Preoccupation with the identification of the Mexican species of Rhynchosia resulted in a previous paper (Turner, 2011) which recognized R. texana as an amphitropical taxon occurring in both North America and South America. Grear (1978) and Fortunato (1983) also recognized the inter-continental distribution of the latter but preferred to treat these as but varieties of an expanded R. senna; so treated the correct varietal name becomes R. senna var. texana (T. & G.) M.C. Johnston. When treated as a more restricted bicentric species the name becomes R. texana T. & G.

In Grear's study, and that of Fortunato, I was surprised to note that they also recognized another amphitropical disjunct, namely Rhynchosia diversifolia var. prostrata Burkart, the latter typified by material from Argentina, but also occurring in Brazil, Uruguay and, surprisingly, northeastern Mexico (Nuevo Leon and Tamaulipas). Intuition and preliminary examination of the materials concerned led me to believe that the relatively rare Mexican specimens might represent a novel taxon, but after examining material of the taxa concerned, I concluded, like both Grear and Fortunato, that the variety does have a bicentric distribution (Figs. 1, 2) and, as with R. texana, I prefer to treat this at the specific level:

The species is named for its amphitropical or bicentric distribution, occurring both north and south of the New World Tropics. In South America the taxon is relatively common. In Mexico it is relatively rare, having been collected at only three sites as listed in Grear's treatment (by Berlandier 3131, ‘‘River de l'aronsara near de Galian,'' May 1834 [at least 5 sheets mounted], and Mueller & Mueller 501, in 1933 [4 sheets], both from the vicinity of Monterrey, and a single sheet from the vicinity of San Jose, Tamaulipas, ‘‘Mesa de Tierra.'' Bartlett 10468, 19 Jul 1930). So far as known, no subsequent collections have been assembled from the area, in spite of much collecting by recent professionals.

In my taxonomic assessment of the above taxon, I examined material from all of the Mexican sites examined by Grear (as noted in the above), and numerous sheets from South America (LL-TEX), also examined by the latter worker. Among the sheets examined, I found no morphological evidence to suggest that natural hybrids between the two taxa might be occurring, although the occasional intermediate might be expected in regions of overlap. In short, I agree with Grear that the varietal taxon, like R. texana, has a bicentric distribution. Much of my conviction stems from the comments of Grear, who notes:

The characters used to divide Rhynchosia diversifolia into two varieties are given in Grear's key [p. 90]:

Fig.1.

Distribution of Rhynchosia bicentrica in North America.

Grear notes further [p. 91]:

Like the typical variety, Rhynchosia diversifolia var. prostrata is most closely related to R. senna and exhibits essentially the same kind of disjunct distribution as shown by R. senna var. angustifolia [= R. texana] in Mexico and South America. Although the occurrence in Mexico is restricted, the explanation for the amphitropical distribution of R. senna var. angustifolia applies to this case as well.

Rhynchosia diversifolia var. diversifolia has often been confused with R. senna var. senna, as indicated by the synonymy of the former. Where sympatric they appear to hybridize to some degree. Nevertheless, they are strikingly different in fruit and leaflet characters and the occasional cross is rare. Each retains it own peculiarities in areas of sympatry.

Fig. 2.

Distribution of Rhynchosia bicentrica in South America.

It should be emphasized that Grear, accepted the rare hybrid within the Rhynchosia senna complex, but in areas of sympatry within the R. diversifolia complex could not account for a single intermediate, hence my acceptance of var. prostrata as a distinct species, R. bicentrica. No doubt my long-time familiarity with hybrid populations of legume species in Texas and elsewhere (e.g., Baptisia spp., Turner, 1963) has colored my thinking, but I believe my biological intuition as to the specific status of the taxon concerned is sound.

Fortunato (1983) provided a somewhat more expanded, detailed, key to the two varieties of Rhynchosia diversifolia, but did not take issue with Grear's comments regarding the lack of intermediates in regions of sympatry, nor did she cite specimens suggestive of such contamination.

Indeed, Burkart (1987), in his exceptional treatment of Rhynchosia for Entre Rios, Argentina provided an excellent illustration of the varieties concerned, the differences between them quite striking. He noted further that Furtunato (1983) called to the fore (from among many sheets examined) only two specimens examined by her that seemed to be intermediate in habit, but had all the other characters that defined var. prostrata. In short, to my knowledge, no one has found a true intermediate between the taxa concerned, in spite of the fact they are sympatric over a large region, are represented by numerous collections in herbaria, and have been examined by several exceptional scholars of the complex.

Distribution of Rhynchosia bicentrica on the two continents is shown in Figs. 1 and 2, this mostly based upon specimens at LL-TEX, and the distribution maps provided by Grear; the latter worker also provided a distribution map for R. diversifolia (his Fig. 13).

Finally, one of the reviewers of the present paper suggested that I might best summarize my paper with the following: