Bruno Corrias’s research while affiliated with University of Sassari and other places

The hybrid origin of the western Mediterranean orchid Dactylorhiza insularis was demonstrated by genetic markers. Allozyme data showed that throughout its range D. insularis has an allotriploid constitution and reproduces apomictically. The parental species of D. insularis were identified as D. romana and D. sambucina ; they contributed 2 alleles and 1 allele, respectively, at the allozyme loci studied. The maternal species of D. insularis was D. romana , as inferred from cpDNA ( trn L(UAA) intron). High genetic similarities were found when comparing present populations of D. romana and D. sambucina with their respective genomes 'frozen' in D. insularis . Dactylorhiza insularis showed fixed (or nearly fixed) heterozygosity at 11 out of the 19 loci studied, and poor genetic variation: eight multilocus genotypes were detected at allozyme level. No multilocus genotype differs from the most similar one by more than one allele substitution. All D. insularis individuals showed the same cpDNA haplotype (I), regardless of their geographic origin and multilocus genotype. The I haplotype is similar, but not identical to that found in D. romana (R). No recurrent formation of D. insularis was observed in hybrid zones between D. romana and D. sambucina , where diploid sexual hybrids (F1; Fn, backcrosses) were detected. Available data agree with a single origin for D. insularis , which possibly occurred in the present postglacial, when D. romana and D. sambucina , expanding from their glacial refugia, came into contact. The genetic homogeneity found between D. romana and D. markusii , both from their locus classicus, indicates that the latter is a junior synonym of D. romana ; on the other hand, D. romana and D. sambucina are well differentiated species (D _Nei = 0.59).

Genetic divergence between population samples ofOrchis laxiflora and ofO. palustris from various European locations was studied by electrophoretic analysis of 25 enzyme loci. An average genetic distance of DNei = 1.24 was found between the two taxa, with 12 out of 25 loci showing alternative alleles (diagnostic loci). Genetic heterogeneity was observed within bothO. laxiflora andO. palustris, when northern and southeastern populations were compared, being lower in the former taxon (D = 0.06), than in the latter (D = 0.16). Karyologically, 2n = 36 was found for bothO. laxiflora andO. palustris. O. laxiflora andO. palustris produce hybrids, described asO. ×intermedia. Genotype analysis of several sympatric samples showed the presence of hybrid zones, including F1 hybrids and, in low proportions, recombinant classes, putatively assigned to Fn and backcrosses, as well as a few introgressed individuals of both taxa. These data indicate that hybrids are only partially fertile, with a very limited mixing up of the two parental gene pools; this is also shown by the lack of significant lowering of genetic distances when sympatric and allopatric heterospecific samples are compared. Accordingly,O. laxiflora andO. palustris form a syngameon; nevertheless they can be considered as good taxonomic species, with virtually distinct gene pools, which evolve independently. The genetic variability inO. laxiflora andO. palustris is remarkably low ( e = 0.05 and e = 0.02, respectively). In particular, nearly complete absence of polymorphic loci was found inO. palustris from northcentral Europe. Two hypotheses are considered to explain the low genetic variability of this endangered species.

The genetic differentiation ofOrchis papilionacea from Italy, Corsica, Sardinia, Sicily, and Greece was analyzed by isozyme studies. No significant differences at any of the 28 loci tested were observed between samples of the putative subspeciespapilionacea from Italy,grandiflora from Sicily andheroica from Greece, or by comparing sympatricpapilionacea-like andgrandiflora-like subsamples from Sardinia and Corsica. The genetic relationships found among the populations studied reflect more their geographic origin than their taxonomic assignment;papilionacea, grandiflora, andheroica are, therefore, at least in the study area, not to be considered as subspecies, but as simple morphs, controlled by few genes (possibly two) with Mendelian inheritance. The genetic divergence observed among the samples studied is very low and consistent with that of geographic populations of a single cohesive gene pool, linked by a high gene flow (Nm = 5.9).

Genetic studies have revealed in Southern Corsica, near Bonifacio, not only the existence of sporadic F₁ hybrids between Orchis longicornu Poiret and O. morio L., but of a hybrid zone where both parental taxa are more or less introgressed. Our data indicate complete interfertility between the two taxa, which should therefore be considered as subspecies of the same biological species.

Data are presented on genetic variation at 27 enzyme loci of the Green-Winged orchid,Orchis morio, in 18 population samples from Italy. The existence in Italy of two subspecies, i.e. subspp.morio andpicta, is not supported by allozyme data. No genetic heterogeneity was found betweenmorio-like andpicta-like samples and specimens. Moreover, morphological transition between the two forms was observed in different Italian populations. The parameters of genetic variability estimated forO. morio populations are consistent with those found among monocotyledon plants, and among those outcrossing, animal-pollinated and with wind-dispersed seeds. Genetic diversity of ItalianO. morio is mostly within populations. Correspondingly, low values of interpopulational genetic distance were found. This appears to be due to high levels of gene flow, which were estimated with different methods. The lack ofO. longicornu from Italian samples, as well as of any hybrid withO. morio (F1, backcrossed or recombinant individuals) is demonstrated on the basis of genetic data. It is concluded that recurrent reports ofO. longicornu from Italy are due to confusion withO. morio or with otherOrchis species.

Data are given on the genetic structure, morphology and chorology of the Long-Spurred orchid, Orchis longicornu Poiret (1789) from Sardinia and the occurrence in this island of the morphologically similar Orchis morio L. (1753), often recorded for Sardinia, is investigated. The genetic analysis of 27 enzyme lo ci in population samples from locations where both species had been recorded showed that: (i) in the population samples from Sardinia the polymorphic loci are in Hardy-Weinberg equilibrium; (ii) these samples are genetically poorly differentiated from each other (average Nei's D = 0.01): less than 2% of the overall genetic variation observed is attributable to differences between populations (GST = 0.015); (iii) a high rate of gene flow was estimated between Sardinian populations: Nm ≈ 4, possibly owing to seed dispersal by wind; (iv) when Sardinian samples are compared with O. morio from continental Italy, significant differences in genetic variation were observed: average H_e = 0.16 in the former, 0.12 in thc latter; (v) the genetic distance found between Sardinian populations and those of O. morio from continental ltaly is relatively high: average Nei's D = 0.18, average Rogers' D = 0.22; (vi) highly significant differences in allele frequencies were found at a number of loci ( Mdh-2, Sod-3, Pgm-1, Gpi-1, Gpi-2 ) between populations from Sardinia and continental Italy, each giving a probability of correct identification that varies from 0.90 to 1 (diagnostic loci). These findings provide evidence that O. morio is not present in the material genetically analyzed from Sardinia, which includes a single species, apparently corresponding to O. longicornu . The slight heterogeneity observed at some loci in the Sardinian population samples can be attributed in some cases (e.g. Mdh-1, Adh, Est-6 ) to local differences, possibly adaptive; in others (e.g. Pgm-1 and Gpi-l ) to genetic drift effects. The hypothesis that some alleles ( Gpi-1 ¹⁰⁰, Dia ¹⁰⁷, NADHdh-2 ⁹⁶) found in O. morio and recorded at low frequency in a few Sardinian populations reflect palaeointrogressive phenomena (owing to sporadical immigrant individuals of O. morio having diluted their genes in O. longicornu genome through multiple generations of backcrosses) is not sufficiently supported by the available data. The genetic data are in agreement with the results from morphological studies. These involved the examination of about 1500 specimens (both alive and dried) from all over Sardinia, and their comparison with herbarium specimens of O. longicornu from patria typica (Algeria) and of O. morio from continental Italy. The typification of O. longicornu is given. The specimens from Sardinia correspond well, on a morphological basis, to the Algerian ones and therefore were all assigned to O. longicornu . Several characters were considered, differentiating O. longicornu from O. morio both in fresh and dried specimens. However, some of them, currently used in diagnostic keys, show much overlap between the two species. This explains the quotations of O. morio from Sardinia, that appear to be misidentifications of O. longicornu . A description of O. longicornu from Sardinia is provided, and its distribution in the island is defined, on the basis of (i) a critical revision of literature rccords, and (ii) the examination of fresh and dried Sardinian specimens. O. longicornu is shown to be widespread throughout Sardinia, living in many different habitats from 0 to 1500 m a.s.l.

Si descrive la Serapias nurrica e l' Orchis mascula .