Switching by Lepsiella vinosa (Gastropoda) in South Australian mangroves.

In South Australian mangroves Lepsiella vinosa feeds on two species of barnacles, Balanus amphitrite and Elminius modestus, and only attacks other types of prey if no barnacles are available. Laboratory experiments indicate that L. vinosa can be trained to either species and switches in terms of Murdoch's (1969) model.The distribution of barnacles is heterogeneous and distinct prey patches are found in the mangroves. In high density patches, that is over 25 B. amphitrite per metre length of pneumatophore, L. vinosa concentrated its attacks on B. amphitrite irrespective of the relative abundance of the two species. In moderate density patches, that is less than 25 B. amphitrite per metre length of pneumatophore, L. vinosa fed on the two species as expected on the basis of relative abundance. There was, however, also an association between changes in the absolute density of B. amphitrite and the proportion it formed in the diet.Prey selection in the moderate density patches was not random and was influenced by previous feeding history. L. vinosa which fed on the more abundant species, E. modestus, selected this species at a higher rate than expected in subsequent surveys. This result suggests that "ingestive conditioning" took place in the field, and provides some support for the switching hypothesis.Field cage experiments indicate that L. vinosa shows switching provided that the density of the preferred prey was not too high. L. vinosa had a weak preference for B. amphitrite over E. modestus, c was calculated to be 1.26, when equal numbers of each species were present. The preference appears to be consistent between individual L. vinosa as few individuals fed predominantly on only one prey species.The switching response in L. vinosa was asymmetrical with regard to the two prey species. L. vinosa switched to B. amphitrite when it was more abundant irrespective of prior training. L. vinosa trained to E. modestus, or unfed, prior to the experiment switched to E. modestus when it was more abundant, but, those trained to B. amphitrite did not. L. vinosa also showed asymmetry in its feeding rate. Those trained to E. modestus, or unfed, attacked prey at the average rate observed in the field irrespective of the prey presented. L. vinosa trained to B. amphitrite showed a marked reduction in feeding rate when B. amphitrite was less abundant.Absolute density as well as the relative density of the two prey species influenced prey selection. Higher predation on B. amphitrite was found, despite a fixed relative abundance of one B. amphitrite to five E. modestus, when the barnacle density was increased from 11.2 to 56 per metre of pneumatophore.