The arenophilic system of anomalodesmatan clams comprises multicellular glands, muscular papillae, and radial lines of adhesive secretion, which glue sand grains and other extraneous particles to the external surface of the shell. The presence and morphology of these organs have been established in an increasing number of taxa, but arenophilic glands were not previously known to occur in the deep-sea, carnivorous family Cuspidariidae. Dissections, histological sections and ultrastructural examination of members of the cuspidariid genera Cuspidaria and Cardiomya revealed a functional arenophilic system positioned around the siphonal apparatus. Glands are contained in muscular papillae at the posterior tip of the siphonal sheath, from where they deposit their adhesive secretion onto the shell rostrum. Discovery of an arenophilic system in cuspidariid clams adds support to the idea that the glands are a synapomorphy of Anomalodesmata, secondarily absent in only one of the major component branches (the clade joining Thraciidae, Cleidothaeridae, and Myochamidae). Moreover, similarity between the arenophilic papillae we observed on the siphonal sheath of Cuspidaria, Cardiomya, and Myonera, and literature reports on the crown of tentacles of Multitentacula is puzzling and highlights the need for further studies of the latter genus.

Successions of erroneous taxonomic identifications in biodiversity studies are an undesirable but not rare phenomenon. Such cascades of errors frequently involve taxa with few, polymorphic diagnostic characters, such as representatives of the cuspidariid genus Cardiomya. Four species of the genus are believed to live in Brazilian waters: C. cleryana (d'Orbigny, 1842), C. ornatissima (d'Orbigny, 1853), C. striata (Jeffreys, 1876) and C. perrostrata (Dall, 1881). Many reported occurrences of these species are based on misidentifications, but the error cascade of incorrect records of C. perrostrata is by far the most extensive. Although the species has been reported to live in Brazilian waters for about half a century, all previous records seem to be based on a succession of mistaken identifications. This paper redescribes C. perrostrata based on an analysis of the type series, several lots of museum vouchers and new material collected in the Campos Basis (Rio de Janeiro, Brazil). Ontogenetic changes in shell outline and ultrastructural details of the hinge plate, periostracal surface and larval shell are described and illustrated for the first time. Based on a critical appraisal of the literature, we present an updated geographic and bathymetric distribution of the species. Cardiomya perrostrata displays a gradual increase in the relative length of the rostrum during its ontogeny and in larger specimens the rostrum is more prominently curved upwards. The larval shell is of type 2D, measuring about 215-235 μm in length. The periostracum is minutely pitted and a lithodesma is present. Its distribution is Western Atlantic, ranging from 42°N to 30°S, in depths of 18 to 1646-1829 m. Cardiomya knudseni (Allen & Morgan, 1981), a possibly related amphi-Atlantic species whose type material has been lost, is regarded herein as a nomen dubium.

No abstract