Rhynchospora tenuis together with close allies form a large species complex, which is widely distributed from Central America and the West Indies southwards to southern Brazil and northern Argentina. Observations made on the ample collections of the complex from its total range proves that only R. graminea can be separated specifically from the rest of the complex. In R. graminea the lower glumes of spikelets are conspicuously coriaceous and the very small style-bases are much narrower than the achene body, whereas in the remainder of the complex the glumes are membranous and the well-developed style-bases are almost as wide as the body of achene.
Cyperologists have not agreed as to whether or not R. emaciata is specifically distinct from R. tenuis. As far as the type specimens are concerned, the distinction between the two entities consists mainly in the inflorescences, which are decompound with many elongated corymb rays. Namely, in R. emaciata the ample corymbs are decompound with elongated rays, and the spikelets are solitary at the apex of raylets, whereas in R. tenuis the relatively small corymbs are semicompound to compound with short rays, and the spikelets are mostly paired or digitate in three or four. Lindman, Clarke, and Chodat regard R. emaciata as a variety of R. tenuis, but without justification. Kükenthal in his monograph claims that the two are specifically distinct for the following reason. In R. emaciata its somewhat larger spikelets are solitary and range from 5 to 6 mm in length, its culms are obtusely trigonous without grooves on the sides, and its lighter-colored achenes are crowned with an entire-margined style-base; whereas in R. tenuis the slightly smaller spikelets are clustered and range from 4 to 5 mm in length, its culms are channelled on the sides, and its darker-colored achenes bear a style-base that is lobed at base.
In my observation, the characters of spikelets, achenes, and culms are not always combined as mentioned by Kükenthal. The divisions made by Kükenthal on the basis of the spikelet length, and whether the spikelets are solitary or clustered, are artificial and invalid, since such characters are completely clinal marking no discontinuity at all. The shape of the style-bases are very variable making it impossible to draw a line between entire and lobed bases. Whether the culms are channelled or convexed on sides appears to be the matter of the age of plants, i e, the young plants have either acutely triquetrous culms with often channelled sides, while in the more aged plants the culms tend to be obtusely trigonous to subterete. Rhynchospora emaciata and R. tenuis merely represent two conditions of the same taxon, i e, a well-developed status with copiously branched inflorescences and a small status with less-branched inflorescences. These two forms, which are connected with varying intermediate forms, show no geographical separation either. Ecologically, the R. emaciata type may be found more often in marshes of peaty soil, while the R. tenuis type more commonly occurs in savanna grasslands of especially sandy soil. Rhynchospora spruceana, which was originally separated by its broader leaf-blades and longer spikelets, is not discernible as it completely falls under the category of the R. emaciata type. Not seeing any character to separate these three entities I unite them under R. tenuis.
In southern Brazil and adjoining eastern Bolivia, Paraguay, and northeastern Argentina occurs a variant of R. tenuis which can easily be segregated from typical form by its much smaller habit and densely flowered corymbs, Sharp differences between this variant and the typical form exist in the floral details. In the variant its shorter anthers are only 0.8 to 1 mm long, its spikelets are curved at maturity and bear thinly membranous glumes, while in typical R. tenuis the much longer anthers are 2 to 3 mm long, and straight spikelets bear thickly membranous to chartaceous glumes. The spikelets in the variant are much shorter than those of typical R. tenuis at 2 to 4 mm vs 4 to 9 mm in length, but this range difference does not make a clear boundary. The corymbs in the variant tend to be more spaced occupying the upper 2/3 to 1/2 of the culms, while in typical R. tenuis the corymbs are close together on the apical part of the culms.
Nees in 1842 has already described this variant from Sao Paulo under the name R. capillaris var minor. The Nees’s name, however, is invalid because the binomial, R. capillaris, is an invalid name. A new name, subsp austro-brasiliensis, is given to this variant.
I recognize R. riparia, but as a subspecies of R. tenuis. Subspecies riparia differs from subsp tenuis primarily in its thickish canaliculate leaf-blades and congested inflorescence with pale-brown sessile or subsessile spikelets, in contrast to thin flattish leaf-blades and open corymbose inflorescence with brown to castaneous peduncled spikelets. As to the floral parts, anthers and achenes of subsp riparia are generally smaller than those of subsp tenuis at 1.8-2 mm vs 2-3 mm long, and 0.6-0.8 mm long vs 0.8-1.0 mm long, respectively. In subsp riparia, under emerged and submerged conditions the culms may continue to grow without producing an inflorescence, thus becoming the shape of a stolon. Short secondary flowering culms then arise from nodes of fallen culms. This phenomenon has not been noted in subsp tenuis. Both subspecies, however, are very similar to one another not permitting specific distinction.