Lack of pollinators selects for increased selfing, restricted gene flow and resource allocation in the rare Mediterranean sage Salvia brachyodon.

Range contraction and habitat fragmentation can cause biodiversity loss by creating conditions that directly or indirectly affect the survival of plant populations. Fragmented habitats can alter pollinator guilds and impact their behavior, which may result in pollen/pollinator limitation and selection for increased selfing as a mechanism for reproductive assurance. We used Salvia brachyodon, a narrowly distributed and endangered sage from eastern Adriatic, to test the consequences of range contraction and habitat fragmentation. Molecular data indicate a severe and relatively recent species range reduction. While one population is reproductively almost completely isolated, moderate gene flow has been detected between the remaining two populations. The high pollen-to-ovule ratio and the results of controlled hand pollination indicate that S. brachyodon has a mixed mating system. Quantitative and qualitative differences in the community and behaviour of flower visitors resulted in limited pollination services in one population where no effective pollinator other than pollen and nectar robbers were observed. In this population, self-pollination predominated over cross-pollination. Various environmental factors, in which plant-pollinator interactions play a pivotal role, have likely created selection pressures that have led to genetic and phenotypic differentiation and different resource allocation strategies among populations.

Population size as a major determinant of mating system and population genetic differentiation in a narrow endemic chasmophyte.

BACKGROUND: Mating system is one of the major determinants of intra- and interspecific genetic structure, but may vary within and between plant populations. Our study model included all known populations of Moehringia tommasinii (Caryophyllaceae), a narrow endemic plant inhabiting rock crevices in the northwestern Adriatic, and some populations of co-occurring and widespread M. muscosa, an ecologically divergent relative with an overlapping flowering period. We performed reciprocal crosses within and between taxa and used molecular markers to assess the extent of gene flow within and between populations and taxa. Using coefficient of inbreeding, population size, seed weight, pollen-to-ovule ratio, and flower display size, we also looked for evidence of a selfing syndrome. RESULTS: A surprisingly high variation in mating systems was observed among populations of M. tommasinii. These populations exhibited genetic structuring, with their size positively correlated with both seed weight and pollen production. Although a selfing syndrome could not be confirmed as the majority of selfing resulted from allogamous treatments, the occurrence of selfing was notable. In the presence of M. muscosa, at a site where both species coexist closely, a distinct pattern of fruit production was observed in M. tommasinii following various pollination treatments. Molecular and morphometric data provided evidence of hybridization followed by local extinction at this site. CONCLUSIONS: Population size proved to be the most important factor affecting the mating system in genetically structured populations of M. tommasinii. Lighter seeds and lower pollen production observed in populations with pronounced selfing do not provide enough evidence for the selfing syndrome. Detected gene flow between M. tommasinii and the sympatric M. muscosa suggested weak reproductive barriers between the taxa, which could pose a conservation problems for the former species. Hybridization leading to local extinction may also resulted in floral polymorphism and disruption of mating patterns of M. tommasinii.

Diversity and Typology of Land-Use Explain the Occurrence of Alien Plants in a Protected Area.

Plant life history and functional characteristics play an important role in determining the invasive potential of plant species and have implications for management approaches. We studied the distribution of 24 alien plant taxa in a protected area in relation to different land-uses by applying ordination analyses and generalized linear models. Taxa richness is best explained by the presence of built-up areas, followed by residential areas, marshlands, and agricultural lands with semi-natural formations. The diversity of land-use within the grid cell proved to be an important explanatory factor, being the only significant variable explaining the richness of wood perennials and vines. The richness of annual herbs and seed-dispersed taxa is explained by a similar set of variables, with the exception of residential areas. The richness of invasive species is explained only by agricultural land and the diversity of land-use. The richness of taxa with predominant vegetative dispersal is best explained by built-up, marshland, and seminatural areas along with land-use diversity. When we consider only the presence of plant groups within grid cells, the results are similar. The results of similar studies may provide an important tool for defining sustainable practices and overall conservation management in protected areas.

Understanding biogeographical patterns in the western Balkan Peninsula using environmental niche modelling and geostatistics in polymorphic Edraianthus tenuifolius.

The Balkan Peninsula represents one of the three southern European glacial refugia where biodiversity persisted throughout the climatically unstable Quaternary. This study considered the 'refugia within refugia' model, which assumes the environmental heterogeneity over time and space in larger refugia. To better understand patterns shaped during the Quaternary climatic oscillations, the present and last glacial maximum (LGM) environmental conditions and current morphological variability of Edraianthus tenuifolius, an endemic plant of the western Balkans with a well-known therphical structure, were considered. Potential present and LGM distributions were studied through environmental niche modelling using 161 data of occurrences and six bioclimatic variables, hindcasting the model to LGM conditions using three different global circulation models. To test the geographical variability of the reproductive region, 41 characters of 667 inflorescences from 35 populations within the distribution range were measured. Geographical patterns, using geostatistics together with univariate and multivariate statistical approaches, were analysed. The environmental niche model suggested the current potential distribution in correspondence to its known occurrences. The hindcast to LGM conditions suggested two separate areas of environmental suitability, one in the present-day northern Adriatic coast of Croatia (Istrian Peninsula, Kvarner) and another on the present-day south-eastern Adriatic coast (southern Dalmatia, Montenegro and northern Albania). Morphological variability showed a similar pattern, where southern populations separated from northern populations according to a major split in the central part of its distribution range (central Dalmatia). On other levels, stronger barriers were predicted to separate northern populations from the eastern Istrian Peninsula and the Kvarner area. The results suggested congruent biogeographical patterns to the already known phylogeographical structure. Both environmental niche modelling and the geographical variability of morphological characters suggested spatial partitioning, indicating the potential presence of two separate refugia during the LGM.

Presence and Potential Distribution of Aedes albopictus and Aedes japonicus japonicus (Diptera: Culicidae) in Slovenia.

In Slovenia, two invasive mosquito species are present, Aedes albopictus (Skuse, 1895) (Diptera: Culicidae) and Aedes japonicus (Theobald, 1901) (Diptera: Culicidae). In this study, we examined their actual distribution and suitable habitats for new colonizations. Data from survey of species presence in 2013 and 2015, bioclimatic variables and altitude were used for the construction of predictive maps. We produced various models in Maxent software and tested two bioclimatic variable sets, WorldClim and CHELSA. For the variable selection of A. albopictus modeling we used statistical and expert knowledge-based approach, whereas for A. j. japonicus we used only a statistically based approach. The best performing models for both species were chosen according to AIC score-based evaluation. In 2 yr of sampling, A. albopictus was largely confined to the western half of Slovenia, whereas A. j. japonicus spread significantly and can be considered as an established species in a large part of the country. Comparison of models with WorldClim and CHELSA variables for both species showed models with CHELSA variables as a better tool for prediction. Finally, we validated the models performance in predicting distribution of species according to collected field data. Our study confirms that both species are co-occurring and are sympatric in a large part of the country area. The tested models could be used for future prevention of invasive mosquitoes spreading in other countries with similar bioclimatic conditions.

Testing the efficiency of nested barriers to dispersal in the Mediterranean high mountain plant Edraianthus graminifolius (Campanulaceae).

Due to strong spatial heterogeneity and limited Pleistocene glaciation, the Balkan Peninsula is a major European biodiversity hot spot. Surprisingly little, however, is known about patterns and processes of intraspecific diversification of its biota in general and of high-altitude species in particular. A well-suited system to test hypotheses with respect to various isolating factors acting at different geographic scales and to explore full-range phylogeographic patterns on the Balkan Peninsula is Edraianthus graminifolius (Campanulaceae), distributed in the western Balkan mountain systems, the southwestern Carpathians and the Apennine Peninsula. To this end, we used a dense population sampling and employed amplified fragment length polymorphism (AFLP) markers and plastid DNA sequences supplemented by ecological niche modelling. The strongest splits were inferred to separate southern and northern Balkan populations from the central ones, from where range extension occurred to the Carpathians and, in more recent times, once or twice to the Apennine Peninsula. The three genetic groups in the western Balkan Peninsula were remarkably congruent among molecular markers, suggesting that the barriers to gene flow acted over long time periods facilitating allopatric differentiation. Each main group of Balkan populations contained genetically and geographically distinct subgroups, which likely are the result of local refugia during warmer periods. Evidently, the topographically highly complex and during the Last Glacial Maximum only locally glaciated Balkan Peninsula is a hot spot of species richness and endemism as well as a sanctuary of intraspecific genetic diversity, even if the underlying causes remain insufficiently understood.