Department of the Environment

About us | Contact us | Publications

Header imagesHeader imagesHeader images



Key departmental publications, e.g. annual reports, budget papers and program guidelines are available in our online archive.

Much of the material listed on these archived web pages has been superseded, or served a particular purpose at a particular time. It may contain references to activities or policies that have no current application. Many archived documents may link to web pages that have moved or no longer exist, or may refer to other documents that are no longer available.

State of the Marine Environment Report for Australia: The Marine Environment - Technical Annex: 1

Compiled by Leon P. Zann
Great Barrier Reef Marine Park Authority, Townsville Queensland

Ocean Rescue 2000 Program
Department of the Environment, Sport and Territories, Canberra, 1995

ISBN 0 642 17399 0

Temperate subtidal hard substrata

Michael J. Keough* & Alan J. Butler

Department of Zoology
University of Melbourne
Vic. 3052*
Department of Zoology
University of Adelaide, GPO Box 498
Adelaide, SA 5011


Hard surfaces are common features of many subtidal habitats along most of the coastline of temperate Australia. Natural substrata range from large rock surfaces, to small patch reefs and boulders and to small biogenic surfaces such as bivalve shells. Near major cities there are extensive artificial surfaces, such as piers, breakwaters and various marina facilities.

Hard surfaces generate important spatial heterogeneity in subtidal areas and provide attachment space for a wide diversity of sessile organisms. Some of these organisms, especially large brown algae, also contribute important physical structure for reef habitats which are in turn used as habitat by a wide range of mobile animals, particularly fish.

These habitats are important to humans in a number of ways. They form the base of substantial recreational fisheries, especially down the east coast of Australia (Kingsford, Underwood & Kennelly 1991) and are popular in summer for diving and snorkelling activities. Just as in coral reef habitats, the sessile organisms on hard surfaces are potential sources of important natural products (eg van Altena & Miller 1989, Davis, Butler & Altena 1991).

The animals and plants in this habitat are extraordinarily diverse in global terms. Many groups, such as red and brown algae (Womersley 1990), ascidians (Kott 1985, 1990a, 1990b), bryozoans (Bock 1982) and crustaceans (Poore 1990) have much higher species richness in southern Australia than in comparable temperate habitats elsewhere. Often, these groups are more diverse in temperate areas than in the tropics (eg bryozoans, red algae). It is also possible that the processes in these local temperate reef

habitats are different from those occurring outside Australia or Australasia. Steinberg (Estes & Steinberg 1988, Steinberg & van Altena 1992) has suggested that levels of herbivory and plant defensive compounds differ between Australia and North America, and Jones and Andrew (1990) and Holbrook, Schmitt and Ambrose (1990) have discussed the roles of herbivorous fish and sea urchins in the dynamics of subtidal algal communities in different geographic regions. While some of these lines of evidence are convincing, Underwood and Fairweather (1986) have cautioned that broad-scale comparisons may be confounded by methodological differences.

Our knowledge of temperate subtidal hard substrata is considerably less than that of corresponding intertidal areas, largely because subtidal areas are less accessible and have a shorter history of experimental approaches. The scarcity of quantitative descriptions of the flora and fauna has been highlighted elsewhere (Underwood & Kennelly 1990, Underwood, Kingsford & Andrew 1991, Kennelly & Underwood 1992), and we will not deal with it in much detail other than to emphasise that general statements must be treated with considerable caution.

Approaches: experimental versus observational

As with other habitats, ecologists have taken a variety of approaches to understanding reef systems. We divide the approaches broadly into two categories: observational and experimental.

The observational approach has a long history, developing from simple qualitative description to elaborate methodologies that involve formal sampling (Andrew & Mapstone 1987 review the considerations in sampling properly) and attempts to describe recognisable 'communities', 'assemblages', 'biocoenoses', etc, sometimes with the aid of multivariate statistical techniques. Such approaches seek explanations for the observed patterns by noting correlated variables, such as depth, light intensity, turbidity and exposure to wave action. An example of this approach for British subtidal reefs is Hiscock's (1985).

A neglected aspect of descriptive (or observational) approaches is the need to follow particular sites through time, in order to gain not only a static picture but an understanding of the dynamics of the assemblage. When this has been done (eg Kay & Butler 1983; Underwood, Kingsford & Andrew 1991, Kennelly & Underwood 1992) it was found that, although the overall composition of the assemblage was fairly constant over time, substantial small-scale change was constant. An appreciation of this phenomenon is crucial background to experimental studies of processes, to assessing the state of the system and ultimately to making management decisions.

The experimental approach begins with observations of patterns and correlated variables. From them develop possible explanations for the patterns - call these hypotheses - and then tests of the hypotheses by means of manipulative experiments, generally carried out in the field. There need to be controls in such experiments, the fundamental idea being to compare the system with and without some alteration to the variables under study.

Both approaches have their place, but we would argue that to understand the working of the system - hence to understand the effects of human activities on it and be able to manage it - experiments are essential. There is a large literature arguing this case both by those in the tradition of experimenting on few variables at a time (eg Underwood 1985, 1990) and by those in the tradition of pattern-recognition using powerful multivariate methods, who nevertheless recognise the distinction between recognising patterns and assigning causes (eg Gray et al. 1990, Warwick & Clarke 1991, Clarke 1993).

There has only been limited work of either sort on Australian temperate subtidal reefs, and its geographic extent is small. Studies that allow us to identify the important processes in these habitats are much rarer than the habitat descriptions, and experimental work has been done mainly in the gulfs of South Australia, in Port Phillip Bay, Victoria and off the coast of central New South Wales only. Because of the different quality of the information coming from the studies, we discuss them separately below, and summarise the major studies showing processes on Table 1.

Dominant assemblages - patterns

Descriptions of the floral or faunal assemblages occurring on subtidal hard substrata are geographically patchy. For some areas, such as central regions of South Australia, reasonably comprehensive surveys have been performed that allow the major habitat types to be mapped (eg Shepherd & Sprigg 1976). Elsewhere, including much of the Great Australian Bight and exposed coasts of western Victoria, there is almost no published information. The early descriptions were largely the result of work by Shepherd and Womersley (Shepherd & Womersley 1970, 1971, 1976, Shepherd 1981, Shepherd & Womersley 1981, Shepherd & Sprigg 1976) who provided qualitative descriptions of many reefs in southern Australia. The most useful broad-scale quantitative data come from the studies along the coast of New South Wales by Underwood and colleagues (Underwood & Kennelly 1990, Underwood, Kingsford & Andrew 1991, Kennelly & Underwood 1992) who focused on plants and mobile invertebrates. Most of the available data consist of single surveys, so there is little information about temporal variation on short (seasonal) or long (annual) time scales. We know of almost no long time series for whole assemblages (but see Kay & Butler 1983 and Kennelly & Underwood 1992) and the few long-term data sets for single species are based on commercially exploited species (McShane, Beinssen & Foley 1986, Shepherd 1990, Prince & Shepherd 1992). As an example, Colman, Keough, Quinn, Gwyther and Smith (1991) described the temporal and spatial extent of all known long-term subtidal monitoring programs in Victoria, a process that occupies little space in their report. It is not our purpose here to provide a comprehensive review of these data: this is currently being done in various forms by a range of State authorities such as the Land Conservation Council (Victoria), and the Department of Primary Industries (South Australia).

Table1: Extent of knowledge about processes controlling structure of assemblages on temperate subtidal hard substrata in Australia. Data are drawn from only experimental ecological studies (see text). This table does not cover information not yet published in refereed journals and books (departmental reports, theses, etc).

Keys: Habitat: P, pilings, other human-made structures; R, rocky reef. Organisms: A, sessile animals; K, canopy-forming plants, typically kelp bed; M, motile animals; U, understorey algae (where one or few species of these are named). Process: C, competition; D, dispersal; G, genetics; H, human impacts; L, light; N, natural products, chemical ecology; O, oceanographic features including upwellings, terrigenous nutrient inputs; P, predation; T, turbidity; W, wave action, other disturbance, patch dynamics. Time period: given in years if a pattern-descriptive study. Space: given as number of separate locations (eg different bays). (PPB = Port Phillip Bay)

Edithburgh, central SA
Kay & Keough 1981, Kay & Butler 1983, Keough 1984b, Davis et al. 1991, Butler 1986, 1991, Keough 1983
Keough 1984a, Chernoff, 1987, Pitcher & Butler 1987
Gulf St Vincent, SA
Butler 1986, 1991
Rapid Bay, SA
Kay & Butler, 1983, Keough & Butler 1979
Gulf St, SA
Clavelina (=Podoclavella)moluccens is
Davis 1987a,b, 1988a,b, Vincent, Davis & Butler 1989
4/ many
Steinberg & van Altena 1992, Steinberg 1989
Portsea, PPB, Vic
Russ 1980, 1982, Fletcher & Day 1983
Popes Eye, PPB, Vic
Parma victoriae
Jones & Norman 1986, Norman, & Jones 1984
Central NSW
Fletcher 1987
Central NSW
Kennelly 1983, 1987a,b,c, 1989
Central NSW
Stocker 1991, Stocker & moseleyi Underwood 1991
Kirkman 1981, 1984

1. Small bivalve shells as substrata (Pinna bicolor, Chlamys asperimma)

2. Problems with design (see Underwood & Kennelly 1990).

The lack of knowledge referred to above is a serious impediment, but it can be placed into context by noting that similar complaints have been raised about the quality and quantity of information available for subtidal reefs in other parts of the world, including such well-studied areas as central California and parts of Chile, Canada and New Zealand (Foster 1990, and other papers in Chapman & Underwood 1990).

One of the major problems in producing any comprehensive description of assemblages is the level of taxonomic resolution available. For many surveys, the taxa that are reported depend on the taxonomic expertise readily available to the investigators. This expertise may not correspond to the patterns of species richness among higher taxa at a particular site (Keough & Quinn 1991). The causes of these problems are well known. For many taxa, the majority of species remain undescribed, even though the taxa are acknowledged to be speciose. In other taxa, it is possible that only a few systematists in Australia are capable of identifying the organisms to generic or specific level, and specimens (eg sponges, amphipods, red algae) must be sent to these specialists. This has the effect of either increasing the budget of the work due to consulting costs, creating a bottleneck as the relevant systematists are overwhelmed by the size of the task, or forcing the investigators to classify organisms only to higher taxonomic levels. We note that a number of authors have advocated the use of higher taxa, usually in the form of functional groups (Jackson 1979,Choat & Schiel 1982, Littler & Littler 1984) and have demonstrated the utility of this approach (Keough 1984a, 1984b, Butler 1986, Warwick 1988, Butler 1991). This approach was claimed to be unsuccessful in Kennelly's and Underwood's (1992) study, because it hid the species-specific patterns of variation present in this assemblage. However, some of their pooled higher groups were very broad (eg sessile animals, mobile invertebrates) and may have consisted of a number of what other authors would consider to be functional groups.

The oldest general scheme for describing spatial patterns was devised by Shepherd and Womersley (1970, 1971) using concepts developed from intertidal zones of rocky shores. They recognized three important zones, based mainly on depth/exposure combinations, and characterised by the algal groups that dominate each zone. A detailed summary of this scheme is provided by Womersley and King (1990). Other authors (reviewed in Underwood, Kingsford & Andrew 1991; and see Sanderson & Thomas 1987) have found it necessary to erect a number of other classes of habitats and assemblages.

Habitat-forming plants

The dominant feature of many open coast reefs is the presence of large, canopy forming brown algae. These plants, primarily kelps such as Ecklonia, Macrocystis and Durvillea, occur from shallow subtidal depth to 15 m or so, their actual depth distribution varying with the transparency of the water to light. Beneath their canopies is an assemblage of smaller plants, primarily red and brown algae, and a range of animals. In deeper water, these large algae give way to smaller, turfing species, primarily red algae, mixed in with sessile animals. Red algae and animals may also be important in shallower water, where light, wave action, or sedimentation restrict the presence of canopy-forming brown algae. Womersley and King (1990), Underwood and Kennelly (1990), and Underwood, Kingsford and Andrew (1991) provided very good overviews of this topic. An important feature in these shallow-water habitats is the presence of 'barrens', areas dominated by encrusting coralline algae and high densities of herbivorous invertebrates.

We urge some caution in interpreting this discussion of spatial patterns. A number of recent authors have raised doubts about the utility of broad-scale generalisations about spatial patterns (Foster 1990, Schiel 1990, Underwood, Kingsford & Andrew 1991, Kennelly & Underwood 1992). Underwood, Kingsford and Andrew (1991) cautioned that simple habitat classifications are not generally useful. Some habitats - especially Ecklonia forests and barrens - exist as small interspersed patches, so a broad classification at a large spatial scale will be forced to ignore the small-scale patchiness. Such a classification, using 'average' or 'most common habitat' descriptors, would have no predictive value at smaller spatial scales. The patterns that initially appeared to exist may have been the artifact of a small data base. In other areas, especially the west coast of the United States of America, the strength of generalisations has weakened as more detailed quantitative information has become available (Foster 1990).

Diversity/taxonomy: knowledge and gaps

The best-studied groups of algae are the greens and browns, less so the red algae. They are especially diverse in southern Australia with high levels of endemism at the species level (Womersley 1990). Even though they have received much attention, they are still far from being described completely. Importantly, many species cannot be identified in the field, thus placing serious limitations on our ability to make rapid, comprehensive floral surveys. It is important to retain voucher specimens as a basis for the names reported in such surveys.

Small-scale distributions

The structural complexity of the algal assemblages varies with the degree of exposure to water movement. In general, lower energy environments have canopies that are less complex structurally than high energy environments. Low energy environments may have no macroalgal canopy, or one dominated by species of Sargassum or Cystophora, while kelps provide the structure on more exposed coastlines. This variation is related to some combination of wave action, light and sedimentation, and it is hard to separate these factors.

Biogeography/large-scale patterns

At larger spatial scales, there are geographic differences in algal assemblages, particularly in the dominant brown algae. These differences appear more related to latitude than longitude. Down the east coast of Australia, the dominant canopy-forming species are Ecklonia radiata and Phyllospora species. These taxa - but particularly Ecklonia - are common within Port Phillip Bay and Western Port in Victoria, central South Australia and in south-western Western Australia (Kirkman 1984, Clayton & King 1990). At greater latitudes - particularly open coasts of Victoria, south-eastern South Australia and Tasmania - other kelps are found, including Macrocystis angustifolia, M. pyrifera (Tasmania only) and Durvillea species. Macrocystis and Durvillea do not extend westwards far beyond the Victoria-South Australia border.

These distributional differences are important for understorey plants and animals, because the kelps are morphologically very different from each other; and one of the most important effects of kelps is thought to be their action in attenuating water movement on exposed coasts (Eckman, Duggins & Sewell 1989, 1990, Eckman & Duggins 1991). The hydrodynamic environment provided by Macrocystis is very different from that provided by Ecklonia. We would therefore expect major structural differences between the kelp forests of southern Victoria and Tasmania, and those from lower latitudes. The direct role of kelps in fish recruitment, and their indirect influences on benthic communities, have been studied most intensively along the west coast of the United States of America, but such experimental work has not been done in Australia.

The barrens habitat is rare from east Gippsland through to Western Australia. Jones and Andrew (1990) attributed this to the presence of particular sea urchins. They suggested that Centrostephanus rodgersii is the primary species responsible for maintaining barrens along the eastern coastline while the other common urchin, Heliocidaris erythrogramma, cannot create or maintain barren habitats. We note that Sanderson and Barrett (1989) suggested that Heliocidaris is capable of maintaining barrens, although the evidence for such a suggestion is unclear. The dominant urchins in open coast reef environments vary from C. rodgersii and H. erythrogramma in New South Wales, to Heliocidaris alone in South Australia, to a mixture of Heliocidaris, Tripneustes species and Echinometra mathaei in south-western Western Australia (Jones & Andrew 1990). Jones and Andrew suggested that Echinometra is also capable of maintaining an environment with patches of barrens.


In open coast environments, sessile animals occur in greatest abundance in well-shaded parts of reef, such as vertical rock faces, underhangs and caves and undersides of boulders. They are relatively uncommon on upward-facing rock surfaces. The dominant sessile animals are modular or colonial, including sponges, ascidians, soft corals and hydroids, and bryozoans. In more sheltered environments animals may occur in more open habitats. These distributions presumably reflect the reduced growth rates of plants in those areas. In low energy environments, the sessile animals are predominantly solitary or unitary; these assemblages correspond to those referred to as 'fouling communities'. The numerically dominant groups are barnacles, solitary ascidians, bivalve molluscs and polychaetes.

Mobile animals occur in most areas, but are not well described. Most of these organisms use the structure provided by the plants, or are herbivorous, but in most cases the exact nature of the relationship is unclear. Some associations may be facultative and others obligate. Work overseas suggests a major influence of the physical structure of the macroalgal assemblage on mobile animals in influencing recruitment (eg Carr 1989, Eckman, Duggins & Sewell 1990) or interactions between established organisms (Holbrook, Schmitt & Ambrose 1990,Schmitt & Holbrook 1990a). The most prominent mobile animals are reef-associated fish, sea urchins, and other commercially-exploited species such as abalone and lobster.


The degree of taxonomic resolution is more uneven than for the plants. Despite considerable systematic effort, some groups such as amphipods and sponges have a large number of undescribed species. In other taxa, very common species remain either undescribed or in a state of systematic confusion (eg Celleporaria species: see Bock 1982, also Hutchings 1982, Plate 24.2). Some of these groups have very high species richness in temperate Australia. The larger mobile groups, particularly the fish, echinoderms, and molluscs, are much better known.

Unfortunately, the smaller mobile organisms (crustaceans, opisthobranch molluscs) and the majority of the sessile animals (especially some colonial ascidians and sponges) are almost impossible to identify in the field. It is unlikely in the near future that we will be able to produce detailed descriptions of the fauna of particular locations. Most of the existing literature lumps these groups into higher taxa, often corresponding to phyla or classes; so we can provide only very broad ideas of biogeographic patterns.

Small-scale distributions

As we mentioned earlier, there is a scarcity of detailed quantitative descriptions of faunal assemblages. The best details come from studies in a couple of areas only. A range of papers provides descriptions at levels ranging from species to functional groups. These data, from central South Australia, come from a variety of natural and artificial surfaces. The other major data set is from rocky reefs in central New South Wales. At one of the most intensively studied sites, Edithburgh (South Australia), the composition of the assemblage of sessile animals varies considerably with the nature of the substratum. Large, stable surfaces are dominated by sponges and tunicates, while smaller, isolated surfaces have many bryozoans and unitary organisms. The presence of this small-scale variation makes it difficult to interpret or generalise from spot censuses at other localities.

Biogeography/large-scale patterns - variation among sites

Different taxa show degrees of large-scale variation, ranging from local variation in structure of assemblages to the consistent presence of some widely-distributed taxa. Many animal groups are mixtures of species with restricted and wide distributions.

Given the taxonomic limitations, it is difficult to comment on sessile organisms. Many sessile animals do have restricted spatial distributions (eg Kott 1985) but there are some broad similarities in the best-studied sessile assemblages (Portsea in Victoria; Edithburgh and Rapid Bay in South Australia). Even at these sites, the dominant species at a particular site may be rare or absent from others (eg Kay & Keough 1981, Russ 1982).

Among mobile animals there are similar patterns. In addition to the sea urchin example discussed earlier, there are some regional differences in other echinoderms. For example, the Tasmanian asteroid fauna includes a number of species that are restricted to that region (Dartnall 1980). The major reef-associated fish also vary; Jones and Andrew (1990) provided representative lists of herbivorous fish from temperate Australia and New Zealand; and other families show some differences, both longitudinally and latitudinally (Last, Scott & Talbot 1983; Hutchins & Swainston 1986).

A major cautionary note can be derived from Butler's (1986) description of sessile assemblages at a number of piers in South Australia. He found substantial differences between assemblages at piers separated by as little as 5 km, with different species dominating the assemblages. Kennelly's and Underwood's (1992) is another of the few studies covering such spatial scales and reporting important differences at small scales. In attempting to assess gaps in our knowledge, or provide an overview, it must be remembered that we are integrating a range of studies that are spatially and temporally well separated and such integrations may be of limited use if there is a large amount of variation on smaller spatial scales. This situation is particularly important where there are large areas with only one or a few quantitative descriptions of the fauna or flora (eg Western Australia: see Kirkman 1984; Hatcher 1989).


Subtidal reefs are important fish habitats, and this is perhaps the most widely perceived attraction to sport divers. In dimly lit places such as caves, beneath overhangs, on vertical rock faces and the pilings of well- shaded jetties and other human-made structures, brown and green algae become rare, even the red algae being reduced in variety and abundance. The space becomes dominated by sessile animals. In addition to the fish, in both plant- and animal-dominated assemblages there are many motile, benthic animals, only a few of which have been studied (Keough & Butler 1979, Edgar 1982, Keough 1984a, Fletcher 1987, Andrew & Underwood 1989, Watts, Johnson & Black 1990, Andrew 1991, McShane & Smith 1991, Steinberg & van Altena 1992