Cuscuta chinensis

Stefanović, Mihai Costea Ian Spence Saša, 2011, Systematics of Cuscuta chinensis species complex (subgenus Grammica, Convolvulaceae): evidence for long-distance dispersal and one new species, Organisms Diversity & Evolution (New York, N. Y.) 11 (5), pp. 373-386 : 378

Cuscuta chinensis and Cuscuta applanata : one species with disjunct distributions

In the identification key of his monograph, Yuncker (1932) separated C. chinensis and C. applanata as follows: C. chinensis —“flowers mostly 3–3.5 mm long, styles commonly stoutish and not especially exserted”; C. applanata —“flowers mostly about 2–2.5 mm long; styles commonly slender and exserted”. In the text, however, he described C. chinensis flowers to be 2–3.5 mm long with “stoutish or slender” styles, while C. applanata had 2- to 3-mm-long flowers (the thickness of the styles was not mentioned). After a detailed examination of the types and numerous herbarium specimens we can affirm that the morphological distinction between C. chinensis and C. applanata is tenuous at best. Although we did not perform a morphometric analysis, the morphology and micromorphology of the flowers are remarkably similar in both C. chinensis and C. applanata ( Figs. 1 , 2 ; Costea 2007 -onwards). Flowers and seeds tend to be slightly larger in C. chinensis , but the ranges of variation overlap with those of C. applanata (particularly of the flowers; see the description of C. chinensis ). Indeed, without knowing their geographical provenence, many herbarium specimens of these two taxa would be difficult, if not impossible, to identify as C. chinensis or C. applanata . Together with the phylogenetic data, the morphological similarity strongly suggests that C. chinensis and C. applanata represent one single species with two geographical varieties. Therefore, C. applanata is retained as a variety of C. chinensis , and a new nomenclatural combination is proposed.

Cuscuta chinensis clade is one of the eight major groups of subg. Grammica that have originated and diversified in Mexico and the adjacent areas ( Stefanović et al. 2007), and all other species of this complex are restricted to Mexico and southern US. Therefore, long-distance dispersal is the mostly likely explanation for the presence of var. chinensis in Australasia, Indo-Malaysia and the Asian part of Palearctic. However, in the absence of more data, the timeline and route of dispersal are difficult to hypothesize. Several other strongly supported cases of long-distance dispersal have been uncovered by the phylogeny of Cuscuta subg. Grammica ( Stefanović et al. 2007) , but nearly all these events are likely to be older because they are associated with completed speciation events. For example C. victoriana and C. tasmanica are endemic to Australia yet they belong to a large Mexican and Central American clade of subg. Grammica (“clade G”; Stefanović et al 2007). Similarly, C. hyalina from Eastern Africa, India and Pakistan is the only species of the C. umbellata complex, a clade otherwise distributed from the southern US to northern South America ( Costea and Stefanović 2010). While these, presumably older, long-distance dispersal events also await an explanation, at the opposite end of the time scale are cases of dodder species that have managed to spread over large geographical areas in the last two centuries following anthropogenic pathways. One such example is C. campestris . This weedy species is inferred to be North American in origin but has dispersed successfully worldwide despite the protection and quarantine legislative measures adopted by most countries (reviewed by Costea and Tardif 2006). Yet there is no indication that populations of C. campestris found on different continents have acquired sufficient genetic and/or morphological divergence to justify even the recognition of infraspecific taxa ( Costea et al. 2006a). Instances of putative long-distance dispersal events that involve allopatric varieties are less common and not as distant geographically as the C. chinensis case. For instance, C. obtusiflora var. obtusiflora is encountered in South America while C. obtusiflora var. glandulosa and the majority of the clade members (“clade B”, Stefanović et al 2007) are centered in North America. In the case of C. chinensis , more sampling from Asia- Australia and the use of faster evolving molecular markers would be necessary to recover the routes of migration that led to its extensive geographical distribution.

Figures

Fig. 1 a–i Morphology of dissected calyx in species of Cuscuta chinensis complex. a,d Cuscuta chinensis var. chinensis. b,e C. chinensis var. applanata. c,f C. alata. g C. potosina. h C. azteca. i C. yucatana. Bars 1 mm

Fig. 2 a–g Scanning electron micrographs of multicellular protuberances on the calyx of Cuscuta chinensis species complex. a C. chinensis var. chinensis. b C. chinensis var. applanata. c–e C. alata. f