Smelling home: the use of olfactory cues for settlement site selection by coral reef fish larvae

Most coral reef fish larvae hatch with well-developed sensory systems, and are thought to utilize a range of sensory cues for navigation in the pelagic environment and in settlement site selection. While visual, auditory and olfactory cues are all known to be important, our understanding of how, when and where each sense is used is incomplete. Olfaction is believed to be particularly important in habitat selection. However, reef fish larvae may respond to a range of different olfactory cues at different developmental stages, with olfaction involved in everything from navigation towards reefs to choice of specific settlement sites. This thesis combined novel laboratory and field-based research techniques to explore the use of olfactory cues for orientation and settlement site selection in coral reef fish larvae, using anemonefish as a model group.\ud \ud It is known that olfactory cues are used in selection of specific microhabitats by coral reef fishes, but olfaction may also be relevant on larger spatial scales, to orientate towards specific reefs. In Chapter 2, I tested whether newly recruited anemonefish (Amphiprion percula) can distinguish between different reef habitats using olfactory cues and assessed whether attraction to certain cues is consistent with field-based settlement patterns. A. percula recruits displayed strong preferences for olfactory cues in water collected adjacent to reefs with vegetated islands compared with submerged reefs where no islands were present. These fishes also showed a strong preference for water treated with the olfactory cues of leaves from rainforest plants. The two anemones utilized by Amphiprion percula are found almost exclusively on reefs associated with vegetated islands, therefore leaves could provide a good olfactory cue to locate appropriate reefs on which to settle.\ud \ud Strong preferences for terrestrial leaf odour and the ability of A. percula to distinguish between reefs suggests that the use of terrestrial borne olfactory cues for island reef identification may be significant to a variety of island-dwelling reef fishes. In Chapter 3, I tested the generality of these findings for other fishes that are found predominantly on reefs associated with vegetated islands. Eight island-associated species, from 3 families (Chaetodontidae, Pomacentridae and Labridae) were tested for island odour preferences and the use of terrestrial cues in island identification. Juveniles of all species displayed a preference for both island reef cues and terrestrial leaf litter, highlighting the potential importance of terrestrial cues in the island-reef ecosystem. The strong preferences for these cues across a wide range of fishes highlights the need for integrated terrestrial and marine management for the conservation of coral reef fishes.\ud \ud Although settlement-stage reef fish larvae possess olfactory preferences involved in settlement site selection, it is not known at what stage the ability to distinguish olfactory cues becomes established. In Chapter 4, the responses to specific olfactory cues were tested throughout the entire larval stage for two species of anemonefish (A. percula and Amphiprion melanopus). A clear preference for untreated water over reef-based cues occurred during the first three days post-hatching. However, approximately halfway through the larval phase, at seven days post-hatching, larvae displayed a preference for reef-based olfactory cues, which strengthened until nine days post-hatching. These results indicate that reef fish larvae are capable of distinguishing olfactory cues almost immediately after hatching, but their responses to reefbased cues changes during development; initially newly hatched larvae are repelled by reefbased cues, which would tend to take them into the pelagic environment. These findings suggest that olfaction may be important in both dispersal and settlement of larval reef fish.\ud \ud Recent studies show that many reef fishes settle on natal reefs, which may be due to imprinting on specific reef cues during early development. In Chapter 5, I examined the olfactory preferences of imprinted larval anemonefish in the laboratory and then tested the effect that innate preferences have on imprinting in the natural system. Laboratory reared larval A. percula exhibited the ability to imprint on chemical cues after a 12-hour window of exposure. Chemical cues which larvae displayed a strong innate preference to (e.g. host anemones) resulted in only a slight increase in olfactory preference after imprinting. In the field, larvae with the opportunity to imprint on the olfactory cues of one of two host-anemones did not display a significantly increase in the probability of returning to that species of host anemone once settlement stage was reached. The capacity for anemonefish to imprint on different chemical stimuli suggests it may be important in nature. While strong settlement preferences are exhibited, which may be refined by imprinting, settlement site selection ends up being largely governed by availability.\ud \ud In the natural system, settlement-stage larvae are exposed to a magnitude of different odour cues and must respond appropriately to the information received. In Chapter 6, I determined the olfactory decision-making ability of three species of anemonefish, testing whether predation risk moderates settlement site selection. Settlement preferences, in all species, were conditional to the choices available, with larvae always choosing the host anemone or low-risk option compared with a non-host anemone species or high-risk areas indicated by the olfactory cues of a predator. These results demonstrate that olfactory cues can be used by larval reef fishes for complex risk assessment during settlement site selection. However, locating the correct habitat appears to be the most important factor when selecting a settlement site.\ud \ud Although emphasis is placed on selecting the correct habitat, the importance of predator recognition cannot be discounted. It has recently been demonstrated that the ability of fish larvae to discriminate between olfactory cues of settlement habitat is impaired by elevated levels of dissolved CO₂. In Chapter 7, eggs and larval A. percula were reared in water treated with increased CO₂ levels. I tested the effect elevated CO2 has on predator recognition in both newly hatched and settlement-stage coral reef fish larvae. Larvae, regardless of age, normally displayed an innate avoidance of odour cues indicating the presence of a predator. Newly hatched larvae exposed to the elevated CO₂ treatment exhibited expected patterns of olfactory discrimination. However, settlement-stage larvae treated with elevated CO₂ were incapable of discriminating between olfactory cues of predators and non-predators. The disruption of larval behaviour could have profound effects on population replenishment and connectivity patterns of many marine species.\ud \ud This thesis provides strong evidence for the importance of olfaction in the decisions made by reef fish larvae, and highlights the complexity and sophistication of this remarkable sensory mechanism. Future research into the way chemical cues interact with other important sensory information will significantly improve our understanding of the critical factors that give coral reef fish larvae the ability to find their way home