| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | SEARCH RESULT |
|
|
|
|||||||
|
|||||||||
|
(American Journal of Botany. 2009;96:498-506.) doi: 10.3732/ajb.0800054 © 2009 Botanical Society of America, Inc. |
What's this? |
Population Biology |
2 Department of Evolution, Ecology and Organismal Biology, Ohio State University, Columbus, Ohio 43210 USA
Received for publication 12 February 2008. Accepted for publication 13 October 2008.
ABSTRACT
Cultivated plants that cannot survive on their own often have maladaptive domestication traits. Unharvested crop seeds may generate feral populations, at times causing serious weed problems, but little is known about the evolution of ferality. We explored the potential for cultivated radish, Raphanus sativus, to become feral, given that closely related taxa (e.g., R. raphanistrum and crop–wild hybrids) are well-documented weeds. First, we measured the population growth of five experimental, cultivated, self-seeding radish populations in Michigan, USA, for three generations. Three late-flowering populations went extinct, and two others apparently hybridized with local R. raphanistrum. A common garden experiment showed that the two surviving populations had earlier flowering, smaller root diameters, and greater individual fecundity than did nonhybridized populations. We also used artificial selection to measure the evolutionary potential for earlier flowering. After two generations of strong selection, two of three lineages flowered earlier and produced more seeds than control lineages, but insufficient genetic variation prevented dramatic evolution of crop phenotypes. In summary, it seems unlikely that radishes could spontaneously become feral in our study area without gene flow from R. raphanistrum. Applying these approaches to other cultivated species may provide a better understanding of mechanisms promoting the evolution of feral weeds.
Key Words: Brassicaceae • fecundity • ferality • Michigan • naturalized • population dynamics • Raphanus sativus • selection experiments • United States • volunteers
Crop ferality and volunteerism are well-known phenomena in agriculture (Ogg and Parker, 1989
; Gressel, 2005). Volunteer populations occur when unharvested seeds from a previous crop germinate and grow in and around agricultural fields. For example, canola harvesting can drop more than a thousand seeds per acre, and volunteers often compete with subsequent crops (Gulden and Shirtliffe, 2003). By colonizing field margins, home gardens, ditches, disturbed areas, and other unmanaged habitats, feral plants can reproduce independently of the agricultural management upon which they depended as crop plants (e.g., Heenan et al., 2004; Lumaret et al., 2004; Ulloa et al., 2006). Here, we explore evolutionary pathways that might lead to the evolution of feral radish (Raphanus sativus L.).
Weedy species are known to evolve rapidly in response to local selection pressures, and the recent development of genetically engineered crops, especially herbicide-resistant cultivars, highlights the possibility that volunteer and feral plants may create new environmental and crop–weed management risks in agricultural systems (e.g., Smyth et al., 2002). Many cultivated plants have naturalized in introduced areas (e.g., Mack and Erneberg, 2002; Heenan et al., 2004; Tabuti et al., 2004), sometimes becoming invasive (e.g., ornamental plants: Kowarik, 2005; weedy rye: Stump and Westra, 2000). Infestations of feral rye (Secale cereale L.) in wheat fields in the western United States can reduce yield by 30–70% and result in major economic losses annually (Pester et al., 2000). By identifying crops that can produce feral populations, we may anticipate and perhaps avoid this ecological and economic damage.
The extent to which cultivated plants can persist as feral populations is related to heritable variation in domestication traits, such as lack of seed dispersal or seed dormancy, which may limit the plants’ ability to persist as free-living populations. For example, canola, cranberry, olive, and many horticultural taxa are relatively undomesticated, and volunteer individuals can easily give rise to feral populations (Pessel et al., 2001; Bervillé et al., 2005; Kowarik, 2005). In rice (Oryza sativa L.), few mutations are needed for the evolution of weedy rice from crop plants (e.g., Li et al., 2006). Other crops such as maize (Zea mays L. subsp. mays) and soybean [Glycine max (L.) Merr.] are so strongly domesticated that their progeny rarely persist beyond first-generation volunteers (Gressel, 2005).
In addition to spontaneous mutations and purging of domestication traits by natural selection, gene flow from wild relatives can provide a stimulus for the evolution of feral populations. Gressel (2005) described two mechanisms by which cultivated plants can become feral—endoferality and exoferality. These terms refer to dedomestication without gene flow from other taxa vs. dedomestication that is facilitated by gene flow from other taxa. Both mechanisms could contribute to the evolution of ferality, and they may be difficult to distinguish, especially when breeding practices or spontaneous gene flow from wild relatives allow wild genes to enter the crop gene pool. In sunflower (Helianthus annuus L.), for example, off-type crop plants with weedy traits appear to be the result of pollen contamination from wild sunflower in commercial seed nurseries, and seeds from off-type sunflowers are more likely to give rise to feral populations than those from normal volunteer plants (Reagon and Snow, 2006). In general, the persistence of feral populations and the loss of deleterious domestication traits can depend on heritable variation in these traits, the strength of selection, and rates of gene flow from other taxa such as weedy relatives (Mekuria et al., 2002; Bond et al., 2004).
Several observational studies document the persistence of feral populations in Brassica species (e.g., Crawley and Brown, 1995; Pessel et al., 2001; Heenan et al., 2004). More recently, workers have adopted an experimental approach to understanding the potential for crop ferality (Lutman et al., 2003; Harker et al., 2005), but little is known about the potential for ferality to evolve from volunteer populations. Here, we asked whether a widely planted radish cultivar (‘Red Silk’) can persist as feral populations and evolve more wild-type characteristics. First, we measured population growth rates of five self-seeding, experimental populations over three years in Michigan, USA. Using a common garden experiment, we tested for heritable differences in several phenotypic traits among the surviving populations and compared them to a nonselected control lineage (Fig. 1). Second, we used an artificial selection experiment to measure how much heritable variation the crop cultivar possesses for a key life-history trait, early flowering. We also tested for correlated effects of earlier flowering by examining the lifetime fecundity and other traits of selected and control lineages in a common garden (Fig. 1). We suggest that this combination of natural and artificial selection experiments provides useful insights into the mechanisms involved in the spontaneous evolution of feral weeds.
|
Study system
Radish (Raphanus sativus L., Brassicaceae) is a variable and ancient crop that is cultivated around the world for its edible root, seed sprouts, seed pods, leaves, and as forage for livestock (Crisp, 1995; Snow and Campbell, 2005). Many cultivated varieties of R. sativus (i.e., European and daikon) are late-flowering annuals or biennials that produce an enlarged, edible hypocotyl/root. In contrast, its wild relative, R. raphanistrum, produces a thin tap root before flowering early in the growing season as an annual weed. The relatively early flowering of R. raphanistrum may ensure that plants reproduce before frost, drought, or herbivory, and before crop fields are tilled, sprayed with herbicides, or harvested. Raphanus raphanistrum (wild radish or jointed charlock) is among the world’s worst agricultural weeds, especially in small grain and vegetable crops, and is distributed worldwide (Cheam and Code, 1995; Holm et al., 1997; Warwick and Francis, 2005).
Hybrids between R. sativus and R. raphanistrum are highly successful weeds in some environments, including coastal dunes, roadsides, and agricultural areas in California (Panetsos and Baker, 1967; Klinger and Ellstrand, 1994; Hegde et al., 2006). European cultivated radishes reportedly became naturalized along the California coast (Panetsos and Baker, 1967), but it is not clear whether this process required hybridization with weedy R. raphanistrum or whether endoferality had evolved directly from cultivated radishes. According to Hegde et al. (2006), the "pure" populations of feral R. sativus originally reported no longer exist in California. Ridley et al. (2008) found that hybridization between cultivated radish and R. raphanistrum has been bidirectional, with both taxa acting as pollen recipients during initial hybridizaton events.
In northern lower Michigan, where the current study was conducted, R. raphanistrum occurs as an annual weed in alfalfa, oats, potato, sugar beet, and other agricultural fields, often leaving a persistent seed bank (L. G. Campbell and A. A. Snow, personal observation). Cultivated radish is grown in home gardens and commercial fields, with the potential to hybridize with the weed when the crop is grown for seed or when fields are neglected and the rosettes bolt to produce flowers. Indirect evidence for crop-to-wild gene flow comes from the presence of crop-specific flower colors (flowers with white or pink-hued petals), as reported for a Bay City, Michigan, population of R. raphanistrum by Kercher and Conner (1996). In this population, 55% of the plants had crop-specific flower colors, and we have often observed lower frequencies of putative hybrids in other populations (A. A. Snow, personal observation). In a companion study to the one described here, we followed the population dynamics and evolution of four wild R. raphanistrum and four hybrid R. raphanistrum x R. sativus populations in Michigan, USA (described in Campbell et al., 2006). We found that all hybrid populations persisted for at least three years and had similar intrinsic growth rates to wild R. raphanistrum. Therefore, it is clear that feral radish can become dedomesticated via exoferality, as postulated by Panetsos and Baker (1967) and Hegde et al. (2006). To date, however, we are not aware of studies in which the potential for endoferality has been examined in radish.
Natural selection experiment
Seed source
In 2001, we cross-pollinated 100 crop plants in a greenhouse, using the self-incompatible Raphanus sativus cultivar Red Silk (Harris-Moran Seed Co., Modesto, California, USA), to create first-generation (G1) plants. This "European" cultivar is widely planted in the USA and is similar to other widely consumed cultivars such as "Cherry Belle" and "French Breakfast". Plants were grown under controlled conditions, including a 16-h daylight schedule with a 23–28°C/20–25°C day/night temperature range. A long-day photoperiod encourages radish flowering (Erwin et al., 2002), and these conditions were maintained during the artificial selection experiment described later. Maternal parents were randomly assigned pollen donors from concurrently flowering plants.
Population establishment
In 2002, we established five first-generation radish populations (C1–C5), one in each of five meadows in Emmett and Cheboygan counties, Michigan (MI) (Fig. 1). The experimental populations were separated from local populations of R. raphanistrum by >1 km to restrict unintended gene flow via pollinators or seed dispersal. Flower color provided a useful indicator of whether wild R. raphanistrum genes had inadvertently entered the experimental populations. Raphanus sativus is homozygous for the dominant, white petal allele, whereas R. raphanistrum is homozygous for the recessive, yellow carotenoid pigment (Panetsos and Baker, 1967; Kay, 1976; Campbell et al., 2006). Therefore, the wild-type flower color could be detected in homozygotes but not in heterozygotes at this locus. Inheritance of pinkish petal hues that blend with white or yellow colors is more variable and complex (Irwin et al., 2003). In the context of this study, pink-flowered plants were grouped with white-flowered ones, and bronze-flowered plants were grouped with yellows (as in Snow et al., 2001).
We planted G1 seedlings in peat-filled biodegradable pots at the University of Michigan Biological Station (UMBS) in Pellston, MI. At each site, we transplanted 60 seedlings to recently tilled 15 x 15 m fertilized plots (as in Campbell et al., 2006). Between 50–60 plants survived the growing season in each population although not all surviving plants flowered. No resident wild radish plants emerged from the seed bank at these plots in 2002 or 2003. Each spring, plots were tilled, fertilized, and hand-weeded for approximately two weeks to simulate agricultural management and promote population persistence. Otherwise, these self-seeding populations were exposed to naturally occurring weather conditions, competing plants, herbivores, pathogens, and pollinators (Lee and Snow, 1998; Campbell et al., 2006). By 2003, two populations (C2, C3) had declined in size to <10 seedlings. In a futile attempt to promote population persistence, we added 25 G1 seedlings from the original seed source to each population (C1–C5) in 2003. Therefore, it is possible that more than one generation persisted in each population (this could also occur due to seed dormancy).
Population surveys
Population size was estimated from direct counts when <1000 plants were present or subsampling larger populations. For the latter, we determined the average number of plants in 49 1-m2 quadrats per site and multiplied this value by the total area.
We did not expect to find any yellow-flowered plants in the experimental populations because the crop was homozygous for white flower color. However, two of the five populations (C4 and C5) inadvertently acquired plants with yellow flowers and other crop–wild hybrid traits from R. raphanistrum, as described later. When we first observed yellow-flowered plants in 2004, we began assessing the frequency of yellow-flowered plants in all populations during peak flowering (25 June–4 July). Yellow-flowered plant frequencies were based on direct counts or subsamples as described earlier, and allele frequencies for this recessive trait were estimated using Hardy–Weinberg expectations. Because the yellow petal allele is recessive, we could not determine the original level or timing of this putative introgression, and we cannot rule out the possibility that C1–C3 also experienced gene flow from R. raphanistrum. Based on flower color frequencies and the presence of hybrid phenotypic traits in the common garden experiment (discussed next), we refer to C4 and C5 as "introgressed" populations and C1–C3 as "non-introgressed" populations, with the caveat that the evidence for this distinction is tentative.
Common garden
To compare fitness-related traits of the populations that were still present in 2004 (C1, C4, C5) and to measure how much each population evolved over the course of three growing seasons, we used a common garden experiment that included unselected control lineages from the artificial selection experiment described later (Fig. 1). In 2004, we collected one seed per fruit of three fruits from haphazardly selected branches on 30 plants per experimental population. The garden included five G4 individuals from each field population and control lineage in each of 10 blocks, resulting in 150 plants derived from field populations and 150 control plants in a complete block design randomized across populations and lineages.
Seeds were planted in 300 mL of peat in fiber pots (3–10 May 2005) with four oat seeds (Avena sativa) in a greenhouse at UMBS. Cultivated oats were added as a standard competitor, and they were thinned to one oat plant per pot. Once the radish seedlings developed their first true leaves, each fiber pot was moved into a 4-L plastic pot (Hummert International, Earth City, Missouri, USA) filled with a mixture of 3.7 L of local sandy soil surrounding the fiber pot. The plastic pots were placed on level, recently tilled soil and were separated by 30 cm to minimize competition between experimental plants (roots could grow through the pot bottom). Seedlings that died within one week after transplanting were replaced, and all plants survived to reproduce. Plants were watered daily for the first month and every other day until 31 August. Insect herbivory was minimal due to two early applications of an insecticide (0.0033% esfenvalerate, Scotts Miracle-Gro Co., Marysville, Ohio, USA). Aphids were present at low densities later in the season but did not colonize any plant heavily. Pollinators were abundant throughout the experiment, as in Lee and Snow (1998). Plants were individually harvested at senescence until the first hard frost when all remaining plants were harvested (16–20 September).
For each plant, we recorded age at flowering, pollen fertility, stem diameter, flower number, seed production, and aboveground vegetative biomass. Age at flowering was calculated as the difference between the dates of emergence and flowering. Pollen fertility was quantified because hybrids were present in two of the populations, and hybrids between Raphanus raphanistrum and R. sativus are known to be heterozygous for a reciprocal translocation that affects chromosome pairing during meiosis (Panetsos and Baker, 1967). Typically, F1 crop–wild hybrids produce approximately 40–50% aborted pollen grains, whereas wild R. raphanistrum produce less than 20% aborted pollen grains (e.g., Snow et al., 2001; Campbell et al., 2006). We collected pollen from two flowers per plant and assessed pollen fertility using a compound microscope and Alexander’s stain (Alexander, 1969) to count the proportion of aborted grains in stained samples of at least 100 grains (Campbell et al., 2006). As a correlate of plant size, we measured stem diameter at the widest point at the base of each plant at harvest. We also counted the total number of flower pedicels and fruits per plant. Fruit set was calculated as number of fruits produced divided by the total number of flowers. Number of seeds per plant was estimated by multiplying the number of fruits by the average number of seeds per fruit, estimated from 10 randomly chosen fruits per plant. Finally, we weighed the vegetative biomass (g) of the competing oat plants.
Percentage fruit set required no transformations prior to analysis. Number of seeds, flowers per plant, and oat biomass were log-transformed, and pollen fertility was arcsine-square-root transformed. All analyses were performed using SPSS (v.13, SPSS, Chicago, Illinois, USA). To test for differences in lifetime fecundity and relative competitiveness among the experimental field populations and the control lineages, we ran a mixed model ANOVA for each trait. Population was a fixed effect and block was a random effect. Variance of the random effect was estimated using restricted maximum likelihood.
Artificial selection experiment
Seed source and selection
The G2 generation was the base generation for the selection experiment and was produced from 100 G1 crop x G1 crop crosses in the same room and under similar conditions to those of the parental generation (in a greenhouse at OSU, Fig. 1). We randomly assigned G2 plants to one of two selection treatments (earlier flowering vs. the unselected control) and one of three replicate lineages per selection treatment, for a total of six lineages. The three replicate lineages allowed us to exclude drift as the evolutionary cause of the changes in flowering phenology. Each lineage in the G2 generation included 140 plants.
Plants were grown in Cone-tainers (Steuwe and Sons, Corvallis, Oregon, USA), filled with standard potting soil (PRO-MIX BX peat, Premier Horticulture Ltd., Rivière-du-Loup, Quebec, Canada). The replicate lineages served as blocks in the experiment and pairs of selection treatments of each replicate occupied two adjacent greenhouse benches. We randomly repositioned plants within the greenhouse every two weeks to reduce the effects of environmental variation in the greenhouse.
Each generation, we recorded the dates of germination and flowering for each plant. To produce the following generation, we imposed truncation selection for early flowering plants in three lineages by selecting 10% of each lineage that flowered earliest. For the three control lineages, we randomly selected 10% of the plants, regardless of flowering times. Selected plants were cross-pollinated in a complete diallel. For the second and third generations, every lineage was initiated with at least 225 seeds. Two weeks later, we reduced the size of every lineage to the size of the smallest lineage to maintain equally sized lineages. Each G2, G3 and G4 lineage included 140, 130, and 50 plants produced by 100 G1, 16 G2, and 13 G3 parental plants.
Common garden
G4 plants from the early and control lineages were grown in a common garden to assess differences in flowering time, fecundity, and other phenotypic traits. Control lineages were expected to represent variation in the evolutionary trajectory of randomly mated populations that have not experienced selection. If these control lineages had adapted to greenhouse conditions, we assumed this had minor effects, if any, on measured traits because germination rates were high and plants were mated randomly in each generation.
Plants from the artificial selection experiment were planted in the same common garden described earlier, using identical methods, and were interspersed with plants from the field populations. Each of the 10 blocks included five individuals per artificial selection lineage, for 150 early flowering and 150 control plants in the complete randomized block design. To represent each lineage, we haphazardly collected one seed per fruit from three fruits per plant.
Data collection and analysis
As before, we noted age at flowering, stem diameter, flower number, seeds per fruit, and seed number. All plants survived to reproduce. To determine if plants responded to selection, we compared the age at flowering of selected and nonselected paired replicate lineages using a repeated-measures ANOVA of G4 plants. To explore if plants exhibited a correlated response to selection in plant size, we compared the stem diameter of selected and nonselected paired replicate lineages using a repeated-measures ANOVA. In both analyses, selection treatment was the repeated measure, and replicate and block were random effects. Age at flowering and stem diameter were log-transformed prior to analysis. To test for differences in lifetime fecundity in selected and nonselected paired replicate lineages, we compared seeds per fruit, flower number, and seed number using a multivariate, repeated-measures ANOVA. Selection treatment was the repeated measure, and replicate and block were random effects. Number of seeds and flowers per plant were log-transformed.
RESULTS
Population dynamics of experimental field populations
By 2005, three of the five populations (C1, C2, C3) were effectively extinct (no flowering adults, Fig. 2). One of these populations, C1, lasted longer and grew to 149 plants in 2003, declining to only nine plants in 2005. In the two populations that persisted through 2005, peak population sizes were 7839 plants for C4 and 502 plants for C5. Both of these populations appeared healthy at the end of the study (Fig. 2). We observed plants with nonflowering rosettes with large roots in all of the populations, and this domestication trait undoubtedly retarded population growth rates. Populations C4 and C5 flowered earlier than C1 in 2004, unlike in 2002, when these three populations flowered synchronously (L. G. Campbell and A. A. Snow, personal observation).
|
Phenotypes of field populations in a common garden
For many phenotypic traits, plants from population C1 were intermediate between plants from populations C4 and C5 (which included yellow-flowered plants) and those from the nonselected control lineages created in the artificial selection experiment (Table 1; for the ANOVA, see Appendix S1 in Supplemental Data with online version). Unlike the plants in the field populations, all plants in the common garden initiated flowering before the end of the growing season. Perhaps field populations were less likely to bolt than common garden plants because they were not confined to pots and could reach larger rosette sizes. In the common garden, plants from population C1 initiated flowering 3 days earlier than the control population and 5–7 days later than populations C4 and C5 (F3,27 = 83.92; P < 0.001). Plants from C4 and C5 had significantly smaller stem diameters than either C1 or the control populations (F3,27 = 67.62, P = 0.001; posthoc comparison: P < 0.001), but did not differ from each other (posthoc comparison: P = 1). The competing oats of control populations were largest, whereas the oats of C4 were significantly smaller than those of population C1 and the control population (F3,27 = 13.61, P < 0.001). This suggests that plants from population C4 were more competitive than those from C1. Plants from population C1 did not evolve significantly different lifetime seed production than the plants from the control lineages (P > 0.05; Table 1).
|
Phenotypic evolution in artificial selection lineages
The radish cultivar Red Silk responded somewhat to two generations of strong selection for earlier flowering, but the results are equivocal (Fig. 3). Overall, selection for early flowering advanced age at flowering (selection effect: F1,2 = 40.27, P < 0.001). Specifically, posthoc testing revealed that two of the three early flowering replicate lineages flowered significantly earlier (5–6 days) than control lineages. However, in the third replicate lineage, control and early flowering lineages did not differ in date of flowering, suggesting that the third lineage in the early selection treatment did not possess heritable variation for age at flowering (selection x replicate effect: F2,18 = 15.18, P < 0.001). Age at flowering was also affected by the main effect block (F9,18 = 2.86, P = 0.004) and the selection x block interaction (F9,18 = 2.05, P = 0.04). Therefore, the cultivar apparently possesses a limited amount of heritable variation for earlier flowering.
|
Correlated fitness consequences of early flowering
Early flowering plants were much more fecund than control plants in the common garden (Fig. 3; online Appendix S2). Early replicate 1 produced 93% more seeds than control replicate 1 and early replicate 2 produced 87% more seeds than control replicate 2 (selection x replicate effect: F2,18 = 7.098, P < 0.001). These large differences may be due to 72–78% higher flower production (selection effect: F1,2 = 4.36, P = 0.039; selection x replicate effect: F2,18 = 24.32, P < 0.001). Unlike the other two replicates, the lifetime fitness of early replicate 3 did not differ significantly from its control replicate 3 for any trait.
DISCUSSION
Persistence of feral radish populations
Our efforts to establish experimental populations of feral R. sativus in Michigan in the absence of crop–wild hybridization were unsuccessful. The three apparently nonintrogressed populations did not survive for more than three generations, whereas the two putatively introgressed populations grew vigorously. Compared with nonselected control plants, the most persistent nonintrogressed population (C1) evolved somewhat smaller roots and earlier flowering, but not greater fecundity. Many plants in populations C1, C2, and C3 had swollen roots and failed to flower or set fruit before the growing season ended, allowing other weedy annuals to invade the field plots. In a milder climate than Michigan, such as the coast of California, incipient feral populations with late-flowering and biennial phenotypes might be able to persist, but the delayed flowering trait was clearly maladaptive in Michigan.
To our knowledge, nonhybridized feral Raphanus sativus populations occur in Japan and Brazil, and they probably occurred along the California coastline before being displaced by crop–wild hybrids (Panetsos and Baker, 1967; Hegde et al., 2006, Ridley et al., 2008). In Japan, feral R. sativus populations occur in roadside ditches near fields of cultivated daikon radish, where hybridization with wild plants was not observed (Yamaguchi and Okamoto, 1997). These biennial, feral daikon plants have small, branched roots and are considered by farmers to both "contaminate" and enhance the genetic diversity of landraces (Yoon and Pyo, 1977). In southern Brazil, feral R. sativus weeds, apparently derived from cultivated forage radishes, have developed resistance to herbicides (Heap, 2008; G. Theisen, Fundacep Fecotrigo, Cruz Alta, Brazil, personal communication). Since 1980, no-till agriculture has predominated in this region, and a R. sativus forage cultivar known as forrageiro was planted widely as a cover crop. Where forrageiro occurred as a volunteer weed in summer crops, it was often treated with various herbicides. By 2001, feral R. sativus acquired resistance to several acetolactate synthase (ALS)-inhibiting herbicides in the southern state of Rio Grande do Sul. Based on photographs of the flowers and fruits of these plants (Heap, 2008), and information provided by G. Theisen, feral R. sativus does not appear to have the constricted fruits or yellow flower pigmentation common in R. raphanistrum, suggesting that these are nonhybridized feral populations.
Hybridization as a stimulus for ferality
Two of our experimental populations appeared to have hybridized with R. raphanistrum based on the presence of yellow-flowered plants and other traits such as smaller root diameters, earlier flowering, and lower pollen fertility, which is characteristic of crop–wild hybrids (Panetsos and Baker, 1967). The source of R. raphanistrum genes in these populations is unknown. We suspect our own movements among flowering populations of wild, hybrid, and feral plants, which were sometimes visited sequentially on the same day, might be the source of pollen contamination. Other possible sources of pollen contamination seem unlikely. We have never detected evidence of heterozygous crop parents in the thousands of crop and crop–wild hybrids that we have examined; therefore, it is unlikely that the original crop seeds included a few crop–wild hybrids. Also, the original greenhouse crosses were performed under strict isolation to prevent contamination by wild pollen, and no hybrid phenotypes were seen in the first generation or the next two generations, when we examined the flower color of >3000 crop plants grown in the artificial selection study. Because no seedlings of R. raphanistrum emerged during the 2002–2003 summers, based on the lack of yellow-flowered plants, it seems unlikely that dormant seeds were present at the field sites. Natural cross-pollination also seems improbable because local populations of R. raphanistrum were uncommon and distant, as were constructed hybrid populations (Campbell et al., 2006).
Regardless of their source, the unintentional presence of wild alleles in two of our populations permitted us to explore the weed potential of apparently hybridized feral populations. It is intriguing that the two populations with the yellow petal allele (C4 and C5) exhibited earlier flowering and greater fecundity than the C1 population and also had the most rapid population growth in the experimental plots. Many other factors could have influenced population growth rates, such as differences in weedy competitors, soil fertility, pathogens, or deer browsing. Thus, it is not possible to draw strong conclusions from the fact that the only healthy and persistent populations were the two that had wild-type traits. Nonetheless, we suggest that heritable phenotypic differences among the five field populations, due primarily to the presence of wild alleles, may have contributed to the observed differences in population persistence vs. extinction.
Along with other published studies (Panetsos and Baker, 1967; Klinger and Ellstrand, 1994), our results suggest introgression of wild alleles facilitated population persistence and higher growth rates in feral radish populations and allowed the rapid evolution of earlier flowering and increased fecundity as compared to nonintrogressed feral populations. Introgressed populations had lower pollen fertility than nonintrogressed plants, but reduced male fitness is not expected to impede population growth under natural pollination conditions where seed set is not pollen-limited (Stanton et al., 1986). Our results suggest that infrequent hybridization events can act as an external "trigger" for the evolution of weedy feral populations of R. sativus. Therefore, we hypothesize that exoferality has contributed more to dedomestication than endoferality in regions where R. raphanistrum is present.
Hybridizing crop–wild radish populations have been reported in Europe (Clapham et al., 1987; Stace, 1991), Asia (Yamaguchi and Okamoto, 1997), and North America (Panetsos and Baker, 1967). Wild R. sativus var. hortensis f. raphanistroides, a biennial plant, is common in Japan and Korea along beaches, cliffs, abandoned fields, and other ruderal areas, and sometimes occurs adjacent to cultivated daikon radishes (Yamagishi and Terachi, 2003), but these populations may be derived from R. raphanistrum rather than R. sativus (Hegde et al., 2006). Hybridized populations have been particularly well studied in California. Independent populations of feral Raphanus sativus and wild R. raphanistrum first appeared in California in the 19th century (Panetsos and Baker, 1967). These taxa hybridized so extensively that distinct populations of R. raphanistrum seem to have disappeared (Hegde et al., 2006).
Heritability of earlier flowering in radish cultivars
To complement results from naturally evolving populations, we examined the potential for earlier flowering to evolve using an artificial selection experiment. Age at flowering is a highly heritable trait in R. raphanistrum, R. sativus, and crop–wild hybrids (Rabbani et al. 1998; Mazer and Schick, 1991a,b; Campbell et al., 2008). Delayed flowering is a domestication trait that is maladaptive in short growing seasons, and we wanted to determine whether plant breeders had essentially eliminated heritable variation for earlier flowering in the Red Silk cultivar. The evolution of earlier flowering may be one of the first traits required for the establishment of volunteer and feral radish populations (Yamaguchi and Okamoto, 1997; Snow and Campbell, 2005).
Only two of the three lineages responded to selection for earlier flowering and their response was modest—they flowered about 5 days earlier than unselected controls. This lack of response again supports the argument that hybridization seems to be essential for the dedomestication of cultivated radish, rather than ferality via selection on standing variation within the crop. We found this difference in flowering date was correlated with smaller root diameters and a surprisingly large (
90%) increase in flower number and lifetime fecundity relative to control lineages (Fig. 3). Reasons for greater fecundity in the two early-flowering lineages are unclear and do not appear to be related to frost damage or other changes that occurred late in the growing season.
The choice of crop cultivar may influence the available genetic variation upon which selection can act in feral populations. In a survey of 10 radish cultivars, we found that daikon, podding, and forage radishes flowered 8–18 d earlier than the European small-rooted radishes (Campbell, 2007). It is perhaps not coincidental that the reported occurrences of nonhybridizing feral radish involve similar types of daikon and forage radishes (Yamaguchi and Okamoto, 1997; Snow and Campbell, 2005). In California, the co-occurrence of diverse radish cultivars, each possessing unique traits and genetic diversity (Ellstrand and Marshall, 1985; Muminovic et al., 2005; Hegde et al., 2006; Ridley et al., 2008), also may have promoted ferality in radish.
Conclusions
Our results suggest that the Red Silk cultivar of Raphanus sativus is not capable of establishing feral populations in Michigan without the aid of gene flow from R. raphanistrum. In other regions, volunteer radish populations might be considered "incipient" weeds. Some of these populations may evolve into vigorous weeds given an advantageous combination of environmental conditions and genetic diversity, as occurred in no-till rotations in Brazil (Snow and Campbell, 2005), and this process may be accelerated by high rates of gene flow from diverse cultivars or weedy R. raphanistrum. Given that agriculture disperses seeds over great distances, feral populations of R. sativus may eventually encounter favorable conditions for invasion and rapid expansion. However, for cultivated European radishes to evolve endoferality, key mutations and strong selection may be needed to reduce frequencies of deleterious crop traits such as delayed flowering and increase frequencies of "weedy" traits.
FOOTNOTES
1 Thanks to K. Mercer, P. Sweeney, Y. Teklu, M. Reagon, S. Su, J. Gao, D. Campbell, and C. Ridley for comments on this manuscript; the Ginop, Reimann, Sterzik, Jarman, and Romanik families for sharing their farmland; Harris-Moran, Ohio for generously donating seed; J. Leonard, the staff of UMBS, N. Marsh, and T. Waite for help with the experiments; and our many student researchers for their hard work. USDA grant 2002-03715, UMBS and The Nature Conservancy fellowships, and NSF grant 0508615 supported this work. 
3 Author for correspondence (e-mail: lgc{at}rice.edu); Department of Ecology and Evolutionary Biology, Rice University, 6100 Main St., Houston, Texas, 77005, USA; phone: 713-348-3054; fax: 713-348-5232
LITERATURE CITED
Alexander, M. P. 1969. Differential staining of aborted and nonaborted pollen. Stain Technology 44: 117–122.[Web of Science][Medline]
Ashiq Rabbani, M., Y. Murakami, Y. Kuginuki, AND K. Takayanagi. 1998. Genetic variation in radish (Raphanus sativus L.) germplasm from Pakistan using morphological traits and RAPDs. Genetic Resources and Crop Evolution 45: 307–316.[CrossRef][Web of Science]
Bervillé, A., M. H. Muller, B. Poinso, AND H. Serieys. 2005. Ferality—Risks of gene flow between sunflower and other Helianthus species. In J. Gressel [ed.], Crop ferality and volunteerism, 209–230. CRC Press, Boca Raton, Florida, USA.
Bond, J. M., R. J. Mogg, G. R. Squire, AND C. Johnstone. 2004. Microsatellite amplification in Brassica napus cultivars: Cultivar variability and relationship to a long-term feral population. Euphytica 139: 173–178.[CrossRef][Web of Science]
Campbell, L. G. 2007. Rapid evolution in a crop–weed complex (Raphanus spp.) Ph.D. dissertation, Ohio State University, Columbus, Ohio, USA.
Campbell, L. G., A. A. Snow, AND C. E. Ridley. 2006. Weed evolution after crop gene introgression: Greater survival and fecundity of hybrids in a new environment. Ecology Letters 9: 1198–1209.[CrossRef][Web of Science][Medline]
Campbell, L. G., A. A. Snow, P. M. Sweeney, AND J. M. Ketner. 2008. Rapid evolution in crop–weed hybrids under selection for divergent life histories. Evolutionary Applications [Online Early doi:10.1111/j.1752-4571.2008.00051.x].
Cheam, A. H., AND G. R. Code. 1995. The biology of Australian weeds. 24. Raphanus raphanistrum L. Plant Protection Quarterly 10: 2–13.
Clapham, A. R., T. Tutin, AND D. M. Moore. 1987. Flora of the British Isles. Cambridge University Press, New York, New York, USA.
Crawley, M. J., AND S. L. Brown. 1995. Seed limitation and the dynamics of feral oilseed rape on the M25 motorway. Proceedings of the Royal Society of London, B, Biological Sciences 259: 49–54.
Crisp, P. 1995. Radish, Raphanus sativus (Cruciferae). In J. S. Smartt, and N. W. Simmonds [ed.], Evolution of crop plants, 86–89. Longman Scientific & Technical, Harlow, UK.
Ellstrand, N. C., AND D. L. Marshall. 1985. Interpopulation gene flow by pollen in wild radish, Raphanus sativus. American Naturalist 126: 606–616.[CrossRef][Web of Science]
Erwin, J. E., R. M. Warner, AND A. G. Smith. 2002. Vernalization, photoperiod and GA3 interact to affect flowering of Japanese radish (Raphanus sativus Chinese Radish Jumbo Scarlet). Physiologia Plantarum 115: 298–302.[CrossRef][Medline]
Gressel, J. 2005. Crop ferality and volunteerism. CRC Press, Boca Raton, Florida, USA.
Gulden, R. H., AND S. J. Shirtliffe. 2003. Harvest losses of canola (Brassica napus) cause large seedbank inputs. Weed Science 51: 83–86.[CrossRef][Web of Science]
Harker, K. N., G. W. Clayton, R. E. Blackshaw, J. T. O’Donovan, E. N. Johnson, Y. Gan, F. A. Holm et al.. 2005. Glyphosate-resistant wheat persistence in western Canadian cropping systems. Weed Science 53: 846–859.[CrossRef][Web of Science]
Heap, I. M. 2008. International survey of herbicide resistant weeds. Website http://www.weedscience.org/in.asp [accessed 18 November 2007].
Heenan, P. B., R. G. Fitzjohn, AND M. I. Dawson. 2004. Diversity of Brassica (Brassicaceae) species naturalised in Canterbury, New Zealand. New Zealand Journal of Botany 42: 815–832.[Web of Science]
Hegde, S. G., J. D. Nason, J. M. Clegg, AND N. C. Ellstrand. 2006. The evolution of California’s wild radish has resulted in the extinction of its progenitors. Evolution 60: 1187–1197.[Web of Science][Medline]
Holm, L. G., J. Doll, E. Holm, J. Pancho, AND J. Herberger. 1997. World weeds: Natural histories and distributions. John Wiley, New York, New York, USA.
Irwin, R. E., S. Y. Strauss, S. Storz, A. Emerson, AND G. Guibert. 2003. The role of herbivores in the maintenance of a flower color polymorphism in wild radish. Ecology 84: 1733–1743.[CrossRef][Web of Science]
Kay, Q. O. N. 1976. Preferential pollination of yellow-flowered morphs of Raphanus raphanistrum by Pieris and Eristalis spp. Nature 261: 230–232.[CrossRef][Web of Science]
Kercher, S., AND J. K. Conner. 1996. Patterns of genetic variability within and among populations of wild radish, Raphanus raphanistrum (Brassicaceae). American Journal of Botany 83: 1416–1421.[CrossRef][Web of Science]
Klinger, T., AND N. C. Ellstrand. 1994. Engineered genes in wild populations: Weed–crop hybrids of Raphanus sativus. Ecological Applications 4: 117–120.[CrossRef][Web of Science]
Kowarik, I. 2005. Urban ornamentals escaped from cultivation. In J. Gressel [ed.], Crop ferality and volunteerism, 97–122. CRC Press, Boca Raton, Florida, USA.
Lee, T. N., AND A. A. Snow. 1998. Pollinator preferences and the persistence of crop genes in wild radish populations (Raphanus raphanistrum, Brassicaceae). American Journal of Botany 85: 333–349.[Abstract]
Li, C. B., A. L. Zhou, AND T. Sang. 2006. Genetic analysis of rice domestication syndrome with the wild annual species, Oryza nivara. New Phytologist 170: 185–193.[CrossRef][Web of Science][Medline]
Lumaret, R., N. Ouazzani, H. Michaud, G. Vivier, M. F. Deguilloux, AND F. Di Giusto. 2004. Allozyme variation of oleaster populations (wild olive tree) (Olea europaea L.) in the Mediterranean Basin. Heredity 92: 343–351.[CrossRef][Web of Science][Medline]
Lutman, P. J. W., S. E. Freeman, AND C. Pekrun. 2003. The long-term persistence of seeds of oilseed rape (Brassica napus) in arable fields. Journal of Agricultural Science 141: 231–240.[CrossRef]
Mack, R. N., AND M. Erneberg. 2002. The United States naturalized flora: Largely the product of deliberate introductions. Annals of the Missouri Botanical Garden 89: 176–189.[CrossRef][Web of Science]
Mazer, S. J., AND C. T. Schick. 1991a. Constancy of population parameters for life history and floral traits in Raphanus sativus L. I. Norms of reaction and the nature of genotype by environmental interactions. Heredity 67: 143–156.[CrossRef][Web of Science]
Mazer, S. J., AND C. T. Schick. 1991b. Constancy of population parameters for life-history and floral traits in Raphanus sativus L. 2. Effects of planting density on phenotype and heritability estimates. Evolution 45: 1888–1907.[CrossRef][Web of Science]
Mekuria, G. T., G. Collins, AND M. Sedgley. 2002. Genetic diversity within an isolated olive (Olea europaea L.) population in relation to feral spread. Scientia Horticulturae 94: 91–105.[CrossRef]
Muminovic, J., A. Merz, A. E. Melchinger, AND T. Lubberstedt. 2005. Genetic structure and diversity among radish varieties as inferred from AFLP and ISSR analyses. Journal of the American Society for Horticultural Science 130: 79–87.
Ogg, A. G., AND R. Parker. 1989. Control of volunteer crop plants. Washington State University Extension Publication no. EB1523, Pullman, Washington, USA.
Panetsos, C. A., AND H. G. Baker. 1967. The origin of variation in "wild" Raphanus sativus (Cruciferae) in California. Genetica 38: 243–274.[CrossRef][Web of Science]
Pessel, F. D., J. Lecomte, V. Emeriau, M. Krouti, A. Messean, AND P. H. Gouyon. 2001. Persistence of oilseed rape (Brassica napus L.) outside of cultivated fields. Theoretical and Applied Genetics 102: 841–846.[CrossRef][Web of Science]
Pester, T. A., P. Westra, R. L. Anderson, D. J. Lyon, S. D. Miller, P. W. Stahlman, F. E. Northam, AND G. A. Wicks. 2000. Secale cereale interference and economic thresholds in winter Triticum aestivum. Weed Science 48: 720–727.[CrossRef][Web of Science]
Reagon, M. R., AND A. A. Snow. 2006. Cultivated Helianthus annuus (Asteraceae): Volunteers as a genetic "bridge" to weedy sunflower populations in North America. American Journal of Botany 93: 127–133.
Ridley, C. E., S.-C. Kim, AND N. C. Ellstrand. 2008. Bi-directional history of hybridization in California wild radish, Raphanus sativus (Brassicaceae), as revealed by chloroplast DNA. American Journal of Botany 95: 1437–1442.
Smyth, S., G. G. Khachatourians, AND P. W. B. Phillips. 2002. Liabilities and economics of transgenic crops. Nature Biotechnology 20: 537–541.[Web of Science][Medline]
Snow, A. A., AND L. G. Campbell. 2005. Can feral radishes become weeds? In J. Gressel [ed.], Crop ferality and volunteerism, 193–208. CRC Press, Boca Raton, Florida, USA.
Snow, A. A., K. L. Uthus, AND T. M. Culley. 2001. Fitness of hybrids between weedy and cultivated radish: Implications for weed evolution. Ecological Applications 11: 934–943.[CrossRef][Web of Science]
Stace, C. A. 1991. New flora of the British Isles. Cambridge University Press, Cambridge, UK.
Stanton, M. L., A. A. Snow, AND S. N. Handel. 1986. Floral evolution: Attractiveness to pollinators increases male fitness in a hermaphroditic angiosperm. Science 232: 1625–1627.
Stump, W. L., AND P. Westra. 2000. The seedbank dynamics of feral rye (Secale cereale). Weed Technology 14: 7–14.[CrossRef][Web of Science]
Tabuti, J. R. S., S. S. Dhillion, AND K. A. Lye. 2004. The status of wild food plants in Bulamogi county, Uganda. International Journal of Food Sciences and Nutrition 55: 485–498.[CrossRef][Web of Science][Medline]
Ulloa, M., J. M. Stewart, E. A. Garcia, S. Godoy, A. Gaytan, AND S. Acosta. 2006. Cotton genetic resources in the western states of Mexico: In situ conservation status and germplasm collection for ex situ preservation. Genetic Resources and Crop Evolution 53: 653–668.[CrossRef][Web of Science]
Warwick, S. I., AND A. Francis. 2005. The biology of Canadian weeds. 132. Raphanus raphanistrum L. Canadian Journal of Plant Science 85: 709–733.[Web of Science]
Yamagishi, H., AND T. Terachi. 2003. Multiple origins of cultivated radishes as evidenced by a comparison of the structural variations in mitochondrial DNA of Raphanus. Genome 46: 89–94.[Medline]
Yamaguchi, H., AND M. Okamoto. 1997. Traditional seed production in landraces of daikon (Raphanus sativus) in Kyushu, Japan. Euphytica 95: 141–147.[CrossRef][Web of Science]
Yoon, H. M., AND H. K. Pyo. 1977. Studies on the inheritance of bolting, flowering time and other traits in radish (Raphanus sativus L.). Korean Journal of Breeding 9: 45–57.
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | SEARCH RESULT |
50 per lineage.