Arguments for a narrow species concept in Rubus sect. Corylifolii

A detailed inventory of Rubus sect. Corylifolii from the province of Halland in southwestern Sweden has been performed and the chromosome number of all distinct taxa and deviating morphotypes has been determined using flow cytometry. Once putative hybrids between Corylifolii and R. caesius or R. idaeus are recognized, essentially all stands of Corylifolii in Halland can be satisfactorily identified, including six taxa with a restricted distribution. Two of the latter have been noticed before: R. sordirosanthus (H. Hyl.) Ryde, found in over 100 localities and R. cordatiformis (Neuman) Ryde (comb. nov.) in almost 80 localities, but only in restricted areas, 23 and 7 km across. The other four have even smaller distributions (1–6 km across) and occur only in 3–15 localities. Clearly, they are local species in the terminology of Weber. Such species are currently usually left undescribed, which in the case of Halland would mean that 25% of the Corylifolii would remain without a name. Such a restriction leads to a number of problems: chaos in botanical collections and provincial floras, wide-spread species potentially being overlooked, red-listing becoming problematic, changes in the occurrence of local species becoming impossible to follow over time, and the Corylifolii treated differently from other apomictic groups. This study shows that a proper description of all local Corylifolii species would solve these problems. It is also argued that by doing so, the total number of Swedish species would not become particularly large, still remaining much lower than in other apomictic genera, like Hieracium, Taraxacum and Ranunculus. Therefore, the four local species are here described as R. onsalaënsis Ryde, R. rugulosus Ryde, R. polybracteatus Ryde and R. soendrumensis Ryde.