Replacement of an inherited stretch receptor by a newly evolved stretch receptor in hippid sand crabs

Primary sensory neurons that are motoneuron-like in morphology and often nonspiking (transmit afferent signals as graded depolarizations) characterize an unusual type of stretch receptor in decapod crustaceans. Nonspiking and spiking receptors occur in similar positions at homologous joints in different species and have been presumed to be homologous, the spiking one considered "primitive". To better understand the evolutionary origin of these stretch receptors and why some are nonspiking, we examined the spiking telson-uropod stretch receptors in the spiny sand crab Blepharipoda occidentalis (Albuneidae) and the squat lobster Munida quadrispina (Galatheidae) and compared them with the nonspking telson-uropod stretch receptor of the mole sand crab Emerita analoga (Hippidae). The position, morphology and responses to stretch of the sensory neurons, and the ultrastructure of the elastic strand portion of the receptor are similar in M. quadrispina and B. occidentalis, except that in B. occidentalis the receptor muscles are substantially smaller and the extracellular matrix of the elastic receptor strand is both more extensive and more organized, reminiscent of the ultrastructure of E. analoga's nonspiking receptor. We conclude that the spiking telson-uropod stretch receptors of albuneids and galatheids are homologous. The differences in the ultrastructure of their receptor strands imply that the efficiency of coupling receptor length change to deformation of the dendritic termini increases in the order M. quadrispina < B. occidentalis < E. analoga. The spiking and nonspiking telson-uropod stretch receptors differ anatomically in three major respects that appear to preclude their homology. (1) The receptor strands are on opposite sides of a conserved muscle. (2) The sensory somata are in different regions of the sixth abdominal ganglion: a lateral cluster of somata for the spiking sensory neurons and two medial clusters, one anterior, one posterior, for the nonspiking sensory neurons. (3) The neuropil projections of the sensory neurons are different. We conclude that the hippid's nonspiking telson-uropod stretch receptor evolved de novo and not by modification of the ancestral anomuran telson-uropod stretch receptor (which Hippidae have lost).