of the effects of population and season on clump size and number
of ramets per genet (top) and means + S.E for each season (bottom). G-G — Gvaot Goral population,

T-A — Tel Arad population.

Source of variation DF F F

Clump size Ramets per genet
Population 1 2.91F 2.38 ns
Error 109
Season 2 0.59 ns 1.82 ns
Season * Population 2 0.91 ns Gel5**
Error 218
Season Clump size (cm) Ramets per genet

G-G T-A G-G TA

2006-7 23.142.6 19.8+2.5 18.342.5 15.0+2.5
2007—8 25125 22.6+2.8 £3023 21.842.8

** pb < 0.01; f p < 0.10; ns not significant.

Table 4. Life table for two populations of J. atrofusca during the seasons 2005—6, 2006—7 and 2007-8.
The table does not account for soil seed bank present at the start of observations. Numbers are for the

whole plot.
Population/stage Season
2005—6 2006—7 2007—8
Gvaot Goral (G-G)
Seeds - 3238 2318
Juveniles 357 142 90
Non-reproducing adults I 16 70
Reproducing adults 43 46 35
Fecundity (seeds/repr. plant) AI 50.4 16.4
Tel Arad (T-A)
Seeds - 992 199
Juveniles 25 Pia) 22
Non-reproducing adults 66 oD 67
Reproducing adults 19 38 19
Fecundity (seeds/repr. plant) aluZ D125 8.4

— = not estimated

Complex ex situ - in situ approach for conservation of endangered plant species... 151

Effect of rhizome initial size on growth and flowering

The results of this experiment (Table 5) clearly show that sexual reproduction (i.e.,
production of a flower) in . atrofusca depends on the rhizome weight and two size-re-
lated parameters, namely length of the leaves and base diameter. The minimal rhizome
weight for flowering appears to be around 2.7 g-3.0 g, but probability of flowering for
such plants is less than 10% (Fig. 5). The optimal rhizome weight with reasonably high
probability of flowering (around 50%) is above 4 g (Fig. 5).

Creation of quasi in situ living collections

Two years after planting, survival rates in two living collections were equally high,
approximating 100%. Percent of flowering plants was substantially higher in Tel Beer

Table 5. Results of multiple logistic regression testing effect of five predictor variables on probability of

flowering.
Parameter Wald Statistics p
Intercept 0.24 0.6210
Rhizome weight 30.76 < 0.0001
Leaf length 13.46 0.0002
Leaf width Q=15 0.7008
Base diameter 4.20 0.0403
Number of leaf-fans 1.26 0.2618
[5] Number of plants in experiments
40 —@— Prportion of plants flowered l
35 0.9
0.8
4% 30
c 0.7
ue =
a 25 06 9
‘O =
ve 1 20 0.5 ve)
Oo
4 04
= 15 y
= 0.3
Z 10
0.2
é LI |
4 1;
2 t

Rhizome weight (g)

Figure 5. Proportion of plants flowering and total number of plants of different rhizome weight classes

in the experiment described in the text.

152 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

Sheva National Park than Tel Arad National Park, while a difference in percentage of
plants that produced mature fruits was less pronounced (Fig. 6). Plants of native geo-
graphic origin had no advantage at either location.

A high percentage of flowering plants at one location (Tel Beer Sheva National
Park) indicates high potential seed productivity in the living collection. The low per-
centage of plants that set fruits appears to result from low numbers of pollinators in
this area. Therefore the genetic refuges can be a source of seeds for relocation once
artificial pollination is provided.

Large scale relocation experiment

High survival rates were observed in the first year after introduction at all locations,
and the highest number of reproducing plants was observed at Ofakim and Lahav
North (Fig. 7). At Har Amasa, plant above-ground biomass was browsed by grazing
livestock, thus, assessment of reproduction was not possible. Two years after the intro-
duction a difference in plant survival rates among the locations started to become obvi-
ous (Fig. 7). Grazing at Har Amasa again prevented assessment of plant reproduction.

The most unexpected and counterintuitive result was zero reproduction at two
experimental sites established in close proximity to the natural populations, G-G and
T-A. In both cases experimental locations were established on an adjacent hill slope.
This indicates the importance of microscale conditions for relocation success. At the

Tel Beer Sheva Tel Arad
Flowering ; Fruiting
9 |
Tessie
Okey a se |
(00) em Em 2st |
Oo 1
80. = O 3
7 60° = :
40. : : =
. O00
0 T 1 ! 1 T El T
NM FOS N ™M FOS
(37) (21) (26) (67) (30) (18) (28) (61)

Populations

Figure 6. Population name, number of genetically distinct individuals (in parentheses), and percentage
of flowered and fruited plants two years after creation of two quasi in situ collections in Tel Beer Sheva and
Tel Arad National Parks, respectively.

Complex ex situ - in situ approach for conservation of endangered plant species... 153

40 survival 2007
BO ee ee ad SEE SE Bia annon--------- =
20 - &£&
10 +- 2 ca Bogue mai #3 “
W”) A : a ey
ni} 0
a 54
+ 4 gm reproduction
3 al
2 |
| + ; :
0 Fi T T T T T 1
Ofakim Tel Beer Nevatim Lahav Tel Arad Har
Sheva North Amasa
survival 2008
”
Y
Cc
@& [J Goral origin - large
ao. i: | reproduction [| Arad origin - large
6- [| Goral origin - small
45 Arad origin - small
2 =| :
0 7 = = ia T 1 mal 1 1
Ofakim Tel Beer Nevatim Lahav TelArad Har Gvaot
Sheva North Amasa Goral

Introduction sites

Figure 7. Survival and reproduction of [. atrofusca one year after experimental introduction. Ten large
(> 20 g) and 30 small (5—10 g) rhizomes of Goral Hills and Arad Valley origin were introduced at each site.

154 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

_ N w A

Oo (o) fo) (o) Oo
|
=

Lis
¥ | 34567 8 9 10111213 1415161718192021 22
c L] Total
MOODS Wrst an. li aR aaa eT ha veut aE nabs an ee ae
a 60 Mi Flowered
50 2008 GH Fruited

7 8 9 101112131415 161718192021 22
Introduction plots

Figure 8. Survival and reproduction of /. atrofusca one year after experimental introduction at Lahav
North Reserve. Sixty-two rhizomes of Goral Hills origin with equal representation of different size classes

were introduced at each introduction plot.

Lahav North Reserve both survival and reproduction of plants were consistently high
during two years of observations. This strongly supports our decision made in 2006 to
start experimental microscale relocation at the Lahav North Reserve. At the Ofakim
site reproduction was high in the first year but dropped dramatically in the second
year after planting.

Small scale relocation experiment

The number of plants observed at the 22 microsites in the Lahav North Reserve one
year after introduction ranged from 25 to 52 plants (out of 62 introduced) with no
significant difference between microsites (G-test, Gay = 9.2, p > 0.05; Fig. 8). The mi-
crosites did not differ either in the number of plants that set fruits (G-test, Sess = 32.4,
p > 0.05), but differed in numbers of flowering plants (G-test, G,, ,, = 36.8, p < 0.05).

Two years after the introduction, the range of surviving plants per microsite was
between 33 and 53, generally higher than the previous year records. This indicates
that some plants were not counted in the first year census, perhaps due to rhizomes
dormancy. As in the first year, no microsite difference was observed for plant survival
= 20.0, p > 0.05), but numbers of reproducing plants
= 74.2, p < 0.001; Fig. 8).

in the second year (G-test, Cai

were significantly different among microsites (G-test, G,. ,,

Complex ex situ - in situ approach for conservation of endangered plant species... PD

34°51'30 34°52'0 34°52'30"E

31°23'30"N
0o"N

31

31°23'0"N

34°51'30"E 34°52'0"E 34°52'30"E

Figure 9. Map of the Lahav North Reserve with 22 microsites at which identical sets of 1. atrofusca

rhizomes were planted.

Contrary to the first year, no flower set fruit at any microsite in the second year. All the
flowers were consumed by grasshoppers and caterpillars, indicating the important role
of biotic interactions at the Lahav North site.

Discussion
Distribution in the Negev and population demography

The observed differences in stage structure (i.e. the frequency of life-cycle stages)
between the two populations correspond to two types of demographic behavior, the
“invasive” or “dynamic” (Gvaot Goral) and “normal” or “stable” (Tel Arad) (Rabot-
nov 1969, 1985, Oostermeijer et al. 1994). The former is characterized by a higher
proportion of immature plants relative to the adults, while in the latter the adults
predominate and the young individuals are low in number. These two population
types are usually associated with different succession stages of the local vegetation
community, but in the two J. atrofusca populations studied no difference was ap-
parent with respect to the succession stage. One major difference between the two
population locations is in aridity, and the observed difference in a proportion of im-
mature plants appears to be due to higher survival of seedlings (although survival of
juveniles does not differ) at less xeric Gvaot Goral site. Nonetheless, it is too early to

156 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

draw conclusions about the long-term dynamics of these two populations. The latter
requires a multi-year census coupled with records of annual rainfall and assessment
of grazing pressure.

The two groups are separated from each other by a distance of ca 20 km with hard-
ly any gene flow between them. Different environmental conditions (soil, rainfall) and
anthropogenic impact (intensive grazing vs agriculture) may have caused differential
selective responses in the two regions. Therefore an /. atrofusca conservation strategy
must be based on the assumption that ecologically important (i.e. adaptive and caused
by biotic/abiotic environmental variation) exists within /. atrofusca in the Negev and
regional criterion (Goral Hills vs Arad Valley range subdivision) is a first approxima-
tion of this variation. This assumption has several implications for this species ex and
in situ conservation.

Ex situ implications

If plants in two regions are adapted to different environmental conditions, sampling
and maintenance of living collections must be done for each region separately. Mix-
ing or physical proximity of plants having different regional origin must be prevented.
If, due to logistical limitations, plants are maintained in the same location, measures
must be taken to prevent spontaneous hybridization (e.g. removal of immature fruits).
On the other hand, interbreeding of plants originating from different populations
within the same region, is desired to decrease risk of inbreeding depression and self-
incompatibility. The latter negative effects were detected in fragmented populations of
I. bismarckiana, a close relative to 1. atrofusca (Segal et al. 2007).

In our study, plants from five populations of J. atrofusca in the Negev were planted
at two National Parks, creating two duplicates of the same living collection. After
careful study of region-specific environmental conditions, anthropogenic impacts and
population demography we concluded that we should divide our collection based on a
regional criterion. In spite of the initial proximal planting of plants from two regions,
no spontaneous hybridization occurred during two years of collection maintenance
because of the precautions described above.

After removal and re-planting of populations representing the Goral Hills area,
into their region-specific location in Tel Beer Sheva National Park, and populations
from the Arad Valley area into their region-specific location at Tel Arad National
Park, the next step in applying quasi in situ conservation, using plants in the living
collections for seed propagation. In a case of low fruit set due to limited availability
of natural pollen vectors (Eucera bees; Sapir et al. 2005) randomly applied artificial
pollination should be performed. The seeds obtained can be treated to reduce strong
innate dormancy, and germinated in mass. Young plants with rhizomes acceding 4 g
can be used for in situ actions, which should be performed with plants of proper
regional origin.

Complex ex situ - in situ approach for conservation of endangered plant species... 157

In situ implications

Rapid destruction of /. atrofusca natural environment in the Northern Negev due to
heavy anthropogenic impact on the one hand, and lack of nature reserves that contain
populations of J. atrofusca in the Negev, on the other hand, leaves very limited options
for conservation of this species. Declaration of new protected areas in the northern Ne-
gev is very problematic because of economic, demographic and political issues. There is
virtually no vital alternative to relocation, i.e. introduction of the species into presum-
ably suitable protected areas with no previous records of the species. At the same time,
the choice of such areas for /. atrofusca in the Negev is limited.

It is too early to draw conclusions from our relocation experiments, started in
2006, about factors limiting species distribution. At least several more years are needed
for reliable conclusions, because among-year fluctuations, as well as long-term effects
should be considered. However, some general considerations about a choice of reloca-
tion material can be done even at this stage. Using regional subdivision as a guideline
for successful relocation, creation of a new population within Goral Hills or Arad Val-
ley region should be done using material from the same region. If, however, a new pop-
ulation location can not be ascribed to one of these regions, possible options include
material of single regional origin (either Goral or Arad) or a mix of two. Although lack
of local adaptation is not an issue here, hybridization of two ecotypes may result in a
disruption of coadapted gene complexes, high genetic load and low average fitness of
plants in the new population. As a result relocation success may be low. Without reloca-
tion experiments, it is impossible to decide which of two plant origins is more suitable.

Conclusions

We conclude that the proposed approach assessing ecological importance, and using
this information for both, ex and in situ conservation is suitable for endangered species
that are distributed over areas with complex and variable ecological conditions. We
hope that detailed guidelines developed from the above approach for: (1) representa-
tive sampling of populations; (2) collection maintenance; and (3) utilization for in
situ actions will be used as a tool for efficiently solving specific conservation problems.

Acknowledgments

We would like to thank Israel Nature and Parks Authority for financial support and
various help during fulfillment of this project. We thank A. Adout, L. Burdeniy, G.
Dror, A. Dvir, M. Gvoa, D. Hawlena, S. Issacharoff, A. Katz and M. Shushan for help
in field and nethouse work, E. Even-Haim and A. Gvoa for help in creation of living
collections, K. Lev for help in maps preparation.

158 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

References

Aguilar R, Quesada M, Ashworth L, Herrerias-Diego Y, Lobo J (2008) Genetic consequences
of habitat fragmentation in plant populations: susceptible signals in plant traits and meth-
odological approaches. Molecular Ecology 17: 5177-5188.

Allen WH (1994) Reintroduction of endangered plants. BioScience 44: 65-68.

Arafeh RMH, Sapir Y, Shmida A, Iraki N, Fragman O, Comes HP (2002) Patterns of genetic
and phenotypic variation in Iris haynei and I. atrofusca (Iris sect. Oncocyclus = the royal
irises) along an ecogeographical gradient in Israel and the West Bank. Molecular Ecology
11: 39-53.

Atlas of Israel (1985) Atlas of Israel, Survey of Israel, Tel Aviv (3rd edition).

Avishai M, Zohary D (1980) Genetic affinities among the Oncocyclus irises. Botanical Gazette
141: 107-115.

Barrett SC, Kohn JR (1991) Genetic and evolutionary consequences of small population size in
plants: implications for conservation. Pp. 3-30 In: Falk DA, Holsinger KE (Eds) Genetics
and Conservation of Rare Plants. Oxford University Press, New York.

Blecher M (2007) Iris atrofusca of the Nothern Negev: survey, preliminary results. Israel Nature
and Parks Authority.

Bottin L, Cadre S Le, Quilichini A, Bardin P, Moret J, Machon N (2007) Re-establishment
trials in endangered plants: A review and the example of Avenaria grandiflora, a species on
the brink of extinction in the Parisian region (France). Ecoscience 14: 410—419.

Brown ADH, Briggs JD (1991) Sampling strategies for genetic variation in ex situ collections
of endangered plant species. Pp. 99-122 In: Falk DA, Holsinger KE (Eds) Genetics and
Conservation of Rare Plants. Oxford University Press, New York.

Brown ADH, Marshall DR (1995) A basic sampling strategy: theory and practice. Pp. 75-111
In: Guarino L, Ramantha Rao VR (Eds) Collecting plant genetic diversity: technical guide-
lines. CAB International Wallington, UK.

Cranston DM, Valentine DH (1983) Transplant experiments on rare plants species from Upper
Teesdale. Biological Conservation 26: 175-191.

Danin A (1988) Flora and vegetation of Israel. Pp. 129-157 in Y. Yom-Tov and E. Tchernoy,
eds. The Zoogeography of Israel. Junk, Dordrecht.

Danin A (2004) Distribution atlas of plants in the flora Palaestina area. Israel Academy for Sci-
ence and Humanities, Jerusalem.

Ellstrand NC, Elam DR (1993) Population genetic consequences of small population size: im-
plications for plant conservation. Annual Review in Ecology and Systematics 24:217—242.

Falk, D. A. 1987. Inegrated conservation strategies for endangered plants. Natural Areas Jour-
nal 7: 118-123.

Feinbrun-Dothan N (1986) Flora Palaestina. Israel Academy of Sciences and Humanities, Jerusalem.

Given DR (1987) What the conservationist requires of ex situ collections In: Bramwell D, Ha-
mann O, Heywood VH, Synge H (Eds) Botanic gardens and world conservation strategy.
Academic Press, London.

Hedrick PW, Kalinowski ST (2000) Inbreeding depression in conservation biology. Annual
Review in Ecology and Systematics 31: 139-162.

Complex ex situ - in situ approach for conservation of endangered plant species... 159

Heywood VH, Iriondo JM (2003) Plant conservation: old problems, new perspectives. Biologi-
cal Conservation 113: 321—335.

Holland PG (1980) Transplant experiments with trout lily at Mont St. Hilaire, Quebec Journal
of Biogeography 7: 261-267.

Hufford KM, Mazer SJ (2003) Plant ecotypes: genetic differentiation in the age of ecological
restoration. Trends in Ecology and Evolution 18: 147—155.

Husband BC, Campbell LG (2004) Population responses to novel environments: implications
for ex situ plant conservation In: Guerrant EOJ, Havens K, Maunder M (Eds) Ex situ plant
conservation: Supporting species survival in the wild. Island Press, Washington.

Jaffe S (1988) Climate of Israel. Pp. 129-157 In: Yom-Tov Y, Tchernov E (Eds) The Zoogeog-
raphy of Israel. Junk, Dordrecht.

Joshi J, Schmid B, Caldeira MC, Dimitrakopoulos PG, Good J, Harris R, Hector A (2001) Local
adaptation enhances performance of common plant species. Ecology Letters 4: 536—544.

Keller LE, Waller DM (2002) Inbreeding effects in wild populations. Trends in Ecology and
Evolution 17: 230-241.

Kushnir T (1949) Leaf spot diseases of /ris in Palestine. Palestine Journal of Botany, Jerusalem
series 4: 230-233.

Lande R, Barrowclough GF (1987) Effective population size, genetic variation, and their use in
population management. Pp. 87-124 In: Soule ME (Ed) Viable populations for manage-
ment. Cambridge University Press, Cambridge.

Linhart YB, Grant MC (1996) Evolutionary significance of local genetic differentiation in
plants. Annual Review of Ecology and Systematics 27: 237—277.

Maunder M, Havens K, Guerrant EOJ, Falk DA (2004a). Ex situ methods: a vital but underused
set of conservation resources. Pp. 3-20 In: Guerrant EOJ, Havens K. Maunder M (Eds) Ex
situ plant conservation: Supporting species survival in the wild. Island Press, Washington.

Maunder M, Hughes C, Hawkins JA, Culham A (2004b) Hybridization in ex situ plant col-
lections: conservation concerns, liabilities, and opportunities. Pp. 325-364 In: Guerrant
EOJ, Havens K, Maunder M (Eds) Ex situ plant conservation: Supporting species survival
in the wild. Island Press Washington.

Menges ES (2008) Restoration demography and genetics of plants: when is a translocation suc-
cessful? Australian Journal of Botany 56: 187—196.

Millar CI, Libby WJ (1991) Strategies for conserving clinal, ecotypic, and disjunct population
diversity in widespread species. Pp. 149-170 In: Falk DA, Holsinger KE (Eds) Genetics
and conservation of rare plants. Oxford University Press, New York.

Morton JK (1982) Preservation of endangered species by transplantation. The Canadian Bo-
tanical Association Bulletin 15:32.

Newman D, Pilson D (1997) Increased probability of extinction due to decreased genetic effec-
tive population size: Experimental populations of Clarkia pulchella. Evolution 51: 354-362.

Newman D, Tallmon DA (2001) Experimental evidence for beneficial fitness effects of gene
flow in recently isolated populations. Conservation Biology 15: 1054-1063.

Oostermeijer JGB, Vantveer R, Dennijs JCM (1994) Population structure of the rare, long-
lived perennial Gentiana pneumonanthe in relation to vegetation and management in the

Netherlands. Journal of Applied Ecology 31: 428-438.

160 Sergei Volis, Michael Blecher & Yuval Sapir / BioRisk 3: 137-160 (2009)

Rabotnov TA (1969) On coenopopulations of perennial herbaceous plants in natural coenoses.
Vegetatio 19: 87—95.

Rabotnov TA (1985) Dynamics of plant coenotic populations. Pp. 121-142 In: White J (Ed)
The population structure of vegetation. Handbook of vegetation science, Junk, Dordrecht.

Sapir Y, Shmida A, Fragman O (2003) Constructing red numbers for endangered plant species
- Israeli flora as a test case. Journal for Nature Conservation 11: 91-107.

Sapir Y, Shmida A, Fragman O, Comes HP (2002) Morphological variation of the Oncocyclus
irises (/ris : Iridaceae) in the southern Levant. Botanical Journal of the Linnean Society
139: 369-382.

Sapir Y, Shmida A, Ne’eman G (2005) Pollination of Oncocyclus irises (Iris : Iridaceae) by night-
sheltering male bees. Plant Biology 7: 417-424.

Schaal B, Leverich WJ (2004) Population genetic issues in ex situ plant conservation. Pp. 267—
285 In: Guerrant EOJ, Havens K, Maunder M (Eds) Ex situ plant conservation: Support-
ing species survival in the wild. Island Press, Washington.

Segal B, Sapir Y, Carmel Y (2007) Fragmentation and pollination crisis in the self-incompatible
Tris bismarckiana (Iridaceae), with implications for conservation. Israel Journal of Ecology
and Evolution 52: 111-122.

Shimshi D (1979/80) Two ecotypes of Jris atrofusca Bak. and their relations to man-modified
habitats. Israel Journal of Botany 28: 80—86.

Shmida A, Pollak G (2007) Red Data Book: Endangered Plants of Israel. Israel Nature and
Parks Authority.

Tallmon DA, Luikart G, Waples RS (2004) The alluring simplicity and complex reality of
genetic rescue. Trends in Ecology & Evolution 19: 489—496.

Uma Shaanker R, Ganeshaiah KN (1997) Mapping genetic diversity of Phyllanthus emblica:
Forest gene banks as a new approach for in situ conservation of genetic resources. Current
Science 73: 163-168.

Uma Shaanker R, Ganeshaiah KN, Nageswara Rao M, Ravikanth G. (2001) A new approach
to conservation of genetic resources of forest trees: promise and processes. Pp. 263—271 In:
Uma Shaanker R, Ganeshaiah KN, Bawa KS (Eds) Forest genetic resources: Status, threats
and conservation strategies. Oxford and IBH Publishing Co. Pvt. Ltd

Uma Shaanker R, Ganeshaiah KN, Nageswara Rao M, Ravikanth G. (2002) Forest gene banks
— a new integrated approach for the conservation of forest tree genetic resources. Pp. 229—
235 In: Engels JMM, Brown AHD, Jackson MT (Eds) Managing plant genetic resources.
CABI Publishing, Nosworthy. UK.

Volis S, Mendlinger S, Ward D (2002) Differentiation in populations of Hordeum spontaneum
along a gradient of environmental productivity and predictability: life history and local
adaptation. Biological Journal of the Linnean Society 77: 479-490.

von Bothmer R, Selberg O (1995) Strategies for the collecting of wild species. Pp. 93-111 In:
Guarino L, Ramantha Rao VR (Eds) Collecting plant genetic diversity: technical guide-
lines. CAB International Wallington, UK.