Am. J. Bot. Join the BSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Montes-Hernandez, S.
Right arrow Articles by Eguiarte, L. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Montes-Hernandez, S.
Right arrow Articles by Eguiarte, L. E.
Agricola
Right arrow Articles by Montes-Hernandez, S.
Right arrow Articles by Eguiarte, L. E.
(American Journal of Botany. 2002;89:1156-1163.)
© 2002 Botanical Society of America, Inc.

Population Biology

Genetic structure and indirect estimates of gene flow in three taxa of Cucurbita (Cucurbitaceae) in western Mexico1

Salvador Montes-Hernandez2,3,4 and Luis E. Eguiarte3

2Programa de Recursos Genéticos, Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (INIFAP), Apartado Postal 112, Celaya, Guanajuato 38000, México; 3Departamento de Ecología Evolutiva, Instituto de Ecología, Universidad Nacional Autónoma de México, Apartado Postal 70-275, México, D. F., 04510, México

Received for publication October 9, 2001. Accepted for publication January 29, 2002.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Cultivated squash (Cucurbita argyrosperma ssp. argyrosperma and C. moschata) are important in the Mexican traditional agroecosystem. They are typically cultivated within maize fields where adjacent populations of a wild, close relative, C. argyrosperma ssp. sororia, occur. Consequently, there are ample opportunities for gene flow between domesticated and free-living Cucurbita populations. We used allozymes to examine genetic variation and gene flow among these three Cucurbita taxa in the state of Jalisco in Western Mexico. Twelve polymorphic allozyme loci were used to calculate genetic diversity for 16 populations of Cucurbita. We found high levels of genetic variation: polymorphism of 0.96, mean allelic diversity of 2.08, average expected heterozygosity 0.407, and little differentiation among conspecific populations (D = 0.081; FST = 0.087; Nem = 5.22). These findings indicate that Cucurbita possess a high pollen dispersal potential, a somewhat surprising result considering they have specialist pollinators. Unweighted pair group method with arithmetic means (UPGMA) analysis of allozymes suggests the existence of at least two distinct groups of populations, one consisting of both subspecies of C. argyrosperma and another consisting of C. moschata.

Key Words: Cucurbita • Cucurbitaceae • cultivated squash • gene flow • genetic diversity • genetic structure • wild gourd


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
The ecological and cultural conditions under which traditional agriculture is carried out in Mexico have helped to preserve large amounts of diversity in locally cultivated species (Bye, 1993 ; Hernández, 1993 ; Souza et al., 1997 ). This variation is complemented by the presence of their wild relatives within the agroecosystems (Harlan, 1976 ). These wild relatives have many important influences on their cultivated relatives, and store great amounts of genetic variation (Doebley, 1992 ; Escalante et al., 1994 ; Colunga et al., 1999 ), which may be of interest for future crop improvement programs (Doebley, 1990 ; Wilson, 1990 ; Jarvis and Hodgkin, 1999 ). In addition, they can provide information on the domestication process (Doebley, 1990 ), and they represent a potential source of gene flow with their genetically engineered relatives (Doebley, 1990 ; Ellstrand and Hoffman, 1990 ; Wilson, 1990 ; Hancock, Grumet, and Hokanson, 1996 ; Hokanson, Hancock, and Grumet, 1997 ; Ellstrand, Prentice, and Hancock, 1999 ).

In Mexico, squash (Cucurbita spp.) is a very important traditional crop and a great deal of genetic variability has been maintained. Cucurbita spp. are of great nutritional and economic importance for Mexicans; they are used both as an immature fruit (zucchini type) and as a mature fruit and as seeds (Whitaker and Bohn, 1950 ; Lira, 1995 ; Merrick, 1995 ).

The cultivated species are monoecious, and they are pollinated by various wild specialist bees, mainly Peponapis spp. and Xenoglossa spp. (Hurd, Linsley, and Whitaker, 1971 ). In the traditional mesoamerican agricultural systems, called "milpa," maize–beans–squash or maize–squash are cultivated together. In these milpa fields, it is common to find wild relatives growing near the cultivated plants. Thus, it is likely that natural cross-fertilization between wild and cultivated species occurs, and indeed, such gene flow has been documented in maize (Wilkes, 1977 ; Doebley, 1990 ), beans (Escalante et al., 1994 ; Beebe et al., 1997 ), and squash (Kirkpatrick and Wilson, 1988 ; Wilson, 1990 ; Wilson, Lira, and Rodriguez, 1994 ).

In the southwestern part of the state of Jalisco, Mexico, the milpa agroecosystem includes cultivated maize (Zea mays L. ssp. mays) and two cultivated taxa of squash, C. argyrosperma Huber ssp. argyrosperma and C. moschata (Duch. ex Lam.) Duch. ex Poir., growing near their wild relatives, Z. mays ssp. parviglumis Iltis & Doebley or teosinte, and the wild gourd Cucurbita argyrosperma ssp. sororia (L.H. Bailey) Merrick & Bates.

Experimental evidence has revealed that the taxa involved in the present study can be crossed to produce fertile seeds (Merrick, 1990 , 1991 ; Wessel-Beaver, 2000 ). A close relationship exists among C. argyrosperma ssp. argyrosperma (mean diversity [D] = 0.02 [0.00–0.06]) and C. argyrosperma ssp. sororia populations (D = 0.01 [0.00–0.06]; Decker, 1986 ), although C. moschata populations are more distant (D = 0.24 [0.16–0.32]; Wilson, 1989 ; Wilson, Doebley, and Duvall, 1992 ). On the other hand, data on the genetic diversity in extant populations indicate a close relationship between C. argyrosperma ssp. argyrosperma and C. argyrosperma ssp. sororia (average D = 0.03) and greater differentiation between C. argyrosperma ssp. argyrosperma and C. moschata (average D = 0.22; Wilson, 1989 ; Merrick, 1991 ). A study of genetic diversity in Cucurbita revealed that C. moschata has greater genetic diversity (mean expected heterozygosity, H = 0.052) than C. argyrosperma (0.039), although the sample size was small in both species (8–20 individuals; with 14 loci) (Decker-Walters et al., 1990 ). Decker (1986) and Decker-Walters et al. (1990) present evidence that the two species have introgressed. In the common zucchini, the C. pepo L. complex, genetic diversity and heterozygosity (H) are moderately high (D = 0.17 and H = 0.089; Decker and Wilson, 1987 ) and alleles from the cultivated species have been found in wild populations. This has been interpreted as evidence of gene flow among wild and cultivated populations (Decker and Wilson, 1987 ; Kirkpatrick and Wilson, 1988 ; Wilson, 1990 ).

Our main objective was to determine the levels of genetic variation within and among taxa of Cucurbita in Jalisco, Mexico, and to analyze their relationships. Specifically, we assessed (1) the levels of gene flow within and among populations of two different species of squash (both of which are cultivated in the traditional milpa agroecosystem) and a wild type that grows in adjacent fields, (2) the distribution of allozyme variation within and among populations of the wild and cultivated squashes, and (3) the genetic relationships among populations of the three taxa.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 LITERATURE CITED
 
Study sites and plant material
For the allozyme analyses, seeds were collected from plots of cultivated squash in the southwestern part of the Jalisco State in Mexico. In these sites C. argyrosperma ssp. argyrosperma and C. moschata are sown together, but the weedy C. argyrosperma ssp. sororia is also present. For electrophoretic survey, we used seeds from all three taxa and a population of C. pepo as an outgroup. We chose six localities in five municipalities (Fig. 1). The first three taxa were sampled within four of the localities (San Miguel, San Lorenzo, El Chante y Los Parajitos; Table 1). The CAA8 population of C. argyrosperma ssp. argyrosperma and the C. pepo population were collected in Tepec, outside of the study region in an area where weedy Cucurbita species are not found (Table 1). In all, we sampled a total of 16 populations including 6 of C. argyrosperma ssp. argyrosperma, 4 of C. moschata, 5 of the wild type C. argyrosperma ssp. sororia, and 1 of C. pepo (Table 1). Landraces (traditional and local crop varieties) were sampled from seeds donated from several local farmers. An average of 4–5 fruits per population of wild type squash were collected haphazardly near the milpas; all of the seeds from each population were combined and 45–50 seeds from each batch were randomly sampled and sown in organic growing medium (1 : 1, humus : fertile soil). These seedlings were left to grow in a greenhouse and young leaf material was harvested for electrophoresis (see below). On average, protein from 45 seedlings per population, 720 in total, were subjected to electrophoretic analyses.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 1. Locations of the six collection sites of the 16 Cucurbita populations

 

View this table:
[in this window]
[in a new window]
  		Table 1. Accessions of Cucurbita used in this study

 
Isozyme electrophoresis
Starch gel horizontal electrophoresis was used to estimate genetic variation in allozyme loci within and among selected populations of Cucurbita. Young leaf material of all sampled individuals was ground with an electric drill, using 3–4 drops of an extraction buffer, composed of a 3 : 1 mixture of the Veg II buffer of Pitel and Cheliak (1984) and the extraction buffer of Yeh and O'Malley (1980) . The extract was absorbed on 15 x 2 mm Whatman 1 mm filter paper wicks and stored in an ultracold freezer (–80°C) until electrophoresis.

Several tray-gel buffer systems were tested to determine which provided the best resolution for each enzyme locus, and several loci of many enzymes were screened to determine which were polymorphic. Alleles and loci were assessed following the techniques outlined in earlier studies of Cucurbita spp. (Decker, 1986 ; Decker and Wilson, 1987 ; Kirkpatrick and Wilson, 1988 ; Decker-Walters et al., 1990 , among others). Putative loci were assigned sequentially, with the most anodally migrating isozyme designated as 1, the next 2, etc. Likewise, alleles were assigned sequentially with the most anodally migrating allele designated as 1, the next 2, etc. Two gel and electrode buffer systems were used to assay the following allozymes: phosphoglucose isomerase (Enzyme Commission number (E.C.) 5.3.1.9, PGI, two loci), leucine aminopeptidase (E.C. 3.4.11.1, LAP, one locus), esterase (E.C. 3.1.1, EST, one locus), and peroxidase (E.C. 1.1.1.7, APX, three loci; cathodal and anodal) in 11% starch gels (375 mL) ran in LiOH buffer pH 8 (Soltis et al., 1983 ), at 60 mA for 6–7 h. We analyzed isocitrate dehydrogenase (E.C. 1.1.1.42, IDH two loci), malate dehydrogenase (E.C. 1.1.1.37, MDH one locus), phosphoglucomutase (E.C. 2.7.5.1, PGM, one locus), and malic enzyme (E.C. 1.1.1.40, ME, one locus) on the same gels run with a histidine-citrate buffer pH 6.5 system (Cardy, Stuber, and Goodman, 1980 ) at 30 mA for 7–8 h. Allozyme markers for 12 enzyme loci for the four taxa were consistent and interpretable.

Data analysis
Allelic frequencies in each population were calculated from isozyme phenotypes. Allelic variation within and across all populations was quantified by calculating the following statistics for each locus in each population: percentage of polymorphic loci (p), polymorphic index (PI; proportion of polymorphic loci per population), mean number of effective alleles per locus (A), mean observed heterozygosity (Ho; direct estimate), Hardy-Weinberg expected heterozygosity (He), and the fixation index (F; Wright, 1965 ). All estimates were calculated with BIOSYS (Swofford and Selander, 1981 ).

The distribution of genetic variation within and among populations was estimated by calculating Wright's (1951) F statistics (FST, FIS, and FIT) from variance components using Weir and Cockerham's (1984) estimators {theta}, f, and F. These tests assume that the populations are in mutation-drift balance and share the same time of divergence from a common ancestor. Here we derive F statistics both within taxa and among all populations of the three taxa. For the second analyses, we must therefore assume that the three taxa share a recent common ancestor and that the time of divergence of each taxa from this ancestor is approximately the same. Given the inter-crossability of the taxa and the high similarity of their isozyme profiles, this is probably a reasonable assumption. The deviation of fixation index (F) from zero for each locus in each population (in total 192 tests; 12 loci per 16 populations) was tested using a chi-square test (Li and Horvitz, 1953 ), and 95% confidence intervals (CI) were estimated by 1000 bootstraps with the program TFPGA (Tools for Population Genetic Analyses; Miller, 1997 ). For each population in each taxon, the number of loci that exhibited an excess or deficit of heterozygotes based on Hardy-Weinberg (H-W) expectation was noted.

The rate of interpopulation gene flow within taxa and among all populations was estimated using the procedure based on Wright's FST, where FST = (1/4Nem + 1) (Wright, 1951 ). This method estimates Nem, which is the average number of migrants into a population per generation. In this case, we used our estimate of {theta} to estimate Nem following Crow and Aoki (1984) , Nem = [(1/{theta}) – 1]/4{alpha}, where {alpha} = (n/n – 1)2, and n = number of subpopulations. The estimate of Nem assumes that there are equal quantities of migration between all populations. Although this assumption is probably not satisfied, given the fact that we have approximately equal numbers of populations per taxa and representatives of each taxa in each locality, making this assumption is also reasonable.

The genetic divergence between populations was determined using Nei's genetic distances (D; Nei, 1978 ), which is an index of genetic similarity. Dendograms for all taxa together were constructed using UPGMA (unweighted pair group method using arithmetic means) and the confidence of each node was tested with bootstraps. The majority of the analyses were carried out using the computer program TFPGA (Miller, 1997 ), unless indicated otherwise.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 LITERATURE CITED
 
Genetic diversity
We found that the eight enzymes coded for 12 loci (Table 2), all of which were polymorphic. The allele frequencies at each locus in each of the 16 populations of the four Cucurbita taxa studied are given in Table 2. The average effective number of alleles per locus is 2.08 (Table 3). Twenty-two of 31 alleles occurred in all Cucurbita populations (Table 2). Four of the 16 populations harbored a private allele, each one in a distinct taxon, including one in the outgroup taxon, C. pepo. In C. argyrosperma ssp. argyrosperma there were five rare alleles, and one population, CAA11, had two rare alleles and a single private allele (Table 2). Mean percentage of polymorphic loci, p, was slightly higher in C. argyrosperma ssp. sororia (1.0) than in C. moschata (0.97) or in C. argyrosperma ssp. argyrosperma (0.93; Table 3). On average, 96% of the loci were polymorphic in each population and the mean expected heterozygosity was 0.407. The mean expected heterozygosity was higher again in C. argyrosperma ssp. sororia (He = 0.426) and lower in C. argyrosperma ssp. argyrosperma (He = 0.391), while the average expected heterozygosity in C. moschata was intermediate, 0.416 (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 2. Allele frequencies at 12 polymorphic loci for five C. argyrosperma ssp. sororia (numbers 1–5), six C. argyrosperma ssp. argyrosperma (numbers 6–11), four C. moschata (numbers 12–15), and one C. pepo (number 16) populations. For population locations, refer to Table 1

 

View this table:
[in this window]
[in a new window]
  		Table 3. Levels of intrapopulation allozyme variation in four Cucurbita taxa

 
Of the 192 {chi}2 tests of the deviation of F from the expected value per locus per population, 94 loci–population combinations had an F value significantly different from zero. Of these 94 tests, 39 were positively different from zero, and 55 were negatively different from zero ({alpha} ≤ 0.05–<0.005; Tables 3, 4). Sixty percent of the populations of C. argyrosperma ssp. sororia, 75% of C. moschata, and 50% of C. argyrosperma ssp. argyrosperma had average within-population F values that were significantly negative, indicating an excess of heterozygotes (Table 3). In both subspecies of C. argyrosperma, the mean F values averaged across all populations were positive or nearly zero (0.006 in C. argyrosperma ssp. sororia; –0.046 in C. argyrosperma ssp. argyrosperma), while in C. moschata the average F values were negative (–0.174). The number of loci in each Cucurbita population with F values significantly different from zero ranged from three in population CAA7 to 11 in population CAA6 (Table 3). The average number of loci that had F values significantly different from zero was similar in the three taxa (C. argyrosperma ssp. sororia = 6.4, C. argyrosperma ssp. argyrosperma = 5.5, and C. moschata = 6.0; Table 3), although when the F values were calculated over all 12 loci, only one population (CMO12) showed a significant departure from H-W equilibrium (95% CI; data not shown). The loci that exhibit the greatest departure from H-W equilibrium were Apx2 (11 populations) and Pgi2 (16 populations; Table 4).


View this table:
[in this window]
[in a new window]
  		Table 4. Wright's (1965) F statistics for levels of gene flow within and among 16 populations of four taxa of Cucurbita (with jackknife means and standard deviations)

 
Similarly, analyses of Wright's F statistics for each locus showed highly significant departures from zero for many loci (Table 4). However, values of FIS and FST calculated by taxon or locality are low, implying that there are small deficiencies in homozygotes globally and among populations. FIS and FIT values were significantly different from zero for nine and eight loci, respectively; four loci were negatively significant for FIS and three for FIT, averaged over all Cucurbita populations (Table 4). Eleven of the 12 loci had significant FST values (Table 4). Table 5 presents F statistics evaluated by taxon (CAS, CAA, or CMO) and then by locality (San Miguel, San Lorenzo, El Chante y Los Parajitos). The analyses by locality included one population of each taxon resident in that locality. All of the FIS values were negative except the one calculated for taxon CAS and the taxa of San Lorenzo, although none of the values were significantly different from zero. Neither the FIT values calculated by taxon nor by locality were significantly different from zero. On the other hand, all of the FST values were positive and were significantly different from zero based on 95% CI.


View this table:
[in this window]
[in a new window]
  		Table 5. Wright's (1965) F statistics and levels of gene flow (Crow and Aoki's 1984, with standard deviation in parentheses) within and among populations of different number and taxa in Cucurbita and localities. Confidence intervals (95% CI) for means of F statistics values are based on 1000 bootstrap samples. Values for which the CI did not include zero are significant and are in bold face. For taxon names refer to Table 1.

 
Wright's FST statistics also reflected little genetic differentiation among populations in C. argyrosperma ssp. sororia, C. argyrosperma ssp. argyrosperma, C. moschata, and C. pepo (Table 4). The average FST value for all taxa was 0.111 (Table 4), and within-taxon mean values were 0.040 in C. argyrosperma ssp. sororia, 0.096 in C. argyrosperma ssp. argyrosperma, and 0.077 in C. moschata, with an overall average FST value for all 15 Cucurbita populations (without the outgroup) of 0.087 (Table 5). Thus, between 90.3% and 95.9% of the total allozyme variation in a taxon was found within populations.

Relatedness between populations
Nei's genetic distance between populations of the three taxa ranged from 0.024 to 0.228 (Table 6). The range of mean genetic distances was lower between C. argyrosperma ssp. argyrosperma and C. argyrosperma ssp. sororia = 0.069, than between either subspecies and C. moschata (Table 6). Clustering of populations based on D shows that individuals of both subspecies of C. argyrosperma were interlaced and, with the exception of SL populations, appear to cluster more closely based on locality than on taxonomic identity (Fig. 2), while C. moschata's populations form a clearly separate group with a well-supported basal node (Fig. 2).


View this table:
[in this window]
[in a new window]
  		Table 6. Matrix of average Nei's (1978) genetic distance (D; range in brackets) for all sampled populations of the four Cucurbita taxa, C. argyrosperma ssp. sororia (CAS), C. argyrosperma ssp. argyrosperma (CAA), C. moschata (CMO), and C. pepo (CPE)

 


View larger version (22K):
[in this window]
[in a new window]
  		Fig. 2. Dendogram based on UPGMA of allozyme data for four taxa of Cucurbita: C. argyrosperma ssp. sororia (CAS), C. argyrosperma ssp. argyrosperma (CAA), C. moschata (CMO), and C. pepo (CPE), which was used as an outgroup. The numbers in the nodes show the proportion of similar replicates for 1000 bootstraps. For taxon and localities names, refer to Table 1 .

 
Levels of interpopulation gene flow
The values of estimated gene flow for each locus were very variable and nearly all were greater than one (Table 4). The average Nem values averaged over all loci and taxa was 4.49, but ranged from 0.56 to 19.75 at different loci (Table 4). Within-taxon values were 5.74 for C. argyrosperma ssp. argyrosperma, 18.85 for C. argyrosperma ssp. sororia, and 11.54 in C. moschata (Table 5). Nem values averaged by locality ranged from 3.01 in Los Parajitos to 22.46 in San Miguel (Table 5). Average Nem values over loci within taxa and localities are high considering the variation among loci (Tables 4, 5).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
LITERATURE CITED
 
Levels of genetic variation
The per population average polymorphism (p = 0.99), allelic diversity (A = 2.5), and heterozygosity (He = 0.409) are high in comparison with values reported for other outcrossing, animal-pollinated, cultivated seed plants (p = 0.34; A = 2.67; He = 0.205; Hamrick and Godt, 1989, 1997 ) and are comparable to the highest values reported for cultivated outcrossing plants (p = 0.95; A = 1.34; He = 0.225; Jenczewski, Prosperi, and Ronfort, 1999 ). Our estimates were also high compared with other Cucurbita studies. For example, Decker-Walters et al. (1990) found p = 19.3, A = 2.24, He = 0.068 for C. pepo and p = 11.5, A = 1.43, He = 0.039 in C. maxima, even though other Cucurbitaceae taxa have higher He values (He = 0.225; Akimoto, Fukuhara, and Kikuzawa, 1999 ). Nevertheless, we have to point out that our estimates may be upwardly biased because we used polymorphic loci previously reported in other studies.

Inbreeding
The main genetic effects of inbreeding within populations is to increase the levels of homozygosity relative to those expected under conditions of random mating (Brown, 1979 ; Hamrick, 1989 ). Thus, comparisons of genotypic frequencies within populations with those expected under H-W equilibrium conditions can be used to detect historical levels of inbreeding or outcrossing in natural populations. In our Cucurbita populations, individual fixation index values showed significant departures from H-W in several cases and most values were slightly negative, indicating a small excess of heterozygotes and little inbreeding (mean FIS = –0.071, not significantly different from 0). These results certainly do not rule out the possibility of occasional selfing, but they do argue against persistent selfing within these populations.

The combination of high genetic diversity within populations, little differentiation among populations, and high levels of gene flow indicate that these species are predominately outcrossing. Kirkpatrick and Wilson (1988) also found high outcrossing rates and low inbreeding in the cultivated C. pepo and in the weedy C. texana (Scheele) Gray [= C. pepo ssp. overifera (L.) Decker var. texana (Scheele) Filov].

Genetic differentation
There was little genetic differentiation among the taxa. Nei's mean genetic distance values were low, ranging from 0.046 to 0.122. These distance values are lower than other D estimates for Cucurbita and within the range of values reported for conspecific seed plant populations (Decker, 1986 ; Decker and Wilson, 1987 ; Wilson, 1989 ; Merrick, 1991 ). For instance, in south-central Africa populations of C. moschata a study by Gwanama, Labuschagne, and Botha (2000) using random amplified polymorphic DNA (RAPD) markers reported D values from 0.31 to 0.41.

Our values of Wright's FST (average = 0.111) were lower than the reported average for animal-pollinated outcrossing seed plants (Hamrick, 1989 ) or for outcrossing cultivated seed plants (Hamrick and Godt, 1997 ) (FST mean values = 0.187 and 0.234, respectively). Both the D and the FST values suggest that there is substantial gene flow among populations of the three taxa we studied, as we will discuss later.

Our UPGMA analysis shows the existence of two major groups, Cucurbita argyrosperma (including both subspecies, C. argyrosperma ssp. argyrosperma and C. argyrosperma ssp. sororia) and C. moschata. However, we must caution that the bootstrap values in the tree are low, likely caused by the few differences between Cucurbita populations. In particular, populations of C. argyrosperma ssp. argyrosperma and C. argyrosperma ssp. sororia are more differentiated by locality than taxonomic identity, suggesting that their taxonomic classification should be reconsidered. However, it is possible that the similarity of these subspecies at allozyme loci does not reflect their differences at other loci determining morphological or reproductive characters important for species definition or that high levels of gene flow are eroding differentiation between them. This latter possibility is supported by the observation that the only population of C. argyrosperma that was clearly differentiated from another population was population CAA8 of C. argyrosperma ssp. argyrosperma collected from Tepec. This population occurs in a locality where the wild gourd, C. argyrosperma ssp. sororia, is not present. More genetic analyses and reproductive studies would be required before a firm conclusion could be reached about the taxonomic status of these two subspecies.

The distribution of allozyme variation in these three Cucurbita taxa includes high genetic variation within populations and low divergence among populations, consistent with the hypothesis that the taxa share a recent common ancestor and that there is ongoing gene flow between them. Our observation of homozygote deficiency in some populations could be due to high outcrossing rates or mating among genetically different individuals in large populations (disassortative mating) or because gene flow introduces variation into populations before random mating balances it. Cucurbita argyrosperma ssp. sororia, C. argyrosperma ssp. argyrosperma, and C. moschata exist as extensive populations throughout the Jalisco region; they are often cultivated in adjacent plots, and the total number of individuals can range from a few hundred to several thousand. Given that several pollinating bee species move among the plants in the plots, visiting all Cucurbita taxa simultaneously (data not shown), high levels of interpopulation and intertaxa gene flow is not surprising.

Gene flow within wild-cultivated plant complex
Gene flow can be an important evolutionary force. A small amount of gene flow is capable of counteracting other evolutionary forces such as mutation, drift, and selection (Slatkin, 1987 ). Previous studies have shown evidence of gene flow between populations of C. pepo and C. texana (Kirkpatrick and Wilson, 1988 ) and C. argyrosperma with C. fraterna L. H. Bailey (Wilson, Lira, and Rodriguez, 1994 ). The low values of FST reported here suggest that there is exchange of genes among populations, even among those located several kilometers apart within the agroecosystem. Our very low FST estimates are lower than the average reported for outcrossing seed plants (Hamrick, 1989 ). In particular, we found high rates of gene flow in San Miguel among populations within 1.5 km of each other.

The great potential of pollen dispersal is likely to contribute to the high rates of gene flow among cultivated and weedy Cucurbita populations. Kirkpatrick and Wilson (1988) found outcrossing rates of 5% in C. pepo and C. texana, at a distance of 1300 m. Arias and Rieseberg (1994) demonstrated gene dispersal from 2 to 7% at 800 m and 2% at 1000 m in the Helianthus annuus crop/weed complex. Jenczewski, Prosperi, and Ronfort (1999) reported evidence for the occurrence of crop/weed gene flow in Medicago sativa. In addition, in a separate study we found that 62% of farmers exchange squash seeds among themselves (Montes-Hernández, Merrick, and Eguiarte, unpublished manuscript).

One might expect local farmers to keep C. argyrosperma ssp. sororia, C. argyrosperma ssp. argyrosperma, and C. moschata distinct by selection of fruits and seeds to sow in the next season. However, it is possible that squash farmers could have difficulty in distinguishing among the taxa if one or more squash fruits are F1 or part of the progeny of a cross between a landrace and a weedy type, particularly with C. argyrosperma ssp. argyrosperma (José Elizondo Anguiano, squash farmer from El Chante, personal communication).

Although genetic variation of many species could be reduced by human activities, some crop species have increased genetic variation due to certain selective practices inherent in their cultivation (Bye, 1993 ; Hernández, 1993 ). The continuous presence of wild populations of C. argyrosperma ssp. sororia and the milpa tradition of sowing together seeds of C. argyrosperma ssp. argyrosperma and C. moschata with corn adjacent to these wild populations probably both contribute to the high levels of gene flow observed here.

Relevance in the light of release of genetically modified organisms
Gene flow between domesticated plants and their wild relatives may have two potentially harmful consequences, the evolution of increased weediness and the increased likelihood of extinction of wild relatives. Recent concern exists for the possibility of gene transfer between transgenic cultivated species and their wild relatives (Ellstrand, Prentice, and Hancock, 1999 ). However, before this risk can be assessed, studies that examine the extent of introgression between the wild and transgenic plants are necessary (Payne, 1997 ). Our study suggests that crop to weed gene flow regularly occurs between wild and cultivated populations of squashes in Mexico. Mexico is the origin and center of diversity for Cucurbita, and we need to be extremely cautious in the field testing and wide-scale usage of transgenic plants of squash in Mexico.


   FOOTNOTES
 
1 The authors thank Aldo Valera-Vazquez and Mariano Corona-López for assisting in the field seed collections; Aldo Valera-Vazquez and Gumersindo Sánchez-Montoya for assisting in the laboratory; Sara V. Good-Avila, Laura C. Merrick, Erika Aguirre-Planter, and Juan Nuñez-Farfan for reading a previous version of the manuscript; and James F. Hancock and an anonymous reviewer for comments on the manuscript. This study was supported by McKnight Foundation (USA) Milpa project, PAEP-UNAM project 1998-2335, Conacyt project 27938-N, and by the Instituto de Ecología at the Universidad Nacional Autónoma de México, and the Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias (INIFAP) in Mexico. Back

4 Author for reprint requests (smontes{at}miranda.ecologia.unam.mx) . Back


   LITERATURE CITED
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Akimoto J. T. Fukuhara K. Kikuzawa 1999 Sex ratios and genetic variation in a functionally androdioecious species, Schizopepon bryoniaefolius (Cucurbitaceae). American Journal of Botany 86: 880-886[Abstract/Free Full Text]

Arias D. M. L. H. Rieseberg 1994 Gene flow between cultivated and wild sunflowers. Theoretical and Applied Genetics 89: 655-660[ISI]

Beebe S. O. Toro A. V. González M. I. Chacón D. G. Debouck 1997 Wild-weed-crop complex of common bean (Phaseolus vulgaris L., Fabaceae) in the Andes of Peru and Colombia, and their implications for conservation and breeding. Genetic Resources and Crop Evolution 44: 73-91

Brown A. H. D. 1979 Enzyme polymorphism in plant populations. Theoretical Population Biology 15: 1-42

Bye R. 1993 The role of humans in the diversification of plants in Mexico. In T. P. Ramammoorthy, R. Bye, A. Lot, and J. Fa [eds.], Biological diversity of Mexico: origins and distribution, 707–731. Oxford University Press, New York, New York, USA

Cardy B. J. C. W. Stuber M. N. Goodman 1980 Techniques for starch gel electrophoresis of enzymes from maize (Zea mays L). Institute of Statistics Mimeo Series Number 1317. North Carolina State University, Raleigh, North Carolina, USA

Colunga-García-Marín P. J. Coello-Coello L. E. Eguiarte D. Piñero 1999 Isozymatic variation and phylogenetic relationships between henequen (Agave fourcroydes) and its wild ancestor A. angustifolia (Agavaceae). American Journal of Botany 86: 115-123[Abstract/Free Full Text]

Crow J. F. K. Aoki 1984 Group selection for a polygenic behavioral trait: estimating the degree of population subdivision. Proceedings of the National Academy of Sciences, USA 81: 6073-6077[Abstract/Free Full Text]

Decker D. S. 1986 A biosystematic study of Cucurbita pepo. Ph.D. dissertation, Texas A&M University, College Station, Texas, USA

Decker D. S. H. D. Wilson 1987 Allozyme variation in the Cucurbita pepo complex: C. pepo var. ovifera vs. C. texana. Systematic Botany 12: 263-273

Decker-Walters D. S. T. W. Walters U. Posluszny P. G. Kevan 1990 Genealogy and gene flow among annual domesticated species of Cucurbita. Canadian Journal of Botany 68: 782-789

Doebley J. 1990 Molecular evidence for gene flow among Zea species. BioScience 40: 443-448[CrossRef][ISI]

Doebley J. 1992 Molecular systematics and crop evolution. In P. S. Soltis, D. E. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants, 202–222. Chapman & Hall, New York, New York, USA

Ellstrand N. C. C. A. Hoffman 1990 Hybridization as an avenue of escape for engineered genes: strategies for risk reduction. BioScience 40: 438-442[CrossRef][ISI]

Ellstrand N. C. H. C. Prentice J. F. Hancock 1999 Gene flow and introgression from domesticated plants into their wild relatives. Annual Review of Ecological and Systematics 30: 539-563[CrossRef][ISI]

Escalante A. M. G. Coello L. E. Eguiarte D. Piñero 1994 Genetic structure and mating systems in wild and cultivated populations of Phaseolus coccineus and P. vulgaris (Fabaceae). American Journal of Botany 81: 1096-1103[CrossRef][ISI]

Gwanama C. M. T. Labuschagne A. M. Botha 2000 Analysis of genetic variation in Cucurbita moschata by random amplified polymorphic DNA (RAPD) markers. Euphytica 113: 19-24[CrossRef][ISI]

Hamrick J. L. 1989 Isozymes and the analysis of genetic structure in plant populations. In D. E. Soltis and P. S. Soltis [eds.], Advances in plant sciences series, vol. 5, Isozymes in plant biology, 87–105. Dioscorides Press, Portland, Oregon, USA

Hamrick J. L. M. J. W. Godt 1989 Allozyme diversity in plant species. In A. H. D. Brown, M. T. Clegg, M. A. Kahler, and B. S. Weir [eds.], Plant population genetics, breeding, and genetic resources, 43–63. Sinauer, Sunderland, Massachusetts, USA

Hamrick J. L. M. J. W. Godt 1997 Allozyme diversity in cultivated crops. Crop Science 37: 26-30[Abstract/Free Full Text]

Hancock J. F. R. Grumet S. C. Hokanson 1996 The opportunity for escape of engineered genes from transgenic crops. HortScience 31: 1080-1085[Free Full Text]

Harlan J. R. 1976 Genetic resources in wild relatives of crops. Crop Science 16: 329-333[Abstract/Free Full Text]

Hernández X. E. 1993 Aspects of plant domestication in Mexico: a personal view. In T. P. Ramammoorthy, R. Bye, A. Lot, and J. Fa [eds.], Biological diversity of Mexico: origins and distribution, 733–753. Oxford University Press, New York, New York, USA

Hokanson S. C. J. F. Hancock R. Grumet 1997 Direct comparison of pollen-mediated movement of native and engineered genes. Euphytica 96: 397-403[CrossRef][ISI]

Hurd P. D., Jr. E. G. Linsley T. W. Whitaker 1971 Squash and gourd bees (Peponapis, Xenoglossa) and the origin of the cultivated Cucurbita. Evolution 25: 218-234[CrossRef][ISI]

Jarvis D. A. T. Hodgkin 1999 Wild relatives and crop cultivars: detecting natural introgression and farmer selection of new genetic combinations in agroecosystems. Molecular Ecology 8: S159-S173

Jenczewski E. J.-M. Prosperi J. Ronfort 1999 Evidence for gene flow between wild and cultivated Medicago sativa (Leguminosae) based on allozyme markers and quantitative traits. American Journal of Botany 86: 677-687[Abstract/Free Full Text]

Kirkpatrick K. J. H. D. Wilson 1988 Interspecific gene flow in Cucurbita: C. texana vs. C. pepo. American Journal of Botany 75: 519-527[CrossRef][ISI]

Li C. C. D. G. Horvitz 1953 Some methods of estimating the inbreeding coefficient. American Journal Human Genetics 5: 107-117

Lira S. R. 1995 Estudios Taxonómicos y Ecogeográficos de las Cucurbitacea Latinoamericanas de Importancia Económica. Systematic and Ecogeographic Studies on Crop Genepools 9. International Plant Genetic Resources Institute, Rome, Italy

Merrick L. C. 1990 Systematics and evolution of a domesticated squash, Cucurbita argyrosperma, and its wild and weedy relatives. In D. M. Bates, R. W. Robinson, and C. Jeffrey [eds.], Biology and utilization of the Cucurbitaceae, 77–95. Cornell University Press, Ithaca, New York, USA

Merrick L. C. 1991 Systematics, evolution and ethnobotany of a domesticated squash, Cucurbita argyrosperma. Ph.D. dissertation, Cornell University, Ithaca, New York, USA

Merrick L. C. 1995 Squash, pumpkins, and gourds: Cucurbita (Cucurbitaceae). In J. Smart and N. W. Simmonds [eds.], Evolution of crop plants, 2nd ed., 97–105. Longman, London, UK

Miller M. P. 1997 Tools for population genetic analyses. Department of Biological Sciences, Northern Arizona University, Flagstaff, Arizona, USA

Nei M. 1978 Estimation of average heterozygosity and genetic distance from small numbers of individuals. Genetics 89: 583-590[Abstract/Free Full Text]

Payne J. 1997 Regulating transgenic plants: the experience of USDA in field testing, wide-scale production, and assessment for release in centers of origin. In J. A. Serratos, M. C. Willcox, and F. Castillo-González [eds.], Gene flow among maize landraces, improved maize varieties, and teosinte: implications for transgenic maize, 85–90. CIMMYT (Centro Internacional de Mejoramiento de Maíz y Trigo), México, D.F

Pitel J. A. W. H. Cheliak 1984 Effect of extraction buffers on characterization of isoenzymes from vegetative tissues of five conifer species: a user's manual. Information report PI-X-34. Petawawa National Forest Institute, Canadian Forestry Service, Chalk River, Ontario, Canada

Slatkin M. 1987 Gene flow and the geographic structure of natural populations. Science 236: 787-792[Abstract/Free Full Text]

Soltis D. E. C. H. Haufler D. D. Darrow G. J. Gastony 1983 Starch gel electrophoresis of ferns: a compilation of grinding buffers, gel and electrode buffers, and staining schedules. American Fern Journal 73: 9-27[CrossRef][ISI]

Souza V. J. Bain C. Silva V. Bouchet A. Valera E. Marquez L. E. Eguiarte 1997 Ethnomicrobiology: do agricultural practices modify the population structure of the nitrogen fixing bacteria Rhizobium etli Biovar phaseoli?. Journal of Ethnobiology 17: 249-266

Swofford D. L. R. B. Selander 1981 BIOSYS-1: a Fortran program for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72: 281-283[Abstract/Free Full Text]

Weir B. S. C. C. Cockerham 1984 Estimating F-statistics for the analysis of population structure. Evolution 38: 1358-1370[CrossRef][ISI]

Wessel-Beaver L. 2000 Cucurbita argyrosperma sets fruit in fields where C. moschata is only pollen source. Cucurbit Genetics Cooperative Report 23: 62-63

Whitaker T. W. G. W. Bohn 1950 The taxonomy, genetics, production and uses of the cultivated species of Cucurbita. Economic Botany 4: 54-81

Wilkes H. G. 1977 Hybridization of maize and teosinte in Mexico and Guatemala and the improvement of maize. Economic Botany 31: 254-293[ISI]

Wilson H. D. 1989 Discordant patterns of allozyme and morphological variation in Mexican Cucurbita. Systematic Botany 14: 612-623[CrossRef][ISI]

Wilson H. D. 1990 Gene flow in squash species. BioScience 40: 449-455[CrossRef][ISI]

Wilson H. D. J. Doebley M. Duvall 1992 Chloroplast DNA diversity among wild and cultivated members of Cucurbita (Cucurbitaceae). Theoretical and Applied Genetics 84: 859-865[ISI]

Wilson H. D. R. Lira I. Rodriguez 1994 Crop-weed gene flow: Cucurbita argyrosperma Huber and C. fraterna L.H. Bailey (Cucurbitaceae). Economic Botany 48: 293-300[ISI]

Wright S. 1951 The genetical structure of population. Annals of Eugenics 15: 323-354[ISI]

Wright S. 1965 The interpretation of populations structure by F-statistics with special regard to systems mating. Evolution 19: 395-420[CrossRef][ISI]

Yeh F. C. D. M. O'malley 1980 Enzyme variation in natural populations of Douglas-fir, Pseudotsuga menziesii (Mirb.) Franco, from British Columbia. I. Genetic variation patterns in coast populations. Silvae Genetica 29: 83-92




This article has been cited by other articles:


Home page
 Crop Sci.Home page
M. Ferriol, B. Pico, P. F. de Cordova, and F. Nuez
Molecular Diversity of a Germplasm Collection of Squash (Cucurbita moschata) Determined by SRAP and AFLP Markers
Crop Sci., March 1, 2004; 44(2): 653 - 664.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Bot.Home page
K. Birnbaum, R. DeSalle, C. M. Peters, and P. N. Benfey
Integrating gene flow, crop biology, and farm management in on-farm conservation of avocado (Persea americana, Lauraceae)
Am. J. Botany, November 1, 2003; 90(11): 1619 - 1627.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Montes-Hernandez, S.
Right arrow Articles by Eguiarte, L. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Montes-Hernandez, S.
Right arrow Articles by Eguiarte, L. E.
Agricola
Right arrow Articles by Montes-Hernandez, S.
Right arrow Articles by Eguiarte, L. E.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS