Three hypotheses to discern the strong positive association between juvenile fish and mangrove habitat were tested with field and laboratory experiments. Artificial mangrove structure in the field attracted slightly more juvenile fish than areas without structure. Artificial structure left to accumulate fouling algae attracted four-times the total number of juvenile fish than areas without structure or areas with clean structure. Community composition of fish attracted to structure with fouling algae was different when compared with areas with no structure or clean structure; five species were attracted by structure with fouling algae whilst two species were associated with structure regardless of fouling algae. Algae were linked to increased food availability and it is suggested that this is an important selection criteria for some species. Other species were apparently attracted to structure for different reasons, and provision of shelter appears to be important. Predation pressure influenced habitat choice in small juvenile fish in laboratory experiments. In the absence of predators, small juveniles of four out of five species avoided shelter but when predators were introduced all species actively sought shelter. Large fish were apparently less vulnerable to predators and did not seek shelter when predators were added to their tank. Feeding rate was increased in the mangrove habitat for small and medium-sized fish compared with seagrass beds and mudflats indicating increased food availability or foraging efficiency within this habitat. Larger fish fed more effectively on the mudflats with an increased feeding rate in this habitat compared with adjacent habitats. The most important aspect of the mangrove habitat for small juvenile fish is the complex structure that provides maximum food availability and minimises the incidence of predation. As fish grow a shift in habitat from mangroves to mudflat is a response to changes in diet, foraging efficiency and vulnerability to predators.

A total of 53 species of juvenile fish were caught over a 2 yr study period in 2 mangrove lined estuaries in Moreton Bay, eastern subtropical Australia. Comparing juvenile fish communities among mangrove forests, seagrass beds and mudflats identified significant differences in species richness and abundances of juveniles. Seagrass communities comprised distinct species of resident and nonresident fish species of little economic importance. Mangrove forests and mudflats had many shared species (but mangrove forests were dominated by smaller or younger juveniles in greater abundances; Laegdsgaard unpubl. data). Mudflat hab~tats appear to be transition zones between juvenile and adult habitats. Only 4 species were exclusive to seagrass whereas 27 species were exclusive to the mangrove/mudflat habitat. Juveniles of 7 of the 10 commercially harvested fish species in Moreton Bay were found in greatest numbers in mangrove forests. Salinity, temperature and turbidity were similar in all habitats so could not account for differences in habitat choice of juvenlle fish. Most juvenile fish in mangroves during summer were nonresidents and species richness and abundance were highest in summer and lowest in winter There were significant differences among sites and years in the numbers of species and individuals; however, the trends were similar and demonstrated clearly that mangrove sites within Moreton Bay play a more important role and have greater potential as nursery habitats than do adjacent habitats. Preferential selection of mangrove habitats by juvenile fish, particularly commercial specles, indicates a need for conservation.

Abstract. Reproduction of Heliocidaris erythrogramma (Valenciennes) and H. tuberculata (Lamarck) was compared through examination of oogenesis, spermatogenesis and monthly measurement of the gonad index. These species occur sympatrically in the Sydney region. Their reproduction was examined at two sites near Botany Bay, New South Wales, Australia, from February 1989 through January 1990. H. erythrogramma produces buoyant, 450 #m-diam eggs and the sperm have a head region 10 #m in length. By contrast, H. tuberculata produces negatively-buoyant, 95 #m-diam eggs and the sperm have a head region 4 #m in length. Histochemical examination of the gonads revealed that periodic acid Schiff-positive (PAS +) material stored in the nutritive phagocytes appears to support vitellogenesis in both species. In H. tuberculata this material is utilized in the formation of PAS + yolk oligolecithal eggs, whereas in H. erythrogramma the PAS + material appears to be converted to lipid yolk in macrolecithal eggs. H. erythrogramma had a seasonal reproductive pattern with a 3 mo summer spawning period, whereas both populations of H. tuberculata had a 9 mo breeding period characterized by the continual presence of nutrient reserves and vitellogenic oocytes which rapidly replaced spawned ova. Spawning ceased only for 3 mo over the summer. Due to the 9 mo spawning of H. tuberculata it is not clear what factors serve to cue reproduction in this species.

It is clear that saltmarshes are a unique and important component of the coastal biosphere of Australia. Their contribution ranges from stabilisation of fine sediments and providing an excellent protective buffer between land and sea, to their diverse blend of terrestrial and marine fauna. Further, saltmarsh plants are highly specialised and adapted to fill a harsh niche allowing them to act in roles that other vegetation types cannot. Saltmarsh habitats are recognised for their importance to migratory waders under the Ramsar convention, but it is becoming increasingly evident that they are also important to a variety of commercially valuable fish and native mammal species. Activities that are detrimental to saltmarshes continue and need to be addressed in order to conserve remaining saltmarsh areas. In general, urbanisation of the catchment has lead to filling of saltmarshes, tidal restriction, use by recreational vehicles, grazing, trampling and increased sedimentation and nutrient runnoff allowing colonisation and invasion of mangroves. These disturbances have a number of ecological consequences ranging from weed infestation to complete changes in the species composition and ecology. Reversing the disturbance is not always simple and can require extensive groundwork to be successful. Rehabilitation of existing saltmarsh areas has been a successful means to enhance this habitat. In general, it requires relatively little effort to remove weeds and fence off areas to regenerate naturally. Saltmarsh areas have been shown to respond well to this type of manipulation. Restoration and creation require substantial effort and planning to ensure a successful outcome. However, given the right environmental combinations of elevation, tide and salinity, saltmarsh will establish and grow. To speed the process transplantation of saltmarsh plants can be considered either from donor sites or plants propagated in green houses.

Summary Efforts to restore coastal saltmarsh habitats through transplantation of saltmarsh species from donor sites are becoming more frequent. As sections removed are generally small there is little information on the ability of these areas to recover naturally after removal. The present study provides an understanding of the potential of dominant saltmarsh species in NSW (Saltcouch, Sporobolus virginicus and Samphire, Sarcocornia quinqueflora) to regenerate naturally after such small‐scale removal (25 cm × 25 cm plots) at separate sites for each species. The increase in percentage cover of vegetation within the denuded plots was measured over 21 months along a gradient within the marsh (between the mangrove and terrestrial boundaries) and then compared to non‐denuded plots. Recovery of Saltcouch was slow with an estimate of 5–6 years to reach the surrounding cover levels. Similarly, recovery of Samphire on upper levels of the shore is estimated at 4–5.5 years. On the lower level on the shore, however, Samphire had reached a cover that was within the range of the surrounding vegetation by the end of the 21‐month study period.