Orchis×colemanii hybridization: Molecular and morphological evidence, seed set success, and evolutionary importance

Flora 207 (2012) 753–761 Contents lists available at SciVerse ScienceDirect Flora journal homepage: www.elsevier.de/flora Orchis × colemanii hybridization: Molecular and morphological evidence, seed set success, and evolutionary importance Alessia Luca a , Francesca Bellusci a , Bruno Menale b , Aldo Musacchio a , Giuseppe Pellegrino a,∗ a b Department of Ecology, University of Calabria, I-87036 Rende, Italy Botanic Garden of Naples, University Federico II, I-80139 Naples, Italy a r t i c l e i n f o Article history: Received 14 March 2012 Accepted 15 July 2012 Keywords: Food-deceptive orchids Hybrid zones Homoploid hybrid speciation Reproductive barriers a b s t r a c t Despite highly specialized pollination strategies, hybridization is a common phenomenon among Mediterranean deceptive orchids. Food-deceptive species sire a progeny of F1 unfertile plants, which work as a late post-zygotic barrier. Conversely, when pre-zygotic barriers of sexually deceptive (Ophrys) species are absent, the hybrids are fertile and an extensive introgression may occur. Here, we have performed molecular analysis and hand pollination treatments to characterize a hybrid zone of two food-deceptive species, O. mascula and O pauciflora. Hybrids (called O. × colemanii) have shown different amounts of parental nrDNA, strongly supporting that they are F2 and/or successive hybrid generations. Comparable high levels of reproductive success have been detected in natural conditions and in experimental crosses suggesting the absence of effective reproductive barriers either between hybrids, either between hybrids and parental species. In light of ecological and distributional features of O. × colemanii across its distribution range, we hypothesize that these populations have originated by secondary contact in the periglacial belt of Apennines. Moreover, the rare and localized O. pauciflora could benefit a genetic enrichment by hybridizing with a widespread related species. O. × colemanii is not a dead end population, but may have a role as potential reserve of adaptive variability and is an unusual stage along the speciation process. © 2012 Elsevier GmbH. All rights reserved. Introduction Orchids exhibit highly specialized flowers and pollination strategies, so botanists have assumed that pollinator specificity is the main pre-mating barrier for the maintenance of orchid species identity (Darwin, 1859). Consequently, it has been supposed that pollinator fidelity is the major selective force for determining floral shift and speciation (Van der Pijl and Dodson, 1966). In contrast with this hypothesis, Mediterranean terrestrial orchids have revealed high levels of natural hybridization (Willing and Willing, 1977) and a large diffusion of unspecialized pollination systems, that have been put in relation with the prevalence of deceptive pollination systems (Van der Cingel, 1995). Indeed, most of the Euro-Mediterranean orchids are food deceptive, their flowers do not offer energetic rewards but mimic the color of rewarding species growing in the same habitats. Because of peculiar eco-geographical heterogeneity of the Mediterranean region, distribution areas of orchids are often overlapping or intermingled. In addition, most of the deceptive orchids have similar ecological needs, so several species may live in the ∗ Corresponding author at: Department of Ecology, University of Calabria, via Bucci 6/B, I-87036 Rende (CS), Italy. Tel.: +39 984 492958; fax: +39 984 492986. E-mail address: giuseppe.pellegrino@unical.it (G. Pellegrino). 0367-2530/$ – see front matter © 2012 Elsevier GmbH. All rights reserved. http://dx.doi.org/10.1016/j.flora.2012.07.006 same habitat, bloom in the same period, and share the same pollinator fauna (Dafni, 1984). For these reasons, hybrid zones of Mediterranean orchids are scattered across their overlapping ranges, and are usually narrow, with a variable number of both parental species and hybrid individuals. At present, relatively few molecular studies have been carried out on hybrid zones of Mediterranean deceptive orchids. Recently, it has been assessed, with nrDNA sequences and AFLP markers, that a hybrid swarm of Anacamptis morio and A. papilionacea (Moccia et al., 2007) and of Orchis purpurea and O. simia (Bateman et al., 2008) consisted of F1 hybrids, suggesting their role as post-mating reproductive barrier. Conversely, an extensive introgressive hybridization has been revealed by an AFLP analysis of a hybrid zone between Ophrys lupercalis and O. iricolor, suggesting a clear signal of low floral isolation (Stökl et al., 2008). In addition, Xu et al. (2011) have demonstrated by hand pollination experiments that post-pollination barriers were effectively absent among Ophrys sphegodes, O. exaltata and O. garganica, providing that strong floral isolation prevents significant interspecies gene flow among taxa. These findings are consistent with those obtained by large-scale experimental crosses, which have pointed out that speciation in Mediterranean food-deceptive orchids has been driven by the insurgence of post-mating barriers, whereas sexually deceptive species have evolved pre-mating barriers (Scopece et al., 2007). 754 A. Luca et al. / Flora 207 (2012) 753–761 In this paper, we have expanded currently available information on hybrid zones formed by the food-deceptive species Orchis mascula and O pauciflora, which give origin to a hybrid named O. × colemanii. As originally reported (Cortesi, 1907), the hybrid zone of this species pair is consisting of a hybrid progeny with a recognizable morphology and a continuous floral color variation. Moreover, the hybrid zone does not consist of only F1 generation and, thus, not represent a late post-zygotic barrier as predicted to occur between food-deceptive orchids (Cozzolino et al., 2006; Pellegrino et al., 2000). Here we have examined a large population consisting of an admixture of parental and hybrid plants, occurring on the southern slope of Pollino Massif (Calabria region, Italy), with the main purpose of understanding the biological and evolutionary significance of the O. × colemanii hybrid zone. The goal of the present study was to characterize the genetic structure of the plants in a hybrid zone between Orchis mascula and O. pauciflora using a combination of nuclear ribosomal DNA (nrDNA) and chloroplast DNA (cpDNA) and to obtain information on reproductive biology of the examined taxa evaluating fruit and seed production. We used nrDNA analysis to identify and assign each presumed hybrid individual to a specific hybrid class (F1 , F2 or backcrosses) and cpDNA markers to assess maternal lineage of hybrids. Reproductive success was evaluated on plants left in natural conditions, whereas reproductive success of any possible bidirectional cross combinations between hybrid plants and both parental species was checked by hand pollination. Finally, levels of seed viability were established by measuring the percentage of seeds with embryo. Findings are discussed in light of the current hypothesis on the biological significance and/or evolutionary potential trajectories of plant hybrid zones. Materials and methods Study area and orchid species studied The study site is a mixed settlement of several food-deceptive orchid species, occurring on poor calcareous soils at 1400 m a.s.l., on the southern slope of Mt. “Manfriana” (Pollino National Park, Calabria region, southern Italy). In this site populations of Orchis pauciflora Tenore, O. mascula L. and their hybrid progeny, known as O. × colemanii Cortesi, co-exist, while other orchid species, O. quadripunctata Cyrillo ex Tenore and Dactylorhiza sambucina (L.) Soò, co-occur in low density. Orchis pauciflora and O. mascula have an identical chromosome number (2n = 42) (D’Emerico et al., 2002) and are phylogenetically closely related. Bateman et al. (2003) even included O. mascula in the yellow-flowered O. pauciflora group, separating it from the other purple-flowered species. Orchis pauciflora and O. mascula have similar flower morphology (convex trilobate lip, median lobe longer than lateral lobes, cylindrical, horizontal to ascendant spur without nectar) but the former has 2–8 (−15) yellow flowers and is 10–30 cm tall while the second has 15–50 red-purple flowers and is 20–60 cm tall (Delforge, 2005). They show a different distribution area, with O. mascula being a widespread species occurring on the European continent from the Canary islands to Anatolia, from North Africa to Scandianavia, while Orchis pauciflora is a narrowly distributed species with central and eastern Mediterranean distribution, along the Apennine and Balkan peninsula up to the Greek islands including Crete (Fig. 1). Their hybrid, O. × colemanii, can be morphologically variable, in term of habit, outer tepal shape, spur size and, in particular, flower color. O. × colemanii specimens show flower color polymorphism. In fact, its flower color ranges from yellowish to crimson-red to purplish (Del Prete and Miceli, 1981; Nazzaro et al., 1995). Orchis mascula and O. pauciflora are non-model mimic plants that exploit nectar-seeking bumblebee queens or solitary bees by providing general floral signals (Cozzolino et al., 2005; Nilsson, 1983), and by producing scent bouquets (Salzmann et al., 2007). The pollination biology of O. mascula has been extensively studied in the Sweden part of its distribution area by Nilsson (1983) who found that it was mainly pollinated by naïve Bombus queens, Psithyrus females and solitary bees of the genera Eucera, Andrena and Osmia searching for nectar during their first exploratory forays after hibernation. Recently, it has been reported that in Crete island insects belonging to the genera Apis and Bombus were the most fre- Fig. 1. Distribution area of Orchis mascula and O. pauciflora. Gray shadow and dotted circumference show the distribution area of widespread O. mascula, occurring from the Canaries islands to Anatolia, from North Africa to Scandinavian peninsula; dark shadow and black line indicate distribution area of narrowly distributed O. pauciflora, occurring along the Apennine and Balkan peninsula up to the Aegaean islands and Crete. A. Luca et al. / Flora 207 (2012) 753–761 quent pollinators, and among Bombus only queens were observed to pollinate O. pauciflora (Valterová et al., 2007). To enlarge information on the feature of O. × colemanii hybrid zones across the entire distribution range we invited members of the “Italian Group for the Research on Wild Orchids (GIROS)” to send us descriptions of O. × colemanii settlements found during their field trips. Phenotypic trait measurements At the peak of the blooming season we measured phenotypic traits on the second and third flowers from the bottom of the inflorescence of 15 individuals of each taxon, and used the average values from these two flowers in statistical analyses. Floral traits were measured to the nearest 1 mm using a digital caliper and were replicated on both collected flowers. Flower number was evaluated as the total number of opened flowers. Plant height was the distance from the ground to the top of the highest opened flower. Spur length was the distance between the spur mouth and the spur tip. Labellum length was the distance between the labellum tip and the spur mouth. Labellum width was the distance between the edges of the two lateral lobes. Labellum anthocyanin concentration (purple pigment) was estimated extracting the anthocyanins with 0.5-ml methanol/0.1% HCl, and determining the absorbance at 510 nm. Labellum carotenoid concentration (yellow pigment) was estimated similarly, using methylene chloride for extraction and measuring absorbance at 450 nm (Bradshaw et al., 1998). The data matrix was analyzed with Data Desk 6.3 (Velleman, 1997) and SPSS 14.0 (Norušis, 2005). Molecular analysis To characterize the genetic structure of plants from the hybrid zone, we applied internal transcribed spacers (ITSs) of nuclear ribosomal DNA (nrDNA), a powerful tool in investigating the occurrence and extent of hybridization and introgression (Pellegrino et al., 2001; Pellegrino et al., 2005; Rieseberg and Carney, 1998), and chloroplast DNA (cpDNA) for its strictly maternal inheritance in orchids (Cafasso et al., 2005). One leaf of 46 plants of O. × colemanii, 15 of O. mascula, 15 of O. pauciflora and three of the two other co-occurring orchid species were sampled and stored in silica gel. Genomic DNA was extracted using a slight modification of (cetyltrimethyl ammonium bromide) CTAB protocol of Doyle and Doyle (1987). Approx. 0.5 g of each leaf were separately pestled in a 2 ml-eppendorf vial using 500 ␮L of standard CTAB buffer, incubated at 60 ◦ C for 30 min, extracted twice adding 500 ␮L chloroform-isoamyl alcohol (24:1), precipitated with isopropanol and washed with 250 ␮L of ethanol 70%. DNA was re-suspended in 50 ␮L of distillated water. The nuclear ribosomal internal transcribed spacers (ITS1 and ITS2) and the chloroplast non-coding spacer psbK–psbI were amplified by polymerase chain reaction (PCR) using universal pairs of primers as described in Pellegrino et al. (2001) and in Chase et al. (2007), respectively. PCRs were carried out in a total reaction volume of 100 ␮L, containing approx. 10–20 ng of DNA, 100 ␮L of reaction buffer 1×, 2 mM MgCl2 , 100 mM of each dNTP, and 2.5 Units of BioTaqTM DNA Polymerase (Bioline Inc., Boston, MA, USA), and 0.2 mM of each primer (MWG-Biotech AG, Ebersberg, Germany), The thermocycling profile consisted of an initial denaturation step at 94 ◦ C for 3 min, followed by 30 cycles with 30 s at 94 ◦ C, 30 s at 55 ◦ C, and 2 min at 72 ◦ C. PCRs were performed on a PTC-100 Thermal Cycler (MJ Research Inc., Watertown, MA, USA). PCR fragments were purified by QIAquick PCR purification kit (Qiagen S.p.A., Milan, Italy) to remove unincorporated primers and dNTPs. Amplification 755 products were electrophoretically separated on a 2% agarose gel (Methaphore, FMS), compared to a 100 base pair (bp) ladder (Pharmacia Biotech) as the molecular weight marker, stained with ethidium bromide and photographed using a Kodak digital camera. Plastidial and nuclear amplified fragments of three individuals for each parental species and the other two sympatric orchids were sequenced in both directions using a modification of the Sanger dideoxy method as implemented in a double stranded DNA cycle sequencing system with fluorescent dyes. Sequence reactions were then run on a 373 A Applied Biosystems Automated DNA sequencer (Applied Biosystems, Foster City, CA, USA). Nuclear sequences were examined using GeneJockey to find a restriction site that would distinguish them using Polymerase Chain Reaction-Restriction Fragment Length Polymorphism (PCRRFLP). This approach allows the examination of a heterozygous individual (e.g., a hybrid) without the necessity of cloning and subsequently sequencing several ITS clones (heterozygous individuals give overlapping traces from direct sequencing that are often difficult to interpret). Restriction enzyme TaqI, which cuts at 5′ -TC/GA-3′ , differentiated the putative parental taxa due to the presence of a C/T substitution about 24 base (TCGA in O. pauciflora, CCGA in O. mascula) pairs into the ITS 2 sequence; while SmaI, which cuts at 5′ CCC/GGG-3′ , showed a nucleotide substitution A/G, about 195 base pairs (CCCGAG in O. pauciflora, CCCGGG in O. mascula) into the ITS 2 sequences. Sequences of other sympatric orchids, O. quadripunctata and D. sambucina, did not show these restriction sites. Thus, the PCR fragments of all samples (100 ng) were digested in a final 20 ␮L volume with the selected restriction endonuclease (1 U/(g DNA), according to the manufacturer’s instructions (Fermentas), in particular incubated for 3 h at 30 ◦ C for SmaI and 65 ◦ C for TaqI. The fragments were electrophoretically separated on a 3% low melting agarose gel (Methaphore, FMS), compared to a 100 base pair (bp) ladder (Pharmacia Biotech) as the molecular weight marker, stained with ethidium bromide and photographed using a Kodak digital camera. The relative amounts of DNA were estimated on digital photos analyzing them with the Biomax 10 image analysis software (Kodak Digital Science, EDAS, USA). Pollen transfer To ascertain if fruit developed by hybrid plants could have been produced by pollen transferred by different donors, that is from hybrid plants and/or from parental species, we marked 5 individuals O. × colemanii and left them in natural condition. Profiles of the ITS-containing fragments were assessed for each plant following the protocol described above. In June, capsules were collected and seeds of the central part were used for molecular analysis. Seeds were observed under a binocular microscope, and approx. 50 viable seeds (which means seeds with embryo) from each capsule were collected and transferred into single 2 ml-eppendorf to extract their DNA. Ribosomal DNA was amplified and the PCR fragments of all samples were digested using SmaI and TaqI, electrophoretically separated and photographed following the protocol described above. Reproductive success and hand pollination In accordance with our goals, we performed field experimental crosses to gain information on the existence of pre- and postzygotic barriers, as hybrid sterility or fertility selection (Lau et al., 2005; Wolf et al., 2001), and on the reproductive success of parental species and hybrids. To test natural reproductive success we marked at the beginning of flowering period 20 plants of O. × colemanii and 50 plants of each parental species and left them in natural condition. 756 A. Luca et al. / Flora 207 (2012) 753–761 Hand pollination treatments were conducted to evaluate the levels of reproductive fitness reached by any bi-directional possible mating between parental species and hybrids. To this end, 10 plants of O. × colemanii and of each parental species with unopened flowers were bagged with a fine-meshed cloth to exclude pollinators. For hand-pollination, the cover was removed and five randomly selected flowers on each plant were marked with cotton threat and manually pollinated using a toothpick with the pollinia of the same (intrataxon crosses) and of the other taxon (intertaxon crosses). After treatments, plants were bagged again to prevent any further natural pollination or predation. In addition, two flowers on each plant were covered without manipulation to test for spontaneous autogamy. In June, the number of produced fruits was counted for both spontaneous and experimental crosses, and the ratio between the number of fruit/flowers (of the treated flowers in an inflorescence) was determined. Ripe fruits were collected and stored in silica gel in order to prevent their degradation. Capsules were opened longitudinally with a razor blade. To ascertain the presence of viable embryos, at least 1000 seeds for each fruit were removed from the center of the capsule and observed under an optical microscope (100×). Seeds were assigned to two categories (viable and unviable seeds) due to presence or absence of viable embryos. We quantified the pre-zygotic reproductive isolation following the form: RIpre-zygotic = 1 − % fruit set in interspecific crosses . % fruit set in intraspecific crosses Post-zygotic reproductive isolation was defined as the proportion of viable embryos (i.e., viable seeds) produced in interspecific experimental pollinations, relative to the proportion of viable embryos produced in intraspecific crosses. That is: embryo mortality = 1 − % viable seeds produced in interspecific crosses . % viable seeds produced in intraspecific crosses Values of these indices vary from zero (no isolation) to one (total isolation) (Moyle et al., 2004). We evaluated hybrid sterility by performing manual crossing experiments between hybrids and their parental species. Hybrid sterility indices were calculated at fruit production and viable seed production stage; the former index was defined as the ratio of fruits produced in pollination attempts performed using hybrid pollen on both parental plants: HSf = 1 − number of fruits produced . number of pollinated flowers The second index was defined as the ratio of viable seeds produced in the same previous crosses: HSs = 1 − viable seeds . total number of counted seeds Values of these indices vary from zero (no isolation) to one (total isolation): Scopece et al. (2008). Fisher exact tests were used to compare the rate of fruit set between the different experiments. The statistical program package SPSS (version 10, SPSS Inc. Chicago, USA) was used. Results Orchis x colemanii survey We have received a relevant contribution from members of the “Italian Group for the Research on Wild Orchids (GIROS)”, who signaled us many localities, not reported in the scientific literature, where are occurring O. × colemanii zones. Reports were completed by details on the main features of the site and of orchid settlement. This information has allowed us to ascertain that many narrow hybrid zones are occurring across the entire distribution area of O. pauciflora (Appendix 1). In general, it has been highlighted that most of the hybrid zones are located on the calcareous slopes of Apennine chain, usually above 1000 m elevation. Interestingly, the co-occurrence of parental species has been ever observed. Few exceptions have noticed that O. pauciflora is mixed with hybrid plants and O. mascula plants occur nearby, within few hundred meters. Phenotypic trait measurements Morphological analysis showed that hybrids exhibited phenotypic characters more or less intermediate between the two parental species (Fig. 2). As regards a structure of relevant diagnostic value, the labellum size (width and length) of the hybrid plants (14.58 mm ± 0.188; 13.34 mm ± 0.178) was intermediate between O. mascula (13.22 mm ± 0.208; 13.87 mm ± 0.198) and O. pauciflora (15.41 mm ± 0.235; 13.01 mm ± 0.169) – Fig. 2(D and E), as spur length in O. × colemanii (16.12 mm ± 0.342) was intermediate between parental species (13.48 mm ± 0.428 in O. mascula and 19.21 mm ± 0.318 in O. pauciflora – Fig. 2F). In addition O. × colemanii showed a continuous flower color variation (Fig. 3) ranging from red-purple flowers of O. mascula to yellow flowers of O. pauciflora. There were hybrids with high value of anthocyan and low value of carotenoid (more similar to O. mascula flowers) and hybrids with high value of carotenoid and low value of anthocyan (more similar to O. pauciflora flowers) and a lot of hybrids with intermediate concentrations of both pigments (Fig. 3). Molecular analysis The ITS-containing fragments obtained from the parental species and hybrids were approximately 280 (ITS 1) and 300 (ITS 2) bp in length. As expected, ITS 2 of O. pauciflora and O. mascula differ in the presence of different recognition sites for the restriction enzymes TaqI and SmaI. Indeed, ITS 2-containing fragments digested with TaqI showed a single restriction site in O. pauciflora (with two fragments approx. 180 bp and 120 bp long) and no site in O. mascula (Fig. 4a). The ITS 2-containing fragments digested with SmaI showed a single restriction site in O. mascula (with two fragments approx. 160 bp, and 140 bp long) and no site in O. pauciflora (Fig. 4b). All the 46 individuals of O. × colemanii in the study exhibited a direct additive inheritance of these profiles, their digested fragments produced the combination of diagnostic profiles obtained for both O. pauciflora and O. mascula (Fig. 4a and b). However, the restriction profiles differ in parental band intensity among hybrids. About half of hybrids (24 individuals) displayed a balanced proportion (1:1) of nrDNA of both parental species, while the remaining (22 individuals) showed a higher amount of ribosomal DNA from one parental species than from the other one. In detail, 6 and 11 specimens showed a preponderance (approximately 3:2–2:1) of O. mascula and O. pauciflora, respectively, 3 had a preponderance (approximately 7:1) of O. mascula and 2 of O. pauciflora. Chloroplast DNA amplification revealed a length polymorphism in the psbK. Indeed, the psbK amplified fragment of O. mascula was approximately 500 bp long while that of O. pauciflora was approximately 480 bp long. Thanks to this difference, we established that 19 out of 46 hybrids possessed the O. pauciflora plastidial DNA and the remaining 27 that of O. mascula. At the same time, we did not find evidence of introgression into the parental species. Also in this case there is no correlation between maternal inheritance and flower color. A. Luca et al. / Flora 207 (2012) 753–761 757 Fig. 2. Morphometric variation among Orchis mascula (white box), O. × colemanii (gray box) and O. pauciflora (black box). (A) flower number; (B) plant height (cm); (C) inflorescence height (cm); (D) labellum width (mm); (E) labellum length (mm) and (F) spur length (mm). The outlined central box depicts the middle 50% of the data extending from upper to lower quartile; the horizontal bar is at the median. Vertical bars indicate standard errors. Pollen transfer Three out of five individuals of the marked O. × colemanii showed a preponderance of O. mascula nrDNA, while two had approximately equal proportion (1:1) of parental nrDNA. From these five specimens were collected a total of 23 capsules (48.2% of flowers). PCR performed on DNA extracted from seeds gave amplification, and thus PCR fragments of all samples were digested with the selected restriction endonucleases. Restriction analysis showed that all samples had the diagnostic profiles obtained for parental species. Four of the 10 capsules of two plants showing equal proportion of parental nrDNA had O. mascula nrDNA preponderance, 4 of O. pauciflora and two equal proportions; 9 of the 13 capsules of three plants having more O. mascula nrDNA showed O. mascula preponderance, while 4 had an equal proportion (Table 1). Thus, 13 flowers received pollinia from O. mascula, 8 from O. pauciflora and two from F1 hybrids. Reproductive success Natural levels of fruit set in open-pollinated populations were 52.1% for O. mascula, 50.7% for O. pauciflora and 48.2% for O. × colemanii. There was no significant difference in fruit set between species (2 = 0.21, df = 2, P = 0.84). Fruit set percentages derived from artificial hybridizations (80–86.7%) were slightly higher (Fisher exact test: 2 = 2.27, P < 0.100) than those obtained from F2 hybrid generations (62.5%), and from artificial backcrosses (75%): Table 2. If we consider directionality of artificial backcrosses, there was no significant differences (Fisher exact test: 2 = 0.27, P > 0.99) between artificial hybridizations and artificial backcrosses (90%) in which O. × colemanii plants gave the pollinia (Table 2). Finally, percentages of viable seeds observed were statistically not different among all hand manipulations ranging from 85.9 (O. pauciflora × O. mascula) to 97.9 (O. × colemanii × O. mascula): Table 2. Hand manipulation experiments showed the absence of pre- and post-zygotic reproductive barriers; indeed pre-zygotic isolation index (RIpre-zygotic) was 0.19 indicating the absence of pre-zygotic barriers. Moreover O. × colemanii was not affected by some sort of hybrid mortality or sterility: values of hybrid mortality (0.24) and hybrid sterility (0.28) showed the absence of post-zygotic barriers. Discussion Molecular studies have shown that homoploid hybrid populations of Mediterranean orchids are composed of two main types of progenies. In particular, the hybrid swarms of food deceptive orchids (e.g., Anacamptis and Orchis) consist quite exclusively of F1 individuals, which have usually an uniform, intermediate morphology. Because most of them are unfertile, F1 hybrids function as a late post-zygotic barrier (Moccia et al., 2007; Bateman et al., 2008; Jacquemyn et al., 2012). Differently, sympatric species of Serapias have been proved to undergo introgressive hybridization (Bellusci et al., 2010), while sympatric populations of Ophrys have shown both F2 hybrid generations (Stökl et al., 2008) and no introgressive hybridization (Xu et al., 2011). In this scenario, we may affirm that the features of the hybrid zone of O. × colemanii are totally different from those of other Mediterranean food deceptive orchid hybrid zones. Our molecular and morphological analyses have confirmed the hybrid origin of all 758 A. Luca et al. / Flora 207 (2012) 753–761 Table 1 Ribosomal DNA amount in fruits of Orchis × colemanii. Restriction analysis of DNA extracted from seeds of 23 fruits collected from five specimens of O. × colemanii showing different proportion of nrDNA. O. × colemanii specimens 2 3 Parental nrDNA ratio 1:1 Preponderance of O. mascula Fruits 10 13 Fig. 3. Comparison of pigments, anthocyan (A) and carotenoid (B) extracted from labellums of Orchis mascula, O. pauciflora and O. × colemanii. Parental nrDNA in fruits Preponderance of O. mascula Preponderance of O.pauciflora 1:1 4 9 4 0 2 4 the specimens reputed to be O. × colemanii, accounting once again for the clear morphological difference existing between the hybrid progenies and the parental plants. More relevantly, the molecular approaches have proved that the hybrids consist of several classes of hybrids, since hybrid specimen posses either a balanced amount of parental nrDNA either several unbalanced combinations of both parental DNAs (Table 1). Another striking feature of O. × colemanii is the absence of effective pre- and post-zygotic reproductive barriers either between hybrids either between them and both parental species. In this respect, we have found that of three fruits ripen on one and the same plant each contained a different seed set: one identical to the maternal rDNA combinations, others with changed combinations, clearly produced by pollen carried in any direction among all the co-occurring hybrid or parental plants (Table 1). The morphological distinctness and the continuous flower color variation of O. × colemanii are typical features observed also by other authors (Cozzolino et al., 2006; Nazzaro et al., 1995; Pellegrino et al., 2000) and by Italian amateur orchidologists in almost all known populations. Moreover, plants of the hybrid zones regularly grow on arid, calcareous slopes upwards 1000 m above sea level and co-occur with parental species (Appendix 1). So, it is reasonable to argue that these many hybrid zones are stable and of ancient, contemporaneous origin across the entire area of overlap of the parental species’ distribution ranges. In this perspective, we will consider two main evidences. Firstly, the greatly different distribution and ecological ranges of parental species (see Fig. 1) suggest their longstanding divergence. Indeed, they occur separately in two large Mediterranean isles with different geo-climatic history; precisely, in Sardinia has been found only O. mascula and in Crete only O. pauciflora (Delforge, 2005). Secondly, it is not trivial that phylogenetic analyses of Orchidinae show an early divergence between O. pauciflora and the other members of the O. mascula group (Bateman et al., 2003). In accordance, evaluation of genetic distances based Fig. 4. Additive profile of nrDNA in Orchis × colemanii and parental species. Gel electrophoresis of ITS 2 TaqI digestions (A) and ITS 2 SmaI digestions (B) showing the presence of characteristic fragments of ITS region of Orchis mascula (lane M), O. × colemanii (lane C) and O. pauciflora (lane P). Molecular 100 bp ladder (line L). A. Luca et al. / Flora 207 (2012) 753–761 Table 2 Fruit set in artificial hybridization (crosses between Orchis mascula and O. pauciflora), backcrosses (crosses between parental species and O. × colemanii), and F2 hybrid generations (crosses between O. × colemanii specimens). In the first column for each hand-pollination the species above is pollinia donor, the species below is pollinia receiver. (NP, number of plants observed; NF, number of flowers observed; FP, number of fruits produced; %F, percentage of fruits produced; %E, percentage of seeds with embryo.) NP NF FP %F %E Artificial hybridization O. mascula × O. pauciflora O. pauciflora × O. mascula 4 4 15 15 12 13 Mean 80.00 86.65 83.33 86.30 85.90 86.10 Artificial backcrosses O. mascula × O. × colemanii O. × colemanii × O. mascula 4 4 10 10 O. pauciflora × O. × colemanii O. × colemanii × O. pauciflora 4 4 10 10 6 9 Mean 6 9 Mean 60.00 90.00 75.00 60.00 90.00 75.00 94.10 97.90 96.00 96.90 93.65 95.28 F2 hybrid generations O. × colemanii × O. × colemanii 3 8 5 62.50 95.35 on ITS sequences has shown that these two species have a genetic distance higher than those of many other Mediterranean orchid species pairs (Scopece et al., 2007). Overall, these evidences suggest that the origin of so many narrow hybrid zones of O. × colemanii could be of ancient nature. There is an old, large consensus on the relevant role played by geo-climatic changes and human disturbance in the insurgence and establishment of hybrid zones (Comes and Kadereit, 1998; Stebbins, 1959). Moreover, it has been observed that current distribution of species and hybrid zones in both the Old and the New World continents may have originated during the Pleistocene glaciations (Barton and Hewitt, 1985). Noticeable, Apennines were affected by glaciations and glacier valleys and moraines are present until the southernmost Pollino Massif (Acquafredda and Palmentola, 1986). On average, the ice border was at 1200–1300 m a.s.l., the same altitude at which occur most of O. × colemanii populations. Thus, it is reasonably to hypothesize that these populations still live in the same places where they have originated by secondary contact which occurred in the periglacial belt of Apennines. The theory of adaptive speciation predict that hybrid zones could became larger or narrower under the influence of introgressive hybridization or reinforcement of reproductive isolation, respectively (Mayr, 1942). Thus, to explain the evident stability of many hybrid zones appropriate hypotheses have been proposed. In particular, the dynamic equilibrium hypothesis assumes the existence of an equilibrium between gene flow and selection against the hybrids, while the bounded superiority hybrid hypothesis retains that hybrids are more fit than parental species in the restricted regions where they occur (Moore, 1977). In this perspective, we think that features of O. × colemanii are apparently incompatible with the equilibrium dynamic hypothesis, given the apparent lack of gene flow and hybrid selection. Conversely, their persistence could be accounted by the hypothesis of bounded hybrid superiority. Indeed, the high comparable levels of reproductive success, found in all the experimental crosses, strongly suggest a relaxed pre-zygotic selection against hybrids. Similarly, it has been proved that hybrids are regularly visited by pollinators independently from the emission of scent with an intermediate composition in respect to parental taxa (Salzmann et al., 2007). On the other hand the existence of a hybrid superiority could be undetectable as long as ecological conditions are stable. Even an accurate scrutiny of the literature has confirmed the rarity of a plant hybrid zone with the overall features like that of O. × colemanii. Indeed, we are able to report only a narrow hybrid zone between two species of Pitcairnia (Bromeliacae), growing on 759 the Pão de Açucar in Rio de Janeiro (Brazil), in which hybrids have shown a fruit set, seed set and seed germination fitness equivalent to the parental taxa (Wendt et al., 2001). Since both taxa are two narrow endemics, the authors suggested that hybridization could have a positive role contributing to the expansion of individuals on the slope of the mountain. An alternative hypothesis, worthy to be discussed, is the likelihood that O. × colemanii is a nucleus for a totally new species formation or even a yet incipient species. Previous authors have considered O. × colemanii either a quite stable taxon of hybrid origin (Del Prete and Miceli, 1981) and have listed it as an endemic form (Nazzaro et al., 1995). On the one hand, although underestimated, homoploid speciation was assumed to be the source of many plant speciation events (Gross and Rieseberg, 2005), in which a long time is needed to bring about appropriate reproductive barriers. We notice that although the high hybrid fertility could sustain such possibility, no ecological preference of O. × colemanii separating its habitats from those of the parents has been documented and, also, information received by orchids amateurs indicates that one parental species (O. pauciflora) is ever co-occurring and O. mascula is easily found nearby. A previous study (Salzmann et al., 2007) showed that O. × colemanii exhibits a different odor bouquet from those of the two parent species, which could lead to a shift of pollination. It seems that just a change in pollinators may play a key role, being the driving force of speciation by homoploid hybridization (Chase et al., 2010). It has recently been demonstrated, especially in orchids with nectar spurs (Paun et al., 2007) or sexual deceptive orchids, as in the case of the genus Ophrys (Vereecken et al., 2010), that differences in the composition of floral odors have created differences in the attraction of different groups of insects, thus creating strong relationships between orchids and pollinators. In conclusion, this study demonstrates that the hybrid zone of O. × colemanii might have a biological and evolutionary significance different from those attributed to hybrid zones of other Mediterranean deceptive orchids, although its own actual significance is very difficult to be envisaged. Certainly, it does not appear a dead end population and, so, could represent a potential reserve of adaptive variability, as seem to be typical of zones with several hybrid generations (Anderson, 1948; Rieseberg, 1995). In any case, O. × colemanii is an unusual case of frequently occurring fertile hybrids with a continuous phenotypic variation between the two parental species, an interesting step along the speciation process. As recently stressed by Mallet (2008), hybridization and introgression may often lead to a continuum of phenotypic and genotypic variation over a large geographical scale on the one hand and, locally in sympatry on the other hand. The occurrence of such a continuum is seen as evidence that the original vision of Darwin (1859, 1877) on speciation might be reevaluated and that the speciation process is occurring all around us. Acknowledgements We are grateful to members of the “Italian Group for the Research on Wild Orchids (GIROS)” for their valuable help informing about O. × colemanii settlements. Appendix A. Appendix 1 Description of Orchis × colemanii settlements. Word file showing population name, elevation (m above sea level), number of hybrid, presence of parental species, color and collector of known populations of O. × colemanii. 760 A. Luca et al. / Flora 207 (2012) 753–761 Region Populations Elevation m a.s.l. Number of hybrids Parental species color Collector Abruzzo Basilicata Prati Tivo (TE) Moliterno (PZ) Castelluccio (PZ) Monte Manfriana (CS) Sassano (SA) Passo Padula (SA) Monte Maio (FR) Prato di Campoli (FR) Monte Ode (RI) 1200 900 1200 1200 1200 900 900 1200 900 10 20 10–20 ∼200 ∼200 10 10 5 3 From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish Romolini Romolini Romano Gargano Nazzaro Romolini Romolini Arrighi D’Elia From yellowish to purplish From yellowish to purplish Romolini Klaver Purplish Klaver From yellowish to purplish More purplish From yellowish to purplish Purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish Purplish From yellowishto purplish Klaver Klaver Klaver Klaver Antonetti Antonetti Mazzoni G.Pacifico G.Pacifico G.Pacifico Romolini Antonetti G.Pacifico M.Pacifico Marchetti Antonetti Antonetti Antonetti Antonetti Antonetti Antonetti Mazzoni G.Pacifico G.Pacifico Viviani Antonetti Klaver Cosoli Bizzarri Bizzarri Bizzarri Bizzarri Calabria Campania Latium Marche Tuscany Umbria 800 700 10 6 Monte Vermenone (MC) 1220 10 Monte Nerone (PU) Monte Petrano (PU) Monte Catria (PU) Monte Paganuccio (PU) Campocecina (MS) Sassalbo (MS) Monte Borla (MS) Foce Pianza (MS) Massa (MS) Carrara (MS) S. Maria Giudice (LU) Stazzema (LU) 1200 950 950 975 1100 850 1300 1200 1200 850 1000 950 ∼400 ∼50 ∼500 10–20 ∼100 ∼100 10 ∼20 5 ∼30 5 ∼50 Both Both Both Both Both Both Both Both O. pauciflora (O. mascula at 500 m) Both More O. pauciflora than O. mascula More O. pauciflora than O. mascula Both Both Both Both Both Both Both Both Both Both Only O. pauciflora Both Monte Piglione (LU) Monte Matanna (LU) Monte Nona (LU) Monte Prana (LU) Monti Pisani (LU) Prato Fiorito (LU) Monte Gabberi (LU) Minucciano (LU) Pescaglia (LU) 1100 1150 1000 1100 900 1100 1100 800 900 ∼100 ∼50 ∼50 ∼50 10 ∼100 10 1 5 Both Both Both Both Only O. pauciflora Both Both Both Both From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish Yellowish From yellowish to purplish From yellowish to purplish From yellowishto purplish From yellowish to purplish 900 ∼200 Both From yellowish to purplish 1000 1000 1200 1000 ∼200 ∼800 ∼100 ∼150 Both Both Both Both From yellowish to purplish From yellowish to purplish From yellowish to purplish From yellowish to purplish Monte Flavio (RM) Monte Pallone (AN) Monte Cucco (PG) Monte Macchialonga (PG) Monte il Monticello (PG) Monte di Campi (PG) Monte Lungo (PG) References Acquafredda, P., Palmentola, G., 1986. Il glacialismo quaternario nell’Italia meridionale. Biogeografia 10, 13–18. Anderson, E., 1948. Hybridization of the habitat. Evolution 2, 1–9. Barton, N.H., Hewitt, G.M., 1985. Analysis of hybrid zones. Annu. Rev. Ecol. Syst. 16, 113–148. Bateman, R.M., Hollingsworth, P.M., Preston, J., Yi-Bo, L., Pridgeon, A.M., Chase, M.W., 2003. Molecular phylogenetics and evolution of Orchidinae and selected Habenariinae (Orchidaceae). Biol. J. Linn. Soc. 142, 1–40. Bateman, R.M., Smith, R.J., Fay, M.F., 2008. Morphometric and population genetic analyses elucidate the origin, evolutionary significance and conservation implications of Orchis × angusticruris (O. purpurea × O. simia), a hybrid orchid new to Britain. Bot. J. Linn. Soc. 157, 687–711. Bellusci, F., Pellegrino, G., Palermo, A.M., Musacchio, A., 2010. Crossing barriers between the unrewarding Mediterranean orchids Serapias vomeracea and S. cordigera. Plant Species Biol. 25, 68–76. Bradshaw, H.D., Otto, K.G., Frewen, B.E., McKay, J.K., Schemske, D.W., 1998. Quantitative trait loci affecting differences in floral morphology between two species of monkeyflower (Mimulus). Genetics 149, 367–382. Cafasso, D., Widmer, A., Cozzolino, S., 2005. Chloroplast DNA inheritance in the orchid Anacamptis palustris using single-seed polymerase chain reaction. J. Hered. 96, 66–70. Chase, M.W., et al., 2007. A proposal for a standardised protocol to barcode all land plants. Taxon 56, 295–299 (17 authors). Chase, M.W., Paun, O., Fay, M.F., 2010. Hybridization and speciation in angiosperms: a role for pollinator shifts? BMC Biol. 8, 45–47. Comes, P.T., Kadereit, J.W., 1998. The effect of quaternary climatic changes on plant distribution and evolution. Trends Plant Sci. 3, 432–438. Cortesi, F., 1907. New or rare orchids. Annali Botanici 5, 539–545. Cozzolino, S., Schiestl, F.P., Müller, A., De Castro, O., Nardella, A.M., Widmer, A., 2005. Evidence for pollinator sharing in Mediterranean nectar-mimic orchids: absence of premating barriers? Proc. Roy. Soc. Lond. B 272, 1271–1278. Cozzolino, S., Nardella, A.M., Impagliazzo, S., Widmer, A., Lexer, C., 2006. Hybridization and conservation of Mediterranean orchids: should we protect the orchid hybrids or the orchid hybrid zones? Biol. Conserv. 129, 14–23. Dafni, A., 1984. Mimicry and deception in pollination. Annu. Rev. Ecol. Syst. 15, 259–278. Darwin, C., 1859. On the Origin of Species by Means of Natural Selection or the Preservation of Favoured Races in the Struggle for Life. J. Murray, London. Darwin, C., 1877. The Various Contrivances by which Orchids are Fertilised by Insects. J. Murray, London. Delforge, P., 2005. Orchids of Europe North Africa and the Middle East. Timber Press, Portland. Del Prete, P., Miceli, C., 1981. Orchis colemanii Cortesi natural hybrid between Orchis mascula (L.) L. and Orchis pauciflora Ten. Giorn. Bot. Ital. 115, 396–397. D’Emerico, S., Cozzolino, S., Pellegrino, G., Pignone, D., Scrugli, A., 2002. Heterochromatin distribution in selected taxa of the 42-chromosomes Orchis s.l (Orchidaceae). Caryologia 55, 55–62. Doyle, J.J., Doyle, J.L., 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem. Bull. 19, 11–15. Gross, B.L., Rieseberg, L.H., 2005. The ecological genetics of homoploid hybrid speciation. J. Hered. 96, 241–252. Jacquemyn, H., Brys, R., Honnay, O., Roldàn-Ruiz, I., Lievens, B., Wiegand, T., 2012. Nonrandom spatial structuring of orchids in a hybrid zone of three Orchis species. New Phytol. 193, 454–464. A. Luca et al. / Flora 207 (2012) 753–761 Lau, C.P.Y., Ramsden, L., Saunders, R.M.K., 2005. Hybrid origin of Bauhinia blakeana (Leguminosae: Caesalpinioideae), inferred using morphological, reproductive, and molecular data. Am. J. Bot. 92, 525–533. Mallet, J., 2008. Hybridization, ecological races and the nature of species: empirical evidence for the ease of speciation. Phil. Trans. R. Soc. Lond. B 363, 2971–2986. Moccia, M.D., Widmer, A., Cozzolino, S., 2007. The strength of reproductive isolation in hybridizing food deceptive orchids. Mol. Ecol. 16, 2855–2866. Moyle, L.C., Olson, M.S., Tiffin, P., 2004. Patterns of reproductive isolation in three angiosperm genera. Evolution 58, 1195–1208. Moore, W.S., 1977. An evaluation of narrow hybrid zones in vertebrates. Q. Rev. Biol. 52, 263–267. Mayr, E., 1942. Systematics and the Origin of Species. Columbia University Press, New York. Nazzaro, R., Menale, B., Di Novella, N., 1995. Le Orchidaceae della zona occidentale del Vallo di Diano (Salerno). Webbia 50, 25–35. Norušis, M.J., 2005. SPSS 14.0 Advanced Statistical Procedures Companion. Upper Saddle, Prentice Hall, New York. Nilsson, L.A., 1983. Anthecology of Orchis mascula (Orch.). Nord. J. Bot. 3, 157–179. Paun, O., Fay, M.F., Soltis, D.E., Chase, M.W., 2007. Genetic and epigenetic alterations after hybridization and genome doubling. Taxon 56, 649–656. Pellegrino, G., Caputo, P., Cozzolino, S., Menale, B., Musacchio, A., 2000. Molecular characterization of a hybrid zone between Orchis mascula and O. pauciflora in Southern Italy. Biol. Plant. 43, 13–18. Pellegrino, G., Cozzolino, S., Grünanger, P., Musacchio, A., 2001. Ribosomal DNA (ITS) as a molecular tool in the study of orchid hybridization. J. Eur. Orch. 33, 369–376. Pellegrino, G., D’Emerico, S., Musacchio, A., Scrugli, A., Cozzolino, S., 2005. Confirmation of hybridization among sympatric insular populations of Orchis mascula and O. provincialis. Plant Syst. Evol. 251, 131–142. Rieseberg, L.H., 1995. The role of hybridization in evolution: old wine in new skins. Am. J. Bot. 82, 944–953. Rieseberg, L.H., Carney, S.E., 1998. Plant hybridization. New Phytol. 140, 599–624. 761 Salzmann, C.C., Cozzolino, S., Schiestl, F.P., 2007. Floral scent in food-deceptive orchids: species specificity and sources of variability. Plant Biol. 9, 720–729. Scopece, G., Musacchio, A., Widmer, A., Cozzolino, S., 2007. Patterns of reproductive isolation in Mediterranean deceptive orchids. Evolution 61, 2623–2642. Scopece, G., Widmer, A., Cozzolino, S., 2008. Evolution of postzygotic reproductive isolation in a guild of deceptive orchids. Am. Nat. 171, 315–326. Stebbins, G.L., 1959. The role of hybridization in evolution. Proc. Am. Philos. Soc. 103, 231–251. Stökl, J., Schlüter, P.M., Stuessy, T.D., Paulus, H.F., Assum, G., Ayasse, M., 2008. Scent variation and hybridization cause the displacement of a sexually deceptive orchid species. Am. J. Bot. 95, 472–481. Valterová, I., Kunze, J., Gumbert, A., Luxová, A., Liblikas, I., Kalinová, B., Borg-Karlson, A., 2007. Male bumble bee pheromonal components in the scent of deceit pollinated orchids; unrecognized pollinator cues? Arthropod.–Plant Interact. 1, 137–145. Van der Cingel, N.A., 1995. An Atlas of Orchid Pollination. European Orchids, Balkema, Rotterdam. Van der Pijl, L., Dodson, C.H., 1966. Orchid Flowers: Their Pollination and Evolution. University of Miami Press, Coral Gables. Velleman, P.F., 1997. DataDesk Ver 6.0 Statistic Guide. Data Description, Inc., Ithaca. Vereecken, N.J., Cozzolino, S., Schiestl, F.P., 2010. Hybrid floral novelty drives pollinator shift in sexually deceptive orchids. BMC Evol. Biol. 10, 103–105. Wendt, T., Ferreira Canela, M.B., Gelli De Faria, A.P., Iglesias Rios, R., 2001. Reproductive biology and natural hybridization between two endemic species of Pitcairnia (Bromeliaceae). Am. J. Bot. 88, 1760–1767. Willing, B., Willing, E., 1977. Bibliographie über die Orchideen Europas und der Mittelmeerländer 1744–1976. Willdenowia 11, 1–325. Wolf, D.E., Takebayashi, N., Rieseberg, L.H., 2001. Predicting the risk of extinction through hybridization. Conserv. Biol. 15, 1039–1053. Xu, S., Schlüter, P.M., Scopece, G., Breitkopf, H., Gross, K., Cozzolino, S., Schiestl, F.P., 2011. Floral isolation is the main reproductive barrier among closely related sexually deceptive orchids. Evolution 65, 2606–2620.