BIDENS ANDICOLA H.B.K., Nov. Gen. et Sp. 4:237 (186). 1820; B. andicola H.B.K. vars. normalis and heterophylla O. Kuntze, Rev. Gen. PI. 3^II^: I36. 1898; B. fruticulosa Mey. and Walp., Nov. Act. Nat. Cur. 19 Supplem. I. 271. 1843.

Descript. amplific.- Herba perennis, semi-procumbens vel etiam erecta, valde hispido-pubescens vel fere glabra, ramosa, 2-8 dm. alta, caulibus parce angulatis. Folia 1-7 cm. longa, valde polymorpha; nunc indivisa, ovata, serrata, sessilia vel alato-petiolata, ad apicem obtusa vel subacuta; nunc tripartite vel 1-3-pinnata foliolis ovatis vel lanceolatis vel linearibus et ad apicem sensim vel abrupte apiculatis. Capitula ramos terminantia, longe pedunculata, radiata; pansa ad anthesin 2-4 vel rarius etiam usque ad 5.5 cm. lata, 0.7-1.4 cm. alta. Involucrum perspicue hispidum, bracteis ex-terioribus 8-10, lanceolatis vel lineari-oblongis, ciliatis, supra saepe glabratis, apice plerumque obtusis, quam interioribus lanceolatis dense hispidis plerumque multo brevioribus. Flores ligulati saepius 8, lutei, ligula elliptico-oblanceolati, apice plerumque minute 3-denticulati, 1.2-2.5 cm. longi. Achaenia tenuiter linearia, inferne sensim attenuate, obcompresso-quadrangularia, sulcata, supra plus minusve erecto-hispida, fusco-nigra, corpore 0.7-1.4 cm. longa et 0.4-1 mm. lata et paleas demum superantia, apice bi- (vel pauca tri-) aristata, aristis tenuibus, brunneo-stramineis vel rubescentibus, re-trorsum hamosis, 1.7-3 mm. longis.

BIDENS ANDICOLA var. DECOMPOSITA O. Kuntze, I.c.; B. macrantha Griseb., Abhandl. Goett. I9:I38 I874; B. grandiflora Balb. var. breviloba 0. Kuntze, I.c.-Folia 2-3-pinnatisecta, usque ad I dm. longa, achaeniis superne valde attenuato-elongata.

For many years the identity of the South American Bidens andicola has been obscured for herbarium workers by the great multiplicity of foliage forms encountered. WEDDELL, as early as 1856 (Chloris And. 1870) described it as a polymorphous plant ("Plante polymorphe et très repande dans la chaine, mais presque exclusivement alpestre"). Later, OTTO KUNTZE, who like WEDDELL had collected in South America, commented upon the variability of the leaves ("Eine robuste Art mit einfach oder mehrfach ternatisecten Blättern, mittelgrossen gelben Strahlblüthen, ziemlich grossen Blüthenköpfen, äusseren zottig behaarten Involucralbracteen etc., aber in Bezug auf Blatttheilung wie manche andere Bidens-Art sehr variabel"; Rev. Gen. Pl. 3^II^: 136. 1898). In herbaria the numerous foliage forms are seen to simulate corresponding forms of B. triplinervia H.B.K. (B. humilis H.B.K., B. crithmifoliac H.B.K., etc.), and this has led often to confusion between the two species. Recently I was enabled, through the courtesy of OTTO BUCHTIEN (cf. SHEREFF, BOT. GAZ. 76: 151. 1923), to study a great number of specimens collected by him and displaying a wide range of variation. From these (all in Herb. Field. Mus.) and many others, totalling more than two hundred specimens, the preceding descriptions are drawn. It was found that sometimes, in poorly developed material, distinction from B. triplinervia is apparently impossible. In well developed material, however, the distinctions are usually very definite, B. andicola being coarser, its thicker heads having commonly about eight instead of commonly about five rays[4], etc. B. andicola has the paleae shorter than the mature achenes and this character separates it from the surprisingly similar aggregation of Mexican forms (Purpus 1547, 1548, 2637, 4135, 5089, 5620; Rose and Painter 6666, 7949; Pringle 4915; E. W. Nelson 3220, etc.) that in late years have passed erroneously under the name B. daucifolia DC. In the latter[5] the paleae are usually very blackish above and commonly surpass the mature achenes.

Occasionally a form of B. andicola is found with the leaves highly compound and the achenes strongly narrowed above, somewhat like those of Cosmos. If it were not for various connecting forms this would seem to be specifically distinct. KUNTZE, who himself collected specimens of it, referred at least one of them, a plant from Cochabamba, Bolivia (Herb. N.Y. Bot. Gard.) to B. andicola, naming it var. decomposita. In a careless moment he named a precisely identical form from between Cochabamba and Rio Juntas, Bolivia (Herb. N.Y. Bot. Gard.) B. grandiflora Balb. var. breviloba, although B. grandifjora is a Mexican species and is not known to occur in South America.

[4] Unfortunately, B. triplinervia produces at times an 8-rayed form. Discussion of this form must be deferred until a later date.

Sherff, E. E. (1926). Studies in the Genus Bidens. VII. Botanical Gazette, 81(1), 25-54.>>