Brett W. Molony

The Effect of Impoundments on the Structure and Function of Fish Fauna in a Highly Regulated Dry Tropics Estuary

Ross River flows through the Townsville/Thuringowa urban area in north Queensland, Australia, which has a dry tropical climate characterized by high inter-annual rainfall variation. Unregulated rivers in the Ross catchment basin deliver freshwater flows to their estuaries during both strong and weak wet seasons. The construction of a series of dams and weirs on Ross River means the wet-dry cycle is accentuated, leading to constant marine salinities throughout the estuary becoming the norm, with a lack of freshwater flow for five or more years at a time. The fish fauna of Ross River estuary was sampled in the post wet and dry seasons during an extremely dry climatic period (1994) and extremely wet climatic period (2000) using a small mesh (6 mm) pocket seine net. The fish fauna seemed to reflect seasonal differences. Catches from 1994 (dry period) were comprised entirely of 88 marine and euryhaline species, while the 69 species captured in 2000 (wet period) included 13 freshwater species. However, the freshwater species in the upper estuary were individuals washed over the weir, rather than part of a functional faunal gradient. During 1994 faunal composition was related more to site identity than to the position of the site along an upstream gradient. In contrast, during 2000 there were clear upstream faunal gradients with compositions in upstream sites heavily influenced by freshwater species, and marine and euryhaline species dominating downstream sites. Patterns of species dominance also varied between years. In contrast, trophic composition showed consistent shifts in both years, from high proportions of herbivores, carnivores and benthoplanktivores in May towards high proportions of benthivores in August. Not only do faunal composition, seasonal faunal change and ecological connectivity seem to be impaired, but ecological processes in the estuary that rely on seasonal freshwater flows are likely to be unable to operate normally in most years. The extreme seasonality in Ross River may serve as a model for many of the changes that will be experienced in dry tropics estuaries under global climate change scenarios of more extreme seasonality.

Effects of feeding history on mobilisation and deposition of body constituents and on growth in juvenile Ambassis vachelli (Pisces: Chandidae)

Young Ambassis vachelli (Richardson) 40 to 50 d old, were used in a laboratory experiment to test the effects of starvation and subsequent re-feeding on body constituents and growth. Fish in three laboratory treatments (fed continuously; fed for 9 d and then starved for 15 d; starved for 9 d and fed for 15 d), were compared to fish from a local field population. Different body stores were mobilised at different times and rates during starvation. Carbohydrates were mobilised from the onset of starvation and were depleted after 3 d. Lipid and protein were mobilised at an increasing rate from the onset of starvation. The mortality in starved fish was relatively high (up to 70%) until re-feeding. Upon feeding, all body stores were restored rapidly in fish that were starved, with carbohydrate levels displaying an “overshoot” (carbohydrate level exceeding normal levels) in comparison to the levels in continuously fed fish. After 15 d of feeding, the starved fish had levels of constituents similar to those in continuously fed fish. Fish that were fed and subsequently starved were able to maintain themselves for at least 15 d prior to death, indicating a better degree of starvation resistance than fish without a history of feeding. This implies that feeding history in the early life of a fish is important in growth and survival but that young fish may have growth regimes flexible enough to survive relatively long periods of starvation.

Otolith increment widths and lipid contents during starvation and recovery feeding in adult Ambassis vachelli (Richardson)

Adult Ambassis vachelli (Richardson) were used to test the effects of starvation and recovery feeding on otolith increment widths, Fultons K and levels of carcass and viscera lipid. Three laboratory treatments (fed throughout the experiment; fed for 41 days then starved for 37 days, starved for 41 days then fed for 37 days), were compared. Fish were also collected from the field throughout the experiment to assess the natural fluctuations of variables examined in this study. Starvation and recovery induced significant changes in Fultons K and lipid contents and were permanently recorded in the widths of otolith increments. Although changes in increment widths did occur in a predictable way, changes in increment widths due to starvation and recovery feeding were not of the same magnitude as changes in lipid levels. Thus, the widths of otolith increments may be determined by feeding regime, but increment widths vary in a more conservative fashion than do lipid levels. Increment widths in field fish showed similar variations compared to fish in the laboratory. Condition and lipid levels, as well as results from a canonical discriminant analysis, indicated that field fish were exposed to intermediate feeding environments, compared to the fed and starved regimes in the laboratory.

Episodes of starvation are recorded in the otoliths of juvenile Ambassis vachelli (Chandidae), a tropical estuarine fish

Juvenile Ambassis vachelli, ≃40 to 50 d old, were used to test the effects of starvation and recovery-feeding on the widths of otolith increments. Three laboratory treatments (fed continuously; fed for 9d then starved for 15 d; starved for 9 d then fed for 15 d), were compared. Fish were also collected throughout the experiment from a field site to assess the natural fluctuations of increment widths. Growth, starvation and recovery were reflected in the widths of otolith increments. The width of increments deposited during starvation were similar in all fish, regardless of prior feeding history, despite the lower mortality recorded in starved fish with a history of intense feeding. Increment-width data were confounded by an ontogenetic decrease in increment widths with age. The results indicate that starvation and recovery are recorded in the increment widths of otoliths and are distinguishable from the effects of ontogenetic development. However, ontogenetic changes in the width of otolith increments confounded the interpretation of otolith-increment series. Therefore, the ontogenetic pattern of otolith increments must be known for each species before valid interpretation of otolith microstructure is possible.

Temporal population dynamics of the swarming shrimp, Acetes sibogae australis, in a tropical near-shore system

Acetes sibogae australis, a small swarming crustacean, is an important component of near-shore systems in eastern Australia. It is considered that densities of A. s. australis are highly variable. However, this variability has never been tested on a fine temporal scale. In this study, densities of A. s. australis within swarms, as well as swarm length and width, were recorded once a week for seven months at one locality in Townsville, Australia, and once a fortnight for seven months at two other localities in the region. Length frequency of a sub-sample of individuals caught at each sample time was also recorded. Density of A. s. australis fluctuated significantly during the seven months, with numbers per replicate varying between 0 and 1000 individuals. Swarm width and length were significantly correlated with density, contributing to large-scale differences in overall abundance. Length frequencies revealed at least three dominant cohorts present during the seven months, although appearance and disappearance of these did not correspond to density changes. Patterns recorded were similar at the three spatially separate localities. The large fluctuations in A. s. australis abundance over time are likely to have important effects on the near-shore system of Townsville.

The relative efficacy of stocking fry or yearling rainbow trout (Oncorhynchus mykiss) into a large impoundment dominated by redfin perch (Perca fluviatilis) in south-western Australia

To compare the efficacy of stocking fry and yearling rainbow trout, Oncorhynchus mykiss, into reservoirs of south-western Australia, 20 000 fry and 1400 yearlings were stocked into Lake Navarino and sampled over a five-month period. Oncorhynchus mykiss fry were collected via seine-nets within 24 h of stocking; however, within 60 h of stocking, no O. mykiss fry could be located, suggesting the total mortality of fry. Within 24 h of stocking, O. mykiss fry were present in the guts of Perca fluviatilis and Salmo trutta. A simple risk analysis of stocking O. mykiss fry, based on gut contents and relative abundances of predatory fishes, indicated that P. fluviatilis has a relative predatory impact on O. mykiss fry more than 100 times greater than other fishes. Perca fluviatilis also has a greater relative predatory impact on freshwater crayfishes than either species of trout. The results demonstrate the advantages of stocking yearling O. mykiss in impounded waters dominated by P. fluviatilis. Existing data of relative abundances of predatory fishes and relative predatory impacts can be used by managers as a powerful and simple tool to assess strategies for stocking fish and the application of other management options.

Coherent patterns of abundance and size of a tropical snapper in dynamic estuary systems

Although tropical estuaries are important juvenile habitats for the snapper, Lutjanus russelli, the extent of this utilisation, and relationships to physical variation are unknown. We used fish trapsto investigate spatial and temporal patterns of distribution, abundance and size of L. russelli within and between three tropical estuaries (Cattle, Barramundi and Alligator Creeks) over a two year period. L. russelli showed no indication of spatial segregation by size along the lengths of the estauries. The probability of capturing L. russelli was similar throughout Barramundi Creek and the seaward parts of Cattle Creek. However, the probability of capture fell markedly in upstream regions of Cattle Creek. These areas where characterised by depressed salinity levels due to seasonal flooding. High probabilities of capture of L. russelli during the wet and post-wet seasons in the whole of Barramundi Creek correlated with the time of smallest modal sizes. Thisreflected the entry of large numbers of small recruits into the trappable population. Temporal patterns in mean length were consistent for both years. The major feature of the size data was a clear pattern of seasonal change in the mean fork length of L. russelli, with largest mean sizes occurring in the pre-wet and the smallest mean sizes in the wet and post-wet seasons. This pattern probably reflected the interaction of four processes; the entry of new recruits into the trappable population, growth and mortality during residence in the estuarine nursery habitats, and migration of large fish away from the estuaries. The three creeks studied provided L. russelli with a harsh and extremely variable environment. Over the three creeks, salinities ranged from extremely hyposaline (0‰) to very hypersaline (55.5‰) and varied up to 39‰ within a location. However, despite the substantial physical differences and the spatial separation of the three estuaries, the patterns of temporal change were clear and consistent. Overall, the annual pattern of change insize seemed to relate more to biological processes of recruitment, mortality and migration than to responses to physical variability.

Challenges of external insemination in a tropical sparid fish, Acanthopagrus berda

Duration of spermatozoa motility and proportion activated were tested under a range of salinities and temperatures in males of the tropical estuarine fish, Acanthopagrus berda, in both laboratory and novel field experiments. A. berda spermatozoa were activated under a wide range of salinities (5–60%) and showed the greatest duration of motility at approximately 30%) in both laboratory and field fish. The duration of activation exceeded 10 minutes at this salinity. However, relatively low activation rates were recorded (50–70%). The long duration of motility, wide range of activation salinities and low proportion of motile spermatozoa may reflect a reproductive style in A. berda of producing at least some motile spermatozoa despite the ambient conditions at spawning. This is discussed in relation to fishes from other habitats and the wide distribution of A. berda.

Resource partitioning and habitat-specific dietary plasticity of two estuarine sparid fishes increase food-web complexity

The diets of the co-occurring sparid fishes Acanthopagrus australis and A. pacificus were investigated to assess dietary overlap and resource partitioning, the extent of spatial variation and implications for the complexity of trophic interactions in tropical estuaries. Both species consumed a wide variety of prey, including bivalves, gastropods, crustaceans and polychaetes. There were clear differences in diet among sites separated by less than a kilometre. Site-specific substrata determined prey type, so diets at sites within a sandy bay were comprised largely of benthic infauna, whereas those at mangrove-lined sites contained substantial amounts of mangrove-associated prey. Although there was general dietary overlap between the two species, their diets differed substantially at any one site when viewed at a fine taxonomic resolution. A. pacificus displayed a very region- and site-specific diet, apparently feeding in different, and spatially separate, food chains at these different locations. Its main prey varied markedly among sites, seemingly because of habitat differences. By feeding in different food chains (and potentially different food webs) at different points in space and time, species such as A. pacificus have the potential both to link food webs operating in a variety of habitats and to add substantially to the complexity of tropical estuarine food webs.

Reproductive periodicity of the sparid, Acanthopagrus pacificus, on a hierarchy of temporal scales

The reproductive periodicity of the sparid, Acanthopagrus pacificus, over four temporal scales is described. Acanthopagrus pacificus had a short spawning season between June and September, and within this, a peak reproductive period from July to early September. During the peak period there were several spawning peaks corresponding to a lunar periodicity, with intense reproductive activity on new and full moons that peaked during the period of the full moon when the tidal range was greatest. At the smallest temporal scale, spawning occurred at night on ebb tides. Because this study draws on data collected in 1991 and 1995, it provides a useful baseline against which to judge future changes in reproductive periodicity.