SynopsisObservations of reproductive features and body measurements were made on wild-caught, freshwater stingrays, Potamotrygon circularis and P. motoro, from the Amazon drainage of western Brazil and southern Colombia. Further observations were made in Detroit's Belle Isle Aquarium on a captive pair of P. motoro and their descendants, which constitute the first known captive breeding colony of potamotrygonids. The gross structure and function of female and male reproductive systems are described. There is no obvious difference between those of the two species. They are aplacentally viviparous, the young being nourished in advanced stages by uterine milk secreted by trophonemata. Size at onset and completion of sexual maturation, breeding season and behavior, gestation period, litter size and sex ratios are discussed. Up to 21 proportional measurements were made on several fetal and postnatal stages of both species. Several proportional changes occur in very early fetal life, but most body proportions undergo only minor changes from advanced fetal through adult stages. A growth curve is proposed for P. motoro based on observations of the captive colony.

SynopsisThe number of venomous caudal spines and their length and position relative to one another were determined in seven species of South American freshwater rays (Potamotrygonidae) and eight marine or euryhaline species of four families from the Caribbean Coast of South and Central America. Most species have two visible spines at certain stages in the shedding-replacement cycle and only one visible spine at other stages (following shedding). If we include the embryological beginnings of the spines before they erupt and become visible, the spine counts of most rays are actually 2 rather than 1 or 2. Since most species apparently follow this pattern, spine counts are of little use in distinguishing between species except in the relatively few that may have only one, or no spines. Eight captive Potamotrygon specimens maintained in simulated tropical temperature conditions over 12 months showed periodic shedding and replacement of spines. The molts were biannual for a given ray but annual for a given spine. They alternated between two spine loci and their cycles were approximately six months out of phase with each other. Recent studies on Dasyatis sabina by others report only one molt per year, with replacement spines forming always posterior to the primary spine rather than alternating between posterior and anterior. Supernumerary spines (counts of more than two, up to five) are also discussed, as are counts of one and zero.