Host-specific performance and host use in the kleptoparasitic marine snail Trichotropis cancellata

Generalist parasites may disproportionately use certain hosts because of different benefits associated with each host species. I measured the growth rate of the marine snail Trichotropis cancellata, a facultative kleptoparasite that can suspension feed and steal food, on different hosts to determine the relative nutritional benefits of each host. The variation in tentacle (feeding structure) area among the hosts studied had the potential to provide parasitic snails with different amounts of nutrition for growth. In field experiments, suspension-feeding snails isolated from potential hosts grew at a similar rate to snails on brachiopods and significantly more slowly than snails on the following polychaete worms: Serpula columbiana (Serpulidae), Pseudopotamilla ocellata (Sabellidae), Schizobranchia insignis (Sabellidae), and Eudistylia vancouveri (Sabellidae). However, choice among worm hosts affected snail growth rates only in the fall, when phytoplankton levels are low. At this time, snails parasitizing the sabellids Schizobranchia and Eudistylia grew more quickly than snails on Serpula. In the spring and summer, with high levels of phytoplankton, Trichotropis grew at similar rates on all worm species tested. Trichotropis spent approximately the same time stealing food from each worm host species, >50% of the time the worms had their tentacles extended (the difference among hosts was not significant). This finding demonstrates that the similarity of snail growth rates on different worm species is not due to the snails compensating for poor hosts (worms that provide food at a slower rate) by spending more time stealing food. Snails in choice experiments preferred live Serpula to empty Serpula tubes, indicating that at least some of the cue(s) snails use to identify hosts are derived from living host tissues. In choice racks containing live Serpula and live Schizobranchia, snails did not choose one host worm significantly more often than the other. Because Trichotropis grows faster on sabellids than serpulids in the fall, I predicted that snails in nature would infect sabellids more often than other species. However, snails were usually distributed randomly among host species. In the few cases where the snails showed a significant preference among host species, proportionally more snails were found on serpulids than on sabellids or sabellarids. This study is the first to quantify under natural conditions the growth benefits of a kleptoparasite across the range of possible hosts, and implies that factors other than growth rate influence host choice specificity in the marine kleptoparasite T. cancellata.