History vs. legend: Retracing invasion and spread of Oxalis pes-caprae L. in Europe and the Mediterranean area.

Oxalis pes-caprae L. is a South African geophyte that behaves as an invasive in the eurimediterranean area. According to a long-established hypothesis, O. pes-caprae may have invaded Europe and the Mediterranean area starting from a single plant introduced in the Botanical Garden of Malta at the beginning of the 19th century. The aim of this work was to test this hypothesis, to track the arrival of O. pes-caprae in different countries of the Euro-Mediterranean area and to understand the pathways of spreading and particularly its starting point(s). Historical data attesting the presence of the plant in the whole Euro-Mediterranean region were collected from different sources: herbarium specimens, Floras and other botanical papers, plant lists of gardens, catalogs of plant nurseries and plant dealers. First records of the plant (both cultivated and wild) for each Territorial Unit (3rd level of NUTS) were selected and used to draw up a diachronic map and an animated graphic. Both documents clearly show that oldest records are scattered throughout the whole area, proving that the plant arrived in Europe and in the Mediterranean region more times independently and that its spreading started in different times from several different centers of invasion. Botanical gardens and other public or private gardens, nurseries and plant dealers, and above all seaside towns and harbors seemingly played a strategic role as a source of either intentional and unintentional introduction or spread. A geographic profiling analysis was performed to analyse the data. We used also techniques (Silhouette, Kmeans and Voronoi tessellation) capable of verifying the presence of more than one independent clusters of data on the basis of their geographical distribution. Microsatellites were employed for a preliminary analysis of genetic variation in the Mediterranean. Even if the sampling was insufficient, particularly among the populations of the original area, our data supported three main groups of populations, one of them corresponding to the central group of populations identified by GP analysis, and the other two corresponding, respectively, to the western and the eastern cluster of data. The most probable areas of origin of the invasion in the three clusters of observations are characterized by the presence of localities where the invasive plant was cultivated, with the exception of the Iberian cluster of observation where the observations in the field predate the data about known cultivation localities. Alternative possible reasons are also suggested, to explain the current prevalence of pentaploid short-styled plants in the Euro-Mediterranean area.