**2. Taxonomic and Phylogenetic Aspects**

*Lasiodiplodia theobromae* (Pat.) Griffon & Maubl. (Dothideomycetes, Botryosphaeriaceae) is the accepted name of the species treated in this paper, prevailing over both the basionym *Botryodiplodia theobromae* Pat. and the teleomorphic name *Botryosphaeria rhodina* (Berk & M.A. Curtis) Arx, after the introduction in mycology of the principle "one species—one name" [7]. Isolated and morphologically identified from a wide range of plant hosts [5,8], it represents the type species of *Lasiodiplodia* which, for many years, was treated as a monotypic genus within the Botryosphaeriaceae [9,10]. However, such a simplified taxonomy was destined to dramatically change with the advent of DNA sequencing. In fact, starting from the year 2004, phylogenetic analyses carried out in the course of studies on *L. theobromae* in novel pathosystems showed the existence of several clades, even within the pool of strains stored in mycological collections [11–14]. Evidence of a higher complexity emerged gradually, to such an extent that more than 30 additional species have been described to date, with some of them, such as *L. endophytica*, *L. gonubiensis*, *L. pseudotheobromae*, *L. thailandica* and *L. venezuelensis*, reported as endophytes [15–23]. Hence, it is likely that several previous findings might be incorrectly classified and that some more recent records are going to be re-examined. The application of high-throughput DNA metabarcoding as a biomonitoring tool is expected to provide a notable contribution in investigations concerning the endophytic occurrence of *Lasiodiplodia* [24].

To further complicate the issue, the existence of hybrid strains has been ascertained [15,25], which is also considered to have affected species identification. As an example, the taxon *L. viticola* Úrbez-Torres, Peduto & Gubler [26] has been shown to be a hybrid between *L. theobromae* and *L. mediterranea*; both these taxa are known on grapevine (*Vitis vinifera*), which most likely represented the venue of the hybridization process [15]. An assumption in biology considers as a species an organism whose population is reproductively isolated from other phylogenetically related populations [27]; hence, the existence of hybrids between several *Lasiodiplodia* spp. may imply that the taxa described so far are not stable. Indeed, further reassessments are to be expected, particularly in consequence of new combinations possibly stimulated by the circulation of plant material hosting genotypes which are potentially capable of hybridizing with autochthonous strains. In order to avoid further misidentifications, the use of multiple genes is recommended when considering the phylogenetic relationships of novel strains, along with direct referencing to the type strains [15,20].

Apart from the variation characterizing the genus *Lasiodiplodia*, phylogenetic relationships have also been evaluated in the species under discussion. Low genotypic diversity was observed in a study considering three populations from different tree species in Venezuela, South Africa and Mexico. A few predominant genotypes were encountered in the first two countries, without evidence of host specificity and in the presence of a very high gene flow between populations from different hosts. The geographic isolation was substantiated by the finding of unique alleles fixed in the different populations. Moreover, the existence of some genotypes that were widely distributed throughout the three countries, coupled with the evidence that pseudothecia are rarely produced in nature, suggests that reproduction is predominantly clonal [8]. A similar conclusion was reached in another phylogeographic study carried out on coconut palm (*Cocos nucifera*) in Brazil, where higher genotypic variation was observed in the northeast in connection with the local higher host diversity and a conjectured repeated introduction from Central Africa, regarded as the possible center of radiation of the species. Differences between genotypes were mainly ascribed to mutations [28].

In Cameroon, cocoa (*Theobroma cacao*) and *Terminalia* spp. are frequently grown together in a peculiar agri-sylvicultural system. A comparison between strains from these two known hosts of *L. theobromae* showed high levels of gene diversity and low genotypic differentiation, in the presence of high gene flow between isolates. The absence of a geographic substructure in these populations across the region where the study was carried out is indicative of the symmetrical movement of the fungus between these hosts. Unlike the case documented on grapevine, no evidence of hybridization was found with the closely related *L. pseudotheobromae*, which also occurs on these plants [29].

Finally, quite a simple genetic structure was once more pointed out in a broader study including strains of more varied origin. In fact, one or two main haplotypes across all genes were identified, and these genotypes were unrelated to both the hosts and the geographic area. Such overall uniformity clearly indicates that large-scale dispersal of *L. theobromae* is essentially derived from commerce and human activities [4].
