Changes in local free-living parasite populations in response to cleaner manipulation over 12 years.

Predation on parasites is an important ecological process, but few experimental studies have examined the long-term impacts on the prey. Cleaner fish prey upon large numbers and selectively feed on the larger individuals of the ectoparasitic stage of gnathiid isopods. Removal of cleaner fish Labroides dimidiatus for 1.5-12.5 years negatively affects coral reef fishes, but the mechanism is unclear. A reduction in local parasite populations or the size of individual parasites would benefit all susceptible fishes. We tested whether cleaner presence reduces local gnathiid populations using 18 patch-reefs distributed between two sites (both at Lizard Island, Great Barrier Reef) which were maintained cleaner-free or undisturbed for 12 years. Using emergence traps (1 m2), free-living gnathiid stages were sampled before and after cleaner fish were removed during the day and night, up to 11 times over the course of the experiment. There were effects of the removal in the predicted direction, driven largely by the response at one site over the other involving 200% more gnathiids, but manifested only in the daytime sampling after 4 months. There was also a main effect (36%) for the shared sample dates at both sites after 12 years. Gnathiid size occasionally differed with cleaner presence, but in no consistent way over time. Contrary to our predictions, changes in free-living gnathiid population numbers and their size structure rarely reflected the changes in fish populations and individuals observed on cleaner-free reefs. Therefore, evidence that this predator alone regulates gnathiids remains limited, suggesting other contributing processes are involved.

Relationship between roving behaviour and the diet and client composition of the cleaner fish Labroides bicolor.

Diet analyses and observations of cleaning behaviour of two cleaner fishes revealed that Labroides bicolor fed more on client mucus, but Labroides dimidiatus fed more on ectoparasites, and that L. bicolor interacted with fewer species (36 species) compared with L. dimidiatus (44 species). The client species which contributed most to the dissimilarity between cleaner species were the dusky farmerfish Stegastes nigricans and bicolor chromis Chromis margaritifer damselfishes, which L. dimidiatus interacted with more often than L. bicolor, and the striated Ctenochaetus striatus and brown Acanthurus nigrofuscus surgeonfishes, which L. bicolor interacted with more; L. bicolor interacted with all parrotfishes (Scaridae) more. These results confirm the importance of repeated interactions and partner choice in determining the nature of interactions in mutualisms.

Female cleaner fish cooperate more with unfamiliar males.

Joint group membership is of major importance for cooperation in humans, and close ties or familiarity with a partner are also thought to promote cooperation in other animals. Here, we present the opposite pattern: female cleaner fish, Labroides dimidiatus, behave more cooperatively (by feeding more against their preference) when paired with an unfamiliar male rather than with their social partner. We propose that cooperation based on asymmetric punishment causes this reversed pattern. Males are larger than and dominant to female partners and are more aggressive to unfamiliar than to familiar female partners. In response, females behave more cooperatively with unfamiliar male partners. Our data suggest that in asymmetric interactions, weaker players might behave more cooperatively with out-group members than with in-group members to avoid harsher punishment.

Host specificity of two species of Gnathia (Isopoda) determined by DNA sequencing blood meals.

Host specificity data for gnathiid isopods are scarce because the parasitic stages are difficult to identify and host-parasite contact is often brief. We examined two common nocturnal species, Gnathia falcipenis and Gnathia sp. C, collected in light traps from two locations at Lizard Island on the Great Barrier Reef, Australia. Engorged third stage gnathiids were photographed and permitted to moult into adults to allow identification. We compared approximately 580 bp sequences of 16S mtDNA from blood meals with host sequences available on GenBank using BLASTn. Where homology was <98%, familial identity was investigated with neighbour-joining trees. All blood meal sequences (n=60) and homologous fish sequences (n=87) from GenBank were used in a Bayesian analysis, which identified all but three sequences to family. The host frequency distributions used by each species were significantly different; only four host families were shared. No gnathiids fed on elasmobranchs, blennies or apogonids, and most fed on host families whose representatives are typically large. Gnathia sp. C showed a distinct predilection for nemipterids. Gnathia falcipenis often parasitised sand-dwelling families, and unlike sympatric diurnal gnathiid species, it also frequently parasitised pomacentrids. We conclude that G. falcipenis and Gnathia sp. C operate as generalist micropredators with preferences.

Endoparasite communities of five fish species (Labridae: Cheilininae) from Lizard Island: how important is the ecology and phylogeny of the hosts?

The parasite community of animals is generally influenced by host physiology, ecology, and phylogeny. Therefore, sympatric and phylogenetically related hosts with similar ecologies should have similar parasite communities. To test this hypothesis we surveyed the endoparasites of 5 closely related cheilinine fishes (Labridae) from the Great Barrier Reef. They were Cheilinus chlorourus, C. trilobatus, C. fasciatus, Epibulus insidiator and Oxycheilinus diagramma. We examined the relationship between parasitological variables (richness, abundance and diversity) and host characteristics (body weight, diet and phylogeny). The 5 fishes had 31 parasite species with 9-18 parasite species per fish species. Cestode larvae (mostly Tetraphyllidea) were the most abundant and prevalent parasites followed by nematodes and digeneans. Parasites, body size and diet of hosts differed between fish species. In general, body weight, diet and host phylogeny each explained some of the variation in richness and composition of parasites among the fishes. The 2 most closely related species, Cheilinus chlorourus and C. trilobatus, had broadly similar parasites but the other fish species differed significantly in all variables. However, there was no all-encompassing pattern. This may be because different lineages of parasites may react differently to ecological variables. We also argue that adult parasites may respond principally to host diet. In contrast, larval parasite composition may respond both to host diet and predator-prey interactions because this is the path by which many parasites complete their life-cycles. Finally, variation in parasite phylogeny and parasite life-cycles among hosts likely increase the complexity of the system making it difficult to find all-encompassing patterns between host characteristics and parasites, particularly when all the species in rich parasite communities are considered.

Cleaning symbioses from the parasites' perspective.

Cleaning behaviour has generally been viewed from the cleaner or client's point of view. Few studies, however, have examined cleaning behaviour from the parasites' perspective, yet they are the equally-important third players in such associations. All three players are likely to have had their evolution affected by the association. As cleaner organisms are important predators of parasites, cleaners are likely to have an important effect on their prey. Little, however, is known of how parasites are affected by cleaning associations and the strategies that parasites use in response to cleaners. I examine here what parasites are involved in cleaning interactions, the effect cleaners have on parasites, the potential counteradaptations that parasites have evolved against the predatory activities of cleaner organisms, the potential influence of cleaners on the life history traits of parasites, and other factors affected by cleaners. I have found that a wide range of ectoparasites from diverse habitats have been reported to interact with a wide range of cleaner organisms. Some of the life history traits of parasites are consistent with the idea that they are in response to cleaner predation. It is clear, however, that although many cleaning systems exist their ecological role is largely unexplored. This has likely been hindered by our lack of information on the parasites involved in cleaning interactions.

Cleaner fish Labroides dimidiatus recognise familiar clients.

Individual recognition has been attributed a crucial role in the evolution of complex social systems such as helping behaviour and cooperation. A classical example for interspecific cooperation is the mutualism between the cleaner fish Labroides dimidiatus and its client reef fish species. For stable cooperation to evolve, it is generally assumed that partners interact repeatedly and remember each other's past behaviour. Repeated interactions may be achieved by site fidelity or individual recognition. However, as some cleaner fish have more than 2,300 interactions per day with various individuals per species and various species of clients, basic assumptions of cooperation theory might be violated in this mutualism. We tested the cleaner L. dimidiatus and its herbivorous client, the surgeon fish Ctenochaetus striatus, for their ability to distinguish between a familiar and an unfamiliar partner in a choice experiment. Under natural conditions, cleaners and clients have to build up their relationship, which is probably costly for both. We therefore predicted that both clients and cleaners should prefer the familiar partner in our choice experiment. We found that cleaners spent significantly more time near the familiar than the unfamiliar clients in the first 2 minutes of the experiment. This indicates the ability for individual recognition in cleaners. In contrast, the client C. striatus showed no significant preference. This could be due to a sampling artefact, possibly due to a lack of sufficient motivation. Alternatively, clients may not need to recognise their cleaners but instead remember the defined territories of L. dimidiatus to achieve repeated interactions with the same individual.

Parasite infection rather than tactile stimulation is the proximate cause of cleaning behaviour in reef fish.

Cleaning behaviour is a popular example of non-kin cooperation. However, quantitative support for this is generally sparse and the alternative, that cleaners are parasitic, has also been proposed. Although the behaviour involves some of the most complex and highly developed interspecific communication signals known, the proximate causal factors for why clients seek cleaners are controversial. However, this information is essential to understanding the evolution of cleaning. I tested whether clients seek cleaners in response to parasite infection or whether clients seek cleaners for tactile stimulation regardless of parasite load. Parasite loads on client fish were manipulated and clients exposed to cleaner fish and control fish behind glass. I found that parasitized client fish spent more time than unparasitized fish next to a cleaner fish. In addition, parasitized clients spent more time next to cleaners than next to control fish, whereas unparasitized fish were not attracted to cleaners. This study shows, I believe for the first time, which is somewhat surprising, that parasite infection alone causes clients to seek cleaning by cleaners and provides insight into how this behaviour evolved.