Flora of New Zealand Series

Gnaphalium (Helichrysum) grandiceps Hook. f. Handbk N.Z. Fl. 1864, 154.

Helichrysum grandiceps Kirk Stud. Fl. 1899, 313.

Type locality: Mount Torlesse Range. Type: K, 3500-5000 ft., J. Haast, 1864.

Stock rather stout, woody; stems decumbent, woody at base, branches ascending; basal lvs not in subrosettes. Lvs densely to somewhat laxly imbricate, spreading, sts recurved, 5-10 × 2-4 mm., often apiculate, 3-5-nerved at base, obovate-cuneate, clad on both surfaces in appressed white to pale buff hairs. Peduncles hardly differentiated from branchlets. Capitula 5-15, congested into dense glomerules 9-15 mm. diam., subtended by up to 15 densely woolly lvs up to 1 cm. long, forming a distinct white ray. Phyll. linear, acute, c. 5 mm. long, scarious with basal stiffening. Achenes c. 1 mm. long, pappus-hairs 3-4 mm. long.

DIST.: S., St. Subalpine to alpine rocks throughout.

HYBRIDISM

Where spp. of Leucogenes and spp. of the subgenus Psychrophyton of Raoulia grow together, one us. finds odd examples of plants that give evidence of hybrid origin in their morphological characters; flowering specimens are rare and no viability tests have been made with the seeds produced. For the most part, at least, the hybrids appear to be sterile. Some forms have been described as spp. and certainly simulate such. An account, under the title "× Leucoraoulia", was given by Allan in T.R.S.N.Z. 68, 1939, 457.

1. Leucogenes grandiceps × R. bryoides

(a) Helichrysum pauciflorum Kirk in T.N.Z.I. 27, 1895, 351. This was based on a specimen collected by L. Cockayne at c. 900 m. on the Craigieburn Range. The habit is that of R. bryoides, but rather more open. The shape and indumentum of the lvs, c. 6-7 mm. long, are also those of R. bryoides, but there are 1-3 basal nerves. The capitula are solitary, not subtended by woolly lvs. The pappus-hairs and the silky hairs of the achenes are those of L. grandiceps. Kirk (loc. cit.) emphasizes the close "external resemblance to H. grandiceps", as does Cheeseman (Man. N.Z. Fl. 1925, 984), who would have placed the specimens he saw under Raoulia but for the habit. The type specimen (W), however, has much more the appearance of R. bryoides than most other examples of the hybrid, which show a linking series both in lf size and branching habit between the two supposed parents.

(b) Raoulia gibbsii Cheesem. in T.N.Z.I. 42, 1910, 216. This was based on specimens collected by F. G. Gibbs on Mount Starveall and Slaty Peak at c. 1200-1500 m. Type locality: "Mount Starveall, Dun Mountain Range, 4000-5000 ft." Type: A, 10109, F. G. Gibbs. These have the more open habit of L. grandiceps; the lvs have the indumentum more like that of R. bryoides, but are larger, up to 8 mm. long, and 1-3-nerved. The capitula are up to nearly 10 mm. diam. The pappus and achenial hairs are nearly those of Psychrophyton. H. pauciflorum thus includes the forms tending towards R. bryoides, and R. gibbsii those towards L. grandiceps. Other forms have been collected, making an intergrading series between the parent forms.

(c) Helichrysum (Leucogenes) grahamii Petrie in T.N.Z.I. 45, 1913, 268. This was described from a specimen collected by P. Graham "in clefts of rocks on Sebastopol Ridge, Sealey Range". The specimen indicates that the plant was a subshrub c. 15 cm. tall, with stout woody deeply descending main root, branches woody at base, branchlets rather flexible. Lvs sessile, appressed at base, 3-nerved (laterals often very faint and short) slightly spreading at tips, c. 6 mm. long, tapering from base c. 2 mm. wide to subacute tip; clad on both surfaces in dense white tomentum, woolly at base, more appressed at tips. Petrie gives: "a few of the uppermost [leaves] longer and broader, but not exceeding the heads. Heads in capitate fascicles of 2-5, terminal, sessile . . . involucral bracts in 3 series, very pale, shining, scarious, linear-oblong, subacute, scarcely radiating, the outer sparingly tomentose on the backs . . . pappus hairs few, slender, barbellate, slightly thickened above. Achene sparsely papillose."

Cheeseman (Man. N.Z. Fl. 1925, 980) doubtfully accepted the plant as a Leucogenes, but remarked: "But the inflorescence, pappus-hairs and achenes are so different to those of Leucogenes that I anticipate that it will ultimately fall into a different genus". Cockayne and Allan (Ann. Bot., Lond. 48, 1934, 49) suggested as a possible parentage L. grandiceps × Helichrysum selago. There is now much better evidence that the origin is probably L. grandiceps × R. bryoides. Wall (Flora of Mount Cook 1925, 33) pointed out the resemblance to H. pauciflorum. W. B. Brockie collected on Mount Cheeseman, Craigieburn Range, plants almost identical with L. grahamii (specimens now in BD). He remarks (in litt. March 1944) "the plant is fairly common on rocky bluffs intermingled with R. bryoides and edelweiss [Leucogenes grandiceps]. Its habit is to form a compact cushion like R. bryoides." The infl. is a densely congested cymose cluster of up to 5 capitula, forming a head c. 1 cm. diam., subtended by c. 15 spreading, densely woolly obovate-spathulate lvs c. 6 × 2 mm. The inner phyll. are 4-5 mm. long, linear-oblong, subacute, the lower half coriac., the upper hyaline, with sparse hairs at junction. Achenes c. 1 mm. long, densely to rather sparsely clad in long silky hairs; pappus-hairs c. 4 mm. long, slightly flattened at base, mostly slightly thickened at tips, some not at all thickened, not barbellate. Characters suggesting L. grandiceps as one parent are thus quite prominent.

2. L. grandiceps × R. goyenii. Cockayne (Rep. Bot. Surv. St. Id 1909, 64) lists "Helichrysum loganii" (Buchan.) Kirk, with a doubt as to the identification, as from "Near summit of Mount Anglem". The type specimen has slender, branching, ascending stems, the branchlets densely clad in imbricating lvs, forming a loosely cushionlike patch. The lvs are wider than those of either supposed parent, with 1-3 nerves. The lf-indumentum is that of L. grandiceps but looser, the lf-apex is that of R. goyenii but more definitely apiculate. I have seen only sterile specimens. Cockayne and Allan (Ann. Bot., Lond. 48, 1934, 49) considered the plant to be a hybrid with the above parentage. Cheeseman (Man. N.Z. Fl. 1925, 972) placed it as Raoulia loganii.

3. On Mount Hutt (Canterbury) forms suggesting the parentage L. grandiceps × R. mammillaris have been collected, but their status requires further study in the field; so too does that of forms suggesting the parentage L. grandiceps × R. grandiflora found on the Bounds Range (Marlborough) and Mount Percival (Canterbury).

4. L. leontopodium × R. rubra. Haastia loganii Buchan. in T.N.Z.I. 14, 1882, 350, t. 30, f. 3 was described from specimens collected on Mount Holdsworth, Tararua Range, at c. 1400 m. Kirk (Stud. Fl. 1899, 310) referred the plant to Helichrysum, but did not exclude the possibility of its being a Raoulia, where later Cheeseman (Man. N.Z. Fl. 1925, 972) placed it. Very uniform specimens answering to the description have been collected in a number of places on the Tararua Ranges. The plant forms cushiony masses of branchlets 8-10 mm. diam., in appearance distinct from either of the supposed parents. The lvs are intermediate in size and in indumentum, 1-3-nerved at the base, the apex more rounded than in R. rubra. The florets have a pinkish tinge; the achenial and pappus hairs are as in Psychrophyton. The viability of the achenes has not been tested, but the intermingling of characters leaves little doubt of the hybrid origin, as given by Cockayne and Allan (Ann. Bot., Lond. 48, 1934, 49.)

5. L. leontopodium × R. grandiflora. V. D. Zotov and M. E. Roberts discovered a single clump on Ruapai Peak, Tararua Range, at c. 1200 m., for which Zotov suggested the above parentage (BD 54290), very probably correctly. Stock rather slender, woody, deeply descending; branches woody at base, branchlets up to c. 7 cm. long, ending in densely cymosely congested compound heads up to 2 cm. diam. and subtended by an involucre of lvs similar to cauline ones. Lvs densely imbricate, sessile, finally spreading, c. 8-12 × 3 mm., lanceolate-oblong, subacute, tomentose on both surfaces; basal portion with lax woolly hairs, 1-3-nerved, limb with appressed felted silvery tomentum. Phyll. up to 6 mm. long, linear, acute, coriac. in lower half, scarious and semi-hyaline in upper half; with sparse long hairs at junction. Pappus-hairs c. 20, slender, flattened at base, c. 4 mm. long, not or slightly thickened at tips. Achenes c. 1 mm. long, sparsely pubescent.

Leucogenes leontopodium × Gnaphalium traversii? Specimens from a large colony near Junction Knob, Central Tararua Range (BD 87800, M. B. Ashwin, 22.1.56) have the general appearance of L. leontopodium with capitula aggregated in dense glomerules subtended by radiating lvs. Compared with normal L. leontopodium growing nearby lvs are smaller, more inclined to be keeled, less evenly arranged, with more fluffy tomentum when young; peduncles short, mostly hidden by basal lvs, where evident rather sparsely lfy; florets, even hermaphrodite ones, nearly all filiform; achenes sparsely pubescent and apparently all infertile; pappus-hairs slender, tapering to a fine point, not at all clavate; phyll. narrow, scarious, brownish, lacking the basal stiffening of L. leontopodium. Several features suggest hybridism with Gnaphalium traversii also abundant nearby.

FL. 11-3. FT. 1-4.

INCERTAE SEDIS

Plants collected by J. K. Forbes on Mt. Terako, Seaward Kaikoura Range (BD 76476 March 1952, 78905-78911 March 1953, 87683 Feb. 1957) form low-growing subshrubs with very widely spreading slender branching stems mostly concealed by broken rock debris, forming very lax patches up to c. 3 m. diam. Lf-bearing branchlets ascending, few together or solitary, up to 3 dm. apart, up to c. 15 cm. long, at first clad in white woolly tomentum. Lvs erecto-patent, imbricate, suborbicular-obovate to broad-spathulate, truncate or slightly rounded at apex; lamina 4-6 × 4-6 mm., on broad flat petiole c. 2-4 mm. long, both densely clad in subappressed soft white tomentum. Capitula 5-10 together in a ± dense glomerule up to 15 mm. diam., ± sunk amongst the upper lvs which do not form a differentiated involucre. Phyll. ∞, in c. 2 series, scarious, thickened in lower half, c. 8 mm. long, ± narrowly lanceolate, tapering to fine tip. Florets c. 10-15, all narrow-tubular, outer ring of 5-6 ♀. Anthers tailed but those examined lacking good pollen. Style-arms long, truncate. Pappus c. = florets, hairs slender throughout and only slightly barbellate towards tip. Achenes mostly undeveloped, the solitary swollen one seen nearly 2 mm. long, not flattened, brown, minutely papillose.

Growth form, general appearance and lf-texture recall Haastia sinclairii which grows nearby, but the glomerular head and tailed anthers suggest comparison rather with Leucogenes, represented by L. grandiceps in the same general area; with the latter genus most other characters fit except the hairless achenes, slender almost smooth pappus-hairs and lack of distinct involucre. The fact that the collections so far made are not quite uniform suggests a possible hybrid origin.