Mite species inhabiting commercial bumblebee (<Emphasis Type="Italic">Bombus terrestris</Emphasis>) nests in Polish greenhouses

Nests of social insects are usually inhabited by various mite species that feed on pollen, other micro-arthropods or are parasitic. Well-known negative effects of worldwide economic importance are caused by mites parasitizing honeybee colonies. Lately, attention has focused on the endoparasitic mite Locustacarus buchneri that has been found in commercial bumblebees. However, little is known of other mites associated with commercial bumblebee nests. Transportation of commercial bumblebee colonies with unwanted residents may introduce foreign mite species to new localities. In this study, we assessed the prevalence and species composition of mites associated with commercial bumblebee nests and determined if the mites are foreign species for Poland and for Europe. The study was conducted on 37 commercial bumblebee nests from two companies (Dutch and Israeli), originating from two greenhouses in southern Poland, and on 20 commercial bumblebee colonies obtained directly from suppliers. The species composition and abundance of mites inhabiting commercial bumblebee nests were determined. Seven mite species from three families were found in nests after greenhouse exploitation. The predominant mite species was Tyrophagus putrescentiae (Acaridae) that was a 100-fold more numerous than representatives of the family Laelapidae (Hypoaspis marginepilosa, H. hyatti, H. bombicolens). Representatives of Parasitidae (Parasitellus fucorum, P. crinitus, P. ignotus) were least numerous. All identified mite species are common throughout Europe, foreign species were not found. Mites were not detected in nests obtained directly from suppliers. We conclude that probably bumblebee nests are invaded by local mite species during greenhouse exploitation.     

Morphological and molecular characterization of the Landes honey bee (<Emphasis Type="Italic">Apis mellifera</Emphasis> L.) ecotype for genetic conservation

A population of honey bees (Apis mellifera mellifera L.) with an annual colony brood cycle adapted to a locally abundant floral source in the Landes region of Southwest France is the subject of genetic conservation efforts. This population is maintained by local beekeepers in an area that experiences both an annual seasonal influx of non-local colonies and the permanent culture of imported stock. However, some colonies native to the Landes do not have the adapted brood cycle and their status as ecotypic are in question. The present study used morphology, mitochondrial DNA and microsatellites to characterize the endemic population and suggests further genetic conservation strategies. These methods yielded different degrees of discrimination of native and imported colonies and provided a powerful suite of tools for local resource managers. Colonies from the Landes could be differentiated from non-local French A. m. mellifera populations using morphometric analysis, and from non-native and reference populations using mtDNA and microsatellites. Seven morphological characters were identified by discriminant analysis as informative for delineating the Landes ecotype from other A. m. mellifera populations. Mitochondrial haplotypes for the population were characterized and five microsatellite loci were found to be informative in characterizing the Landes population. Asymmetric gene flow detected with microsatellite alleles was observed to be 5.5–5.9% from imported to native stocks of honey bees while introgression of native microsatellite alleles into imported colonies was 21.6%.     

Abundance,body size,and morphology of bumblebees in an area where an exotic species, <Emphasis Type="Italic">Bombus terrestris</Emphasis>, has colonized in Japan

An exotic bumblebee species, Bombus terrestris, has colonized in Japan and becomes dominant in some local communities. We examined the effects of land use and bumblebee abundance on the number and body size of bumblebees collected using window traps in a lowland area in the southern Ishikari district, Hokkaido. In 2004, we collected 922 bumblebees of six species using 70 traps at 17 sites. A statistical model fitted to the data demonstrated that dispersion from commercial B. terrestris colonies used in greenhouses positively affected the number of B. terrestris caught by each trap. This exotic species was abundant in sites where paddy fields were prevalent, but three native species, B. hypocrita, B. ardens, and B. diversus, were abundant in sites where farms and woodlands were widespread. The local abundance of B. terrestris was not associated negatively with the number and body size of native bumblebees. Thus, we did not find any competitive interactions between exotic and native bumblebees although habitat conditions seem to be common determinants of the bumblebee populations. A morphological analysis showed that B. terrestris had intermediate tongue length between B. hypocrita and B. ardens.     

Worldwide migration of parasitic mites as a result of bumblebee commercialization

We investigated the status of infestation by a tracheal mite, Locustacarus buchneri, in natural populations of a Japanese native bumblebee species, Bombus hypocrita, collected on Hokkaido Island and in the Aomori prefecture between 1997 and 2001. We also investigated mite infestation in commercial colonies of the European bumblebee, Bombus terrestris, imported from the Netherlands and Belgium, and the Japanese native species, B. ignitus, imported from the Netherlands, between 1997 and 2001. We detected mites in natural populations of the two B. hypocrita subspecies and in the commercial colonies. Analysis of variations in 535 bp sequences of the mitochondrial cytochrome oxidase subunit 1 (CO1) gene showed that the mite haplotypes in the native populations and in the imported colonies did not overlap in 1997–1999, but in 2000–2001 some mites possessing European CO1 haplotypes were detected in the natural populations of Japanese native bumblebees. In addition, many mites possessing Japanese haplotypes were detected in the imported commercial colonies from Europe. Considering the fact that the Japanese native bumblebees, B. hypocrita, were once exported to Europe for commercialization, these results suggest that bumblebee commercialization has caused overseas migration and cross-infestation of parasitic mites among natural and commercial colonies. However, because the Japanese and European CO1 haplotypes were closely related, there was a possibility that the European haplotypes found in the mites in the Hokkaido Island revealed native variation. To clarify the status of mite invasion, further detailed analysis of genetic variation of the mite, using other genetic markers on additional samples, need to be performed.     

MtDNA COI‐COII marker and drone congregation area: An efficient method to establish and monitor honeybee (Apis mellifera L.) conservation centres

Honeybee subspecies have been affected by human activities in Europe over the past few decades. One such example is the importation of nonlocal subspecies of bees which has had an adverse impact on the geographical repartition and subsequently on the genetic diversity of the black honeybee Apis mellifera mellifera. To restore the original diversity of this local honeybee subspecies, different conservation centres were set up in Europe. In this study, we established a black honeybee conservation centre Conservatoire de l'Abeille Noire d'Ile de France (CANIF) in the region of Ile‐de‐France, France. CANIF's honeybee colonies were intensively studied over a 3‐year period. This study included a drone congregation area (DCA) located in the conservation centre. MtDNA COI‐COII marker was used to evaluate the genetic diversity of CANIF's honeybee populations and the drones found and collected from the DCA. The same marker (mtDNA) was used to estimate the interactions and the haplotype frequency between CANIF's honeybee populations and 10 surrounding honeybee apiaries located outside of the CANIF. Our results indicate that the colonies of the conservation centre and the drones of the DCA show similar stable profiles compared to the surrounding populations with lower level of introgression. The mtDNA marker used on both DCA and colonies of the conservation centre seems to be an efficient approach to monitor and maintain the genetic diversity of the protected honeybee populations.     

Geographic variability in mitochondrial introgression among hybridizing populations of Golden-winged (<Emphasis Type="Italic">Vermivora chrysoptera</Emphasis>) and Blue-winged (<Emphasis Type="Italic">V.opinus</Emphasis>) Warblers

The rapidly declining Golden-winged Warbler (Vermivora chrysoptera) is of conservation concern owing in part to hybridization with the closely related Blue-winged Warbler (V. pinus). These species hybridize extensively in eastern North America and over the past century the Blue-winged Warbler has displaced the Golden-winged Warbler from substantial regions of its historic breeding range. A previous study suggested that these genetic interactions result in rapid and asymmetric introgression of Blue-winged Warbler mitochondrial DNA (mtDNA) into Golden-winged phenotype populations within the zones of contact, but more recent and extensive surveys have documented a more complex pattern of genetic interchange between these taxa. We surveyed mtDNA/phenotype associations in 104 individuals of known phenotype drawn from two locations with different histories of contact and found substantial variation between sites in the extent of introgression. Where both species have co-existed for more than a century, we found evidence of bi-directional introgression and the long-term persistence of Golden-winged mtDNA haplotypes. At the leading edge of the northward expansion of Blue-winged Warblers, we found predominantly Golden-winged Warbler mtDNA haplotypes in both Golden-winged and hybrid-phenotype individuals. Across both sites, genetic swamping does not appear to be occurring via the early immigration of Blue-winged Warbler females into populations dominated by Golden-winged Warbler phenotypes. Instead, the differing patterns of mitochondrial introgression may be driven by the relative local population sizes of the parental species coupled with subtle between-species differences in mate choice and habitat preferences.     

Potential host shift of the small hive beetle (Aethina tumida) to bumblebee colonies (Bombus impatiens)

Here we explored the potential for host shift from honeybee, Apis mellifera, colonies to bumblebee, Bombus impatiens, colonies by the small hive beetle, a nest parasite/scavenger native to sub-Saharan Africa. We investigated small hive beetle host choice, bumblebee colony defence as well as individual defensive behaviour of honeybee and bumblebee workers. Our findings show that in its new range in North America, bumblebees are potential alternate hosts for the small hive beetle. We found that small hive beetles do invade bumblebee colonies and readily oviposit there. Honeybee colonies are not preferred over bumblebee colonies. But even though bumblebees lack a co-evolutionary history with the small hive beetle, they are able to defend their colonies against this nest intruder by removal of beetle eggs and larvae and stinging of the latter. Hence, the observed behavioural mechanisms must be part of a generalistic defence system suitable for defence against multiple attackers. Nevertheless, there are quantitative (worker force) and qualitative differences (hygienic behaviour) between A. mellifera and B. impatiens. Received 16 July 2007; revised 16 January 2008; accepted 17 January 2008.     

Spatial and temporal changes in genetic structure of greenhouse and field populations of cabbage looper,Trichoplusia ni

Trichoplusia ni is a subtropical moth that migrates annually from southern California to southern British Columbia, Canada where it invades vegetable greenhouses and field crops. The heated greenhouse environment has altered the natural extinction–recolonization dynamics of T. ni populations, and allows year‐round persistence in some locations. In addition, the extensive use of the biopesticide, Bacillus thuringiensis subspecies kurstaki (Bt) in some greenhouses has selected for resistance. Here we investigated the genetic structure of T. ni populations in British Columbia greenhouses and in field populations in California and British Columbia using amplified fragment length polymorphisms (AFLP) as related to patterns of Bt resistance. The majority of British Columbia field populations were similar to the California field populations, the potential source of migrants. However populations in two geographic areas with high concentrations of greenhouses showed local genetic differentiation. Some of these populations experienced severe bottlenecks over‐winter and following Bt sprays. Greenhouse populations showed a pattern of isolation by distance and a strong positive relationship between genetic differentiation and levels of Bt resistance. These patterns indicate that greenhouses that sometimes support year‐round populations of T. ni and the ensuing strong bottlenecking effects following winter cleanups and Bt application cause genetic differentiation of T. ni populations. Long distance migrants to field populations contribute to genetic homogeneity of these.     

Diploid male production in a rare and locally distributed bumblebee, Bombus florilegus (Hymenoptera, Apidae)

The European bumblebee B. terrestris was recently introduced in Japan for agricultural purposes and has now become naturalized. The naturalization of this exotic species may have great detrimental effects on closely related native Japanese bumblebees. The Japanese bumblebee Bombus florilegus is a rare and locally distributed species found in the Nemuro Peninsula of Hokkaido, Japan. In order to assess its population genetics, we estimated the genetic structure of B. florilegus in 16 breeding colonies (queen, workers, and males) and 20 foraging queens by analyzing microsatellite DNA markers. Of the 36 queens analyzed by genotyping and dissection, 32 had been inseminated by a male. The remaining 4 had not been inseminated at all. Of the 4 nonmated queens, one was triploid. Diploid males were found in 4 breeding colonies. Based on the microsatellite data, it appears that B. florilegus has low reproductive success. Since matched mating and nonmating within local populations are high, the extinction risk is correspondingly higher. Our results suggest that conservation of the Japanese B. florilegus is required in order to protect it from both habitat destruction and the naturalization of alien species. Received 19 July 2007; revised 13 October 2007; accepted 15 October 2007.     

Abiotic stress QTL in lettuce crop–wild hybrids: comparing greenhouse and field experiments

The development of stress‐tolerant crops is an increasingly important goal of current crop breeding. A higher abiotic stress tolerance could increase the probability of introgression of genes from crops to wild relatives. This is particularly relevant to the discussion on the risks of new GM crops that may be engineered to increase abiotic stress resistance. We investigated abiotic stress QTL in greenhouse and field experiments in which we subjected recombinant inbred lines from a cross between cultivated Lactuca sativa cv. Salinas and its wild relative L. serriola to drought, low nutrients, salt stress, and aboveground competition. Aboveground biomass at the end of the rosette stage was used as a proxy for the performance of plants under a particular stress. We detected a mosaic of abiotic stress QTL over the entire genome with little overlap between QTL from different stresses. The two QTL clusters that were identified reflected general growth rather than specific stress responses and colocated with clusters found in earlier studies for leaf shape and flowering time. Genetic correlations across treatments were often higher among different stress treatments within the same experiment (greenhouse or field), than among the same type of stress applied in different experiments. Moreover, the effects of the field stress treatments were more correlated with those of the greenhouse competition treatments than to those of the other greenhouse stress experiments, suggesting that competition rather than abiotic stress is a major factor in the field. In conclusion, the introgression risk of stress tolerance (trans‐)genes under field conditions cannot easily be predicted based on genomic background selection patterns from controlled QTL experiments in greenhouses, especially field data will be needed to assess potential (negative) ecological effects of introgression of these transgenes into wild relatives.     

Population structure,dispersal and colonization history of the garden bumblebee <Emphasis Type="Italic">Bombus hortorum</Emphasis> in the Western Isles of Scotland

New methods of analysing genetic data provide powerful tools for quantifying dispersal patterns and reconstructing population histories. Here we examine the population structure of the bumblebee Bombus hortorum in a model island system, the Western Isles of Scotland, using microsatellite markers. Following declines in other species, B. hortorum is the only remaining long-tongued bumblebee species found in much of Europe, and thus it is of particular ecological importance. Our data suggest that populations of B. hortorum in western Scotland exist as distinct genetic clusters occupying groups of nearby islands. Population structuring was higher than for other bumblebee species which have previously been studied in this same island group (F_st = 0.16). Populations showed significant isolation by distance. This relationship was greatly improved by using circuit theory to allow dispersal rates to differ over different landscape features; as we would predict, sea appears to provide far higher resistance to dispersal than land. Incorporating bathymetry data improved the fit of the model further; populations separated by shallow seas are more genetically similar than those separated by deeper seas. We argue that this probably reflects events following the last ice age when the islands were first colonized by this bee species (8,500–5,000 ybp), when the sea levels were lower and islands separated by shallow channels would have been joined. In the absence of significant gene flow these genetic clusters appear to have since diverged over the following 5,000 years and arguably may now represent locally adapted races, some occurring on single islands.     

Molecular and spatial analyses reveal links between colony‐specific foraging distance and landscape‐level resource availability in two bumblebee species

Foraging distance is a key determinant of colony survival and pollination potential in bumblebees Bombus spp. However this aspect of bumblebee ecology is poorly understood because of the difficulty in locating colonies of these central place foragers. Here, we used a combination of molecular microsatellite analyses, remote sensing and spatial analyses using kernel density estimates to estimate nest location and foraging distances for a large number of wild colonies of two species, and related these to the distribution of foraging habitats across an experimentally manipulated landscape. Mean foraging distances were 755 m for Bombus lapidarius and 775 m for B. pascuorum (using our most conservative estimation method). Colony‐specific foraging distances of both species varied with landscape structure, decreasing as the proportion of foraging habitats increased. This is the first time that foraging distance in wild bumblebees has been shown to vary with resource availability. Our method offers a means of estimating foraging distances in social insects, and informs the scale of management required to conserve bumblebee populations and enhance their pollination services across different landscapes.     

Habitat quality limits gene flow between populations of <Emphasis Type="Italic">Bombus ruderatus</Emphasis> in the South Island,New Zealand

One of the primary reasons for the decline of some bumblebee species has been habitat loss and fragmentation through land use change. Habitat fragmentation can limit connectivity between populations and gene flow between bumblebee populations can be limited by open water and human altered landscapes, however the influence of landscape features on gene flow has only been examined in non-declining species. The ruderal bumblebee, Bombus ruderatus, was successfully introduced to and is now relatively common in New Zealand, providing an opportunity to examine the biology of a species that is now rare in its native range in the UK. In this study, we examine the genetic structuring of B. ruderatus populations in the South Island of New Zealand and we demonstrate that a relatively simple classification of the landscape, into either good or poor foraging habitat at coarse resolution (800 m²), can predict levels of gene flow. We found populations of B. ruderatus as far apart as 160 km showing no significant genetic differentiation. However, this level of gene flow appears to be reliant upon continuous suitable habitat, as other populations <100 km apart were found to be significantly differentiated. These results suggest that corridors of continuous habitat are required to facilitate gene flow over large distances for this species.     

Population genetics of the socially polymorphic ant <Emphasis Type="Italic">Formica podzolica</Emphasis>

Summary. We used microsatellite markers to analyze the hierarchical genetic structure of the North American mound building ant, Formica podzolica. About one-third of all colonies were headed by a single queen (monogynous) whose effective mating frequency was close to one (nestmate worker relatedness r = 0.70), while the remaining colonies were polygynous, with low average nestmate relatedness (r = 0.16). The low worker relatedness found in most polygynous colonies furthermore suggested that the numbers of queens in polygynous colonies of this ant are usually high. Contrary to what has been described from other ants with a queen number dichotomy, we did not find an effect of social form variation on the partitioning of genetic variation above the level of the colony. We found no significant differentiation between the sympatric social forms of F. podzolica, nor did differentiation among populations appear to be affected by colony social organization. These unexpected patterns of genetic structure may have resulted from differences either in the spatial distribution of the social forms or in their social flexibility.Received 12 January 2004; revised 23 February 2004; accepted 10 March 2004.     

Comparison and examination of Bombus occidentalis and Bombus impatiens (Hymenoptera: Apidae) in tomato greenhouses

Experiments were conducted in commercial tomato, Lycopersicon esculentum Miller (Solanaceae), greenhouses to compare the relative foraging effort of two bumble bee species, Bombus occidentalis Greene and Bombus impatiens Cresson, to examine interspecific competition between B. occidentalis and B. impatiens, and to determine whether bumble bee colonies grew to their full population potential in commercial tomato greenhouses. B. impatiens colonies had more brood and workers and made more foraging trips per hour than B. occidentalis colonies. However, B. impatiens returned to the colony without pollen loads and left their colonies without dropping off their pollen loads more frequently than B. occidentalis greenhouse colonies. Our data also suggest that the presence of B. impatiens had a detrimental effect on B. occidentalis populations. Furthermore, B. occidentalis colonies did not grow to their full population potential in tomato greenhouses, with fewer workers in greenhouse colonies than in colonies placed outside in a natural environment, or in colonies that were physically enclosed and protected from external mortality. Together, this study suggests that B. impatiens is a better pollinator than B. occidentalis. It also shows that unknown factors are limiting the size of B. occidentalis colonies in tomato greenhouses.     

Managing honey bees (Hymenoptera: Apidae) for greenhouse tomato pollination

Although commercially reared colonies of bumble bees (Bombus sp.) are the primary pollinator world-wide for greenhouse tomatoes (Lycopersicon esculentum Mill.) previous research indicates that honey bees (Apis mellifera L.) might be a feasible alternative or supplement to bumble bee pollination. However, management methods for honey bee greenhouse tomato pollination scarcely have been explored. We 1) tested the effect of initial amounts of brood on colony population size and flight activity in screened greenhouses during the winter, and 2) compared foraging from colonies with brood used within screened and unscreened greenhouses during the summer. Brood rearing was maintained at low levels in both brood and no-brood colonies after 21 d during the winter, and emerging honey bees from both treatments had significantly lower weights than bees from outdoor colonies. Honey bee flight activity throughout the day and over the 21 d in the greenhouse was not influenced by initial brood level. In our summer experiment, brood production in screened greenhouses neared zero after 21 d but higher levels of brood were reared in unscreened greenhouses with access to outside forage. Flower visitation measured throughout the day and over the 21 d the colonies were in the greenhouse was not influenced by screening treatment. An economic analysis indicated that managing honey bees for greenhouse tomato pollination would be financially viable for both beekeepers and growers. We conclude that honey bees can be successfully managed for greenhouse tomato pollination in both screened and unscreened greenhouses if the foraging force is maintained by replacing colonies every 3 wk.     

Quantifying the relative predation pressure on bumblebee nests by the European badger (Meles meles) using artificial nests

Bumblebee populations are declining. Factors that impact the size and success of colonies act by either limiting resource availability (bottom‐up regulation) or causing mortality, for example, pesticides, disease, and possibly predation (top‐down regulation). The impact of predation has not been quantified, and so, the current study used novel artificial nests as a proxy for wild bumblebee nests to quantify the relative predation pressure from badgers in two habitats: woodland and grassland, and at two nesting depths: surface and underground. Badgers occur across most parts of the UK and are known to predate on bumblebee nests. We found that significantly more artificial nests (pots containing bumblebee nest material) were dug up compared with control pots (pots without bumblebee nest material). This shows that artificial nests have the potential to be used as a method to study the predation of bumblebee nests by badgers. In a location of high badger density, predation pressure was greater in woodland than grassland, whereas no difference was observed in relation to nest depth. Woodland and grassland are shared habitats between bumblebees and badgers, and we suggest that higher predation may relate to activity and foraging behavior of badgers in woodland compared with grassland. We discuss how badger predation in different habitats could impact different bumblebee species according to their nesting behaviors. Understanding the relative impact of badger predation on bumblebee colonies provides key information on how such top‐down regulation affects bumblebee populations.     

Use of selected flowering plants in greenhouses to enhance aphidophagous hoverfly populations (Diptera: Syrphidae)

The addition of floral resources in a crop is the most commonly used conservation biological control strategy. The influence of additional floral resources on the abundance of aphidophagous syrphids has been studied in Mediterranean sweet-pepper greenhouses, in southeast Spain. Sweet alyssum and coriander were the plant species used as flowering plants, distributed in the greenhouse in several monospecific patches. In our first experiment the influence on syrphid pre-imaginal stages (larvae and pupae) was studied and adult stages were studied in a second experiment. A higher number of pre-imaginal syrphids was recorded in two replicated greenhouses where flowers were introduced, compared with two control greenhouses (without additional floral resources). To evaluate the effect on adults, 4 greenhouses were divided into 2 plots in each greenhouse and flowers were introduced in one plot per greenhouse. More hoverfly adults were observed in the plots where flowers had been introduced than in the control. The three most abundant syrphid species found were Eupeodes corollae, Episyrphus balteatus and Sphaerophoria rueppellii. Specimens from these species were dissected, and their pollen content was analysed to assess the food potential of the introduced flowers. The three syrphid species fed on pollen from both the flowering plants, as well as on sweet-pepper pollen. This conservation biological control strategy is an effective method to enhance native syrphid populations in Mediterranean sweet-pepper greenhouses.     

Whole‐genome sequencing reveals small genomic regions of introgression in an introduced crater lake population of threespine stickleback

Invasive species pose a major threat to biological diversity. Although introduced populations often experience population bottlenecks, some invasive species are thought to be originated from hybridization between multiple populations or species, which can contribute to the maintenance of high genetic diversity. Recent advances in genome sequencing enable us to trace the evolutionary history of invasive species even at whole‐genome level and may help to identify the history of past hybridization that may be overlooked by traditional marker‐based analysis. Here, we conducted whole‐genome sequencing of eight threespine stickleback (Gasterosteus aculeatus) individuals, four from a recently introduced crater lake population and four of the putative source population. We found that both populations have several small genomic regions with high genetic diversity, which resulted from introgression from a closely related species (Gasterosteus nipponicus). The sizes of the regions were too small to be detected with traditional marker‐based analysis or even some reduced‐representation sequencing methods. Further amplicon sequencing revealed linkage disequilibrium around an introgression site, which suggests the possibility of selective sweep at the introgression site. Thus, interspecies introgression might predate introduction and increase genetic variation in the source population. Whole‐genome sequencing of even a small number of individuals can therefore provide higher resolution inference of history of introduced populations.     

Introgression from cultivated rice influences genetic differentiation of weedy rice populations at a local spatial scale

Hybridization and introgression can play an important role in genetic differentiation and adaptive evolution of plant species. For example, a conspecific feral species may frequently acquire new alleles from its coexisting crops via introgression. However, little is known about this process. We analyzed 24 weedy rice (Oryza sativa f. spontanea) populations and their coexisting rice cultivars from northern Italy to study their genetic differentiation, outcrossing, and introgression based on microsatellite polymorphisms. A total of 576 maternal plants representing 24 weedy populations were used to estimate their genetic differentiation, and 5,395 progeny (seedlings) derived from 299 families of 15 selected populations were included to measure outcrossing rates. Considerable genetic differentiation (F _st = 0.26) was detected among weedy rice populations, although the differentiation was not associated with the spatial pattern of the populations. Private alleles (28%) were identified in most populations that exhibited a multiple cluster assignments, indicating stronger genetic affinities of some weedy populations. Outcrossing rates were greatly variable and positively correlated (R ² = 0.34, P = 0.02) with the private alleles of the corresponding populations. Paternity analysis suggested that ~15% of paternal specific alleles, a considerable portion of which was found to be crop-specific, were acquired from the introgression of the coexisting rice cultivars. Frequent allelic introgression into weedy populations resulting from outcrossing with nearby cultivars determines the private alleles of local feral populations, possibly leading to their genetic differentiation. Introgression from a crop may play an important role in the adaptive evolution of feral populations.