| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Centre for Legumes in Mediterranean Agriculture, University of Western Australia, 35 Stirling Highway, Crawley, Western Australia 6009, Australia
b The Hebrew University of Jerusalem, Rehovot 76100, Israel
c CSIRO Plant Industry, Private Bag 5, Wembley, WA 6913, Australia
* Corresponding author (Jens.Berger{at}csiro.au)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Including the cultigen, there are nine annual species in the genus Cicer, and these are usually separated into three or four groups on the basis of genetic distance from C. arietinum (Tayyar and Waines, 1996). The primary gene pool of C. arietinum includes C. echinospermum, and C. reticulatum, the putative wild progenitor (Ladizinsky, 1975). A perennial wild Cicer species, C. anatolicum, is also grouped with the primary gene pool species by some authors (Choumane et al., 2000; Kazan and Muehlbauer, 1991). The second closest group consists of C. bijugum, C. judaicum, and C. pinnatifidum (Tayyar and Waines, 1996). The most distantly related annual wild Cicer species from the cultigen are C. yamashitae, C. chorassanicum, and C. cuneatum.
The world's annual wild Cicer species collection is housed in the genebanks of the Consultative Group on International Agricultural Research (CGIAR) system, as well as those registered with the International Plant Genetic Resources Institute (IPGRI, 2001) (Table 1) . Because of the practice of storing duplicate germplasm across several genebanks for the purposes of security, the totals given in Table 1 considerably overestimate the numbers of accessions held in the world collection. Nevertheless, there are less than 140 accessions within any given annual wild Cicer species (Table 1). In comparison, there are more than 2000 accessions in three undomesticated Triticum species (some of the wild progenitors of wheat, Triticum aestivum L.) held in the National Plant Germplasm System of the USDA alone (USDA-ARS, 2001).
|
Breeders around the world have begun to introgress traits from wild species within the primary genepool of C. arietinum (E. Knights, R. S. Malhotra, S. S. Yadav, personal communication, 2001). In the Australian chickpea-breeding program, this approach has been applied since 1986, and in 2000, nine out of 36 entries in advanced trials had C. echinospermum in their pedigree (E. Knights, personal communication, 2001). Because of this interest, it is becoming increasingly important to review the status of the world collection of annual wild Cicer species, to enable effective exploitation of existing genetic resources, and suggest where germplasm is lacking. At present, the world collection is spread over a number of genebanks (Table 1), each with their own system of identifying accessions. This does not promote transparency, and as a result, it is difficult for researchers to know the material that each is using. The problem is exacerbated because of the practice of making selections from original accessions, each of which are subsequently given unique accession identifiers. Consequently, it is possible for researchers ostensibly evaluating a wide range of material to in fact be working on a very limited range of genotypes. This is illustrated by recent work identifying 10 C. bijugum accessions resistant to a range of biotic and abiotic stresses (Singh et al., 1998). In fact, the germplasm was comprised of only two original accessions, one of which (ILWC 7, Table 2) was subsampled repeatedly so that nine of the 10 resistant genotypes came from the same original accession.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The data to compile Table 2 was largely sourced from the internet. The International Plant Genetic Resources Institute (IPGRI, 2001) lists non-CGIAR databases which hold annual wild Cicer species. The databases of the International Crops Research Institute for Semi-Arid Tropics and National Genetic Resources Program, Pullman, USA are both available online (ICRISAT, 2001; USDA-ARS, 2001). The databases of the International Centre for Agricultural Research in the Dry Areas, Australian Temperate Field Crops Collection, and Aegean Agricultural Research Institute were kindly provided by Mr. J. Konopka, Mr. K. Murray, and Prof. Aç
kgöz, respectively.
Missing passport data was augmented wherever possible by contacting the collectors directly. When this was not possible, and sites were described by locality, site descriptors (latitude, longitude, altitude) were obtained by means of a web-based gazetteer (Anonymous, 2002) and an atlas of West Asia (Bartholomew, 1959).
To describe collection site climate, these were matched to their respective nearby weather stations by global information systems (GIS) techniques (ESRI, 2002). The average distance between collection sites and weather stations was 25.0 km, and no significant differences occurred between sites and stations in terms of latitude, longitude, and altitude. Long-term monthly averages for weather stations covering much of the world are available from the FAO (FAO, 2001a). Long-term averages for Israel and Armenia, which are not covered by the FAO (FAO, 2001a) database, were kindly provided by their respective meteorological agencies (http://www.ims.gov.il/ and http://www.meteo.am/; both verified 22 November 2002).
To provide a biological context for collection site climate, phenological estimates were made for each site. Most Mediterranean annuals germinate in the autumn, flower in late winter to early spring, and mature in late spring to early summer (Ehrman and Cocks, 1996; Feinbrun-Dothan, 1986). The phenology of Mediterranean legumes, particularly time of flowering, is strongly influenced by latitude and altitude, with germplasm from southern sites, or low elevations generally flowering earlier than that from northern sites or higher elevations (Berger, 2000; Piano et al., 1996). The flowering period in Cicer species is listed in van der Maesen (1972). These data, augmented by original collection notes (van der Maesen, personal communication, 2001) and field observations in Israel, Jordan, Syria, and Turkey, were used to make phenological estimates for wild populations sampled in ex situ collections. Developmental phases were defined as follows for all accessions except C. cuneatum: (i) germination—October to November, (ii) flowering—February to March or June (based on site latitude and altitude), (iii) pod set—April through May to July through August (based on site latitude and altitude). Flowering was delayed by one month when large differences in elevation (>1300 m) occurred. For example, populations around Jenin (32.47°N, 35.35°E, -190 m) are assumed to flower in February to March, whereas those from Salkhad (32.48°N, 36.70°E, 1447 m) in March to April. Cicer cuneatum, collected exclusively from a single site in northern Ethiopia has a very different phenology compared with the other West Asian species (van der Maesen, 1972), i.e., (i) germination—March, (ii) flowering—September to November, (iii) pod set—November to December.
SPSS V11 (http://www.spss.com/; verified 22 November 2002) was used for all statistical analysis. Principal components analysis was used to describe both actual and potential collection site habitats using site coordinates and altitude, as well as long-term climate averages from nearby weather stations and the phenological estimates described above. Actual collection sites were identified in the ordination by plotting accession principal component scores of a unique symbol for each species.
Potential collection habitats were described on the basis of weather stations located throughout the annual wild Cicer species distribution range. To simplify the presentation of results, hierarchical clustering was performed on principal component scores calculated on all components with eigen-values > 1 (PC1...4). Nine distinct clusters were identified, and confirmed by nonhierarchical K-means clustering (Hand, 1981) to minimize classification errors arising from the imposition of hierarchy.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
In C. chorassanicum, C. cuneatum, and C. yamashitae, there are only one to three original accessions in the entire world collection. The situation is scarcely better in the primary gene pool of the cultigen. There are only 10 original accessions of C. echinospermum, and 18 of the wild progenitor, C. reticulatum (Table 2). Cicer anatolicum, which may be a close relative of the primary gene pool species (Choumane et al., 2000), is only represented by eight original accessions in the world collection. This contrasts strongly with the almost 32 000 accessions of C. arietinum stored at ICRISAT, ICARDA, and USDA-ARS (ICRISAT, 2001; Konopka, 2000; USDA-ARS, 2001). In this context, even the more frequently collected annual wild Cicer species (Table 2): C. bijugum (n = 20), C. judaicum (n = 34), and C. pinnatifidum (n = 28), are very underrepresented in the world collection.
The tendency for original accessions to be subsampled, and selections henceforth given the status of accessions does little to increase the diversity of the world annual wild Cicer collection, and carries the risk of duplication. This practice is widespread among all the annual wild Cicer species, and culminates in C. reticulatum ILWC 21 (Table 2). Since its collection in 1975 by Dr. van der Maesen in Savur, Turkey, ILWC 21 has been subsampled 43 times, and each selection is listed as a separate accession (Konopka, 2000). The original passport data associated with ILWC 21 does not mention the number of plants collected at this particular site (van der Maesen personal communication, 2001). However, it is not unusual for accessions to be based on a very small number of seeds or plants from the original collection site. For example, the USDA-ARS (2001) collection includes a number of accessions in which as little as one seed to four plants were collected (C. bijugum W6 2059, PI 599051; C. reticulatum PI 599052, PI 527934, PI 599092). Clearly in this situation, there would be a real risk of needless duplication if the original accessions were subsampled. To minimize these dangers, but still allow breeders to select individuals from within accessions, perhaps subsamples should be named as selections from the parental line. Thus, the 43 C. reticulatum selections described above could then be identified as ILWC 21.1 to 21.43.
In a small number of cases, original accessions were split because of mixtures in the original collection (Table 3) . These need to be treated with caution. While it is quite probable that C. pinnatifidum or C. reticulatum could be collected alongside C. bijugum in southeastern Turkey, there is no evidence for C. judaicum or C. pinnatifidum having been found in Morocco, Ethiopia, or Afghanistan (Table 4) . Until evidence to the contrary emerges, it is probably safer to assume that at least some of the accessions (Table 3) have resulted from errors in germplasm multiplication, and as such are duplicates of other material.
|
|
|
In C. reticulatum and C. yamashitae, the origin of accessions in the world collection exactly matches the predictions of regional floras (Fig. 1, Table 4). However, in both cases, and particularly the latter, the number of original accessions available is very small. In C. judaicum and C. bijugum, the distribution range of germplasm held in the world collection exceeds expectations based on regional floras. Cicer judaicum has been extensively collected along the Syrian and Lebanese Mediterranean coast, as well as in southeastern Anatolia and northern Jordan (Fig. 1). Ironically, there are very few accessions from southern latitudes in Israel and Palestine, the core of the species' distribution according to regional floras (Table 4). Cicer bijugum has been collected along its predicted distribution range around 36 to 38°N latitude, but also in southern Syria (Fig. 1). However, there is only a single accession from southern Syria and northern Iraq, creating some imbalance in the collection.
Habitat Characteristics of Annual Wild Cicer Species Based on Collection Sites
Approximately 35% of original annual wild Cicer accessions include detailed information regarding collection site characteristics among the passport data (Konopka, 2000; USDA-ARS, 2001). Germplasm was most frequently collected among rocks or rubble, in field margins, and vineyards or orchards, confirming the data from regional floras (Table 4). Parental rock material is commonly limestone or basalt, soil depth varies from shallow to deep (0.1–0.5 m), texture from fine to coarse (130–650 g kg-1 clay), pH from neutral to alkaline (pH 7.3–8.2) (Konopka, 2000).
Most original accessions include the coordinates or location of their collection site, so habitat climate (Table 5) can be readily described by means of long-term monthly averages from nearby weather stations (FAO, 2001a). Principal components analysis based on collection site location, altitude, and climate at germination, flowering, pod set, and maturity demonstrates the range of habitat types from which the annual wild Cicer species have been collected (Fig. 2) . Principal component 1 (X-axis, Fig. 2) explained 47.6% of the dataset variation, and shows that temperatures during germination and winter, and rainfall over the vegetative phase and total growing season were closely related, and contrasted with altitude and longitude. The close relationship between vegetative phase and total growing season rainfall highlights the fact that there is relatively little precipitation during pod set among any of the collection sites (Table 5). Principal component 2 (Y-axis, Fig. 2) explained almost 22% of dataset variation, and was dominated by the effect of latitude and temperatures experienced at pod set. Maturation is delayed at northern latitudes, and pod set may take place in the middle of summer (van der Maesen, 1972), hence the association of the two variables. During pod set, potential evapotranspiration was closely related to maximum temperatures (r = 0.86), but was excluded from the principal components analysis because of missing data.
|
|
Cicer pinnatifidum and C. judaicum have been collected from the most diverse range of habitats of all the annual wild Cicer species. Most C. judaicum accessions are located in two clusters on the lower-right quadrant of Fig. 2, emphasizing their relatively southern origin. The highest concentration of C. judaicum collection sites is found on the positive extreme of PC1 (X-axis, Fig. 2), representing accessions from coastal Lebanon and Syria. This area represents the most mesic of environments among all of the collection sites of the annual wild Cicer species, and is characterized by seasonal rainfall of 964 mm, moderate temperatures during germination (min = 18.6°C), winter (mean = 12.0°C), and pod set (min = 17.5°C, max = 26.1°C). At the other end of the continuum (<-1.0 units along PC2, Y-axis, Fig. 2) is a smaller group of accessions representing southern latitudes in Syria, Israel–West Bank, and Jordan. These habitats are differentiated from those found along the coastal Mediterranean by significantly lower rainfall, and temperatures during germination, winter, and pod set (P < 0.001). There are two C. judaicum outliers (Fig. 2) (ILWC 148: -0.4, -0.1; ILWC 33: -0.7, 1.0) collected from sites where C. pinnatifidum was found, in Gaziantep and Elazig, southeastern Turkey, respectively. Given the similarity between the two taxa, and the fact that C. judaicum has been considered to be a variety of C. pinnatifidum in the past (van der Maesen, 1972), it is possible that these accessions have been misclassified. However, if their identity is confirmed as C. judaicum, it demonstrates that the species has a much wider distribution range than previously indicated (van der Maesen, 1972), and warrants further collection.
Cicer pinnatifidum is found in all quadrants (Fig. 2), confirming the diversity of its collection sites (Table 5). Most accessions are plotted along the positive part of PC2 (Y-axis, Fig. 2), indicating their origins in southeastern Anatolia, where pod set may occur as late as May to July, hence the high temperatures experienced during this phase (min = 17.5°C, max = 30.9°C). Two accessions from Gaziantep are separated from the Anatolian germplasm on the basis of marginally lower latitude and pod set temperatures (P < 0.05), hence their position –0.4 units along PC2 (Y-axis, Fig. 2). C. pinnatifidum has also been collected along the Anti-Lebanon Range, and in central Israel (Fig. 1). These six accessions are clustered around -2.0 units along PC2 (Y-axis, Fig. 2), and characterized by intermediate season rainfall (562 mm), and low temperatures during germination (min = 5.8°C), winter (mean = 5.0°C), and pod set (min = 8.0°C, max = 23.2°C). Finally, there is a single accession from the mesic environment of coastal Lebanon (ILWC 60) plotted along the positive extreme of PC1 (X-axis, Fig. 2). It should be noted that almost all the C. pinnatifidum sites characterized by negative PC2 values correspond to locations where C. judaicum was collected as well; therefore, the issue of possible misclassification again arises.
Implications of the Current State of the Annual Wild Cicer species Collection
A considerable effort has been directed at screening the annual wild Cicer species for resistance to biotic and abiotic stresses. If only a small proportion of the potential genetic diversity is present in ex situ collections, it is likely that only a small fraction of resistant genotypes has been identified. This is largely confirmed by the results reported thus far. A small number of accessions from a variety of species have shown resistance to Ascochyta blight [caused by Ascochyta rabiei (Pass) Labr.] (Singh et al., 1991; Stamigna et al., 2000), Fusarium wilt (caused by Fusarium oxysporum Schlecht ex Fr.) (Infantino et al., 1996; Kaiser et al., 1994; Stamigna et al., 2000), Botrytis gray mold (caused by Botrytis cinerea Pers.) Singh et al., 1991), cyst nematode (Heterodera ciceri Vovlas, Greco & Di Vito) (Di Vito et al., 1996), and cold temperatures at the vegetative stage (Singh et al., 1990; Singh et al., 1995).
A comprehensive screen against these stresses (with the exception of Botrytis gray mold), as well as leaf miner [Liriomyza cicerina (ROND.)] and bruchid (Callosobruchus chinensis L.) predation, performed in 248 accessions of all eight annual wild Cicer species (Singh et al., 1998) enables the potential of the taxa to be compared. Of all the annual wild Cicer species, C. bijugum was resistant against the greatest number of biotic and abiotic stresses (Singh et al., 1998). Resistance to Fusarium wilt, leaf miner, and cyst nematode was relatively common across all species, whereas Ascochyta blight resistance was rare, occurring only in one to seven accessions of C. echinospermum, C. judaicum, C. pinnatifidum, and C. bijugum (Singh et al., 1998). Resistance against multiple stresses was even rarer, reflecting the multiplication of low probabilities. Only two original accessions of C. bijugum (ILWC 79 and 9 selections of ILWC 7) were resistant against Fusarium wilt, bruchid, cyst nematode, and cold temperatures during winter (Singh et al., 1998). Only one accession (C. pinnatifidum ILWC 250) was resistant to Ascochyta blight and three other stresses (Fusarium wilt, leaf miner, cyst nematode). No accessions in any species were resistant to both Ascochyta blight and cold winter temperatures. Given the significant yield penalty these last two stresses in particular place on cultivated chickpea (Singh et al., 1994), it remains imperative to find additional sources of resistance among the annual wild Cicer species.
For many of the stresses listed above, it is difficult to comment on the effect of collection site habitat on accession response. However, there is a direct relationship between the number of original accessions, the diversity of collection site habitats, and the range of responses to biotic and abiotic stresses recorded. Singh et al. (1994) recorded the largest response range in C. pinnatifidum and C. judaicum, the species with the largest number of original accessions collected from the most diverse habitats (Fig. 1 and 2), whereas the opposite was the case for C. yamashitae, C. chorassanicum, and C. cuneatum. Cicer reticulatum, C. echinospermum, and C. bijugum were intermediate, both in terms of stress response (Singh et al., 1994), and accession number and habitat diversity (Fig. 1 and 2). There is some support for these tendencies in morphological and reproductive characters measured in the annual wild Cicer species (Robertson et al., 1997). The widest range of values for plant height, canopy width, and days to flowering were recorded in C. judaicum and C. pinnatifidum (Robertson et al., 1997). Clearly, because of the small number of original accessions in the world collection, there is likely to be an immediate return in terms of increased trait diversity, when a wider range of habitat types is sampled.
Some traits observed in the annual wild Cicer species can readily be interpreted in terms of collection site habitat. Cicer judaicum and C. cuneatum have consistently been characterized by high mortality or susceptibility to damage caused by low temperatures during the vegetative phase, whereas the opposite is the case for C. bijugum, and to a lesser extent, C. echinospermum and C. reticulatum (Singh et al., 1990; Singh et al., 1995; Singh et al., 1998). The first two species were predominantly collected from areas which experience mild winters (Fig. 2, Table 5), whereas this is not the case for any of the cold tolerant species. Clearly there is a strong selection pressure for cold tolerance exerted by the cold winters in Central Turkey which is not evident in the mesic environments of the coastal Mediterranean and montane Ethiopia. It is of interest to note that there are a number of accessions of C. reticulatum (PI 599092, TR 39248, PI 599042, PI 599050) and C. echinospermum (TR 47738, ICCW 12, TR 39241) collected from relatively northern latitudes (Table 2) that have not been tested for cold tolerance, but may provide useful resistance given their origin. The lack of cold tolerance in the Afghani species, C. chorassanicum and C. yamashitae, collected from habitats in which the long-term average winter temperature is only -2.9°C (Table 5), is an interesting anomaly. It is possible that these species avoid the winter cold by delaying germination until March where temperatures fluctuate between 0.6 to 12.6°C and there is abundant rainfall (99 mm) (FAO, 2001a). If this is the case, these species are employing a different strategy than their relatives collected from central Turkey, where average winter temperatures are similar (Table 5). Unfortunately, the lack of germplasm from across the species' native range in Afghanistan, and northern and northeastern Iran (van der Maesen, 1972) makes it impossible to test this hypothesis.
No assessment of tolerance to cold temperatures during the reproductive phase of any annual wild Cicer species has been reported. This is of particular significance for the ongoing development of cultivated chickpea because of the crop's sensitivity to this stress, which delays the onset of podding, and exposes the pod filling stage to terminal drought (Leport et al., 1999). In general, the annual wild Cicer species have retained their cool-season life cycle of autumn germination and spring-summer maturation, as opposed to C. arietinum which evolved as a spring-sown crop in West Asia (Kumar and Abbo, 2001; Srinivasan et al., 1999). As a result, the annual wild Cicer species may be a useful resource for overcoming this stress. Temperatures experienced during pod set were a strong discriminating variable for separating annual wild Cicer collection sites (Fig. 2). Cicer cuneatum (n = 1) collected from Ethiopia, and C. judaicum (n = 5) from southern Syria, Israel, Palestine, and Jordan experience the lowest temperatures during pod set of all the annual wild Cicer species (Fig. 2), and should be assessed for cold tolerance at this growth stage.
The discipline that appears to be least affected by the genetic diversity of ex situ collections appears to be the area of systematics, because differences between taxa are much larger than within (Tayyar and Waines, 1996). As a result, the systematics of the genus remain broadly consistent irrespective of the range of germplasm analyzed, or the technique employed (Ahmad, 1999; Ahmad and Slinkard, 1992; Choumane et al., 2000; Kazan et al., 1993; Labdi et al., 1996; Tayyar and Waines, 1996). However, within taxa, the relative diversity reported changes depending on the quantity and quality of germplasm used. Thus Labdi et al. (1996) found more diversity within many annual wild Cicer species than previously indicated (Ahmad and Slinkard, 1992; Kazan et al., 1993), and attributed these differences to the larger number of accessions analyzed in their study. Consequently it is difficult to comment on absolute levels of diversity within species unless all the accession details are given so that it is clear what is being compared: original material or selections, germplasm from a single area, or a wide range of habitats.
Opportunities for Increasing Diversity
Having defined the ecogeography of the world annual wild Cicer collection and described the implications of the limited diversity available to researchers at this stage, it is important to demonstrate the existence of habitat types currently unsampled in ex situ collections. A comparison between the shaded portion of Fig. 1 and the symbols representing accession collection sites demonstrates the vast, currently untapped areas available for future collection missions. Principal components analysis demonstrates the geographic and climatic diversity of these habitats (Fig. 3)
. Because of the complexity of the relationships between climate and geography, hierarchical clustering was used to identify discrete habitat types with principal components scores one to four (PC1 to PC4). To minimize classification errors arising from the imposition of hierarchy, K-means clustering (Hand, 1981) was used to define nine discrete habitat types identified in the previous hierarchical classification. These clusters were mapped (Fig. 4)
, and characterized, in terms of geography and climate (Table 6)
. PC1 is dominated by minimum temperatures at germination, vegetative and total growing season rainfall, all contrasted with altitude (Fig. 3). PC3 is dominated by the effect of latitude. Podset minimum temperatures, postanthesis rainfall, and particularly winter average temperatures, were loaded on both PC1 and PC3 (Fig. 3). Longitude and podset maximum temperatures were not well modeled in the ordination of PC1 and PC3, reflecting the complexity of the dataset. Shaded areas in Fig. 3 represent habitats of the material currently held in ex situ collections, and demonstrate that only a small proportion of the total habitat diversity available across the species' distribution range has been sampled.
|
|
|
Within Cluster 2, the combination of moderate winter temperatures, low latitude, and low rainfall experienced in southern Syria, central Jordan, and southern Israel should exert a strong selection pressure on early phenology. However, because podset temperatures are cool, with average minima as low as 6.0°C, germplasm collected from these environments may prove to be a useful source of cold tolerance during the reproductive phase. The combination of earliness with cold tolerance at podset is one of the key objectives of breeding programs targeted at Mediterranean-type climates (Singh et al., 1994). Despite this, very few habitats from Cluster 2 are represented in ex situ collections of annual wild Cicer species, even though C. pinnatifidum and C. judaicum occur extensively within the area (Fig. 3, Table 4).
Clusters 3 and 4, largely located in the upper-right quadrant of Fig. 3, represent mesic environments with intermediate altitude, moderate germination, moderate winter and podset temperatures, and high to very high rainfall (Table 6). Cluster 4 is restricted to the upper eastern Mediterranean from Safita (Syria) to Antakya (Turkey), and along the Caspian Sea (Fig. 4) and is characterized by higher rainfall and lower maximum temperatures at podset than Cluster 3 (Table 6). Cluster 3 has a much wider geographic distribution, extending from the eastern Mediterranean through southeastern Anatolia to northern Iraq and Iran (Caspian Sea), and as far north as Armenia (Fig. 4). Although the largest number of annual wild Cicer were collected from Cluster 3 (n = 58), the northern and eastern extremes of the cluster are not represented in the world collection.
Cluster 5 represents stressful environments characterized by cool germination phases and winters, low rainfall, and warm to hot temperatures during podset at intermediate altitudes across southern Anatolia and Armenia, northern Syria, Iraq, and Iran (Table 6, Fig. 4). Although these habitats lie within the distribution range of many of the annual wild Cicer species, and would exert considerable selection pressure for drought tolerance, another important breeding goal in chickpea (Singh et al., 1994), very few accessions have been collected from within this cluster (Table 6).
Cluster 6, largely confined to the negative of PC1 (Fig. 3), comprises habitats characterized by cool to cold germination and winter temperatures with moderate podset temperatures, altitude and rainfall. Clusters 6 and 7 are found mainly throughout inner Anatolia, and to a lesser extent northern Syria and Iran (Table 6, Fig. 4). Cluster 7 also overlaps substantially with Cluster 6 on the ordination of PC1 and PC3 (Fig. 3, 4). However, Cluster 7 habitats are limited to much higher altitudes, with an attendant decrease in temperature throughout the lifecycle, particularly in winter and at podset. Moreover, Cluster 7 habitats receive more rain, and occur more frequently in eastern Anatolia, but also in the anti-Lebanon mountain range. Although 28 accessions have been collected from Cluster 6 and 7, most of the habitat range remains unsampled (Fig. 3 and 4). Given the low winter temperatures experienced in these habitats, (particularly in Cluster 7), and the fact that this area falls within the distribution range of a number of annual wild Cicer species (Table 4), this may be a useful source of cold resistant germplasm.
Cluster 8, on the negative extreme of PC1, represents much drier habitats at similar altitudes to Cluster 7, in northern Iran and central Afghanistan. Despite the considerable longitudinal range in which Cluster 8 habitats occur, only five accessions have been collected from a relatively narrow band in eastern Afghanistan (Fig. 4). Nevertheless, Cluster 8 overlaps the distribution range of C. anatolicum, C. yamashitae, and C. chorassanicum (van der Maesen, 1972). Cluster 9, on the top left quadrant of Fig. 3, is comprised of even drier habitats than Cluster 8, at lower altitudes, largely throughout Afghanistan. Although altitudes remain moderately high, the reduced elevation relative to Cluster 8 is associated with an increase in temperature throughout the lifecycle (Table 6). The combination of low rainfall and relatively high podset temperatures make Cluster 9 habitats particularly stressful, and no germplasm has been collected from here to date.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
kgöz and Mr. Kevin Murray are thanked for providing the passport data associated with annual wild Cicer collections held at the Aegean Agricultural Research Institute and Australian Temperate Field Crops Collection, respectively. The Israel Meteorological Service and the Department of Hydrometeorology of the Republic of Armenia are thanked for their quick response in furnishing long-term climate averages for weather stations throughout their respective countries. S. Abbo acknowledges the support of the Grains Research and Development Corporation, Australia, in providing a Grains Industry Visiting Fellowship. J. Berger and N. C. Turner would like to thank the Australian Centre for International Agricultural Research (ACIAR) for their generous support of research on chickpea adaptation. Received for publication January 2, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Ben-David, S. Lev-Yadun, C. Can, and S. Abbo Ecogeography and Demography of Cicer judaicum Boiss., a Wild Annual Relative of Domesticated Chickpea Crop Sci., April 25, 2006; 46(3): 1360 - 1370. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
), C. cuneatum (
), C. echinospermum (
), C. judaicum (
), C. pinnatifidum (
), C. reticulatum (
), C. yamashitae (+). Shaded areas represent the known distribution of annual wild Cicer species and C. anatolicum. The inset shows an enlargement of those areas of West Asia from which the most annual wild Cicer species have been collected.
), C. echinospermum (
