Hieracium tridentatum (Asteraceae) in Britain

pdf constitutes the


Introduction
Linnaeus (1753) lists only twenty six species in the genus Hieracium. Of these there are several species that are no longer considered to belong to the genus. The study of the genus Hieracium proceeded very slowly but in 1819 the epithet tridentatum was published as a variety of Hieracium vulgatum by Professor Elias Fries (1819, p.77). Fries (1845, p. 6) subsequently raised H. tridentatum to the rank of species.
On this occasion Fries gives no description but refers his reader to Herb. Norm. III:4 (Fries, 1837) (Fig. 1), thus selecting syntypes, one of which should be selected as the lectotype. There are sets of Herb. Norm. in at least fifteen herbaria, so plenty of syntypes are available. The first British mention of H. tridentatum was by Babington (1847, p. 198), he gave a brief description and confirmed the presence of the species in England. Fries (1848, p. 171) published a description in Latin, but by then had decided to include three forms in the species, and named the exsiccata Herb. Norm. III:4 forma dentata.
The next account of H. tridentatum in Britain is contained in Backhouse (1856, p. 67). There is a decent description of the species and a record of distribution in Britain stretching from Teesdale to Kent. The flowering period is given as August and September, which is rather late for the southern parts of the plant's distribution, where it can be in flower in July. Backhouse (1856) refers to Fries (1848, p. 171 (1904, p. 262) published a new account of the genus Hieracium in which species 82 H. rigidum includes var. ε. tridentatum (Fr.), and a note, "Other vars. or forms have been described." This treatment reflects the view held by Dahlstedt (1889), when he issued Hierac. Exs. 3:57 (Fig. 2). Dahlstedt (1894) revised his view of the status of H. tridentatum and raised the taxon to the rank of species together with H. rigidum which had ten subspecies, the sixth being trichocaulon Dahlst. (Fig. 3). In Britain the views of Dahlst (1894) were adopted in modified form by Linton (1905, p. 80-84).
However, Linton (1905) (Linton, 1908), where that species is recorded from six vice-counties and H. rigidum var. trichocaulon from 12 vice-counties. Not until the eleventh edition of the London Catalogue (Roffey, 1925), where both taxa appear as species, did H. tridentatum with 17 vice-counties exceed H. trichocaulon with 14 vice-counties. After the introduction of many new names by Roffey (1925), there was a big drop in interest in the genus. Pugsley (1948, p. 261), was still only able to quote H. tridentatum Fr. (1848) ex parte, as the date of publication at the rank of species, but he increased the distribution to cover 23 vice-counties with numerous localities. H. trichocaulon was only recorded in nine vice-counties. Following Zahn (1922), he also described one variety and one form of H. tridentatum as new taxa.  Looking back at the work of Zahn (1921-23, p. 856), under 'species principales collectivae' 360 H. laevigata Willd., he mentions the place of publication of H. tridentatum, at the rank of species, "Summ. veget. 6", easily over looked and missed by British authors. Zahn placed the three H. tridentatum exsiccata of Fries (1848, p.171)

Post 1950
By 1950 the new journal of the Botanical Society of the British Isles Watsonia had commenced publication and moves were being made to publish a revised British Plant List. The first major publication on British botany for many years was Clapham et al., 1952). The arrangement of the genus Hieracium in this work commenced with section Umbellata, the section containing the one species that occurs in Britain that has the base chromosome number of the genus. Unfortunately the authors or editor of Hieracium in the revised British Plant List did not follow that lead and it was not until Stace (1991) that an arrangement of the genus closer to the 1952 arrangement appeared. Sell & West (1955) published a paper to validate new names and combinations to be used in the new list, but nothing was published to explain the many alterations made in the list. Three years later the new list of Hieracium was published (Sell & West, 1958). Sell & West (1958, p.129) Sell & West (1958) for publication at the rank of species. Sell & West (1958) thus removed H. tridentatum (Fr.) Fr. from the British list and it is still excluded.
Despite the examples of Clapham et al., (1952) and Stace (1991), Sell & Murrell (2006) still decided to start the genus Hieracium with section Sabauda (Fr.) Arv.-Touv. However they have re-arranged the order of the sections and Tridentata is now section 4. The treatment of H. tridentatum (Fr.) Fr. remains much as it was in 1958. H. tridentatum sensu Pugsl. pro parte becomes H. tridentatum auct. and appears under the same three species. Sell & Murrell (2006, p. 251), under H. calcaricola (F. Hanb.) Roffey, state, "True H. tridentatum Fr. is a slender short plant with eglandular involucral bracts which does not seem to occur in our area."

Morphology
As can be seen from the above images, H. tridentatum is not a short plant. The image of H. trichocaulon has one short plant and one much larger plant. The size of the plant in section Tridentata depends entirely upon the growing conditions that the indivdual plant happens to be in, although there is a limit on the maximum size possible. Dahlstedt (1894) gives the stem length of H. tridentatum as "50-70ctm." and for H. trichocaulon "35-90ctm." The statement by Sell and Murrell (2006) that H. tridentatum Fr. is a slender short plant is not very convincing. Babington (1847 p. 198) wrote of H. tridentatum "St. 2 feet high, stout, branching slightly in the upper half, scabrous, somewhat hairy." Backhouse (1856 p. 68) writes "Plant 2 to 5 feet high" and on page 69, "under cultivation frequently 6 feet high". However, it should be noted that Backhouse probably included the taxon now called H. eboracense with H. tridentatum.
The immediately obvious difference between the two species is the type of teeth on the leaves. The margins of median leaves of H. tridentatum are dentate, with three to five teeth about the middle of the leaf, and the margins of the median leaves of H. trichocaulon are denticulate. The phyllaries (involucral bracts) of H. tridentatum have pale green tips and margins, particularly the inner phyllaries, whereas H. trichcaulon has phyllaries with green margins, pale compared with the dark central stripe. Pugsley (1948, p. 261) describes the indumentum of H. tridentatum phyllaries as "the outer obscurely floccose-edged, ± densely clad with longer dark-based and fine, pale pilose hairs, typically eglandular, but usually with ± numerous glandular hairs in British examples." Pugsley (1948, p. 262) goes on to discuss Swedish specimens that suggest the presence or absence of glandular hairs may be an unfixed character. Fries (1862, p. 116) in his last description of H. tridentatum writes "filiformes involucraque pallida undique canofloccosa et breviter pilosa, pilis raro glandulosis," [threadlike involucre pale on all sides greyish white floccose and shortly pilose, hair rarely glandular], which supports Pugsley's observation about the presence of glands and undermines once again the statement of Sell & Murrell (2006 p. 251). Pugsley (1948, p. 266) describes the indumentum of H. trichocaulon phyllaries as "usually efloccose, with numerous unequal, dark based pilose hairs and sometimes many dark (usually shorter) glandular hairs intermixed." Sell & Murrell (2006, p. 251) describe the indumentum of H. trichocaulon phyllaries as "with few to numerous, short to medium, pale simple eglandular hairs, numerous, short and minute, dark glandular hairs and few or no stellate hairs." Pugsley (1948) and Sell & Murrell (2006) describe the indumentum of the phyllaries of H. trichocaulon in much the same way and both allow some variability in the details. Dahlstedt (1894, p. 235) writes of H. trichocaulon "efloccose, v. basi levissime stellata pilis sparsis − densiusculis glandulis brevibus − minutis sparsis − densiusculis obtectae." [efloccose or base very faintly stellate with scattered hairs − short dense glands − minute glands − densely covered]. An examination of images of syntypes shows scattered hairs that extend well up from the base of the phyllaries and an examination of British specimens determined by Sell & West reveal that short and minute glandular hairs are very dense about the base.
Examination of the images of syntypes of H. trichocaulon show that at least in herbarium specimens, the styles start out yellowish and change through fuscous to dark, in much the same way as specimens of H. tridentatum, although often some yellow is retained. Pugsley (1948, p. 263)  Clark (1900, p. 83) records that H. tridentatum appeared for the first time in British literature in Babington (1847), but Babington only gives the distribution as E [England]. The earliest specimen I have so far discovered in OXF was collected by G.C. Druce, in June 1882, near Brockenhurst, New Forest, v.c.11 (South Hants.). Townsend (1883, p. 199) records "Hedgebanks, Brockenhurst," together with 24 other localities in the three Vice-counties 10-12. Townsend also records the first record as "Herb. Bromfield, 1845" [for v.c.10, Isle of Wight], just two years ahead of Babington's publication. However, Pugsley (1948, p. 263) Pugsley (1948, p. 274) was of the opinion that Petiver (1713, pl. 13, 9) is actually H. latobrigorum (Zahn) Roffey, a member of section Foliosa. The name H. latobrigorum was published in 1925 and has now been replaced by H. lanceolatifolium (Zahn) Prain, published in 1921, on the Euro+Med Plantbase and IPNI. Pugsley (1948, pp. 261-2) followed Zahn(1921-3, p. 866) in the treatment of variation in H. tridentatum but as Zahn's account is of the variation in the entire range of the species, Pugsley restricted his account to the variation that is present in Britain. Pugsley places the subvar. glandulosiceps Zahn as a new combination, f. glandulosiceps (Zahn) Pugsley, and states it differs from the type in "phyllaries with numerous dark glandular but no pilose hairs". This form has a restricted distribution: v.c.11(2), 12(1), 16(2), 17(1) and 22(1), a total of just 7 localities. The other taxon described by Zahn that Pugsley includes as British is H. laevigatum subsp. tridentatum var. polyphyllum, which he publishes as H. tridentatum var. polyphyllum (Zahn) Pugsley. This taxon varies considerably from the type and has a wide distribution: v.c.4(1), 9(1), 12 (1)  polyphyllum, none of which appear to have been seen by Sell & West or Sell. In Sweden, Dahlstedt (1894) divided Archieracia into four sections. Section III Rigida Lindbg. included only two species, H. tridentatum Fr. and H. rigidum Hartm. Before 1894 H. rigidum had included a wide range of variation which Dahlstedt split into ten subspecies. In Britain, members of the Botanical Exchange Club were looking out for specimens that fitted into the new subspecies, and Linton (1905) recorded seven of the ten subspecies as present in Britain, but at the rank of variety. All ten subspecies are now published at the rank of species in Britain and Scandinavia, and those that Linton (1905) thought to occur in Britain are listed in Table 1. In Table 1 the number of sheets of H. trichocaulon in the OXF column has been restricted to those sheets confirmed by Sell & West and may increase as further study of the specimens is completed. The British centre of this variation is in southern England and the variation was recorded in detail in Surrey and Sussex. Salmon (1931, p. 418) records five taxa as present in Surrey. Salmon confirms that he personally had seen a dried specimen, or living plants at a locality listed, of every one of the five taxa. Pugsley (in Salmon, 1931) contributed records of H. tridentatum Fr., H. scabrescens Dahlst. and H. rigidum Hartm. Pugsley had obviously not compared British specimens with Scandinavian specimens at that time. Wolley-Dod (1937, p. 265) in Sussex, included four taxa, missing out any mention of H. trichocaulon.. H.W. Pugsley is only mentioned in connection with scabrescens Dahlst. and he does not confirm the records, merely saying "near this" and "(cf.)".

New combinations
In The form subpilosiceps (Zahn) was not know to Pugsley as British, but it seems reasonable to maintain the simple approach taken by Pugsely to f. glandulosiceps (Zahn).

Commentary
As can be seen from the list of taxa in Table 1, there is a group of closely related species which were originally described in Sweden. All except H. rigidum Hartm. and H. friesii Hartm. had syntypes selected by the author. When all eight species are lectotypified, it is possible that British material may need to be re-examined in the light of new knowledge. Figure 1 has an image of specimen UPS V-112720 on the left. There are three labels in the bottom right hand corner, the original label being the lowest one on the right. Fries (1837)  Fr. (1845 p. 6), so that name has priority. The first version of the International Code of Nomenclature was only published in 1906, so Dahlstedt had no guidelines to follow. The taxon was called H. laevigatum subsp. asperatulum Zahn (1921-23, p. 885), a subspecies name that does not appear to have been published at the rank of species. H. asperum Bernh. (1800 p. 136) makes asperum unavailable. The large type written label confirms the information on the printed label below.
Only H. trichocaulon (Dahlst.) Johanss. was recorded as occurring in Britain by Sell & Murrell (2006), and their list has been followed by all subsequent authors. Further work will be carried out to try and determine the correct identity of the remaining unidentified specimens in OXF.