Degree Name

Doctor of Philosophy


Department of Environmental Science


The structure and dynamics of intertidal rockpool, or tidepool, fish assemblages were studied along the south coast of New South Wales (NSW), Australia. This work aimed to fulfil a number of objectives that contribute to three key components in fish ecology. These are: 1) the structure and patterns of variation in rockpool fish assemblages, 2) role of fish movements (recruitment, relocation and homing) in influencing fish population dynamics and 3) the role of structural heterogeneity in structuring fish assemblages.

Since sampling rockpool fishes is difficult owing to the variable-sized rockpools, and their small secretive fishes, the effectiveness of three sampling methods was assessed. Emptying water from rockpools with a battery-powered bilge pump proved the best method, which accounted for significantly more species and individuals than a fish anaesthetic (clove oil) and visual census. Despite being more time consuming and cumbersome, bilge-pumping provides access to crevices and under rocks, thus results in a more accurate census than the other methods trialled.

The rockpool fish assemblages of the NSW south coast comprised 50 species, mainly permanent residents endemic to Australian waters. The main families represented were: Blenniidae, Clinidae, Tripterygiidae, Gobiidae and Gobiesocidae. At the level of family, NSW rockpools fish assemblages were different from those in the United States, Spain, France and Portugal, but similar to Chile and New Zealand. They were surprisingly similar to those in South Africa, to the extent of sharing species. At a regional scale, south coast NSW fish assemblages differed from those to the north in having different relative abundance of the species common in each region. Some species of economic and conservation importance were caught, such as Girella elevata and the protected black rockcod, Epinephelus daemelii. Their presence exclusively as juveniles indicates that rockpools may be possible nursery areas for such species that live their adult lives subtidally.

The rockpool fish assemblages were dynamic across various spatial and temporal scales. Horizontal spatial variability across scales of kilometres to hundreds of kilometres may be due to barriers to larval dispersal (e.g. local currents or upwelling events) or barriers to fish migration (e.g. sandy beaches between headlands). Across scales of metres to kilometres, the presence of different microhabitats within and among locations favours species having specific requirements for a particular habitat. In contrast, vertical spatial variability may be controlled by differences in fish physiology, which appears to broadly separate species by their intertidal residential status. Highly adapted resident species generally occupy rockpools of highest elevations where the environment can be harsh and variable, whereas opportunists and transients possessing few or no adaptations for intertidal life occupy lower rockpools where water chemistry is similar to the adjacent ocean.

Temporal variability in rockpool fish assemblages across scales of months to years was due to seasonal recruitment that occurs from summer to autumn (February to April) for most species studied. During recruitment pulses the abundance of each species varies but the relative change is minimal, resulting in highly stable and persistent assemblages. In the short-term (weeks to months), fish assemblages are resilient to perturbations and recolonise rockpools within weeks. This process was primarily due to sub-adult and adult fishes moving in from nearby rockpools, although many fishes displaced up to 20 m have the ability to home. This homing ability m a y be facilitated by identification of specific topographic cues learned by swimming over a limited geographic range at high tide. Juveniles appear only to be a major contributor to the recolonisation process if perturbations coincide with recruitment periods. The high stability of rockpool fish assemblages both in the long and short term may be a result of only fishes with specialised adaptations being able to utilise vacant habitats in the highly variable rockpool environment. Despite occupying a highly variable environment, the high resilience and stability detected indicates that NSW rockpool fish assemblages are probably regulated by deterministic, rather than stochastic, processes. However, a longer study period is needed to determine the effects of natural stochastic elements these fish assemblages.

Manipulation experiments explored the effects of varying structural complexity of two major habitat types (suspended algae and substratum cover) on rockpool fish assemblages. Fish did not discriminate between different habitat types so long as some shelter was available. The amount of substrate rock cover afforded the most attractive shelter to fish, since most rockpool fishes are benthic. However, increased habitat complexity only increased the numbers of species and fish in a rockpool to a threshold where other factors (possibly food) probably become limiting. Alternatively, the lack habitat specificity of fishes indicates fish may only utilise rockpools as temporary refuges before moving throughout the intertidal zone during high tide. Since rockpool fishes may be still vulnerable to predation at low tide (i.e. terrestrial predators), appear to select rockpools with shelter where predators may have difficulty of access.

By undertaking experiments that manipulated both fish assemblages and habitats in rockpools, the dynamics of rockpool fish populations can be better understood. Establishing a link between regulation processes allowed the development of the "Nearest Neighbour Model", which describes the dynamics of rockpool fish assemblages being primarily dependent upon fish populations and habitats in neighbouring rockpools. This model opposes population regulation of reef and estuarine fish assemblages, which are primarily controlled by recruitment processes, and stochastic availability of resources.

The paucity of data on rockpool fishes in Australia and the increasing potential of anthropogenic impacts on coastal ecosystems highlight an urgent need to document, protect and manage its biological integrity. The most appropriate means of managing intertidal fishes may be the use of intertidal protected areas (IPA) where human activity is restricted. The limited movements and larval dispersal of intertidal fishes and the small scale patchiness of assemblage structure indicate that several small reserves be better than a single large reserve. However, there are many socio-economic issues that need to be carefully addressed before implementation of IPAs.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.