The genus Eremopyrum (Ledeb.) Jaub. and Spach belongs to tribe Triticeae Dumort., (Poaceae). It is represented by five annual species in Iran: E. bonaepartis (Spreng.) Nevski, E. confusum Melderis, E. distans (K. Koch) Nevski, E. orientale (L.) Jaub and Spach and E. triticeum (Gaertner) Nevski (Bor, 1970). These are widely distributed in dry areas from Morocco in the West to western China in the East (Bor, 1970; Frederiksen, 1991). According to the records in the Flora Iranica, these xerophytic weed species are of open, stony slopes, steppe and semi-desert habitats. The genus grows extensively at North Eastern and North Western parts of Iran. Eremopyrum bonaepartis is an important native pasture species. This is the first study of this genus in Iran. Frederiksen (1991) gave systematic notes and the distribution of the genus.
Chromosome base number is x = 7. Diploid and tetraploid (2n = 14 and 28) position has been observed (Sakamoto 1967, 1972 and 1973). The anatomical structure of this genus has not been studied. Morphological characters are capable in distinguishing different taxa of tribe Triticeae (Baum 1976; Frederiksen 1991; Keshavarzi et al., 2002; Terrell, 1993).
Anatomical studies of grasses have provided some important diagnostic features in coastal and inter-coastal parts (Metcalfe, 1960; Ogundipe and Olatunji, 1992; Keshavarzi et al., 2002). Due to the importance of tribe Triticeae and wheat relatives, the purpose of this paper was to investigate the morphological and anatomical properties of Eremopyrum in Iran. Frederiksen (1991) consider 4 species in this genus but there were mentioned 5 species for flora of Iran. The aim of this study was to consider inter and intra species variations.
MATERIALS AND METHODS
Eremopyrum accessions were collected from different localities in Khorasan, Azerbaijan, Quazvin, Semnan and Tehran provinces (Table 1). 10 individuals were gathered at each sampling site. Specimens are deposited in the herbarium of the faculty of Science at Alzahra University.
The plant samples were identified according to Frederiksen (1991) and Bor (1970). Fresh samples were used for morphological measurements (all ten individual of a population). Thirty two quantitative and 32 qualitative characters of vegetative and reproductive parts of specimens were evaluated (Table 2 and 3). We use basal spikelet of each individual for measurements. Samples were fixed in FAA for anatomical studies. We study the anatomical characters of dorsal epidermis in coastal and inter-coastal regions and also leaf transverse sections. The length and width of the stomata were measured with an ocular micrometer using the dorsal leaf epidermis. Totally 34 quantitative and qualitative anatomical characters were evaluated and measured for all 10 individuals of each population (Table 4).
For grouping the species and populations different methods of cluster analysis
including single linkage, UPGMA and Ward as well as ordination base on Principal
Component Analysis (PCA) were performed on standardized data.
||Voucher details of Eremopyrum sampled in this study
||Studied quantitative morphological characters of Eremopyrum
||Studied qualitative morphological characters of Eremopyrum
||Studied Anatomical Characters for species of Eremopyrum
|C = Coastal, IC = Intercostal
Squared Euclidean distance was used as a measure of similarity in cluster
analysis. Uni-Variate and multivariate statistical analyses used SPSS ver. 9
Quantitative morphological characters: From studied characters, lemma and glume nerve number, palae nerve and lobe numbers show no variation. None variant features are omitted from further analysis. By factor analysis, it was shown that 7 factors are responsible for more than 72% of variation in Eremopyrum species. We choose the first three factors that made 53% of variations. The first factor which causes more than 29% of variation comprises characters as lemma awn Length, length of lowermost and uppermost lemma, length of lemma hair, awn length of lowermost and uppermost lemma, have the most important role in morphological variation in Eremopyrum species in Iran. As second factor there are spike width and length and ligules length. Length of caryopsis and width of lemma take part in third factor and cause 9% of variation. There is a great correlation between length of awn in lowermost and uppermost glume, length of lowermost and uppermost glume, length of upper glume hairs and hairs of lemma.
Dendrogam based on quantitative characters reveals clear separation of taxa. UPGMA cluster was in concordance with Ward method. In this Diagram E. distans have a separate branch. Eeremopyrum bonaepartis var bonaepartis and E. orientale comprise another distinct branch. There are confusions between E. confusum var glabrum and E. bonaepartis var sublanuginosum and E. confusum var confusum. PCA diagram shows also these relations between taxa (Fig. 1). As it could be seen there is a clear separation in E. distans and E. orientale due to 32 quantitative morphological characters.
Qualitative morphological characters: The results of qualitative morphological characters are shown in Table 5. One of the most distinct differences between Eremopyrum species in Iran is Length of Spike (Fig. 2). E. triticeum has the shortest spike and E. bonaepartis var bonaepartis have the longest one.
Factor analysis shows that the first 5 factors cause more than 97% of variation in these taxa. Spike shape, the frequency of glume and lemma hairs are the most important qualitative factors (46% of variation). Caryopsis and ligules shape cause more than 20% of variation (2nd factor). The presence of hairs in glume, spike and lemma are responsible for 16% of variation (3rd factor).
Anatomical observations: In coastal region, the walls of long cells were even (E. distans, E. orientale, E. confusum var. glabrum and E. triticeum) or uneven (E. bonaepartis, E. confusum var. confusum). Silica cells were in pair or single in all species. E. triticeum has only single silica cells (Table 6). Except E. distans and E. bonaepartis var. sublanuginosum other species have prickles. Subsidiaries cells are dome shapes in E. triticeum and E. distans and E. orientale but these are parallel sided in E. bonaepartis and E. confusum (Fig. 3 and 4).
PCA diagram for Eremopyrum Species in Iran by Using
Quantitative Characters (Abbreviations: B: E. bonaepartis, C: E.
confusum, D: E. distans, O: E. orientale)
||Results of evaluation of qualitative morphological characters
in Eremopyrum of Iran
|Codes are as mentioned in Table 3
||Result of Qualitative characters study of dorsal epidermis
in Eremopyrum species of Iran
|Codes are as mentioned in Table 4
Spike morphological Variation in Eremopyrum species
of Iran. A = E. distans, B = E. bonaepartis var. bonaepartis,
C = E. bonaepartis var. sublanuginosum, D = E. triticeum,
E = E. confusum var. glabrum, F = E. confusum var. confusum
and G = E. orientale
In inter-coastal region there are single short cells except in E. orientale. Walls of long cells are only even in E. distans and E. triticeum but both types are present in E. orientale, E. bonaepartis and E. distans) Fig. 3 and 4). Prickles are present in E. bonaepartis var. bonaepartis, E. bonaepartis var. sublanuginosum, E. confusum and E. orientale. There are macro hairs in E. distans, E. bonaepartis var. sublanuginosum and E. orientale. Stomata subsidiaries shape in E. distans, E. orientale and E. triticeum are dome shape. Parallel-sided subsidiaries are seen in other two species of Eremopyrum.
Factor analysis shows three first qualitative factors in dorsal epidermis causes more than 76% of variation. As First factor, frequency of prickles and thickening of long cells are important. Frequency of buliform cells (2nd factor) and presence of prickles (3rd Factor) have also important roles in Eremopyrum variations.
Factor analysis of quantitative characters of dorsal epidermis also shows that first three factors are more important causing more than 60% of variation. In first factor width and length of short cells in intercostal region, stomata length, diameter of stomata opening has the major effects. Width of stomata, prickles number and length of long cells in coastal region, cause the variation in 2nd factor. The observed variation in 3rd factor is due to macro-hair number in inter-coastal and coastal regions.
Flag leaf transverse sections show some homogeneity. All species show two layer
bundle sheaths. Inner layer is complete and the cells are smaller than outer
layer. Mesophyll has a radial differentiation. Shapes of transverse sections
show some differences in Eremopyrum species (Fig. 5
and 6).Factor analysis of quantitative characters of transverse
section shows the importance of three first factors. These factors cause more
than 64% of variations. In First factor width of sclerenchyma, length of buliform
cells, number of vascular bundles and length of sclerenchyma have the most important
role. In second factor number of cells of outer layer of bundle sheath, cause
the 20% of variation. Number of cells of inner layer of bundle sheath (3rd factor)
makes the 14% of variation.
Dorsal epidermis in Eremopyrum species of Iran. A and B =
E. bonaepartis var. bonaepartis, C and D = E. bonaepartis var. sublanuginosum,
E = E. confusum var. glabrum, F, G and H = E. confusum var. confusum, I
and J = E. triticeum, K and L = E. distans. Abbreviations: a = Prickles,
b = Subsidiaries, c = Long cell, d = Silica bodies, e = Intercoastal region,
f = Coastal region, g = Long cell with sinuous walls, H = Dome shaped subsidiaries
Dorsal epidermis in Eremopyrum orientale. A and B =
Bojnourd population, C and D = Beshghardash population. Abbreviations:
a = Subsidiaries, b = Long cells, c = Macro hair, d = Silica bodies, e =
Short cell, f = Prickle, g = Coastal region, h = Intercoastal region
||Different shapes of transverse sections in Eremopyrum
species. A = E. orientale, B = E. confusum var.
Mid rib differences among Eremopyrum species of Iran.
A = E. bonaepartis var subalnuginosum, B = E. bonaepartis
var bonaepartis, C = E. confusum var confusum,
D = E. confusum var glabrum, E = E. orientale, F
= E. distans, G = E. triticeum. Abbreviations: a = Grid of
Scleranchyma, b = outer layer of bundle sheath, c = bulliforme cells and
d = differentiated mesophyl
There are a lot of significant differences in Eremopyrum species. Due to factor analysis awn of glume and lemma, spike shape, caryopsis shape, hair in glume and lemma, the presence of bristles in lodicules and the gap of the palae tip are important diagnostic characters. We can use them in providing an efficient identification key for Eremopyrum species in Iran.
Frederiksen (1991) stated that Eremopyrum has only 4 species in world.
She put E. confusum as a separate subgroup of E. bonaepartis.
We found that E. bonaepartis are distinguished from E. confusum
in Iran by awn position. Some reproductive morphological characters as length
of spike, glume and lemma, hair and gap of the tip of palea and the size of
lodicules in E. confusum are more than E. bonaepartis. Spike shape
is very efficient feature in separating Eremopyrum species in Iran. In
identification key of Eremopyrum species of Iran, some characters as
glume awn, length of glume, presence of prickles and macro-hairs, shape of glume
and lemma are very important. Accessions of E. orientale show some variation
in morphology and anatomy. They should be studied further to find whether there
are any genetic variations.
Anatomical characters are efficient in distinguishing some morphologically similar species group as E. orientale, E. distans and E. triticeum. E. confusum and E. bonaepartis are anatomically distinct. Some differences in number of stomata rows and vascular bundles were shown. Factor analysis shows that length of buli-form cells, number of vascular bundles, number of cells of inner and outer bundle sheaths, frequency of prickles, frequency of short cells, width and length of short cells in inter-coastal region and number of prickles are diagnostic anatomical characters.
Morphological and anatomical variations of Eremopyrum species were studied. The values of characters were evaluated. We find five distinct Eremopyrum species in Iran. Due to the observed differences, it seems inefficient to place E. confusum as a separate subgroup of E. bonaepartis. Further study should focus the genetic variation of some taxonomic confusing species as E. confusum and E. bonaepartis.