No abstract

Genetic variation in the southern green stink bug Nezara viridula (Linnaeus) from 11 geographically separated sampling locations (Slovenia, France, Greece, Italy, Madeira, Japan, Guadeloupe, Galapagos, California, Brazil and Botswana) was studied by sequencing 16S and 28S rDNA, cytochrome b and cytochrome c oxidase subunit I gene fragments and random amplified polymorphic DNA (RAPD) analysis. Sequencing revealed 11 distinct haplotypes clustering into lineages A, B and C. Lineage C was characteristic for a single analysed specimen from Botswana. Lineage B was detected in Japan, and it probably arose in Asia. Haplotypes of European and American specimens belonged to lineage A; specimens from France, Slovenia, Madeira and Brazil shared highly similar haplotypes (>99%) from subgroup A1, while all the specimens from Greece, California, Galapagos and Guadeloupe shared a haplotype from subgroup A2. RAPD data were more variable but consistent with mtDNA sequences, revealing the same clustering. They separated the Botswanian specimen from Japanese specimens and from a group of more closely related specimens from Europe and America. Sequence and RAPD results both support the African origin of N. viridula, followed by dispersal to Asia (lineage B) and, more recently, by expansion to Europe and America (lineage A). RAPD analysis revealed two highly supported subgroups in Japan, congruent with mtDNA lineages A2 and B, suggesting multiple colonization of Japan. Invariant sequences at the 28S rDNA combined with other results do not support the hypothesis that cryptic (sibling) species exist within the populations investigated in this study.

Genetic variation in three Croatian donkey populations, Istrian (IS), North Adriatic (NA) and Littoral-Dinaric (LD), was analysed using eight microsatellite loci and by sequence and SSCP analysis of the proximal portion of the mtDNA D-loop region. The analysis of microsatellite loci revealed observed heterozygosities in the range of 0.37 (MPZ002 in LD) to 0.85 (AHT21 in LD) and polymorphic information content values in the range of 0.36 (MPZ002 in NA) to 0.78 (AHT21 in LD). The overall probability of exclusion was 0.991. Two populations (IS and NA) were closely related (Fst=0.0034), whereas genetic distances between IS and LD (Fst=0.021) and NA and LD (Fst=0.027) were higher. Using AMOVA, 97.6% of the total genetic variance was portioned within populations, while 2.7% was portioned between the Littoral-Dinaric population and the Istrian/North Adriatic population group. Sequencing of the proximal part of the mtDNA D-loop region revealed 36 polymorphic sites representing 19 haplotypes which clustered into three haplotype groups (Y, W, Ws). Only the Y haplotype was found in the IS population which is characterized by a large body size. Haplotypes W and Ws were found in the NA and LD populations which include smaller animals. All three haplotypes were found in the LD population, indicating sporadic migration events from the IS into LD donkey population.

While the negative effects of inbreeding and reduced heterozygosity on fecundity and survival are well established, only a few investigations have been carried out concerning their influence on morphological traits. This topic is of particular interest for a small and closed population such as the Lipizzan horse. Thus, 27 morphological traits were measured in 360 Lipizzan mares and were regressed on the individual inbreeding coefficients, as well as on the individual heterozygosity and mean squared distances (mean d(2)) between microsatellite alleles within an individual. Both individual heterozygosity and mean d(2) were based on 17 microsatellite loci dispersed over 14 chromosomes. The results obtained by multivariate analysis reveal significant effects of stud (P <.0001), age at measurement (P <.0001), and mean d(2) (P =.0143). In univariate analyses, significant associations were obtained between length of pastern-hindlimbs and inbreeding coefficient (P <.01), length of cannons-hindlimb and mean d(2) (P <.01), and length of neck and mean d(2) (P <.001). After adjustment of single-test P values for multiple tests (Hochberg's step-up Bonferroni method), only the association of the length of neck and mean d(2) remained significant (P =.0213). Thus, no overall large effects of inbreeding, microsatellite heterozygosity, and mean d(2) on morphological traits were observed in the Lipizzan horse.

Mitochondrial DNA from 49 Lipizzan horses representing 16 maternal lines from the original stud at Lipica was used for SSCP analysis and DNA sequencing. The SSCP analysis of the 444 bp long fragment of the D-loop region extending from the tRNA(Pro) gene to the central conserved sequence block revealed three distinct groups of SSCP patterns. Both ends of the D-loop region (378 bp and 310 bp), which are considered as the most variable regions within the mammalian mitochondrial DNA, were sequenced. According to 49 polymorphic sites identified within the both parts of the D-loop region, the 16 maternal lines were grouped into 13 distinct mitochondrial haplotypes. The minimal difference between two different haplotype DNA sequences was one nucleotide and the maximal 24 nucleotides. The inheritance of mitochondrial haplotypes was stable and no sequence variation potentially attributable to mutation within maternal line was observed. Considerable DNA sequence similarity of Lipizzan mitochondrial haplotypes with the haplotypes from other breeds was observed. Phylogenetic analysis of the sequence data revealed a dendrogram with three separated branches, supporting the historical data about the multiple origin of the Lipizzan breed.