Abstract: As a part of the characterization and the valorization of plant genetic resources of fodder and pastoral interest in Algeria, fourteen natural populations of Hedysarum coronarium L. were subjected to a morpho-physiological study. Ecological factors (altitude and rainfall) of the originating environment of the different populations were considered. Thirty four characteristics related to seed weight, seedling emergence, vegetative development, flowering and formation of fruiting heads were included in the study. The results indicate the existence of variability within the species and between the populations, especially for the rate of seedlings at the second simple leaf stage, the development of the height axis, the end of the bloom and the full fruiting heads characteristics which are significant. This would be linked to the ecological factors of the environment of origin, especially the altitude which would constitute the most important factor. The results of principal components analysis show off some important characteristics, particularly the numbers of leaves and ramifications per plant, which oppose themselves to certain characteristics linked to the breeding (flowering and formation of fruiting heads). H. coronarium L. should be developed, particularly in the isolated and disinherited areas, where they could adapt themselves with advantage.
INTRODUCTION
Sulla (Hedysarum coronarium L.) is a short-lived, perennial legume originating from the Mediterranean where it is used widely for hay, silage and greenfeed (Whyte et al., 1953; In: Douglas, 1984; Kernick, 1978; In: Douglas, 1984). It is also an ornamental (Battandier and Trabut, 1902; Bonnier, 1934) and bees plant (Bonnier, 1934).
Meneghetti et al. (1997) specify that this member of the tribe Hedysareae is distributed in Spain, Algeria and southern Italy.
In Spain, the growth of sulla is restricted to Menorca Island, where it forms monophytic meadows and to the southern province of Cadiz, where it is mainly associated with other natural forage species (Rodriguez-Navarro et al., 1991).
In Algeria, this species is very common in the Tell Constantinois and very scarcely found in the other regions, such like El-Kantara, Alger, Oran (Quezel and Santa, 1962). In many parts of Northeast Algeria, H. coronarium occurs together with other natural forage such as the genera Trifolium, Medicago, Astragalus, Phalaris, Scorpiurus, Avena, Hedysarum, Lolium, Picris, Carthamus, Sonchus, Dactylis, Festuca, Hordeum, Leontodon, Lepturus, Lotus, Onobrychis (Issolah et al., 2001).
Sulla is quite rich in terms of local varieties, especially in Italy (Villax, 1963). Spanish and Italian varieties are in general more sensitive to cold than those of the North Africa, However, a variety with white flowers was reported to be more rustic, but giving lower yield (Villax, 1963).
Several authors throughout the world indicated that Hedysarum coronarium L. is characterized by a stable chromosome number (2n = 16) whereas two chromosome numbers (2n = 16 and 2n = 18) were encountred in Algerian populations of this species (Issolah et al., 2006).
The area of cultivation of sulla includes Portugal, Spain, Morocco, Algeria, Tunisia, Egypt, Italy, former Yugoslavia and Greece (Flores et al., 1997). Hedysarum coronarium is an exotic legume to Australia (Casella et al., 1984). It was introduced into New Zealand in the early 1950s largely for use in soil conservation, but its potential as a forage was more fully realised in 1982 (Watson, 1982; In: Krishna et al., 1990).
In the genus Hedysarum, little is known about the nature of genetic variability in diploid species (Baatout et al., 1991a) and about the taxonomic relationships of the different taxa (Baatout et al., 1991b).
The objective of this study is to understand the heriditary mechanism and quantify the genetic diversity available within Algerian populations of Hedysarum coronarium L. of diverse biogeographical origins using a number of morpho-physiological characteristics. This research follows previous studies conducted on fodder legumes species in Algeria (Issolah and Abdelguerfi, 1998, 1999a, 1999b, 2000, 2001, 2002, 2003; Issolah et al., 1993, 2000, 2001, 2006; Issolah, 2006).
MATERIAL AND METHODS
Collecting missions were undertaken by the National Institute of Agronomic Research (INRAA) throughout the North-East of Algeria in 1998 (Issolah et al., 2001). Several populations of Hedysarum coronarium L. were collected (Issolah et al., 2006). Fourteen ones were subjected to a trial on two consecutive years (1999/2000 and 2000/2001) (Fig. 1).
The populations were sown (13/11/99) at the experimental station of Mehdi Boualem of Baraki (INRAA) at the altitude of 18.5 m, in a subhumid zone. The soil was clayey-mudy with pH of 8.12. The temperature was ranged from 10.9°C (m1) to 21.7°C (M1) for the first year (99/00) and from 11.7°C (m2) to 22.9°C (M2) for the second year (00/01). The annual rainfall was about 412 mm (99/00) and 408.1 mm (00/01), respectively.
After determinating the 1000 seed weight (WTS) of the populations, these are sowed on lines separated of 1.5 m, with 35 individuals per population. The individuals were separated between each others of 40 cm. We used a randomized complete block design with two replications. Thirty four characteristics were evaluated: 1000 seed weight (WTS); date of the emergence of the seedlings (DES); rate of seedlings at the first simple leaf stage (12/12/99: SL1, 22/12/99: SL2); rate of seedlings at the second simple leaf stage (27/12/99: SL3, 03/01/00: SL4, 09/01/00: SL5); rate of seedlings at the third simple leaf stage (17/01/00: SL6, 24/01/00: SL7, 01/02/00: SL8); Number of leaves (composed per three leaflets) per seedling (23/02/00: NL1, 06/03/00: NL2, 16/03/00: NL3); Number of leaflets per plant (26/03/00: NLP); Number of ramifications per plant (23/02/00: R1, 06/03/00: R2, 16/03/00: R3, 26/03/00: R4); winter development (31/03/00): height of axis (HA1) and width of axis (WA1); spring development (24/05/00): height of axis (HA2) and width of axis (WA2); autumnal development (10/01/01): height of axis (HA3) and width of axis (WA3); winter development (28/03/01): height of axis (HA4) and width of axis (WA4); flower bud (FB); first inflorescence (FI); full Bloom (FB1); end of the bloom (EB); duration of the bloom (DB); number of inflorescences per plant (NFP); appearence of the first fruiting head (FH1); full formation of fruiting heads (FH).
| Fig. 1: | Origin areas of fourteen natural populations of Hedysarum coronarium L. in Algeria. A) Bejaia; B) Setif; C) Skikda; D) Guelma; E) Tarf; F) Souk Ahras (Issolah et al., 2006) |
| Table 1: | Ecological characteristics of the originating environment of fourteen natural populations of Hedysarum coronarium L. in Algeria (Issolah et al., 2006 completed) |
The notations concerning the vegetative development were made periodically, at precise dates during the two years (99/00 and 00/01). The notations concerning the bloom and the fruiting heads were made by regular passages (2-3 days), during the first year. In this case, the indicated values are expressed in the number of days after sowing.
The considered characteristics made the subject of statistics treatments (statistical parameters, variance analysis, matrix of correlations, linear regression analysis, principal component analysis) through Statistica (1997), Excel (2000) and Stat-itcf (1988) computer programs. Certain ecological factors such as the rainfall (ANRH, 1993) and the altitude of the natural habitat have been considered (Issolah et al., 2006) (Table 1).
RESULTS AND DISCUSSION
Statistical parameters: Results of statistical parameters of the 1000 seed weight (mother seeds) of the populations, indicate the performance of the population 13/98. As a matter of fact, this one presents the heaviest seeds (6.27 g) whereas the population 28/98 is characterised by the lightest ones (3.89 g). The mean value of the species is 5.02 g. The standard deviation and the coefficient of variation are, respectively 0.6328 and 12.60. Our results correspond to the finding of Villax (1963) who says that the 1000 seed weight in Hedysarum coronarium L. is 5-6 g.
The study of the average weight of one seed within eight species of the genus Hedysarum in Algeria (Abdelguerfi-Berrekia, 1985) showed that H. naudinianum presents the heaviest seeds (17.2 mg). Next come those of H. glomeratum (8.9 mg), H. spinosissimum (8.0 mg), H. pallidum (7.8 mg), H. aculeolatum (7.2 mg), H. flexuosum (5.9 mg), H. carnosum (5.1 mg) and finally H. coronarium (4.9 mg). In the same genus, the mean weight of a seed is comprised between 17.2 and 4.9 mg (Abdelguerfi-Berrekia, 1985). The weight of 1000 seeds (5.3 g) within Hedysarum confertum Desf., which is present in the matorral areas of southeast Spain (Rios et al., 1991), reaveled itself near to that of H. coronarium.
Thus, within the genus Hedysarum, H. coronarium is met among the species with light seeds.
Variance analysis: The results of variance analysis permitted to show variability within Hedysarum coronarium L. (Table 2).
Thus, out of thirty three analysed characteristics, four are significant: the rate of seedlings at the second simple leaf stage (SL3***), the development of the height of axis (OA2**), the end of the bloom (EB*) and the full formation of fruiting heads (FH*).
The population 4/98 was accurately distinguished in regard to important development of the axis height (HA2; HA3), the late appearence of the Flower Bud (FB), the First Inflorescence (FI), the Full Bloom (FB), the end of the bloom (EB), the high number of inflorescence per plant (NFP), the late appearence of the first fruiting head (FH1) and the full formation of Fruiting Heads (FH).
Except the existing variability, we note the importance of the vegetative development of the populations which attains a mean of 222.5 cm (PA4). It can go up to 304 cm in population 18/98. Thus, the covering of the soil revealed itself very interesting in this species.
Matrix of correlations: The matrix of correlations has emphasized very numerous relationships between the morpho-physiological characteristics taken into consideration. The main existing relationships are synthesized as follows (Table 3).
The populations presenting a high 1000 seed weight (WTS), are precocious in terms of first fruiting head appearance.
The populations which seedlings rate at first simple leaf stage (SL1) is high, present also a high rate of seedlings at the second (SL4) and third (SL8) simple leaf stage, a weak development of height axis at the end of winter season (WA1) and during spring period (WA2).
The populations that ensure a good soil re-covering (WA2), present a weak rate of seedlings at the first and second simple leaf stage (SL1; SL4), a high Number of leaflets per plant (NLP), a high number of ramifications per plant (R2; R4), a good initial development in width and height (WA1; HA2), a good final development in height (HA4) and a high number of inflorescences per plant (NFP).
| Table 2: | Results of the variance analysis within fourteen Algerian populations of Hedysarum coronarium L. |
| Min: Mean of a population with the lowest value; Max: Mean of a population with the highest value; Mean: Mean of the species. Signification: *5%; **1%; ***0.1% | |
| Table 3: | Relations between the morpho-physiological characteristics in fourteen (14) Algerian populations of Hedysarum coronarium L. |
| Signification (*): p<0.05 | |
The precocious populations in terms of flower bud appearance, present a high seedlings rate at the second simple leaf stage (SL3), a high number of leaves per seedling (NL1) and a high number of ramifications (R1). They are precocious in regard of appearance of first inflorescence, full bloom, ending bloom, the appearance of first fruiting head and full formation of fruiting heads.
Linear regression analysis: In order to visualize better the relations established between certain morpho-physiological characteristics and some ecological factors (altitude and rainfall) of originating environment of the populations in H. coronarium, we had recourse to regression (Fig. 2 A-F).
The linear model (Fig. 2A) explains at 39.9% the positive correlation between the 1000 seed weight and the altitude of the environment where the populations are originated; this correlation appears significant (y = 0.0011x + 4.5309; R2 = 0.399*) and at 39.03% (Fig. 2B) the negative correlation between the percentage of seedlings at the second simple leave stage and the rainfall of originating environment of populations (y = 0.0655x + 63.606; R2 = 0.3903*).
It shows at 34.83% (Fig. 2C) the negative correlation between the end of the bloom and the altitude that appears significant (y = -0.0081x+198.78; R2 = 0.3483*) and at 30.73% (Fig. 2D) the positive correlation between the end of the bloom and the rainfall of the origin site (y = 0.0236 + 178.99; R2 = 0.3073*).
| Fig. 2: | Linear regression analysis of morpho-physiological characteristics related to ecological factors (altitude, rainfall) of originating environment of H. coronarium populations. A- WTS x altitude, B- Rate of seedlings (second simple leave stage) x rainfall, C- End of the bloom x altitude, D- End of the bloom x rainfall, E- Appearance of the first fruiting heads x altitude, F- Full fruiting heads x altitude |
The linear model (Fig. 2/E) explains at 44.1% the negative correlation which exist between the appearance of the first fruiting head and the altitude of origin site (y = -0.0069x + 185.12; R2 = 0.441*) and at 28.69% (Fig. 2F) the negative correlation between the full formation of fruiting heads and the altitude of the environment of origin (y = -0.0065 + 197; R2 = 0.2869*).
We note that the altitude intervenes on four characteristics (weight of seeds, end of the bloom, appearance of the first fruiting head, full fruiting heads) whereas the rainfall intervenes only on two characteristics (rate of the seedlings at the second simple leave, end of the bloom). We signalize that only one characteristic (end of the bloom) is influenced by both of the altitude and the rainfall. The altitude would constitute the most important ecological factor for its influence on the morpho-physiological variation of the populations within H. coronarium.
Our results show that the populations which present the highest thousand seed weight are originating from the high altitude areas whereas a previous work (Abdelguerfi-Berrekia, 1985) conducted on Algerian populations of the genus Hedysarum showed that it was not possible to establish clear links between characteristics related to fruiting heads, seeds and origin environment. By another way, Abdelguerfi (2001) indicated, in 17 Algerian populations of H. coronarium, that the factors of the environment of origin don’t seem to have an effect on the weight of one hundred joints, the thousand seed weight and the ratio seeds/pods. Nevertheless, the same author signalizes, in H. spinosissimum subsp. capitatum, that the weight of seeds seems to increase with the altitude.
Berrekia and Abdelguerfi (1988) indicate for H. coronarium that the Algerian populations of the sub-humid level are frequently more precocious than those of the humid level. Our present work shows off the importance of the height of the axis (HA2), which the variation revealed itself highly significant, without establishing links with the factors of the origin environment. Nevertheless, a previous study conducted on twenty Algerian populations of H. coronarium indicate that the populations which presented a weak final development of height come from the high altitude areas (Issolah et al., 2001).
In Italy, Sarno et al. (1978) have also shown the influence of high elevation on the diversification of morphological types within H. coronarium.
In Algeria, Abdelguerfi (2001) indicated that in Hedysarum coronarium, the stand type and the development of the orthotropic axis seem to be influenced by high elevation. High-altitude populations have a prostrate stand whereas low-altitude ones have an erected stand. In some species of the genus Hedysarum, some populations were subject to a strong selection pressure from the originating environment and formation of ecotypes seems evident (Abdelguerfi, 2001). In our case, there are not relations between the number of ramifications and the ecological factors (altitude, rainfall) of the originating environment of populations.
Principal component analysis: The analysis of the principal components turns on fourteen populations and thirty three characteristics. The altitude and the rainfall are used as supplementary variables. The plan 1-2 extracts 55.2% from information.
According to the decreasing quality of the representation (Table 4), the most important variables appear as follows: NLP > R2 > R4 > R1 > NL1 > NL2 > R3 > NFP > FI > FH > SL4 > WA2 > SL3 > NL3 > FB > HA2. These sixteen variables have a quality = 52.40%. Six others variables are included between 40.64 and 49.29% and eleven show a rate = 33%.
Globally, the results highlighted the good distribution of the variables through two axis (Fig. 3) and the importance of a certain number of characteristics, in particular the number of leaves and ramifications per plant which are distinguished clearly and are opposed to certain characteristics related to the bloom and the breeding (formation of fruiting heads).
Axis 1 is an axis of precocity and of architecture; the earliest populations with flowering are located on the right and latest ones are on the left.
Axis 2 is an axis of production.
The analysis of the Fig. 3, which represents the dispersion of the populations in the plan of components 1 and 2, shows that populations of the group C (originated from the area of Setif) and those of the group B (originated from the area of Guelma) are stretched along axis 1. Certain individuals are early, others are late. They are heterogeneous populations showing great morphophysiological variation.
The populations of group A (coming from the area of Bejaia) are spread out along the positive part of axis 2. They are especially characterized by a good production of inflorescences.
The cloud of the populations of the group B overlaps with group A, which indicates the existence of individuals or morphologically similar populations.
The population 24/98 (Skikda) with late flowering, is opposed to axis 1 at population 10 (Setif) early and having a high number of leaves (composed per three leaflets) per seedling, a high number of leaflets and ramifications per plant.
| Fig. 3: | Distribution of Algerian populations of H. coronarium in the plan engendered by the first 2 axis of the principal components. Group A: populations No. 2/98, 4/98, 6/98; Group B: populations No. 16/98, 18/98, 20/98, 22/98; Group C: populations No. 8/98, 10/98, 12/98, 13/98; Group D: population No. 26/98; Group E: population No. 24/98; Group F: population No. 28/98 |
Axis 2 opposes the population 4 (Bejaia), characterized by a good spring vegetative development of the height and width axis, a high number of inflorescences per plant and a full late formation of fruiting heads with populations 12/98, 13/98, 8/98 (setif) and 28 (Tarf) presenting the inverse characteristics and with the populations 16/98, 18/98, 20/98, 22/98 (Guelma) having intermediate characteristics.
This analysis highlights the existence of great morpho-physiological variation between the populations coming from the same area, particularly from setif and intraspecific for the characteristics of precocity and production in Hedysarum coronarium L. species.
A previous cytogenetic study conducted on some of these populations revealed that the chromosomic variation (x = 8 and x = 9), established for the first time in H. coronarium L., would be linked to the ecological factors of the originating environment (Issolah et al., 2006). Throughout the length of the chromosomes and the diversity of heterochromatic bands, the Algerian populations would be richer in heterochromatin than Italian populations (Issolah et al., 2006). The environment of the origin of the populations would have an influence on the distribution of the heterochromatin in Hedysarum coronarium (Issolah et al., 2006).
In the same family (Fabaceae), the morphological variability observed in several Algerian species of the genus Trifolium (Issolah et al., 1993) is linked to certain ecological factors (altitude, rainfall, longitude, latitude) of the originating environment of the populations. (Issolah and Abdelguerfi 1998, 1999a, 2000, 2001, 2002; Issolah et al., 2000).
Issolah (1997) and Issolah and Abdelguerfi (1999b) determinated two chromosome numbers (2n = 10 and 2n = 12) in Algerian populations of Trifolium scabrum. The population with 2n = 12 is characterized by a high number of pods by fruiting head and a big size of fruiting heads. Besides, it comes from a marginal area with relatively low altitude and rainfall.
A study conducted on Trifolium glomeratum indicate that early-flowering lines were found in sites with dry, hot springs (short growing seasons), mid-season lines from cooler, moister areas, while the late-flowering lines came from areas with good spring rains and cold winters (Woodward and Morley, 1974).
According to Negri and Veronesi (1987), Italian populations of white clover coming from higher altitudes showed a low degree of spring regrowth and later bloom date.
In native white clover (Trifolium repens) populations in Nova Scotia, Fraser (1991) indicate that the variations in raceme number per plant appeared to be related to latitude of origin.
| Table 4: | Principal component analysis of the morpho-physiological study within fourteen Algerian populations of Hedysarum coronarium L. |
| 1st Column: co-ordinates on the principal axis, 2nd column: square cosinus (quality of the representation) | |
In Trifolium subterraneum, both morphological traits and early dry matter yield are influenced by the environment of origin of the Sicilian populations (Pecetti and Piano, 1993). In another study conducted on the same species, Piano et al. (1993) indicate that native populations have adjusted their adaptative characteristics to the climatic features of the sites where they naturally grow.
Thus, the relations seem to be established between some morpho-physiological characteristics and the originating environment of the populations within certain species of the Fabaceae family.
CONCLUSIONS
The analysis of the morpho-physiological variation of some natural populations of the species Hedysarum coronarium L. in Algeria has for principal objective the understanding of the hereditary mechanism and the diversity of the populations within this species. The existence of variability is clearly demonstrated throughout this study. The results showed the preponderant part of the altitude and the rainfall of the origin environment on a certain number of characteristics linked to the behaviour of the different populations. The established variability would contribute to widen the range of choice in spontaneous populations of Hedysarum coronarium L., particularly for the isolated and disinherited areas, where they could adapt themselves with advantage. More investigations will help us to understand the mechanism of evolution of these populations within the species H. coronarium.
ACKNOWLEDGMENT
We thank Mr. Ali Hamid (ANRH) and Mr. A. Aitéche (INA) for their help during this study.