Crop Science Grow Your Career with CSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in Crop Science
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 ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by del Rio, A. H.
Right arrow Articles by Bamberg, J. B.
Right arrow Search for Related Content
PubMed
Right arrow Articles by del Rio, A. H.
Right arrow Articles by Bamberg, J. B.
Agricola
Right arrow Articles by del Rio, A. H.
Right arrow Articles by Bamberg, J. B.
Related Collections
Right arrow Crop Genetics
Right arrow Potato
Right arrow Plant Genetic Resources
Published in Crop Sci. 44:1170-1177 (2004).
© 2004 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA

CROP BREEDING, GENETICS & CYTOLOGY

Geographical Parameters and Proximity to Related Species Predict Genetic Variation in the Inbred Potato Species Solanum verrucosum Schlechtd

A. H. del Rioa,* and J. B. Bambergb

a Dep. of Horticulture, Univ. of Wisconsin–Madison, 1575 Linden Dr., Madison, WI 53706
b USDA-ARS, Vegetable Crops Research Unit, Inter-Regional Potato Introduction Station, 4312 Hwy. 42, Sturgeon Bay, WI 54235

* Corresponding author (adelrioc{at}wisc.edu).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Identification of ecogeographic parameters associated with genetic variation would help to prioritize populations within conserved species for collection and maintenance. Previous studies comparing accessions of three Solanum species representing different breeding systems, S. fendleri A. Gray (a disomic polyploid selfer), S. jamesii Torr. (a diploid outcrosser), and S. sucrense Hawkes (a polysomic polyploid outcrosser), revealed no significant associations between genetic and ecogeographic variation. Even physical separation did not predict genetic distance. These previous studies are expanded here by investigating the relationship of genetics and geographic parameters for a fourth type of breeding system: a diploid selfer, modeled by wild S. verrucosum Schlechtd (2n = 2x = 24) from Mexico. The objective was to assess whether genetic distances between accessions are predicted by differences in geographic parameters at the natural sites of origin. Twenty-seven S. verrucosum accessions were studied using 152 RAPD markers to estimate the genetic distance between pairs of accessions. The association of genetic variation with proximity of the S. verrucosum accessions to other Mexican wild potato species (some known to be sympatric with S. verrucosum) was also analyzed. Pairwise comparisons showed that the average GD of all pairs of accessions was 12.8% (ranging from 0–29.3%). Parameters with variation significantly correlated with the average GD of S. verrucosum populations were physical separation (0.42), latitude (0.70), longitude (0.51), proximity to S. hjertingii Hawkes (0.80), S. hougasii Corr. (–0.75), and S. demissum Lindl. (–0.50). Significant effects of proximity to S. hjertingii (for example) might be caused by introgression, or the coincidence that all S. hjertingii occur in the extreme NE corner of the S. verrucosum range.

Abbreviations: GD, genetic distance • GR, genetic rarity • NRSP-6, Inter-Regional Potato Introduction Station • RAPD, Random Amplified Polymorphic DNA


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
THE PRIMARY OBJECTIVE of the U.S. Potato Genebank (NRSP-6) is to preserve and safeguard genetic diversity for potential future uses in research and breeding. As with NRSP-6, other potato genebanks worldwide have acknowledged that research to improve conservation methods is required and should be considered a high priority (Bamberg et al., 1995). One factor potentially preventing optimal representation of the species gene pool is inadequate sampling during collections (Altieri and Merrick, 1987). Inadequate sampling may, in part, be attributed to the lack of accurate information about the distribution of genetic diversity among natural populations (IBPGR, 1985). Several studies on this subject agree that the best approach to sampling is to assume an association of genetic diversity with diversity in ecogeographical patterns (Antonovics, 1971; Ferguson et al., 1998; Loveless and Hamrick, 1984; Moeller and Schaal, 1999). Linhart and Grant (1996) conclude that the experimental evidence overwhelmingly supports the generalization that natural selection tailors the genetics of plant populations to their environment.

The advent of DNA markers has provided the power to make objective, quantitative conclusions about the partitioning of genetic diversity (Link et al., 1995). Thus, there is the potential to test for trends in genetic variation along ecogeographic clines within species to identify parameters that predict "hot spots" of diversity (Schoen and Brown, 1991). Such information could have a significant economic impact on genebanks by allowing maximum collection of genetic diversity with the minimum number of samples and expense.

The use of DNA markers to study the association of genetics with geography is complicated by the very complex patterns in the distribution of genetic variation exhibited by plant species (Cobb et al., 1994; del Rio et al., 2001; Fahima et al., 1999; Gallois et al., 1998; Hamrick, 1987; Mitton et al., 1998), necessitating the use of multifaceted analyses. Hawkes (1971) and Yonezawa and Ichihashi (1989) suggested that information from ecology, population biology, genetics, and/or reproductive biology must be obtained to gain a better understanding of the genetic–geographical patterns. Greene et al. (1999a)(1999b) demonstrated that geographic information coupled with geographic information systems (GIS) analysis could increase the precision of sampling protocols in ecogeographically diverse sites. Their technique allows determination of information such as site uniqueness relative to other sites sampled, likelihood that an accession reflects adaptation to the site, and uniqueness of a given accession relative to other accessions of the same taxon collected.

With nearly 200 different species distributed in the Americas from the USA to southern South America (Hijmans and Spooner, 2001), potato species are good plant models for studying the association of genetics and geography. They have a wide range of ecological niches, as well as geographical distributions, and exhibit different genetic characteristics that have strong effects on the organization of diversity (Loveless and Hamrick, 1984): disomic vs. polysomic polyploidy, selfing, outcrossing, and asexual reproduction by tubers. Previous studies have looked for links between ecogeographic parameters and specific traits within potato species. Significant associations have been found between altitude and frost killing temperature in S. acaule Bitter as well as altitude and resistance to potato leafhopper and glycoalkaloid content in S. chacoense Bitter (Hijmans et al., 2003). Spooner et al. (1995) reported a significant correlation between general DNA marker differences and geographic separation for populations within species in the series Etuberosa, a group of highly inbreeding taxa. Previously, we attempted to associate general DNA marker variation within species with ecogeographical differences by evaluating 124 wild potato accessions with different geographical origins (del Rio et al., 2001; del Rio and Bamberg, 2002) and different breeding systems: S. fendleri (a disomic polyploid selfer), S. jamesii (a diploid outcrosser) and S. sucrense (a polysomic polyploid outcrosser). In each case, it was concluded that no significant link exists between the genetic and ecogeographic variables tested. Even physical separation, a common criterion for selection of collection sites in explorations, did not predict genetic distance. Thus, the objective of this investigation was to expand those previous studies by examining a fourth type of breeding system: an inbred diploid, modeled by S. verrucosum from Mexico to assess whether genetic distance between accessions is predicted by differences in geographic parameters.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Samples of all available accessions of S. verrucosum that had location data were obtained from the U.S. Potato Genebank (NRSP-6), Sturgeon Bay, WI, USA. Table 1 provides latitude, longitude, and altitude of the populations sampled to produce the accessions evaluated. Figure 1 gives a visual representation of the geographic distribution of the accessions. More detailed information about the 27 accessions used in this study can be accessed through on-line national and international databases linked to the NRSP-6 homepage, http://www.ars-grin.gov/nr6; verified 16 March 2004.


View this table:
[in this window]
[in a new window]
  		Table 1. Solanum verrucosum plant introductions from the NRSP-6 Potato Introduction Station, their geographic locations in Mexico, genetic clusters, and GR.

 


View larger version (20K):
[in this window]
[in a new window]
  		Fig. 2. Principal component analysis clusters corresponding to physical location (see Table 1 for codes identifying accessions).

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 1. Origin locations of S. verrucosum accessions studied and natural range of some other wild potato species growing nearby.

 
Two replicates of 27 bulked seedlings from each accession were grown and used to generate 152 polymorphic random amplified polymorphic DNA (RAPD) loci by using 54 different primers. The methods for DNA extraction, RAPD assay and data analyses are described in del Rio et al. (1997). Genetic distance (GD), based on RAPD markers, was calculated as the complement of the simple matching coefficient (i.e., the percent loci with common band status). These were developed by means of the NTSYS-pc program (Rohlf, 2002). For each accession, the average GD of all of its pairwise comparisons with other accessions was calculated for a measure of genetic rarity (GR). Tests of association between GR, geographic variables, and proximity to other species were calculated with the Pearson's Correlation function in Microsoft Excel. Regression analyses were performed by PROC REG subroutine of the SAS program (SAS Institute, 1989).

The average physical separation of each accession from all others was calculated and then correlated with its GR. For the other geographic parameters, the real, un-compared value associated with a given accession (i.e., not differences or averages) was tested for correlation with GR. For example, a high positive correlation between GR and altitude would mean that greater GR is associated with higher altitude. GR of each S. verrucosum was also compared to its proximity to the nearest known population of certain other potato species observed to be close to or overlap with that of S. verrucosum. Geographic data already extant in NRSP-6 databases were used for all calculations. Because effects could be obscured by considering the entire set of populations, Principal Component (PC) Analysis (Jolliffe, 2002) was done to identify genetic clusters of populations that could then be independently compared with geographic parameters. PC analysis was calculated by the NTSYS-pc program (Rohlf, 2002).

Variables compared to genetic data were latitude, longitude, altitude, and physical separation—the most basic location-based information recorded by germplasm collectors. These parameters are routinely used by genebanks to determine the geographical representation of the ex situ collection. Reliance on these parameters represents the reasonable assumption that collections relatively close together probably share a similar environment, and that those located farther from each other have limited potential for genetic exchange. The observation that these S. verrucosum accessions originated from the same (or nearly the same) locations as other wild potato species (S. demissum, S. hjertingii, S. hougasii) prompted the inclusion of physical separation from the nearest known site for other species as a additional parameter associated with each S. verrucosum accession. The locations of the nearest reported accessions of other species was determined from genebank germplasm records (Huaman et al., 2000) and herbarium specimen records (as in Hijmans et al., 2002).


   RESULTS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
The average GD of duplicate samples from the same accession was 0.5%. Pairwise comparisons (351 total) among the 27 accessions of S. verrucosum revealed that genetic distance based on 152 polymorphic RAPD markers ranged from 0 to 29% with a mean of 12.8%. Thirteen comparisons between accessions collected at the same site had an average GD of 5.4% with a range of 0 to 14.6%. On the other hand, 24 comparisons between accessions with the greatest geographic separation (between 400 and 450 km) averaged 17%, ranging from 12.8 to 21.4%. The average GR, which equals the average of all pairwise GD, was 12.8%, but ranged from 8.4 to 20.2% (Table 1).

The PC analysis (Fig. 2) revealed clusters with an apparent correspondence to the physical distribution map. These were a "North" cluster of five populations, a "Southeast" cluster of three populations, and a "South" cluster of the remaining 19 populations. Table 1 designates the members of these clusters. As shown in Table 2, average GD of pairs within the North and South clusters were about equal at 14%, but populations within the South cluster were more similar to each other at about 9%. Comparisons between the North and Southeast cluster indicated that populations in these two areas are most different. Considering all populations and clusters, the most rare populations are found in the Southeast cluster (average GD to all = 17.4%), followed by the North cluster (17.1%), and the South cluster (11.0%).


View this table:
[in this window]
[in a new window]
  		Table 2. Average genetic diversity (GD) of paired comparisons within and among North, South, and Southeast clusters identify by Principal Component Analysis. (See Table 1 for cluster members).

 
Table 3 gives the association between the change in geographic parameters and species proximity versus the change in GR within all populations, and within the South cluster. North and Southeast clusters had too few members (five and three, respectively) for within-cluster analysis to have a practical or statistical significance.


View this table:
[in this window]
[in a new window]
  		Table 3. Pearson's correlation coefficients for average genetic distance (GR = genetic rarity) of Solanum verrucosum populations vs. that of geographic variables and proximity to other Solanum species.

 
Considering all populations, both latitude (northward) and longitude (eastward) had significant correlation with GR. It is not surprising that the sum of these (i.e., northeastward), and physical separation per se were also highly correlated with GR. Indeed, the stepwise multiple regression analysis confirmed that latitude and physical distance are the most important parameters (R2 = 0.36).


   DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Most of the parameters in Table 3 had the same overall relationship with GR whether all populations or only South populations were considered. Of the parameters tested, only altitude had no significant relationship to GR. The South cluster (19 of the 27 populations) does not vary much in latitude, and that correlation with GR was not significant, showing that the highly significant effect of latitude is due almost completely to the relative genetic distinction and northern location of the Southeast and North clusters. Latitude and longitude combine for a "northeast" parameter highly correlated with GR (Table 3 and Fig. 3a) . Association of longitude of the South cluster members with GR was barely significant, and overall physical separation was almost significantly negative. The insignificant effect of physical separation in the South cluster is to be expected considering that populations at both the east and west extremes of the range of this cluster have the greatest average physical separation, but the analysis of the longitude association in this group already revealed that greater GR is only slightly associated with eastern populations. So genetic diversity in this cluster does not have a very strong east-to-west pattern, nor does it tend to follow a pattern of dispersion from the center. Indeed, populations in this South cluster are relatively uniform having a much lower average GD ( = GR) than other comparisons within and among other clusters (Table 2).



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 3. Plots of GR within North, South, and Southeast clusters versus (a) Latitude in degrees north of 19 and (b) distance to nearest S. demissum in kilometers.

 
Table 3 shows that Physical Separation and Distance to S. demissum had a significantly positive correlation with GR when considering all the S. verrucosum populations, but a significantly negative one within the South cluster. Figure 3b illustrates how this is possible. North cluster populations have the highest GR, and also very great average separation and distance from S. demissum. So when they are included in the analysis, the significant overall trend is for a positive correlation, despite the fact that the South cluster populations tend to have lower GR if they are farther apart and farther from S. demissum. These cases in which the correlation is positive considering the whole group but negative for a subgroup demonstrates the value of initial clustering of the populations by PC analysis. But the utility of such clustering depends wholly on the goal of the collector—i.e., whether the objective is extending the diversity of the overall set, or capturing the maximum diversity within a specific region (e.g., the origin of the South cluster in this case).

In harmony with the overall effects noted in Table 3, the geographic regions from which the North and Southeast cluster originate would be expected to be the richest sources for collecting additional diversity (Table 2). Proximity to S. demissum has the strongest association of any parameter with respect to the South cluster populations (-0.67) (Table 3). This means the closer a population is to S. demissum, the greater its GR. Note too that the Southeast cluster members are all very close to S. demissum and have high GR. These observations support the idea that the relative distinction of proximal S. verrucosum populations in the South cluster may be due to more intense introgression from S. demissum. A hypothesis of rarity due to introgression is not contradicted by the observation that proximity to S. hougasii is associated with lower GR. Ten of the 19 South cluster populations are found at the same location as a known S. hougasii population. If there were interspecific geneflow to S. verrucosum from S. hougasii at each of these sites, the net effect would be to make these populations more common, in contrast to the effect of S. demissum where only 5 of the 19 populations are sympatric with it. Proximity to S. hjertingii is very strongly associated with GR considering all populations because the members of the North cluster are all much closer to S. hjertingii, and this cluster has a high average GR. While only one member of the North cluster is sympatric with S. hjertingii per se, this population had a higher GR than any other in the North (or anywhere else).

This research sought to detect associations between site-of-origin variables and genetics of accessions of one wild potato species. The ability to do so depends on several factors. Specifically, clear associations may be enhanced by (i) accurate and precise records of geographic coordinates, altitude, etc.; (ii) genetic uniformity of multiple accessions sampled from the same site; (iii) stability of genetic results from a single seed sample from the genebank; (iv) genebank seedlots (the materials tested) that are representative of the in situ population; (v) reliable taxonomy; and (vi) the best choice of parameters characterizing the origin site and genetics of the accessions. These considerations are addressed below with respect to the observed results.

Accurate Records
This analysis did not include the on-site gathering of any geographic data, but depended on existing genebank records. In practice, the choice of materials and variables often are dictated by the existence, precision, and accuracy of site data. We assume reliability of the basic geographic variables for the accessions tested here. This experiment must be interpreted—and was intended—not only to test whether true geographic data predicts germplasm genetics, but rather whether there is a practical opportunity for genebanks to profitably use the imperfect and incomplete data already available.

Genetic Uniformity in Samples from Very Close Collection Sites
If large genetic distances are observed between accessions collected at a single site (identical geographical site characteristics), there is obviously little hope of associating site parameters with GD. This corresponds to the statistical principle of a large experimental error reducing the possible resolution of treatment means. Some previous research has demonstrated that genetic similarity is closely associated with geographical proximity in other plant species (Francisco-Ortega et al., 1993; Hormaza et al., 1994). However, most studies have revealed otherwise (del Rio et al., 2001; del Rio and Bamberg, 2002; Fahima et al., 1999; Gallois et al., 1998; McGregor et al., 2002). Failure to find closer genetic similarities between proximal populations has been pointed out as the likely weak link frustrating detection of an association between geographic data and genetics (del Rio and Bamberg, 2002). This was not true for S. verrucosum. The average GD of accessions from a single site was nearly half that of all pairs of accessions in the study, but only 1/6 of the maximum GD of accessions from different sites. These differences apparently were sufficient to allow significant correlations. Solanum verrucosum accessions from the same collection site averaged a GD of only 5.4% (in contrast to similar studies of accessions of S. fendleri (17%), S. jamesii (16%, del Rio et al., 2001), and S. sucrense (27%, del Rio and Bamberg, 2002).

Intrapopulation homogeneity also increases the chance of stable samples from a single site. If a population is heterogeneous, there is a chance that particular genotypes presented for sampling will differ depending on the particular growing conditions at the time of collection. But when populations are homogeneous, any sample size, pattern, or timing will result in a stable, representative sample. The within-accession uniformity and among-accession distinctions needed to reveal significant correlations may have been promoted by the breeding behavior of S. verrucosum. Inbreeding greatly reduces gene flow and genetic recombination, and maintains linkage disequilibrium. Thus, genetic diversity is reduced within populations, but increased between populations (Loveless and Hamrick, 1984). Although inbreeders may vary more than outcrossers in their among-population heterogeneity (Schoen and Brown, 1991), small niches with a high level of selective pressure (as may be common for wild potato habitats) will select only the genotype genetically tailored to have the exceptional fitness needed to survive. del Rio and Bamberg (2001) confirmed the suspected low heterogeneity within accessions of S. verrucosum compared with other species (S. fendleri, S. sucrense, and S. jamesii) that showed no significant correlations with ecogeographic parameters (del Rio et al., 2001; del Rio and Bamberg, 2002).

Stability of Results from a Single Accession Sample
Variation in the genetic data from a single seedlot from the genebank could result from technical errors in the RAPD protocol (different PCR or scoring results from the same DNA), or different DNA because too-small bulks were taken of seedlots with large intra-accession heterogeneity. These were not a problem in this experiment. The almost perfect fidelity in reproducing the same DNA fragments in separate bulk samples of 27 plants confirms that the sampling method could reliably identify the genetic profile of each accession, and was similar to results obtained in other potato species using the same methods (del Rio and Bamberg, 1998; del Rio and Bamberg, 2000a).

Test Samples Representative of In Situ Populations
Samples tested could be uniform, but still not representative of the genetics of the in situ population or the geographic parameters of the site. Recall that the plants tested were not obtained directly from the wild, but were sexual progeny generated in the genebank from original collections. It is therefore possible that inadvertent selection or drift during genebank preservation made the tested ex situ sample different from the in situ population. Recent evidence, however, suggests that regeneration in the genebank does not significantly alter the population genetic structure, whether the parental material originated in the genebank or in the wild (del Rio et al. (1997) and del Rio and Bamberg, 2000b; respectively). The apparent relative uniformity of S. verrucosum is here again a possible advantage, since there is no hazard of accidental selection in the genebank if genetic variation is absent.

Taxonomy and Introgression
This study was designed with the assumption that the species name S. verrucosum designates a distinct genetic group. Excluded, for example, were accessions of S. macropilosum Corr., a diploid species that perhaps should be considered an unusual form of S. verrucosum (Hawkes, 1990). Obviously, the practical value of characterizing diversity within a species rests on the validity of the assumption that accessions have been grouped into species designations that partition genetic diversity at a more basic level.

Correlations between GD and proximity to other species were sometimes quite high. Is this relationship causal or coincidental? Although these other species are polyploids that would not be expected to easily exchange genes with S. verrucosum, the possibility of introgression has important implications. If, for example, the high GR of S. verrucosum accessions near S. hjertingii is due to introgression from the latter, those high GR accessions are actually not of particular genetic value, because the genes that distinguish those S. verrucosum accessions are likely common in the samples of S. hjertingii in the genebank. Proximity to S. demissum and S. hougasii were both significantly correlated with GD and GR. Whether these significant associations with proximity to other species are due to introgression or simply coincidental with the significant effects of latitude and longitude cannot be determined until the genetics of the relatively unusual S. verrucosum populations is compared with that of the associated species.

Best Choice of Geographical Parameters
The parameters presented in Table 3 are obviously related. Physical separation depends on differences in latitude and longitude. Proximity to other species is also confounded with geographic variables. For example, S. hjertingii occurs only in the extreme northeast and S. hougasii only in the extreme southwest of the S. verrucosum range (Fig. 1). Searching for the parameters or combinations of parameters that have the highest correlations with genetics not only has practical application for collecting additional diversity efficiently, but also may provide clues to the underlying causes for diversity. For example, does the pattern of diversity indicate adaptation to similar environments, dispersion from a single origin, or introgression from other species?

Genetic measurements must correspond to phenotypes and/or true relatedness by descent, and have sufficient variation. RAPD markers have been shown to verify taxonomic groupings in potato based on morphology, and otherwise coincide with known genetic relationships (del Rio and Bamberg, 2000a; Quiros et al., 1993), especially for comparisons within species (Spooner et al., 1995). The variation in the RAPD bands used here was apparently sufficient to allow significant correlations, despite the fact that the range of variation in GD among the S. verrucosum accessions was relatively small (0–29%) when compared with the other species for which no correlations were significant: S. fendleri (0–60%), S. jamesii (1–40%) (del Rio et al., 2001), and S. sucrense (8–44%) (del Rio and Bamberg, 2002).

In conclusion, we found ecogeographical associations with S. verrucosum for the first time, possibly because the inbred and uniform nature of this species improved resolution both in the experimental method and in the genetics of the natural populations themselves. Three genetic clusters could be identified that correspond to physical groups in the North, South, and Southeast locations of the range. This information could be used to enrich the capture of genetic diversity in future collections. However, genetic rarity was also associated with proximity to other potato species. If introgression is the cause of this association, one might make the mistake of pursuing collection of alleles that, while rare in S. verrucosum, are already common in the genebank in other species.


   ACKNOWLEDGMENTS
 
The authors wish to express thanks to the University of Wisconsin Peninsular Agricultural Research Station program and staff for their cooperation. We also would like to thank Robert Hijmans, Karen Williams, Stephanie Greene, Sarah Stephenson, and Adele Douglas for their critical review, constructive comments, and appropriate editing of the manuscript.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Reported use of brand name products does not imply an endorsement by USDA or the University of Wisconsin–Madison.

Received for publication April 23, 2003.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 


Related articles in Crop Science:

THIS ISSUE IN CROP SCIENCE

Crop Science 2004 44: 1109-1112. [Full Text]  




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in Crop Science
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 ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by del Rio, A. H.
Right arrow Articles by Bamberg, J. B.
Right arrow Search for Related Content
PubMed
Right arrow Articles by del Rio, A. H.
Right arrow Articles by Bamberg, J. B.
Agricola
Right arrow Articles by del Rio, A. H.
Right arrow Articles by Bamberg, J. B.
Related Collections
Right arrow Crop Genetics
Right arrow Potato
Right arrow Plant Genetic Resources

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