Francesco Mascia

Assessing the potential invasiveness of Cortaderia selloana in Sardinian wetlands through seed germination study

Abstract The present work is focused on the study of abiotic factors that may favour seed germination and the potential invasiveness of Cortaderia selloana, with particular attention to wetlands. Germination tests were conducted at the Sardinian Germplasm Bank (BG‐SAR), testing different temperatures and percentages of NaCl in order to determine the optimal ecological conditions at which germination occurs and the effect of salt on seed germination and viability, as well as on seedling development. Seeds completely germinated at every temperature tested, but the highest germination rate was found at 25°C. Salinity did not prevent seeds from germinating, but it affected germination rate and seedling vigour.

What does the germination ecophysiology of the invasive Acacia saligna (Labill.) Wendl. (Fabaceae) teach us for its management

Acacia saligna is a phanerophyte native to Australia; it was introduced in the 1950s in Sardinia (Italy) for afforestation mainly in coastal areas and at present it is considered naturalized, becoming invasive in sand dune habitats. Seed biology is one of the most important factors contributing to the invasion success of Australian acacias, for this reason the germination ability is an important issue when trying to clarify the invasion dynamics of A. saligna. Germination tests were conducted at the BG-SAR, testing the effect of different temperatures and percentages of NaCl, on seeds belonging to five accessions from four populations of two coastal habitat types. The optimal temperature range for seed germination was 15–20°C; the increase of salt caused a decrease in the final germination percentages. At 1% of NaCl concentration, germination fell at 25–30°C and at 2% it was relatively low (below 40%) and occurred almost only at 15°C. This work represents a contribution to the knowledge of the seed ecology and germination behaviour of the species, providing new data on the interpopulation and interannual variability, and relating them to the invasion dynamics of A. saligna in the coastal Mediterranean habitats.

Morpho‐colorimetric characterisation of Malva alliance taxa by seed image analysis

Seed morphometric and -colorimetric features describing shape, size and textural seed traits of 28 taxa belonging to the genera Lavatera L. and Malva L., were recorded by means of computer vision techniques. The data were statistically analysed to contribute to the taxonomical treatment of the Malva alliance and to assess some doubtful systematic positions. A clear differentiation between taxa traditionally attributed to Lavatera or Malva was highlighted. Furthermore, the identification system proposed here was able to discriminate among the Lavatera sections, confirming the taxonomic organization of this genus. The results obtained for Malva, both at the species level and among sections, supported this analytical tool as diagnostic for systematic purposes.

Arundo micrantha, a new reed species for Italy, threatened in the freshwater habitat by the congeneric invasive A. donax

Arundo micrantha Lam., a reed species from southern Mediterranean area, has found in Sardinia. Its presence represents the first record for Italy and the third one for western Europe. Investigations on distribution and ecology of the Sardinian population have been carried out, with special focusing on synecology of the species. A. micrantha has found in 40 localities mainly distributed along temporary streams and permanent rivers in the central and southern part of the island. Most of Sardinian populations occurs on alluvial soils in Thermomediterranean bioclimatic context. The autonomy of Sardinian phytocoenoses was characterized with the definition of exclusive floristic taxa, biological and chorological elements, and syndinamic relationships with other assemblages. Comparing the community with other reed beds from the Mediterranean basin, a marked floristic differentiation arose between the phytocoenoses characterized by autochthonous Arundo species and those dominated by the invasive Arundo donax. Finally, management activities are proposed to ensure the conservation of this threatened autochthonous plant community in freshwater habitat.

A NOVEL METHOD FOR PARTITIONING FEATURE SPACES ACCORDING TO THEIR INHERENT CLASSIFICATION COMPLEXITY

The evaluation of the intrinsic complexity of a supervised domain plays an important role in devising classification systems. Typically, the metrics used for this purpose produce an overall evaluation of the domain, without localizing the sources of complexity. In this work we propose a method for partitioning the feature space into subsets of different complexity. The most important outcome of the method is the possibility of preliminarily identifying hard and easy regions of the feature space. This possibility opens interesting theoretical and pragmatic scenarios, including the analysis of the classification error and the implementation of robust classification systems. A first group of experiments has been performed on synthetic datasets, devised to separately highlight specific and recurrent problems often found in real-world domains. In particular, the focus has been on class boundaries, noise, and density of samples. A second group of experiments, performed on selected real-world datasets, confirm the validity of the proposed method. The ultimate goal of our research is to devise a method for estimating the classification difficulty of a dataset. The proposed method makes a significant step in this direction, as it is able to partition a given dataset according to the inherent complexity of the samples contained therein.

Hierarchical Text Categorization Through a Vertical Composition of Classifiers

In this paper we present a hierarchical approach to text categorization aimed at improving the performances of the corresponding tasks. The proposed approach is explicitly devoted to cope with the problem related to the unbalance between relevant and non relevant inputs. The technique has been implemented and tested by resorting to a multiagent system aimed at performing information retrieval tasks.