To sea or not to sea? Flexible life history strategies in diadromous fishes.

A project undertaken at the University of Otago, supervised by Gerry Closs, and research conducted by Andy Hicks.


Movement patterns of pelagic larvae are a mystery, both in terms of the spatial extent of migration and why they do it. Coral, reef-fish, lobsters, abalone, seastars, fanworms and amphidromous fishes all exhibit a pelagic larval phase that has the potential for widespread dispersal. Until recently, dispersal was assumed to be the primary reason for this larval life-history stage, allowing pelagic larvae to drift and connect what would otherwise be isolated populations of adults. Dispersal was seen as a form of bet hedging against unpredictable local conditions, and an ability to colonise isolated habitats was seen as the primary benefit offered by a dispersive larval phase. Increasingly, this perspective is being questioned in multiple taxa with suggested alternative benefits of a pelagic larval period including gaining access to feeding opportunities, breaking cycles of parasitism, and avoiding predation. Under this alternative hypothesis, dispersal is an 'accidental' benefit rather than the primary reason for a pelagic larval period.

Project aims and methods

In this study, we concentrated on patterns of dispersal in amphidromous fishes. Amphidromy is a life history whereby fish spend most of their lives and spawn in freshwater, but rear as pelagic larvae in the sea or sometimes freshwater lakes. It has generally been assumed that the primary function of amphidromy is dispersal of larvae. The frequent observation of fish rearing in lakes, however, challenges the view that larvae disperse widely.

We analysed the otolith microchemistry of galaxiid and bully species around New Zealand and south east Australia. Otoliths are fish earstones comprised of daily deposited layers of calcium carbonate. Distinct differences in the trace element composition of water bodies a fish inhabits are incorporated into the otolith layers and provide a permanent environmental history of a fish’s life. Anlaysing the otolith microchemistry of amphidromous species allowed us to explore the movement patterns of larval stages. Specifically, we determined whether fish spent their larval stages in freshwater lakes. We also explored the extent of exchange among populations in different river and lake systems and whether large scale larval dispersal occurred.

Research findings

Otolith analyses in all species studied suggested the scale of larval movement is far less than currently assumed. Migration histories of all ‘amphidromous’ species completed to date (Galaxias brevipinnis, G. maculatus, G. argenteus, G. fasciatus, Gobiomorphus cotidianus and G. huttoni) included non-diadromous representatives. In other words, all of these species can complete their life cycle in freshwater, rather than requiring larval development at sea. Freshwater larval development occurs in lakes and lagoons, and means the ‘subpopulations’ associated with these lakes and lagoons are relatively isolated in terms of exchange with the regional  ‘metapopulation’.

Limited dispersal was inferred from evidence of distinct larval pools in amphidromous species (Galaxias maculatus, G. brevipinnis, Gobiomorphus huttoni) inhabiting river systems that flowed directly to the sea. Diadromous individuals (sea-going) within river systems shared larval environmental histories more similar to each other than diadromous individuals from other river systems. The river mouths of these systems were separated by less than 20km. These results suggest that even in an open coastal setting, minimal movement of larvae occurs and mixing is far more restricted than expected.

Although freshwater development was observed in all species, any one lake only supported non-diadromous recruitment in a subset of the species inhabiting its catchment. Low productivity, coastal glacial lakes on the West Coast of the South Island, New Zealand (Lake Paringa and Lake Moeraki) supported non-diadromous recruitment in Galaxias brevipinnis but not Galaxias argenteus. High alpine lakes (cool and with low-productivity) also support non-diadromous recruitment in G. brevipinnis. Conversely, high productivity freshwater lakes and lagoons on the South Island (Lake Waihola and Waituna Lagoon) supported non-diadromous recruitment in G. argenteus but not G.brevipinnis. A third species, Galaxias fasciatus, was only found to exhibit high rates of non-diadromous recruitment on the north island of New Zealand. We think these system/species-specific patterns in non-diadromous recruitment relate to different larval habitat requirements of the different species being met by some lakes and not others. This idea has been met with keen interest, and we are pursuing further funding to explore larval habitat requirements of the different species.

Investigations into the population structure of G. brevipinnis with increasing distance inland showed a strong supply-limitation effect. Adult densities throughout most catchments were limited by the number of juvenile fish coming into the system, rather than the amount of habitat available to adults. When evidence of supply-limitation is combined both with evidence for larval retention and the obvious but hitherto neglected idea of species-specific habitat requirements of pelagic larvae, a new picture of amphidromous population dynamics emerges. Namely, the adult density of amphidromous species will be limited by the proximity of high quality larval habitat.

Given our observation of limited dispersal by the larvae of amphidromous fish, we suggest the concept of ‘amphidromy’ needs revisiting. Our preferred explanation is that amphidromy is an ontogenetic migration between adult habitat (usually flowing freshwater) and a pelagic larval habitat (the lake or the sea); salinity is irrelevant.  An ‘amphidromous’ life-history thus revolves around the fecundity advantage conferred to mothers that produce smaller, more numerous offspring when pelagic habitat is available for rearing small larvae. We argue that dispersal did not drive the evolution of amphidromy, and so we should not expect amphidromous larvae to actively move further than the first suitable larval habitat.

Our work will benefit Australia and New Zealand by determining where, how and in what numbers, key species of diadromous fish recruit within regions, catchments and sampled sites. Several species (G. argenteus, G. fasciatus, G. truttaceus & G. huttoni) are listed as vulnerable and of conservation significance in Australia & New Zealand. The research will be applied directly to restoration projects being undertaken by national, state and local government agencies, indigenous and local community groups, enabling identification of key requirements for the sustainable management of fish populations, and improving criteria for monitoring the success or otherwise of restoration projects. Government agencies will benefit significantly through improvements to biomonitoring programs that use the composition of fish communities to assess aquatic ecosystem health due to an improved ability to predict where diadromous fishes should occur. Such fish typically form a significant component of coastal catchment freshwater fish communities throughout New Zealand and south east Australia.


HICKS, A. S., CLOSS, G. P. & SWEARER, S. E. 2010. Otolith microchemistry of two amphidromous galaxiids across an experimental salinity gradient: A multi-element approach for tracking diadromous migrations. Journal of Experimental Marine Biology and Ecology, 394, 86-97.

Project Team

Andy Hicks (Phd student – University of Otago)
Dr Gerry Closs (Supervisor – University of Otago)
Dr Jon Waters (Co-Supervisor – University of Otago)
Dr Bruno David (Collaborator – NZ Environment Waikato)
Dr Rick Stoffels (Collaborator – Murray-Darling Freshwater Research Centre)

For more information please contact Andy Hicks ( or Gerry Closs (