| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Center for Conservation and Breeding of the Agricultural Diversity (COMAV), Polytechnic University of Valencia, Camino de Vera 14, Valencia 46022, Spain
* Corresponding author (fnuez{at}btc.upv.es).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The cultivation of C. moschata soon spread to northeastern Mexico (
3400 BP) and to southwestern USA (900 BP). After the arrival of the Spanish explorers to America, an additional diversification of this species took place in Asia Minor and Japan (Decker-Walters and Walters, 2000). However, no illustration or description gives evidence of C. moschata cultivation in Europe before the late 17th century. This could be due to the poor adaptation of this species to the temperate climates of middle to high latitudes (Paris, 2000). In Africa, C. moschata had become well established as a food crop by the 19th century (Decker-Walters and Walters, 2000).
In Spain, one of the main producers of squashes in Europe (FAOSTAT, 2003), C. moschata is grown mostly in small orchards for self-consumption or sale at local markets. The majority of Spanish production is based on landraces which have been maintained by farmers for centuries, as has occurred in Latin America and Africa (Gwanama et al., 2000; Lira-Saade, 1995). The Center for Conservation and Breeding of Agricultural Diversity (COMAV) of the Polytechnic University of Valencia maintains a germplasm collection with about 250 C. moschata landraces. Most of the accessions have been collected in Spanish provinces, especially in the Canary Islands, where this species is highly appreciated (Decker-Walters and Walters, 2000). In addition, some accessions were collected in Central America, South America, and Africa. COMAV currently houses the European database for the Cucurbitaceae family within the European Cooperative Program for Crop Genetic Resources Network (ECP/GR), developed by the International Plant Genetic Resources Institute (IPGRI) (Diez et al., 2002).
To make this collection useful for breeders and farmers throughout the world, a morphological and molecular characterization of this germplasm is needed. Castetter (1925) established three horticultural types of C. moschata, which are currently used in the commercial trade of North America (Robinson and Decker-Walters, 1997; Whitaker and Davis, 1962). This classification is based primarily on the color, texture, and hardness of the skin and on the shape of the peduncle and fruit. However, the three proposed groups do not encompass all the fruit types that have evolved in tropical America and Asia (Robinson and Decker-Walters, 1997). Subsequent studies concerning the morphological diversity among landraces from different centers of diversity, such as Cuba and Korea, have revealed a great variability of squash types (Chung et al., 1998; Ríos et al., 1997; Youn and Chung, 1998). Wessel-Beaver (1998) also observed considerable morphological variability among different C. moschata landraces from Puerto Rico, especially in the fruit shape.
Only a few molecular analyses have been conducted in this species. RAPD (random amplified polymorphic DNA) markers were used to analyze the genetic diversity among C. moschata landraces from Korea, southern Africa, and other geographical origins (Baranek et al., 2000; Gwanama et al., 2000; Youn and Chung, 1998). In all cases, the accessions grouped according to the agroclimatic regions of origin and not according to morphological traits.
The aim of the present work was to analyze the genetic diversity of the collection of C. moschata landraces held at COMAV, including Spanish, African, South American, and Central American landraces. In addition to the morphological analysis, two molecular marker systems have been employed in this study: (i) SRAP markers (Li and Quiros, 2001), which preferentially amplify open reading frames (ORF) and have been successfully used in diversity analyses of C. maxima Duchesne and C. pepo L. landraces (Ferriol et al., 2003a; 2003b) and (ii) AFLP markers (Vos et al., 1995), which have also been used successfully in diversity analyses of these and other Cucurbitaceae species (Ferriol et al., 2003a, 2003b; García-Mas et al., 2000). The comparison between relationships obtained by both marker systems is discussed.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
DNA Extraction
Genomic DNA was isolated from leaves by the modified CTAB (hexadecyltrimethylammonium bromide) method of Doyle and Doyle (1990). For each accession, 0.5 g of ground leaf tissue from a bulk of 10 plants were suspended in 2.5 mL of extraction buffer [20 mM EDTA, 0.1 M Tris-HCl (pH 8.0), 1.4 M NaCl, 2% (w/v) CTAB and 5 µL of ß-mercaptoethanol]. The suspension was mixed well, incubated at 60°C for 30 min, followed by chloroform-isoamyl alcohol (24:1) extraction, and precipitation with 0.67 vol isopropanol at –20°C. The pellet formed after centrifugation at low speed for 5 min was washed with 76% (v/v) ethanol and 10 mM NH4OAc. The DNA was then suspended in TE buffer. The resulting DNA concentration was measured in a 1% (w/v) agarose gel stained with ethidium bromide using 1-D Manager (2.0; Tecnología para Diagnóstico e Investigación S.A, Madrid; see http://www.tdi.es/programas/1d.htm; verified 11 November 2003), in comparison with a known concentration of Arabidopsis thaliana DNA (AFLP Core Reagent Kit of Invitrogen, San Diego, CA).
SRAP Analysis
The SRAP technique consists of preferential amplification of ORFs by PCR. For this purpose, combinations of two different primers are used. One primer (forward), of 17 base pairs, contains a fixed sequence of 14 nucleotides rich in G and C in the 5' end and three selective bases in the 3' end. This primer amplifies preferentially exonic regions, which tend to be rich in these nucleotides. The second primer (reverse), with 19 base pairs, contains a sequence of 16 nucleotides, rich in A and T in the 5' end and three selective bases in the 3' end. This primer preferentially amplifies intronic regions and regions with promoters, rich in these nucleotides (Li and Quiros, 2001). The usefulness of this technique for detecting polymorphism in ORFs has been previously proved in Cucurbita by sequencing some DNA fragments from C. maxima (Ferriol et al., 2003a) and C. pepo (Ferriol et al., 2003b).
In this assay, 11 different primer combinations were used, with five forward and four reverse primers previously described for analyzing genetic diversity in Brassica and other genera (Li and Quiros, 2001) (Table 2). Each 25-µL PCR reaction mixture consisted of 20 ng genomic DNA, 200 µM dNTPs, 1.5 mM MgCl2, 0.3 µM primer, 2.5 µL 10x Taq buffer, and 1 unit of Taq polymerase (Boehringer Mannheim). Samples were subjected to the following thermal profile: 5 min of denaturing at 94°C, five cycles of three steps: 1 min of denaturing at 94°C, 1 min of annealing at 35°C, and 2 min of elongation at 72°C. In the following 30 cycles, the annealing temperature was increased to 50°C, with a final elongation step of 5 min at 72°C. Separation of the amplified fragments was performed on 12% (w/v) polyacrylamide gels [acrylamide-bisacrylamide (29:1), TBE 1x] at 500 V during 11 h. The gels were stained with AgNO3 for visualizing the SRAP fragments and dried overnight. The fragments between 110 and 950 base pair (bp) were visually scored as present (1) or absent (0).
|
The protocol described previously (Ferriol et al., 2003b) was followed. Six different primer combinations were used (Table 2). Electrophoresis was conducted using an ABI PRISM 310 Genetic Analyzer (PerkinElmer Applied Biosystems, Foster City, CA). Raw data were analyzed with GeneScan 3.1.2 analysis software (PerkinElmer Applied Biosystems) and the resulting GeneScan trace files were imported into Genographer 1.6.0. The AFLP fragments between 60 to 380 bp were scored in Genographer as present (1) or absent (0).
Data Analysis
Principal component analysis (PCA) and principal coordinate analysis (PCoA) were performed with the standardized morphological quantitative data and the molecular data respectively, to obtain a graphical representation of the relationship structure of the characterized accessions.
In the morphological characterization, Euclidean distances among accessions were calculated using the quantitative traits. In SRAP and AFLP molecular analysis, genetic distances (1 – Sij) among genotypes were calculated according to the Nei and Li (1979) similarity coefficient Sij = 2a/(2a + b + c), where Sij is the similarity between two individuals i and j, a is the number of shared bands, b is the number of bands exclusively amplified by i, and c is the number of bands exclusively amplified by j. The distance matrix was subjected to cluster analysis by the Unweighted Pair-Group Method (UPGMA, Sneath and Sokal, 1973). The goodness of fit of the cluster to the data matrix was calculated by means of the cophenetic coefficient. The reliability and robustness of the dendrograms were tested by bootstrap analysis with 1000 replications to assess branch support using PHYLIP 3.6 software. One accession of C. maxima was used as the outgroup.
The Mantel test was used to determine the correlation between the elements of the distance matrices on the basis of the morphological data and of the two different marker types (Mantel, 1967). The statistical analyses were performed with NTSYS-pc (version 2.0).
Gene Diversity (Nei, 1973) was estimated by POPGENE 32. In addition, the discriminatory power of the two molecular marker types was estimated by calculating the effective multiplex ratio (EMR) and marker index (MI) according to Vandemark et al. (2000). EMR is defined as the product of the fraction of polymorphic loci and the total number of polymorphic loci. MI is defined as the product of the gene diversity (Nei, 1973) and the EMR for that marker type.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A high correlation among the seed characters (length, width, thickness, and weight of the seeds) was obtained. Consequently, only the seed length was considered in the PCA. The first component, which accounted for 42.1% of the total variation, fundamentally grouped the accessions according to fruit shape (length/width ratio) and to fruit weight (Fig. 1a and b) . These two characters were inversely correlated. The weight of the flattened and globular fruits was significantly higher than that of the round and elongated fruits (Duncan test, P < 0.05, data not shown). The second component, which accounted for 24.1% of the total variation, grouped the accessions according to the fruit and seed length.
|
Molecular Characterization
The analysis of the 47 C. moschata accessions with 11 SRAP primer combinations identified a total of 148 reproducible fragments (Table 3). Among them, 98 were polymorphic (66.2%), ranging in size from 140 to 950 bp. Between 9 and 21 fragments were amplified per primer combination, with an average of 13.5 bands. The number of polymorphic fragments for each primer combination varied from 6 to 16, with an average of 8.9. An example of a polyacrylamide gel with polymorphic SRAP fragments is shown in Fig. 2
.
|
|
A cluster analysis was performed using the Nei and Li distance (1979) and the UPGMA method. The cophenetic coefficient was 0.86 indicating a good fit. The dendrogram showed a clear separation between the Spanish, Moroccan, Central American, and South American accessions (Fig. 3) . Among the American landraces, the South American accessions clustered separately from the Central American ones. Cluster I included all the accessions from the represented South American countries, Ecuador and Peru (bootstrap = 98%). Within this cluster, the accessions from both countries were intermingled. Cluster II included all the accessions from Central America (bootstrap = 53%), except RD1, from the Dominican Republic, which clustered independently (bootstrap = 98%). Within cluster II, the Cuban accessions clustered together (bootstrap = 64%), while the Guatemalan accessions clustered with RD2, from the Dominican Republic (bootstrap = 62%). Cluster III included all the accessions from the Spanish peninsula, the Canary Islands and Morocco. Within this cluster, some pairs of accessions clustered with high bootstrap values. In three cases, the accessions had a common geographical origin: Canary Islands (CA24 and CA2, bootstrap = 81%, and CA149 and CA213, bootstrap = 99%) and Balearic Islands (B20 and B3, bootstrap = 91%). In the other three cases, the accession pairs did not share geographical origins (bootstrap = 74, 93, and 51%). The commercial cultivar USA4 (Butternut Squash) clustered independently, halfway between the Spanish and the South American accessions. In general, the cluster analysis did not group the accessions according to any of the morphological traits.
|
|
Very low correlation was found between the distance matrices using morphological and SRAP data (Mantel test, r = –0.08; P = 0.23).
For analysis with AFLP markers, 39 accessions were selected according to the different morphological types, geographical origin, and SRAP results. In this analysis, six primer combinations were used (Table 2). A total of 156 reproducible fragments, ranging in size from 60 to 380 bp, were identified, of which 134 (85.9%) were polymorphic (Table 3). Between 14 and 41 fragments were amplified per primer combination, with an average of 26 bands. The number of polymorphic fragments for each primer combination varied from 14 to 29, with an average of 22.3.
The range of dissimilarity obtained varied between 0.069 (between CA116 and CA37, two elongated forms) and 0.48 (between the globular PER1 and the oval CA188). Thirty-one fragments were uniquely amplified in single accessions, from both South America and Spain. In three cases, there was a unique fragment absence. The genetic diversity (Nei, 1973) with AFLP markers was smaller than that obtained with SRAP markers (0.17 ± 0.15). However, the EMR (19.73 ± 4.73) and the MI (2.80 ± 1.47) were greater using AFLP than using SRAP. Therefore, AFLP markers appeared to be more efficient in detecting polymorphisms than SRAP markers.
A cluster analysis was performed by means of the Nei and Li distance (1979) and the UPGMA method. The cophenetic coefficient was 0.84, indicating a good fit. The dendrogram grouped the accessions in four main clusters (Fig. 5) . As with the results obtained with SRAPs, a clear grouping according to geographical origin was observed. Cluster I included the accessions from South America (bootstrap = 99%). Cluster II grouped 3 Spanish accessions, displaying fruits with different morphological traits and coming from different geographical regions (bootstrap = 57%). Cluster III included all the Central American accessions and the commercial cultivar Butternut squash (bootstrap = 88%). The two accessions from Guatemala clustered together (bootstrap = 86%), while RD1, from the Dominican Republic, clustered separately (bootstrap = 96%). The fact that RD1 clusters independently agrees with the results obtained with SRAP markers. This accession was collected in La Altagracia province from a polyculture composed of squash, coffee (Coffea arabica L.), and cacao (Theobroma cacao L. subsp. cacao). RD1 is a localized landrace probably grown for self-consumption, in contrast with RD2, which was collected in a market of Santo Domingo. Cluster IV included the remainder Spanish accessions and the two accessions from Morocco. Within this cluster, three pyriform accessions (Pear-shaped and Dumbbell) from different Spanish regions (B17, V23, and CM51), clustered together, separately from the remaining accessions (bootstrap = 55%). This grouping agrees with that obtained with SRAPs. Likewise, some pairs of accessions clustered together with high bootstrap values. Three of them had the same geographical origin (bootstrap = 98, 50, and 97%), while the other three comprised accessions from different Spanish regions (bootstrap = 57, 83, and 66%).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The majority of the characterized accessions could be grouped according to fruit shape (Wessel-Beaver, 1998), which is a complex and multigenic character (Brown and Myers, 2002). As with some of the morphotypes of C. pepo (Paris, 1989), some C. moschata forms were grown in the pre-Columbian era, including the "Crookneck" fruits (here referred to as Guiro), or the oval to pear-shaped fruits belonging to "Seminole Pumpkin," grown by the Florida natives. Subsequently, other fruit types were reported, such as the "Butternut" type, which was derived in the 1930s from a Crookneck fruit (Mutschler and Pearson, 1987; National Research Council, 1989). The great diversity of forms observed among the Spanish landraces could thus demonstrate heterogeneous origins and histories for these landraces.
The observation that the South American accessions exhibited more primitive traits than the other accessions, including those from Central America, agrees with previous studies (Filov, 1966; Wessel-Beaver, 2000; Whitaker and Davis, 1962). These previous studies suggest that the center of origin and domestication of C. moschata is located in northern South America (Nee, 1990; Sanjur et al., 2002; Wessel-Beaver, 2000).
With both SRAP and AFLP markers, considerable genetic diversity was observed among C. moschata landraces, in agreement with the morphological variability observed. Previous studies with SRAPs indicated greater polymorphism in a germplasm collection of C. pepo landraces (72.7%) (Ferriol et al., 2003b), while in a collection of C. maxima, the polymorphism was smaller (56.8%) (unpublished data). With AFLPs, unlike the results obtained with SRAPs, the polymorphism among the C. maxima (55.34%) and the C. pepo accessions (53.15%) was lower than that obtained among the C. moschata accessions. These results agree with previous studies using other molecular markers. In all cases, greater genetic diversity in C. moschata than in C. maxima was observed. However, the results between C. moschata and C. pepo varied with the study (Baranek et al., 2000; Decker-Walters et al., 1990; Jeon et al., 1994; Jobst et al., 1998; Wilson et al., 1992).
With both SRAPs and AFLPs, some fragments were uniquely amplified from single accessions. Similar results were observed for some Oriental C. moschata landraces using RAPD markers (Jeon et al., 1994), in contrast to the results obtained among African C. moschata landraces with the same markers, where no fragments specific to single accessions could be found (Gwanama et al., 2000).
With both markers, a clear grouping according to geographical origin was obtained. The separation between the two American areas may support the existence of two independent domestications of C. moschata in Central America and South America, as suggested by some authors (Decker-Walters and Walters, 2000; Robinson and Decker-Walters, 1997). In C. pepo, two independent domestications in southern Mexico and the eastern USA have also been suggested. However, the existence of closely related wild species in C. pepo support this hypothesis (Decker-Walters and Walters, 2000).
The separation between the accessions from South America and Central America could also be due to introgressions from other species. In North America, in the region encompassing northwestern Mexico and southwestern USA, C. moschata may have been cultivated alongside C. argyrosperma Huber and C. pepo for centuries. In Mexico, spontaneous hybridization between C. moschata and C. argyrosperma has been reported (Decker-Walters et al., 1990). These authors, using isozymes, also observed different band patterns between the landraces from South America and those from the southwestern USA, northwestern Mexico and Japan. In agreement with these studies, Montes-Hernández and Eguiarte (2002) detected gene flow among C. moschata, C. argyrosperma ssp. argyrosperma and C. argyrosperma ssp. sororia in the traditional Mexican agroecosystems, where the wild related species grow frequently near the cultivated species.
The grouping of the accessions obtained in the PCoA was somewhat different with both markers. These differences could be due to the different information provided by each marker system. While the SRAP markers preferentially amplify ORFs, which may include coding regions of the genome involved in morphological and agronomic traits (Ferriol et al., 2003b; Li and Quiros, 2001), the AFLP markers are thought to amplify both coding and neutral regions of the genome. In fact, recent studies indicate that the AFLP markers obtained with EcoRI/MseI enzymes in tomato [Lycopersicon spp.] are mainly clustered in the centromeres (Bonnema et al., 2002), although it is not clear if this effect also occurs in other vegetables such as cucurbits (Wang et al., 1997). The different information provided by SRAPs and AFLPs is corroborated by the low correlation found between the distance matrices obtained with both markers (Mantel test).
The PCoA performed with SRAPs separated the South American accessions, which show primitive traits. However, grouping according to the fruit shape or other morphological traits could not be observed, as indicated by the low correlation coefficient obtained between the morphological and the molecular genetic distance matrices. Similarly, Decker-Walters et al. (1990) did not observe any intraspecific morphological grouping in C. moschata with other nonneutral markers, such as isozymes. In other species of Cucurbita, with SRAP markers, similar results were obtained (Ferriol et al., 2003a, 2003b). In C. pepo, the accessions belonging to each of the two described subspecies were clearly separated. Within the two subspecies, a certain grouping according to morphotype was observed (Ferriol et al., 2003b). In C. maxima, the accessions were grouped according to the type of use, associated with other valuable agronomic traits (Ferriol et al., 2003a). With AFLPs, the C. moschata landraces did not group according to morphological traits. This result agrees with previous studies using other neutral markers, such as RAPDs, where correspondence could not be observed between the morphology of 22 C. moschata accessions from South Korea and relationships based on data obtained using molecular markers (Youn and Chung, 1998).
Among the Spanish accessions, with both SRAP and AFLP markers, a clear separation between the landraces from the Canary Islands and those from the peninsula was found. One hundred sixteen out of 230 accessions included in the Spanish C. moschata collection held at the COMAV were collected in the Canary Islands, thus confirming the importance of its cultivation in this area. There is no evidence of the existence of C. moschata in Europe before the 17th century, probably because of the poor adaptation of this species to the relatively short summers of temperate climates and the long summer days of middle to high latitudes (Paris, 2000). However, in the more tropical climate of the Canary Islands, C. moschata could have adapted well after its introduction with other squashes by Spanish explorers. This could have led to a different evolution and diversification of the landraces of this species on both the Islands and on the peninsula. Furthermore, the first reports of the arrival of Cucurbita species to Spain from Peru (late 18th century) referred to different shipments to the Botanical Gardens of Madrid and Tenerife (del Campo, 1993). Consequently, different landraces could have been initially introduced to the islands and the peninsula, contributing to a greater differentiation of the landraces in both regions.
The dendrograms showed that some Spanish accessions from different geographical origins clustered together with both markers. Different factors could have led to this grouping. Possibly, only a few accessions were introduced from America into Europe since the 17th century, causing a bottleneck effect, which could have been followed by a secondary differentiation performed by Spanish farmers. In addition, the outcrossing in C. moschata may have contributed to similarities within regions, especially after the introduction of a limited number of accessions. On the other hand, this grouping could also be due to the existence of seed exchanges among farmers. The same phenomenon seems to occur among farmers from Malawi and Zambia, where 40% of the C. moschata seeds are exchanged (Gwanama et al., 2000) and also in Mexico, in the traditional "milpa" farming systems, where the percentage of C. moschata seed exchange is approximately 62% (Montes-Hernández and Eguiarte, 2002).
Both SRAP and AFLP marker systems have proved to be useful for analyzing the genetic diversity of some C. moschata landraces. The knowledge of the diversity of this germplasm will facilitate its use in breeding programs and improve the management of large collections of this species.
Received for publication March 12, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. J. Ariss and G. J. Vandemark Assessment of Genetic Diversity among Nondormant and Semidormant Alfalfa Populations Using Sequence-Related Amplified Polymorphisms Crop Sci., November 7, 2007; 47(6): 2274 - 2284. [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 | |||
); Globular (
); Round (•); Oval (*); Oviform (
); Heart-shaped (
); Cuneiform (
); Oblong (
); Cylindrical (
); Pear-shaped (
); Dumbbell (
); Butternut (
); Guiro (+); Unclassified (°). 
Accessions used for AFLP analysis.
