On the origin of artichoke and cardoon from the <Emphasis Type="Italic">Cynara</Emphasis> gene pool as revealed by rDNA sequence variation

The evolutionary history of artichoke and cultivated cardoon and their relationships to wild allies of the genus Cynara are not fully understood yet. To try resolve the evolutionary patterns leading to the domestication of these two crops, a study of molecular evolution was undertaken. The species C. cardunculus, including artichoke, cultivated and wild cardoon, together with four wild Cynara species were taken into consideration. Internal (ITS) and external (ETS) rDNA transcribed spacers were used as markers of nuclear genome, the psbA-trnH spacer as a marker of chloroplast genome. Sequences were analysed using phylogenetic analysis packages. Molecular data indicate that the whole genus is quite recent and that the domestication of artichoke and cultivated cardoon, crops diverging for reproduction system and use, are independent events which diverge in time and space. As for wild Cynara species, an evolutionary pattern consistent with their present geographical distribution was hypothesized in relation to the climatic changes occurring in the Mediterranean during the last 20 millennia: C. humilis and C. cornigera appeared to have differentiated first, C. syriaca and C. baetica were differentiated in a second period, while C. cardunculus showed to be the most recent and plastic species. The high plasticity of C. cardunculus has not only allowed its nowadays wide distribution, but has also given the potential for domestication. Dedicated to the memory of Richard Neville Lester (1937–2006) who greatly contributed to the understanding of the evolution, domestication, and genetic resources of eggplants as well as to the biosystematics and taxonomy of allied species. IGV Publication N. 78 of the Institute of Plant Genetics, CNR     

Isozyme relationships between cultivated artichoke and the wild relatives

Summary The genetic affinities between the cultivated artichoke Cynara cardunculus L. var. scolym (L.) Fiori [= C. scolymus L.] and its wild relatives were assessed by tests of 20 enzyme systems (28 loci). Six representative cultivars and the following wild taxa were examined: (i) wild cardoon C.cardunculus L. var. sylvestris (Lamk) Fiori, (ii) C. syriaca Boiss., (iii) C. cornigera (Lindley) [=C. sibthorpiana Boiss.], (iv) C. algarbiensis Cosson, (v) C. baetica (Spreng.) Pau [=C. alba Boiss.], and (vi) C. humilis L. Twenty one out of the 28 loci tested were polymorphic (mainly between species). The genetic identity between the cultivars and the wild cardoon forms ranged from 0.92 to 0.96; while that between these two taxa and the five other wild Cynara species ranged between 0.67 and 0.79. This implicates wild C. cardunculus in the wild ancestry of the cultivated artichoke.     

The wild ancestry of the cultivated artichoke

Summary The genetic affinities between the cultivated artichoke Cynara cardunculus L. var. scolymus (L.) Flori (= C. scolymus L.) and its wild relatives were tested by means of a crossing programme. The following wild taxa were involved: (i) wild cardoon C. carcundulus L. var. sylvestris (Lamk) Fiori, (ii) C. syriaca Boiss., (iii) C. cornigera Lindley, (iv) C. algarbiensis Cosson, (v) C. baetica (Spreng.) Pau (= C. alba Boiss.), and (vi) C. humilis L. Only the wild cardoon was found to be fully cross-compatible and fully infertile with the crop. In contrast, all other five wild Cynara species turned to be almost fully-or fully-cross-incompatible with the crop, and the few interspecific F₁ hybrids recovered were partly or almost fully sterile. These finds establish the wild cardoon as the wild ancestor of the cultivated vegetable.     

Genetic variation in wild and cultivated artichoke revealedby RAPD markers

Determination of intraspecific genetic diversity is an important first step for the utilization of genetic resources and canprovide useful information on crop evolution. A living collection of Italian and foreign artichoke varieties and ecotypes is maintained at the Germplasm Institute, Bari, Italy. A total of 32 accessions of cultivated (Cynara cardunculus L. var.scolymus (L.) Fiori), three of wild (Cynara cardunculus var.cardunculus L.) artichoke and two ofcultivated cardoon (Cynara cardunculus var.altilis L.) were analysed in order to studygenetic variation and relationships within the species, using RAPDmarkers. Cultivated accessions were selected according tomorphological variation and geographical distribution available inthe collection. Fifty arbitrary decamer primers were initially tested, 18 ofwhich showed from 3 to 10 unambiguously interpretable fragments. Anintra-accession analysis using 4 varieties and 8 polymorphicprimers revealed that no RAPD variation was detected amongindividuals. Jaccard's similarity index (JSI) was comprisedamong 1 and 0.693 when all accessions were considered, on the otherhand, within the cultivated artichoke, JSI ranged between 1 and0.817. A UPGMA dendrogram based on the similarity matrix showed thatwild artichokes were clearly separated from cultivated accessions.Moreover, within cultivated artichokes, several groups could bedistinguished.     

Cross-species Transferability of Rice Microsatellites in its Wild Relatives and the Potential for Conservation Genetic Studies

Here, we investigated the transferability of 60 microsatellite markers characterized for cultivated rice Oryza sativa L. in three wild Oryza species representing different genome types: O. rufipogon Griff. (AA), O. officinalis Wall. et Watt. (CC), and O. granulate Nees et Arn. ex Watt. (G). The results indicate the 60 rice SSR loci tested produced homologous amplification products to different extents in O. rufipogon (100%), O. officinalis (90%) and O. granulata (73.3%). Proportions of polymorphism for successfully amplified loci ranged from 0.983 via 0.667 to 0.364 in O. rufipogon, O. officinalis and O. granulata, respectively. The utility of these microsatellite markers was tested for the characterization of genetic diversity in 117 genotypes of these four Oryza species. The values of genetic diversity in cultivated rice are higher than the other two wild species O. officinalis and O. granulata, suggesting microsatellites tend to have more variability in the focal species than in non-focal species to which they are applied. However, much lower levels of genetic variation were observed in rice than in its wild progenitor O. rufipogon, which indicates severe loss of genetic variation may reflect the ‘domestication bottleneck’ through which rice passed. The observation that most of the rice microsatellites are able to detect allelic polymorphisms at different extent in Oryza species suggest that rice microsatellite loci should be useful for the analysis of genetic diversity and inter- and intra-specific relationships in the genus. Therefore, high rates of successful cross-amplification of rice microsatellites among Oryza species with different genome types will offer excellent opportunities to investigate the population genetic structure of wild rice species and explore their conservation genetics.     

Genetic diversity of European pear cultivars (<Emphasis Type="Italic">Pyrus communis</Emphasis> L.) and wild pear (<Emphasis Type="Italic">Pyrus pyraster</Emphasis> (L.) Burgsd.) inferred from microsatellite markers analysis

The aim of this study was to identify the group of highly polymorphic microsatellite markers for the identification of six pear cultivars (P. communis) and two individuals of wild pear (P. pyraster). From among 40 tested SSR markers, 19 were selected to profile genetic diversity in pear genotypes due to high polymorphisms. These markers showed high heterozygosity levels (0.5–1) and, on average, 6.4 alleles per marker were found. The set of microsatellite markers employed in this study demonstrated usefulness of microsatellite markers for the identification of pear genotypes. The examined wild forms were represented in this study by only two individuals of P. pyraster. It can be assumed that these forms were distinctly different from the cultivated pear cultivars.     

Genetic characterisation of olive trees from Madeira Archipelago using flow cytometry and microsatellite markers

In this study, native olive plants from Olea maderensis (≡ O. europaea ssp. cerasiformis) and O. cerasiformis (≡ O. europaea ssp. guanchica), wild olives (O. europaea ssp. europaea var. sylvestris) and cultivated olives (O. europaea ssp. europaea var. europaea) were analysed with respect to genome size and microsatellite markers. The mean nuclear DNA content of O. maderensis was estimated as 5.97 ± 0.191 pg/2C, while the remaining studied taxa presented mean genome sizes ranging from 2.99 to 3.18 pg/2C. These data and the obtained simple sequence repeats (SSR) profiles, i.e., with 2–4 alleles in O. maderensis and a maximum of two alleles in the other taxa, enabled the identification of a new ploidy level, tetraploidy, for a species belonging to the Olea genus. Cluster analysis of the microsatellite data revealed a clear separation of each species in different clusters and a high genetic dissimilarity could be observed among genotypes belonging to different species. This work contributed to a better characterization of olive species and the obtained data can be helpful to support taxonomic studies, and to develop germplasm preservation strategies in endangered populations of O. maderensis from Madeira Archipelago.     

Using SSR markers to determine the population genetic structure of wild apricot (<Emphasis Type="Italic">Prunus armeniaca</Emphasis> L.) in the Ily Valley of West China

Genetic structure of three wild populations (Xinyuan, Gongliu and Daxigou) of apricot in the Ily Valley, Xinjiang Uygur Autonomous Region of China, was investigated with microsatellite (simple sequence repeat, SSR) markers. The higher polymorphism and greater transportability of these markers between Prunus species proved SSR markers were much efficient for conducting genetic diversity studies in wild apricot. Nei's gene diversity (He) and Shannon's index of diversity (I) were 0.287 and 0.458, respectively. This indicated that the wild apricot in the Ily Valley still maintained a relatively high level of diversity. The Gst of 0.137 and Fst of 0.164 revealed that genetic variation mainly resided among individuals within populations (83.6–86.3%). Population differentiation could also be found according to the distribution of SSR alleles between the populations. Mantel test showed the genetic distance between populations was significantly correlated to the geographical distance. The modest amount of gene flow (2.684) would reduce the disjunction between wild apricots. The long-distance dispersal of pollen by insects was probably the main way of gene flow between populations. Based on the study of population genetic structure, an effective conservation strategy of the species was discussed.     

Genetic variability in wild populations of <Emphasis Type="Italic">Prunus divaricata</Emphasis> Ledeb. in northern Iran evaluated by EST-SSR and genomic SSR marker analysis

A population genetic analysis based on eight genomic SSR markers and three EST-SSR (expressed sequence tags) markers developed in peach (Prunus persica (L.) Batsch) and Japanese plum (Prunus salicina Lindl.) was carried out in 12 wild populations of cherry plum (Prunus divaricata Ledeb.) sampled along the Iranian coast of the Caspian Sea. A total of 184 alleles (3–31 per locus) were detected with a mean value of 16.7 alleles per locus. None of the loci or populations showed deviation from Hardy–Weinberg equilibrium, and all markers proved to be unlinked. The mean values for the observed and the expected heterozygosity were 0.66 and 0.73, respectively. There was very little genetic differentiation among populations, as was indicated by low overall values of Wright’s F_ST (0.03) and Nei’s G_ST (0.08). An analysis of molecular variance (AMOVA) showed that 96.8% of the total variance was attributable to differences between individuals within populations. Genetic and geographic distances were nevertheless positively correlated, as evidenced by a Mantel test. The high level of genetic diversity and the apparent lack of genetic structure in wild P. divaricata may be attributed to frequent long distance gene flow through frugivorous birds and possibly humans, as has been documented for other Prunus species.     

Wild artichokes of south Italy: did the story begin here?

A collection of wild artichokes (Cynara cardunculus L. var. sylvestris (Lam.) Fiori) was conducted in south Italy. The distribution of the wild populations was not even all over the explored territory, from the coastline to the inner mountains, and in some areas wild artichokes were completely absent. A certain level of morphological variation for height and habit of plants, size and shape of capitula, length, shape, and position of bracteal spines, was observed. It has been noticed that picking capitula from the wild for human consumption is a common habit in southern Italy and Sicily. Historical data and etnobotanic inference suggest the idea that the artichoke could have been domesticated in south Italy, and Sicily in particular, and was spread by the Arabs to other regions of the Mediterranean Basin.     

Analysis of population substructure,genetic differentiation and phylogenetic relationships among selected Asiatic <Emphasis Type="Italic">Vigna</Emphasis> Savi species

Random amplified polymorphic DNA markers were used to study sub-structure and genetic differentiation amongst 31 populations (seven cultivated and 24 wild populations) belonging to 14 Asiatic Vigna species. Ten pre-selected RAPD primers generated 152 polymorphic amplification products. Estimates of polymorphism indices were higher for the wild taxa in comparison to the cultivated forms. F_ST values between populations ranged from 0.111 to 0.801 and Nei’s genetic diversity values between and within species varied from 0.26 to 0.70 and 0.04 to 0.56 respectively. The high F_ST and F_CT values indicated strong subdivision of populations and high differentiation among species. Analysis of molecular variance was performed by grouping the populations conforming to specific species. AMOVA was also performed separately to better resolve the differentiation of species within mungo–radiata complex. Molecular phylogenetic relationships amongst the species of radiata–mungo complex; namely, black gram (V. mungo (L.) Hepper), green gram (V. radiata (L.) Wilczek), V. radiata var. sublobata, V. radiata var. setulosa, V. mungo var. silvestris and V. hainiana, were studied through cluster analyses. Two distinct groups were recognized within the complex, with population samples of V. hainiana forming one cluster. Further, V. hainiana appeared to be equidistant to both V. radiata and V. mungo.     

Assessment of genetic diversity in cultivars and wild accessions of Luohanguo (<Emphasis Type="Italic">Siraitia grosvenorii</Emphasis> [Swingle] A. M. Lu et Z. Y. Zhang), a species with edible and medicinal sweet fruits endemic to southern China,using RAPD and AFLP markers

Genetic variation of wild populations and cultivars of Luohanguo (Siraitia grosvenorii), a plant species endemic to southern China, was assessed using random amplified polymorphic DNA (RAPD) and amplified fragment length polymorphism (AFLP) markers. Based on the results for 130 individuals from seven populations, a high level of genetic diversity of Luohanguo was observed at the species level. The percentage of polymorphic loci (P) was 89.4%, Nei’s gene diversity (H _e) was 0.239, and Shannon’s information index (H _o) was 0.373 based on the combined AFLP and RAPD data. There was a high degree of genetic differentiation, with 45.1% of the genetic variation attributed to differences between the populations. The genetic diversity of the Luohanguo cultivars is much lower than that of wild populations (P = 41.8%, H _e = 0.141, H _o = 0.211), and a distinct genetic differentiation is observed between the cultivars and wild accessions. The pool of genetic variation in the wild populations provides an excellent gene resource for Luohanguo breeding.     

Inter Simple Sequence Repeat (ISSR) Analysis of Diploid Coffee Species and Cultivated Coffea arabica L. from Tanzania

Inter simple sequence repeat (ISSR) markers were used to evaluate levels of genetic similarity among Coffea arabica L. accessions from Tanzania and to estimate levels of genetic similarities in C. arabica and diploid coffee species. The six ISSR primers used generated a total of 82 fragments and the dissimilarity values ranged from 0.21 to 1. Mean dissimilarity values between provenances (0.56–0.85) were higher than within provenances (0.37–0.68). Cluster analysis based on Nei’s genetic distances showed C. arabica provenances grouping based on geographical origin. Two major clusters were formed that constituted of provenances from Kilimanjaro and Arusha in one sub-cluster; Tanga and Morogoro in the other; the second cluster had Mbeya provenances and diploid species, respectively. The implication is that Mbeya provenances are different from the rest of Tanzanian C. arabica. A principal coordinate analysis (PCA), whose first three coordinates explained 43% of the variation, showed similar groupings as in the cluster analysis. A separate cluster analysis of diploid species showed a distinct separation of the three species used. ISSR data gave results similar to previous findings from random amplified polymorphic DNA(RAPD) analysis. The results also confirm the limited diversity present in cultivated C. arabica in Tanzania     

Molecular diversity of Omani wheat revealed by microsatellites: I. Tetraploid landraces

Results of archaeological studies indicate a millennia-old cultivation history for wheat (Triticum spp.) in Oman. However, in spite of numerous collection surveys and efforts for phenotypic characterization of Omani wheat landraces, no attempts have been made using molecular tools to characterize this germplasm. To fill this gap, 29 microsatellite markers revealing 30 loci were used to study the genetic diversity of 38 tetraploid wheat landrace accessions comprising the species T. dicoccon, T. durum and T. aethiopicum. A total of 219 alleles were detected whereby the number of alleles per locus ranged from 2 to 16 with an average number of 7.1 alleles per locus. The highest number of alleles occurred in the B genome with on average 7.9 alleles per locus as compared to the A genome with 6.5 alleles per locus. Heterogeneity was detected for all microsatellites except for GWM 312, GWM 601 and GWM 192B with an average heterogeneity over all primers and lines of 14.4%. Approximately 10% of the accessions contained rare alleles with an average allele frequency <4%. Gene diversity across microsatellite loci ranged from 0.26 to 0.85. The pairwise comparison of genetic similarity ranged from 0.03 to 0.91 with an average of 0.2. Cluster analysis revealed a clear separation of the two species groups T. dicoccon versus T. durum and T. aethiopicum. Within the species clusters regional patterns of subclustering were observed. Overall, this study confirmed the existence of a surprisingly high amount of genetic diversity in Omani wheat landraces as already concluded from previous morphological analyses and showed that SSR markers can be used for landraces’ analysis and a more detailed diversity evaluation.     

Development and use of microsatellite markers for genetic diversity analysis of cañahua (<Emphasis Type="Italic">Chenopodium pallidicaule</Emphasis> Aellen)

Cañahua (Chenopodium pallidicaule Aellen) is a poorly studied, annual subsistence crop of the high Andes of South America. Its nutritional value (high in protein and mineral content) and ability to thrive in harsh climates make it an important regional food crop throughout the Andean region. The objectives of this study were to develop genetic markers and to quantify genetic diversity within cañahua. A set of 43 wild and cultivated cañahua genotypes and two related species (Chenopodium quinoa Willd. and Chenopodium petiolare Kunth) were evaluated for polymorphism using 192 microsatellite markers derived from random genomic cañahua sequences produced by 454 pyrosequencing of cañahua genomic DNA. Another 424 microsatellite markers from C. quinoa were also evaluated for cross-species amplification and polymorphism in cañahua. A total of 34 polymorphic microsatellite marker loci were identified which detected a total of 154 alleles with an average of 4.5 alleles per marker locus and an average heterozygosity value of 0.49. A cluster analysis, based on Nei genetic distance, clearly separated from wild cañahua genotypes from the cultivated genotypes. Within the cultivated genotypes, subclades were partitioned by AMOVA analysis into six model-based clusters, including a subclade consisting sole of erect morphotypes. The isolation by distance test displayed no significant correlation between geographic collection origin and genotypic data, suggesting that cañahua populations have moved extensively, presumably via ancient food exchange strategies among native peoples of the Andean region. The molecular markers reported here are a significant resource for ongoing efforts to characterize the extensive Bolivian and Peruvian cañahua germplasm banks, including the development of core germplasm collections needed to support emerging breeding programs.     

Genetic relationships among Arachis hypogaea L. (AABB) and diploid Arachis species with AA and BB genomes

The cultivated peanut (Arachis hypogaea L.) is an allotetraploid, with two types of genomes, classified as AA and BB, according to cytogenetic characters. Similar genomes to those of A. hypogaea are found in the wild diploid species of section Arachis, which is one of the nine Arachis sections. The wild species have resistances to pests and diseases that affect the cultivated peanut and are a potential source of genes to increase the resistance levels in peanut. The aim of this study was to analyze the genetic variability within AA and BB genome species and to evaluate how they are related to each other and to A. hypogaea, using RAPD markers. Eighty-seven polymorphic bands amplified by ten 10-mer primers were analyzed. The species were divided into two major groups, and the AA and the BB genome species were, in general, separated from each other. The results showed that high variation is available within species that have genomes similar to the AA and the BB genomes of A. hypogaea.     

Close Genetic Proximity Between Cultivated and Wild Bactris gasipaes Kunth Revealed by Microsatellite Markers in Western Ecuador

Bactris gasipaes Kunth (peach palm or Pejibaye) is the only domesticated palm of the Neotropics. The genetic relationships between the crop and its wild relatives are still unclear. We undertook field and laboratory work in order to describe differentiation and relationships between the wild and cultivated populations of the species in Western Ecuador, and their possible interactions. A volumetric study was undertaken on the fruits of both populations, as well as a population genetic analysis in order to clarify these relationships. Fruits from cultivated plants collected in the region of sympatry of wild and cultivated plants in North-West Ecuador showed intermediate volumes between those of reference samples for the wild and cultivated plants in allopatry. Using 8 microsatellite loci, we assessed 83 wild and cultivated individuals from Western Ecuador and cultivated plants from Amazonia and Central America as a reference for the cultivated form. We detected high polymorphism in the wild and cultivated samples and low, but significant level of genetic differentiation between wild and cultivated populations. The cultivated population in North-Western Ecuador showed close genetic proximity with the sympatric wild population, consistent with the volumetric study. These results have implications for hypotheses on evolution of this crop and for strategies of genetic conservation of the wild forms.     

Processes affecting genetic structure and conservation: a case study of wild and cultivated <Emphasis Type="Italic">Brassica rapa</Emphasis>

When planning optimal conservation strategies for wild and cultivated types of a plant species, a number of influencing biological and environmental factors should be considered from the outset. In the present study Brassica rapa was used to illustrate this: to develop Scandinavian conservation strategies for wild and cultivated B. rapa, DNA-marker analysis was performed on 15 cultivated and 17 wild accessions of B. rapa plus 8 accessions of the cross compatible B. napus. The B. rapa cultivars were bred in Sweden and Finland in 1944–1997 and the wild B. rapa material was collected from Denmark, Sweden and United Kingdom. The B. napus accessions were bred within the last 20 years in the Scandinavian countries. Results were based on scoring of 131 polymorphic ISSR markers in the total plant material. A Bayesian Markov chain Monte Carlo (MCMC) approach implemented in NewHybrids demonstrated a clear distinction of B. rapa and B. napus individuals except for three individuals that seemed to be backcrosses. The backcrossed hybrids descended from two Swedish populations, one wild and one escaped. The overall pattern of genetic variation and structure in B. rapa showed that cultivated and wild B. rapa accessions formed two almost separated clusters. Geographical origin and breeding history of cultivars were reflected in these genetic relationships. In addition, wild populations from Denmark and Sweden seemed to be closely related, except for a Swedish population, which seemingly was an escaped cultivar. The study point to that many processes, e.g. spontaneous introgression, naturalisation, breeding and agricultural practise affected the genetic structure of wild and cultivated B. rapa populations.     

PCR Marker-based Analysis of Wild and Cultivated Yams (<Emphasis Type="Italic">Dioscorea</Emphasis> spp.) in Nigeria: Genetic Relationships and Implications for <Emphasis Type="Italic">ex situ</Emphasis> Conservation

Reliable characterization of the variation among wild and cultivated yams in Nigeria is essential for improved management and efficient utilization of yam genetic resources. RAPD and double stringency PCR (DS-PCR) analyses were used to investigate genetic relationships and the extent of redundancy among 30 accessions of two cultivated, and 35 accessions of four wild yam species collected from Nigeria. Twenty-five selected random decamer and two microsatellite primers were used individually and in combination to generate DNA profiles for each accession of the six Dioscorea species. The number of amplified fragments varied from 7 to 18 fragments per primer/primer combination. Different levels of intraspecific genetic diversity were found, with Dioscorea rotundata Poir. being the most variable. Based on identical profiles for the RAPD and DS-PCR primers, 12 duplication groups consisting of a total number of 37 accessions were observed in the present study. An UPGMA analysis grouped the majority of plants according to the species. Cultivated yams belonging to the D. cayenensis–rotundata species complex, which were classified into seven morphotypes/varietal groups, could be clearly separated into two major groups corresponding to D. rotundata Poir. and D. cayenensis Lam. D. cayenensis cultivars exhibited a low level of intraspecific variation and were genetically close to the wild species Dioscorea burkilliana J. Miège. D. rotundata cultivars classified into six varietal groups showed a high degree of DNA polymorphism and were separated into two major groups that appeared most closely related to Dioscorea praehensilis Benth. and Dioscorea liebrechtsiana de Wild. We propose, based on these results, that cultivars classified into D. cayenensis should be considered as a taxon separate from D. rotundata. The implications of intraspecific variability for the ex situ conservation of wild and cultivated yam germplasm in Nigeria are discussed.     

Capturing genetic diversity of wild populations for <Emphasis Type="Italic">ex situ</Emphasis> conservation: Texas wild rice (<Emphasis Type="Italic">Zizania texana</Emphasis>) as a model

Genebanks complement other conservation programs because they preserve genetic diversity needed for future breeding and restoration. We evaluated efficiency of capturing genetic diversity, using endangered Zizania texana (Texas wild rice) as a model for plants with recalcitrant seeds. This perennial aquatic grass is restricted to 4 km of the San Marcos River in Texas. An early conservation collection included plants from stands throughout the river, based on the assumption stands would be unique genotypes. Using microsatellite markers, we found that genetic diversity was concentrated in five of 15 large, demographically stable stands; 96 stands smaller than 2 m² contributed no unique alleles. High heterozygosity and few duplicate genotypes suggested that sexual reproduction occurs more often than presumed. Simulations of stratified sampling of large stands captured all alleles in only 45 individuals, while random sampling along the river captured much less diversity. The early conservation collection captured as much diversity as expected from random sampling. Texas wild rice stands resemble a mainland-island metapopulation; our analyses suggest that stratified sampling maximizes genetic diversity for this population dynamic. Demographic and genetic information is important for validating the design of efficient ex situ collections and guiding in situ conservation.