Search

The cover of seagrasses and macroalgae, landings and fishery-independent measures of spawning stock and recruitment for brown tiger prawns, were monitored immediately following a major cyclone in Exmouth Gulf, Western Australia in March 1999. Anecdotal evidence on the extent of seagrass from 1990 to 1998 suggests that the cyclone caused a major, immediate disruption and loss of the seagrass/macroalgal beds (to ≤2% cover), the critical prawn nursery habitat, and mangroves in the shallow inshore waters of the system. Prawn landings and recruitment to the fishery were not affected in the year of the cyclone, but were markedly lower in the two years immediately afterwards and then increased as the cover of macrophytes increased to over 40% in 2003. Tiger prawn landings and catch rates were not affected in Shark Bay, a system 500 km south of Exmouth Gulf that did not experience cyclonic disturbance. Seagrasses in Exmouth Gulf showed a succession of species from small colonising species (Halophila ovalis and Halodule uninervis) to larger, broad-leaved species (Cymodocea serrulata, Syringodium isoetifolium) only two years after the cyclone. The recruitment and landings of tiger prawns were correlated significantly with the total cover of macroalgae and seagrass. The large loss of seagrass and macroalgae reduced the settling habitat for postlarvae and the nursery habitat for juvenile tiger prawns, probably leading to the lower recruitment to the fishery. These findings suggest that the extent of seagrass and macroalgae are some of the factors defining the productivity of the tiger prawn fishery in Exmouth Gulf.

The cover of seagrasses and macroalgae, landings and fishery-independent measures of spawning stock and recruitment for brown tiger prawns, were monitored immediately following a major cyclone in Exmouth Gulf, Western Australia in March 1999. Anecdotal evidence on the extent of seagrass from 1990 to 1998 suggests that the cyclone caused a major, immediate disruption and loss of the seagrass/macroalgal beds (to ≤2% cover), the critical prawn nursery habitat, and mangroves in the shallow inshore waters of the system. Prawn landings and recruitment to the fishery were not affected in the year of the cyclone, but were markedly lower in the two years immediately afterwards and then increased as the cover of macrophytes increased to over 40% in 2003. Tiger prawn landings and catch rates were not affected in Shark Bay, a system 500 km south of Exmouth Gulf that did not experience cyclonic disturbance. Seagrasses in Exmouth Gulf showed a succession of species from small colonising species (Halophila ovalis and Halodule uninervis) to larger, broad-leaved species (Cymodocea serrulata, Syringodium isoetifolium) only two years after the cyclone. The recruitment and landings of tiger prawns were correlated significantly with the total cover of macroalgae and seagrass. The large loss of seagrass and macroalgae reduced the settling habitat for postlarvae and the nursery habitat for juvenile tiger prawns, probably leading to the lower recruitment to the fishery. These findings suggest that the extent of seagrass and macroalgae are some of the factors defining the productivity of the tiger prawn fishery in Exmouth Gulf.

Despite attempts in many countries, large-scale prawn (shrimp) stock enhancement programmes have only been implemented and continue to operate in Japan (Penaeus japonicus) and China (Penaeus chinensis). In these countries, prawn stock enhancement has been possible because of the capacity to produce large numbers of small individuals for release as a result of aquaculture. While exploring the feasibility and bio-economics of tiger prawn (Penaeus esculentus) stock enhancement in Exmouth Gulf, Australia, a broad framework for stock enhancement research was developed. This framework included developing a bio-economic model and risk assessment for all components of a stock enhancement operation, i.e., hatchery, grow-out, harvest-transport-release, population dynamics of released and wild prawns, and monitoring the fishery. However, predictions from the bio-economic model were made with great uncertainty about the values for post-release mortality of prawns because there were no data from experiments on optimal release strategies. Consequently, data from field studies of growth, and laboratory studies of predation, were used to set values for growth and mortality in Monte-Carlo simulations of the potential success of P. esculentus releases in different habitats. The survival of 2 mm carapace length (CL) prawns to the size at emigration from the nursery habitat was ∼1.9 times higher for individuals associated with seagrass of high (∼100 g m-2) than for those in low (∼10 g m-2) plant biomass. Survival was 19 times higher for prawns in high biomass seagrass than for those on bare substratum. The differences in survival between habitats were reduced greatly when the releases of much larger prawns (10 mm CL) were simulated. These findings highlight the importance of research on developing optimal release strategies for prawns in stock enhancement programmes. In particular, the size, habitat, time and density at release needs to be identified to provide more reliable informatio

Despite attempts in many countries, large-scale prawn (shrimp) stock enhancement programmes have only been implemented and continue to operate in Japan (Penaeus japonicus) and China (Penaeus chinensis). In these countries, prawn stock enhancement has been possible because of the capacity to produce large numbers of small individuals for release as a result of aquaculture. While exploring the feasibility and bio-economics of tiger prawn (Penaeus esculentus) stock enhancement in Exmouth Gulf, Australia, a broad framework for stock enhancement research was developed. This framework included developing a bio-economic model and risk assessment for all components of a stock enhancement operation, i.e., hatchery, grow-out, harvest-transport-release, population dynamics of released and wild prawns, and monitoring the fishery. However, predictions from the bio-economic model were made with great uncertainty about the values for post-release mortality of prawns because there were no data from experiments on optimal release strategies. Consequently, data from field studies of growth, and laboratory studies of predation, were used to set values for growth and mortality in Monte-Carlo simulations of the potential success of P. esculentus releases in different habitats. The survival of 2 mm carapace length (CL) prawns to the size at emigration from the nursery habitat was ∼1.9 times higher for individuals associated with seagrass of high (∼100 g m-2) than for those in low (∼10 g m-2) plant biomass. Survival was 19 times higher for prawns in high biomass seagrass than for those on bare substratum. The differences in survival between habitats were reduced greatly when the releases of much larger prawns (10 mm CL) were simulated. These findings highlight the importance of research on developing optimal release strategies for prawns in stock enhancement programmes. In particular, the size, habitat, time and density at release needs to be identified to provide more reliable informatio

An international review of the Northern Prawn Fishery tiger prawn assessment was carried out in 2001. The review drew attention to the high level of uncertainty in the assessment and recommended that the logbook data be augmented by fishery independent survey data. In response to the review, industry funded a consultancy project in 2002 to investigate and design an integrated monitoring program for the NPF. Following an industry meeting, NORMAC decided to conduct a one-year pilot survey in 2002/03. The project (FRDC 2002/101) was funded through the FRDC, and included a spawning index survey in August and a recruitment index survey in January. The success of the pilot project led to a FRDC-funded monitoring project (FRDC 2003/075) in 2003/04 and this project (FRDC 2004/099) in 2004/05. Two surveys were undertaken during the 2004/05 financial year.

During October to December 1997, we trawled estuarine habitats in the Joseph Bonaparte Gulf (JBG) to determine the distribution of juvenile red-legged banana prawns, Penaeus indicus (H. Milne Edwards, 1837) and white banana prawns, Penaeus merguiensis (de Man, 1888). We made 229 beam-trawls at 185 sites, mostly over a 100-m path (3-min duration). A Global Positioning System (GPS) receiver was used to verify our location. During October to December 1998, we intensively resampled three of the rivers that were sampled in 1997 to confirm the gulf-wide distribution of P. indicus and P. merguiensis and to investigate the microhabitat use of P. indicus. We chose previously sampled and new sites in Forsyth Creek (eastern JBG), the Lyne River (Cambridge Gulf), and the Berkeley River (western JBG). We made 249 trawls at 21 sites, mostly over 100 m. Juvenile banana prawns were abundant in eastern JBG, Cambridge Gulf and western JBG. They were not abundant in southern JBG, although fewer trawls were made there, due to its inaccessibility. In eastern JBG and Cambridge Gulf, over 96% and 73% (respectively) of juvenile banana prawns were P. indicus and they were more abundant there than in the western JBG. Conversely, in the western JBG over 93% of the juvenile banana prawns were P. merguiensis and they were more abundant than in the eastern JBG and Cambridge Gulf. The Lyne River in the northwestern Cambridge Gulf seems to be the transition zone; both P. indicus and P. merguiensis are equally abundant. P. indicus are most abundant on the mangrove-lined muddy banks of waterways within mangrove forests, similar habitats to P. merguiensis. Within these habitats, they were most abundant in gutters and small creeks, rather than rivers and large creeks. Few P. indicus or P. merguiensis were caught in 100 m 2 trawls undertaken midriver (on the channel bottom and on emergent banks), although these habitats may be only 100 m from the mangrove-lined habitats. In all creek and river habitats

During October to December 1997, we trawled estuarine habitats in the Joseph Bonaparte Gulf (JBG) to determine the distribution of juvenile red-legged banana prawns, Penaeus indicus (H. Milne Edwards, 1837) and white banana prawns, Penaeus merguiensis (de Man, 1888). We made 229 beam-trawls at 185 sites, mostly over a 100-m path (3-min duration). A Global Positioning System (GPS) receiver was used to verify our location. During October to December 1998, we intensively resampled three of the rivers that were sampled in 1997 to confirm the gulf-wide distribution of P. indicus and P. merguiensis and to investigate the microhabitat use of P. indicus. We chose previously sampled and new sites in Forsyth Creek (eastern JBG), the Lyne River (Cambridge Gulf), and the Berkeley River (western JBG). We made 249 trawls at 21 sites, mostly over 100 m. Juvenile banana prawns were abundant in eastern JBG, Cambridge Gulf and western JBG. They were not abundant in southern JBG, although fewer trawls were made there, due to its inaccessibility. In eastern JBG and Cambridge Gulf, over 96% and 73% (respectively) of juvenile banana prawns were P. indicus and they were more abundant there than in the western JBG. Conversely, in the western JBG over 93% of the juvenile banana prawns were P. merguiensis and they were more abundant than in the eastern JBG and Cambridge Gulf. The Lyne River in the northwestern Cambridge Gulf seems to be the transition zone; both P. indicus and P. merguiensis are equally abundant. P. indicus are most abundant on the mangrove-lined muddy banks of waterways within mangrove forests, similar habitats to P. merguiensis. Within these habitats, they were most abundant in gutters and small creeks, rather than rivers and large creeks. Few P. indicus or P. merguiensis were caught in 100 m 2 trawls undertaken midriver (on the channel bottom and on emergent banks), although these habitats may be only 100 m from the mangrove-lined habitats. In all creek and river habitats

Several species of prawns, including juveniles of Penaeus merguiensis, will move into mangrove forests when the forests are inundated by flood tides. However, we do not know how extensively the prawns use the forests or whether some parts of the forests are more valuable to the prawns than others. We assessed the distribution of juvenile prawns in 3 different mangrove communities in intertidal forests adjacent to a small creek and a river in northern Australia between December 1993 and February 1995. The 3 mangrove communities were dominated, respectively, by the structurally complex Rhizophora stylosa, and the more open Ceriops tagal and Avicennia marina. We used stake nets (100 m long, 2 mm mesh) to sample discrete areas of the mangrove forests, and fyke nets (2 mm mesh) to sample prawns moving through the forests. The area of each stake net site ranged from 430 to 650 m2 and the distance of the midpoint of each site inland from the creek or river mangrove fringe ranged from 13 to 225 m. A large size range of juvenile P. merguiensis (2 to 21 mm carapace length, CL) was caught in the mangroves and prawns were caught as far as 200 m into the forests. In the creek forest, there was no clear relationship between mangrove community type and prawn catches. The highest densities of P. merguiensis were recorded in the creek and were 28.1 and 27.6 prawns 100m-2 in Rhizophora sp. and Ceriops sp. forest respectively. The highest mean catches were taken 59 m from the creek mangrove fringe. In contrast to the creek, in the river mangrove forest, there was a clear pattern of catches: the number of P. merguiensis caught decreased with distance into the mangroves and at shallower water depths. We have hypothesised that the distribution of juvenile P. merguiensis inside the mangroves depends largely on the local topography and pattern of water currents within each forest. Large numbers of unidentified Metapenaeus spp. and smaller numbers of M. ensis and P. monodon were also record

Several species of prawns, including juveniles of Penaeus merguiensis, will move into mangrove forests when the forests are inundated by flood tides. However, we do not know how extensively the prawns use the forests or whether some parts of the forests are more valuable to the prawns than others. We assessed the distribution of juvenile prawns in 3 different mangrove communities in intertidal forests adjacent to a small creek and a river in northern Australia between December 1993 and February 1995. The 3 mangrove communities were dominated, respectively, by the structurally complex Rhizophora stylosa, and the more open Ceriops tagal and Avicennia marina. We used stake nets (100 m long, 2 mm mesh) to sample discrete areas of the mangrove forests, and fyke nets (2 mm mesh) to sample prawns moving through the forests. The area of each stake net site ranged from 430 to 650 m2 and the distance of the midpoint of each site inland from the creek or river mangrove fringe ranged from 13 to 225 m. A large size range of juvenile P. merguiensis (2 to 21 mm carapace length, CL) was caught in the mangroves and prawns were caught as far as 200 m into the forests. In the creek forest, there was no clear relationship between mangrove community type and prawn catches. The highest densities of P. merguiensis were recorded in the creek and were 28.1 and 27.6 prawns 100m-2 in Rhizophora sp. and Ceriops sp. forest respectively. The highest mean catches were taken 59 m from the creek mangrove fringe. In contrast to the creek, in the river mangrove forest, there was a clear pattern of catches: the number of P. merguiensis caught decreased with distance into the mangroves and at shallower water depths. We have hypothesised that the distribution of juvenile P. merguiensis inside the mangroves depends largely on the local topography and pattern of water currents within each forest. Large numbers of unidentified Metapenaeus spp. and smaller numbers of M. ensis and P. monodon were also record

Fine mesh enclosures (0.9 m 2 in basal area, 1 m high, with 100 μm mesh) and a jet-net retrieval system were developed to test the influence of juvenile prawn stocking density on growth rates in (1) different months (April and October/November) and (2) different types of intertidal seagrass beds in the Embley River estuary of tropical Australia. Small juvenile tiger prawns (3-6 mm in carapace length, CL) were stocked in enclosures at densities of 4-32 prawns per enclosure (4.4-35.5 prawns m -2) on a high biomass seagrass bed (about 70 g m -2 of mostly Enhalus acoroides) and one with low biomass (about 10 g m -2 of mostly Halodule uninervis). After 2-3 weeks in the enclosures, recovery rates, and hence possibly survival, were greater on the high biomass Enhalus than on the low biomass Halodule. However, not all fish and crustaceans could be excluded from the enclosures. Growth rates were twice as fast on the high biomass Enhalus than on the low biomass Halodule. It is likely that the high biomass Enhalus, with its greater surface area, supported more epiphytic flora and fauna and reduced the potential for interference competition between prawns, compared with the low biomass Halodule. Growth rates on Enhalus were significantly faster at a stocking density of 4 prawns per enclosure (1.3 mm CL week -1) than at a stocking density of 16 and 32 prawns per enclosure (both 0.8 mm CL week -1), and did not differ significantly between April and October/November (temperatures were about 30°C at both times). The mean growth rate at 8 prawns per enclosure (1.1 mm CL week -1) did not differ significantly from those at 4, 16 and 32 prawns per enclosure. These results from two seagrass beds suggest that the carrying capacity for juvenile tiger prawns was greater in the high biomass Enhalus than the low biomass Halodule bed.

Coastal habitats are critical to the sustained production of many fisheries. It is important, therefore, that fishery managers obtain accurate estimates of the extent of these habitats. This study investigated three methods of estimating the linear extent and area of mangroves (commercially available topographic data, aerial photographs and Landsat Thematic Mapper satellite imagery) in two regions in northern Australia: the Joseph Bonaparte Gulf (in particular, the Berkeley and Lyne Rivers) with typically narrow fringes of mangroves (<50 m wide), and the Embley River on Cape York Peninsula, with much broader mangrove stands (50-1000 m wide). Ground-truthing verified that aerial photographs provided the most accurate estimates of extents of mangroves in all rivers, because of their high spatial resolution (2 m). Landsat Thematic Mapper imagery gave good estimates of the area of habitats, but, because of the 30 m pixel resolution, it underestimated the linear extent in places where the mangrove fringe was narrow. Topographic data gave good estimates of the extent of mangroves where the forests were more extensive and less linear in shape, but were very poor otherwise at this scale. These findings have implications for the use of remote sensing techniques in ecological studies in these regions.

Coastal habitats are critical to the sustained production of many fisheries. It is important, therefore, that fishery managers obtain accurate estimates of the extent of these habitats. This study investigated three methods of estimating the linear extent and area of mangroves (commercially available topographic data, aerial photographs and Landsat Thematic Mapper satellite imagery) in two regions in northern Australia: the Joseph Bonaparte Gulf (in particular, the Berkeley and Lyne Rivers) with typically narrow fringes of mangroves (<50 m wide), and the Embley River on Cape York Peninsula, with much broader mangrove stands (50-1000 m wide). Ground-truthing verified that aerial photographs provided the most accurate estimates of extents of mangroves in all rivers, because of their high spatial resolution (2 m). Landsat Thematic Mapper imagery gave good estimates of the area of habitats, but, because of the 30 m pixel resolution, it underestimated the linear extent in places where the mangrove fringe was narrow. Topographic data gave good estimates of the extent of mangroves where the forests were more extensive and less linear in shape, but were very poor otherwise at this scale. These findings have implications for the use of remote sensing techniques in ecological studies in these regions.

The juveniles of red-legged banana prawns Penaeus indicus and white banana prawns P. merguiensis are difficult to distinguish. Gel electrophoresis, with the enzyme malate dehydrogenase, clearly separated into species both large and small prawns. Four morphological characters were measured from juvenile prawns (identified using electrophoresis): three of these provided reliable separation into species. The mean ratio of the distance from the anterior-most dorsal rostral tooth (first tooth) to the tip of the rostrum (L 1 ): to the distance between this tooth and the second tooth (L 1,2 ) was higher for P. indicus (3.16) than P. merguiensis (1.56). All juvenile P. indicus had the first tooth posterior to the antennular peduncle, while it was anterior for all P. merguiensis. More than 90% of P. indicus had six rostral teeth behind the eye, while over 90% of P. merguiensis had five teeth behind the eye. Damage to the rostrum prevented measurements of the position of the first tooth in 20 to 30% of juvenile prawns. Combining the measurements of the ratio of L 1 : L 1,2 and the count of the number of rostral teeth behind the eye meant that all prawns could be unambiguously identified.

Artificial collectors and seagrass units have mainly provided qualitative samples of epifaunal abundance or have been difficult and time-consuming to sample. Consequently, they are useful for distinguishing temporal or spatial trends in abundance or they are deployed for several weeks and, as a result, the quantitative samples are cumulative. We developed a portable artificial seagrass unit (ASU) with buoyant plastic artificial seagrass (47 cm long by 15 mm wide strips) that can be retrieved, harvested and re-deployed with 98% catch efficiency in about 5 min by two people from a small boat. They can quickly and easily quantify settlement of crustacean and fish postlarvae over tidal or deil periods. When set for 24 h, postlarvae settled from the plankton during the night and their abundance is the result of a distinct settlement event. When set for longer periods, the numbers of postlarvae may represent several settlement events and post-settlement activities. Crustacean and fish postlarvae and juveniles used ASUs deployed within seagrass in a similar way to natural seagrass. Estimates of juvenile tiger prawn abundance from beam-trawl catches showed similar densities in natural seagrass (2.93 Penaeus semisulcatus de Haan m−2 day) to those in the ASUs (2.40 P. semisulcatus m−2 day−1); their density was significantly lower on bare trays (0.48 P. semisulcatus m−2 day−1). When deployed on bare areas, more epifaunal crustacean postlarvae were collected from the ASUs (e.g. Portunus pelagicus Linnaeus, 1.21 m−2 day−1; caridean shrimp 4.03 m−2 day−1) than from the bare trays (e.g. P. pelagicus, 0.46 m−2 day−1; caridean shrimp 0.78 m−2 day−1). However, greater abundances of the postlarvae of other crustacean taxa were collected from the bare trays than the ASUs (e.g. Sergestes spp. 1.21 and 0.31 m−2 day−1; tiger prawn postlarvae 0.15 and 0.06 m−2 day−1, respectively). Sampling with portable ASUs allows settlement to be assessed temporally (e.g., daily or tidally) or spatially

Artificial collectors and seagrass units have mainly provided qualitative samples of epifaunal abundance or have been difficult and time-consuming to sample. Consequently, they are useful for distinguishing temporal or spatial trends in abundance or they are deployed for several weeks and, as a result, the quantitative samples are cumulative. We developed a portable artificial seagrass unit (ASU) with buoyant plastic artificial seagrass (47 cm long by 15 mm wide strips) that can be retrieved, harvested and re-deployed with 98% catch efficiency in about 5 min by two people from a small boat. They can quickly and easily quantify settlement of crustacean and fish postlarvae over tidal or deil periods. When set for 24 h, postlarvae settled from the plankton during the night and their abundance is the result of a distinct settlement event. When set for longer periods, the numbers of postlarvae may represent several settlement events and post-settlement activities. Crustacean and fish postlarvae and juveniles used ASUs deployed within seagrass in a similar way to natural seagrass. Estimates of juvenile tiger prawn abundance from beam-trawl catches showed similar densities in natural seagrass (2.93 Penaeus semisulcatus de Haan m−2 day) to those in the ASUs (2.40 P. semisulcatus m−2 day−1); their density was significantly lower on bare trays (0.48 P. semisulcatus m−2 day−1). When deployed on bare areas, more epifaunal crustacean postlarvae were collected from the ASUs (e.g. Portunus pelagicus Linnaeus, 1.21 m−2 day−1; caridean shrimp 4.03 m−2 day−1) than from the bare trays (e.g. P. pelagicus, 0.46 m−2 day−1; caridean shrimp 0.78 m−2 day−1). However, greater abundances of the postlarvae of other crustacean taxa were collected from the bare trays than the ASUs (e.g. Sergestes spp. 1.21 and 0.31 m−2 day−1; tiger prawn postlarvae 0.15 and 0.06 m−2 day−1, respectively). Sampling with portable ASUs allows settlement to be assessed temporally (e.g., daily or tidally) or spatially

The juveniles of red-legged banana prawns Penaeus indicus and white banana prawns P. merguiensis are difficult to distinguish. Gel electrophoresis, with the enzyme malate dehydrogenase, clearly separated into species both large and small prawns. Four morphological characters were measured from juvenile prawns (identified using electrophoresis): three of these provided reliable separation into species. The mean ratio of the distance from the anterior-most dorsal rostral tooth (first tooth) to the tip of the rostrum (L 1 ): to the distance between this tooth and the second tooth (L 1,2 ) was higher for P. indicus (3.16) than P. merguiensis (1.56). All juvenile P. indicus had the first tooth posterior to the antennular peduncle, while it was anterior for all P. merguiensis. More than 90% of P. indicus had six rostral teeth behind the eye, while over 90% of P. merguiensis had five teeth behind the eye. Damage to the rostrum prevented measurements of the position of the first tooth in 20 to 30% of juvenile prawns. Combining the measurements of the ratio of L 1 : L 1,2 and the count of the number of rostral teeth behind the eye meant that all prawns could be unambiguously identified.

We investigated predation of juvenile penaeids, in particular the grooved tiger prawn Penaeus semisulcatus and the brown tiger prawn P. esculentus, by sampling prawns and their potential fish predators on 2 intertidal seagrass beds in the Embley River, Queensland, Australia. Despite large differences in above-ground biomass of seagrass, these beds are both critical nursery areas for postlarval [<3 mm carapace length (CL)] and juvenile (>=3 mm CL) tiger prawns. Thirty-seven species of fish were found to eat juvenile penaeids, but 76% of P. semisulcatus and 43% of P. esculentus (numbers) were found in the stomachs of 1 species of fish (Scomberoides commersonianus). Predation levels at the 2 seagrass beds did not appear to be related to the amount of seagrass present; rather the numbers of Penaeus semisulcatus eaten by fish increased with the numbers of P. semisulcatus on the seagrass beds. Compared to the size distributions of tiger prawns resident on the seagrass, postlarval and small juvenile tiger prawns were under-represented in the stomach contents of fish and only 17 of the 287 P. semisulcatus found in fish stomachs were <=4 mm CL. Although we were unable to quantify the component of natural mortality attributable to predation by fish, we have identified the major fish predators of juvenile tiger prawns and provide evidence suggesting that fish are not significant predators of postlarval and small juvenile tiger prawns.

We studied the fortnightly, seasonal and annual variation in abundance of postlarval and juvenile Penaeus merguiensis in the Embley and Mission River estuaries (northeastern Australia) and of juveniles emigrating from the rivers between September 1986 and March 1992. The climate is tropical with distinct wet and dry seasons, and the pattern of salinity and temperature changes in the estuaries closely reflects this seasonal variation. Postlarval and juvenile prawns were caught for up to 8 mo of the year, from before the wet season to just after the wet season (October to May). There was a weak bimodal distribution of catches during this period but the pattern of variation was quite variable from year to year. There were also subtle differences in the seasonal patterns of abundance between the Embley and Mission Rivers, probably due to different impacts of wet season rainfall on the 2 estuaries. In the rivers, environmental variation explained very little of the variation in catches of planktonic postlarvae, benthic postlarvae and juveniles, but increased emigration of juveniles from the rivers was significantly correlated with increased rainfall. The main factor determining the abundance of juvenile P. merguiensis in the estuaries was the supply and successful settlement of postlarvae from offshore areas. The highest densities of prawns caught within the estuaries were near the upper reaches of small creeks rather than in the main rivers. It is likely that these differences are due to differential settlement patterns of postlarvae between the creek and the main river. Although the largest emigration of juvenile prawns from the estuaries occurred during the wet season, lower levels of migration of prawns from the creeks to the main rivers occurred throughout the year. Over the 6 yr of this project, the annual offshore commercial catch of adult P. merguiensis was significantly correlated with catches of prawns emigrating from the Embley River during the wet season.

Postlarval and juvenile tiger prawns (Penaeus esculentus and P. semisulcatus) were sampled by beam trawls in different seagrass communities of the western Gulf of Carpentaria to test the influence of seagrass characteristics on their distribution and abundance. Two sampling regimes were carried out: fortnightly sampling over a 3 month period during the time of peak recruitment to investigate the importance of different seagrass types to tiger prawns; and at one time to investigate the wider geographic patterns of distribution and abundance of tiger prawns. The results from both studies showed that the pattern of distribution and abundance of 1–1.9 mm carapace length (CL) postlarvae differed from those for the 2–2.9 mm CL postlarvae and juvenile tiger prawns (≥3 mm CL). The 1–1.9 mm CL size class consists of individuals that are about to settle and those that have recently settled: this size class was more widely distributed than the larger postlarvae and juveniles and was sometimes found on substrates where no seagrass was present. Larger tiger prawns were found in higher numbers on seagrass beds where the tall, broad-leaved seagrass Enhalus acoroides was the dominant species. Although seagrass biomass was not a consistent linear predictor of juvenile tiger prawn numbers, mean catches of both the 2–2.9 mm CL postlarvae and juvenile P. esculentus were highest when the biomass of seagrass exceeded 100 g m−2. However, these high biomass seagrass beds contribute only 6% to the total extent of seagrasses in the shallow waters (<2.5 m deep) of the Gulf of Carpentaria. Although the numbers of juvenile tiger prawns were lower in the low-biomass seagrass beds, because of their extent, these seagrass beds are the main nurseries for sustaining the production of the valuable Northern Prawn Fishery in Australia.

We investigated predation of juvenile penaeids, in particular the grooved tiger prawn Penaeus semisulcatus and the brown tiger prawn P. esculentus, by sampling prawns and their potential fish predators on 2 intertidal seagrass beds in the Embley River, Queensland, Australia. Despite large differences in above-ground biomass of seagrass, these beds are both critical nursery areas for postlarval [<3 mm carapace length (CL)] and juvenile (>=3 mm CL) tiger prawns. Thirty-seven species of fish were found to eat juvenile penaeids, but 76% of P. semisulcatus and 43% of P. esculentus (numbers) were found in the stomachs of 1 species of fish (Scomberoides commersonianus). Predation levels at the 2 seagrass beds did not appear to be related to the amount of seagrass present; rather the numbers of Penaeus semisulcatus eaten by fish increased with the numbers of P. semisulcatus on the seagrass beds. Compared to the size distributions of tiger prawns resident on the seagrass, postlarval and small juvenile tiger prawns were under-represented in the stomach contents of fish and only 17 of the 287 P. semisulcatus found in fish stomachs were <=4 mm CL. Although we were unable to quantify the component of natural mortality attributable to predation by fish, we have identified the major fish predators of juvenile tiger prawns and provide evidence suggesting that fish are not significant predators of postlarval and small juvenile tiger prawns.

We studied the fortnightly, seasonal and annual variation in abundance of postlarval and juvenile Penaeus merguiensis in the Embley and Mission River estuaries (northeastern Australia) and of juveniles emigrating from the rivers between September 1986 and March 1992. The climate is tropical with distinct wet and dry seasons, and the pattern of salinity and temperature changes in the estuaries closely reflects this seasonal variation. Postlarval and juvenile prawns were caught for up to 8 mo of the year, from before the wet season to just after the wet season (October to May). There was a weak bimodal distribution of catches during this period but the pattern of variation was quite variable from year to year. There were also subtle differences in the seasonal patterns of abundance between the Embley and Mission Rivers, probably due to different impacts of wet season rainfall on the 2 estuaries. In the rivers, environmental variation explained very little of the variation in catches of planktonic postlarvae, benthic postlarvae and juveniles, but increased emigration of juveniles from the rivers was significantly correlated with increased rainfall. The main factor determining the abundance of juvenile P. merguiensis in the estuaries was the supply and successful settlement of postlarvae from offshore areas. The highest densities of prawns caught within the estuaries were near the upper reaches of small creeks rather than in the main rivers. It is likely that these differences are due to differential settlement patterns of postlarvae between the creek and the main river. Although the largest emigration of juvenile prawns from the estuaries occurred during the wet season, lower levels of migration of prawns from the creeks to the main rivers occurred throughout the year. Over the 6 yr of this project, the annual offshore commercial catch of adult P. merguiensis was significantly correlated with catches of prawns emigrating from the Embley River during the wet season.

Postlarval and juvenile tiger prawns (Penaeus esculentus and P. semisulcatus) were sampled by beam trawls in different seagrass communities of the western Gulf of Carpentaria to test the influence of seagrass characteristics on their distribution and abundance. Two sampling regimes were carried out: fortnightly sampling over a 3 month period during the time of peak recruitment to investigate the importance of different seagrass types to tiger prawns; and at one time to investigate the wider geographic patterns of distribution and abundance of tiger prawns. The results from both studies showed that the pattern of distribution and abundance of 1–1.9 mm carapace length (CL) postlarvae differed from those for the 2–2.9 mm CL postlarvae and juvenile tiger prawns (≥3 mm CL). The 1–1.9 mm CL size class consists of individuals that are about to settle and those that have recently settled: this size class was more widely distributed than the larger postlarvae and juveniles and was sometimes found on substrates where no seagrass was present. Larger tiger prawns were found in higher numbers on seagrass beds where the tall, broad-leaved seagrass Enhalus acoroides was the dominant species. Although seagrass biomass was not a consistent linear predictor of juvenile tiger prawn numbers, mean catches of both the 2–2.9 mm CL postlarvae and juvenile P. esculentus were highest when the biomass of seagrass exceeded 100 g m−2. However, these high biomass seagrass beds contribute only 6% to the total extent of seagrasses in the shallow waters (<2.5 m deep) of the Gulf of Carpentaria. Although the numbers of juvenile tiger prawns were lower in the low-biomass seagrass beds, because of their extent, these seagrass beds are the main nurseries for sustaining the production of the valuable Northern Prawn Fishery in Australia.

The microhabitat preferences of juvenile tiger prawns (3-10 mm carapace length), Penaeus esculentus and Penaeus semisulcatus, were tested in the field at Groote Eylandt, in the western Gulf of Carpentaria, Australia. A partitioned apparatus containing live seagrass was used. Both species of prawns selected seagrass (Syringodium isoetifolium) over bare substrate. Juvenile P. esculentus, the most abundant species in this regions were also given paired choices of seagrasses with different leaf morphologies (representing a range of structural complexity) and sediments of different particle size. They selected a seagrass with broad, long leaves (Cymodocea serrulata) over one with narrow, long leaves (S. isoetifolium), which in turn was selected over the seagrasses with narrow short leaves (Halodule uninervis and shortened S. isoetifolium). Predation experiments have shown that juvenile P. esculentus are detected and eaten less often in broad, long-leaved seagrass than in narrow, short-leaved seagrass or bare substrate, so their preference for the former may shelter them from predators. No habitat preference was evident for P. esculentus when offered a choice between sediments consisting mainly of sand (71% sand particles) and silt (60% of silt and clay). The selection by both species of tiger prawn of seagrass over bare substrate, and P. esculentus's selection of seagrass with long, broad leaves, provides an explanation for the distribution of juvenile tiger prawns in the field. Thus, in the seagrass beds around Groote Eylandt, P. esculentus is more abundant in seagrass with broad, long leaves than in seagrass with short, thin leaves. In addition, its distribution in this region is relatively independent of sediment type. Leaf surface area (or habitat structural complexity) appears to be the main determinant of distribution for juvenile P. esculentus.

The microhabitat preferences of juvenile tiger prawns (3-10 mm carapace length), Penaeus esculentus and Penaeus semisulcatus, were tested in the field at Groote Eylandt, in the western Gulf of Carpentaria, Australia. A partitioned apparatus containing live seagrass was used. Both species of prawns selected seagrass (Syringodium isoetifolium) over bare substrate. Juvenile P. esculentus, the most abundant species in this regions were also given paired choices of seagrasses with different leaf morphologies (representing a range of structural complexity) and sediments of different particle size. They selected a seagrass with broad, long leaves (Cymodocea serrulata) over one with narrow, long leaves (S. isoetifolium), which in turn was selected over the seagrasses with narrow short leaves (Halodule uninervis and shortened S. isoetifolium). Predation experiments have shown that juvenile P. esculentus are detected and eaten less often in broad, long-leaved seagrass than in narrow, short-leaved seagrass or bare substrate, so their preference for the former may shelter them from predators. No habitat preference was evident for P. esculentus when offered a choice between sediments consisting mainly of sand (71% sand particles) and silt (60% of silt and clay). The selection by both species of tiger prawn of seagrass over bare substrate, and P. esculentus's selection of seagrass with long, broad leaves, provides an explanation for the distribution of juvenile tiger prawns in the field. Thus, in the seagrass beds around Groote Eylandt, P. esculentus is more abundant in seagrass with broad, long leaves than in seagrass with short, thin leaves. In addition, its distribution in this region is relatively independent of sediment type. Leaf surface area (or habitat structural complexity) appears to be the main determinant of distribution for juvenile P. esculentus.

It is accepted that mangroves are important nursery areas for prawns and fish, including some of major commercial importance, but little is known about how these mobile animals use the mangrove forests. We recorded the distribution of juvenile banana prawns Penaeusmerguiensis and of fish in an intertidal mangrove forest adjacent to a small creek in northern Australia in November 1992 and in March 1993. Four discrete areas of the forest were enclosed with a 100 m long, 2 mm mesh stake net: 2 at the creek mangrove fringe and 2 at further distances into the mangroves. The mean distance of each sampling site inland from the creek mangrove fringe ranged from 13 to 59 m and the area of the sites ranged from 480 to 640 m2. Two mangrove communities, one dominated by the structurally complex Rhizophorastylosa, the other by the more open Ceriopstagal were sampled. A large size range of juvenile prawns and small fish moved at least 43 m into the mangroves at high tide, and the density of prawns near the creek mangrove fringe was inversely related to the maximum tide height. The highest density of juvenile P. merguiensis recorded in the mangroves in November was 18.3 prawns 100 m-2 and in March was 334.5 prawns 100 m-2. Mean fish density over all samples was 83.0 fish 100 m-2 and mean fish biomass was 3.9 g m-2; 55 species of fish were caught during the sampling. P.merguiensis showed no apparent preference for either of the 2 mangrove communities sampled; however, more fish (101 fish 100 m-2) and more fish species (26) were caught at the creek mangrove fringe site than at the other more inland sites; the lowest numbers of fish (27 fish m-2) and species (13) were caught at the inland site (Ceriops). On average, fish caught at the fringe site were also longer and heavier than fish caught at the other sites. By moving well into the mangrove forest, prawns and small fish are probably less vulnerable to predation by larger fish.

It is accepted that mangroves are important nursery areas for prawns and fish, including some of major commercial importance, but little is known about how these mobile animals use the mangrove forests. We recorded the distribution of juvenile banana prawns Penaeusmerguiensis and of fish in an intertidal mangrove forest adjacent to a small creek in northern Australia in November 1992 and in March 1993. Four discrete areas of the forest were enclosed with a 100 m long, 2 mm mesh stake net: 2 at the creek mangrove fringe and 2 at further distances into the mangroves. The mean distance of each sampling site inland from the creek mangrove fringe ranged from 13 to 59 m and the area of the sites ranged from 480 to 640 m2. Two mangrove communities, one dominated by the structurally complex Rhizophorastylosa, the other by the more open Ceriopstagal were sampled. A large size range of juvenile prawns and small fish moved at least 43 m into the mangroves at high tide, and the density of prawns near the creek mangrove fringe was inversely related to the maximum tide height. The highest density of juvenile P. merguiensis recorded in the mangroves in November was 18.3 prawns 100 m-2 and in March was 334.5 prawns 100 m-2. Mean fish density over all samples was 83.0 fish 100 m-2 and mean fish biomass was 3.9 g m-2; 55 species of fish were caught during the sampling. P.merguiensis showed no apparent preference for either of the 2 mangrove communities sampled; however, more fish (101 fish 100 m-2) and more fish species (26) were caught at the creek mangrove fringe site than at the other more inland sites; the lowest numbers of fish (27 fish m-2) and species (13) were caught at the inland site (Ceriops). On average, fish caught at the fringe site were also longer and heavier than fish caught at the other sites. By moving well into the mangrove forest, prawns and small fish are probably less vulnerable to predation by larger fish.

The efficiency with which a small beam trawl (1 x 0.5 m mouth) sampled postlarvae and juveniles of tiger prawns Penaeus esculentus and P. semisulcatus at night was estimated in 3 tropical seagrass communities (dominated by Thalassia hemprichii, Syringodium isoetifolium and Enhalus acoroides, respectively) in the shallow waters of the Gulf of Carpentaria in northern Australia. An area of seagrass (40 x 3 m) was enclosed by a net and the beam trawl was repeatedly hand-hauled over the substrate. Net efficiency (q) was calculated using 4 methods: the unweighted Leslie, weighted Leslie, DeLury and Maximum-likelihood (ML) methods. The Maximum-likelihood is the preferred method for estimating efficiency because it makes the fewest assumptions and is not affected by zero catches. The major difference in net efficiencies was between postlarvae (mean ML q +/- 95% confidence limits = 0.66 +/- 0.16) and juveniles of both species (mean q for juveniles in water <=1.0 m deep = 0.47 +/- 0.05), i.e. the beam trawl was more efficient at capturing postlarvae than juveniles. There was little difference in net efficiency for P. esculentus between seagrass types (T. hemprichii versus S. isoetifolium), even though the biomass and morphologies of seagrass in these communities differed greatly (biomasses were 54 and 204 g m-2, respectively). The efficiency of the net appeared to be the same for juveniles of the 2 species in shallow water, but was lower for juvenile P. semisulcatus at high tide when the water was deeper (1.6 to 1.9 m) (0.35 +/- 0.08). The lower efficiency near the time of high tide is possibly because the prawns are more active at high than low tide, and can also escape above the net. Factors affecting net efficiency and alternative methods of estimating net efficiency are discussed.

The burying and sheltering behaviour of two different sizes of juvenile tiger prawns (shrimp) Penaeus esculentus Haswell was studied during light and dark periods, in three habitats: bare silt substratum; short, thin-leaved seagrass (Halodule uninervis Aschers.); and tall, broad-leaved seagrass (Cymodocea serrulata Ashers & Magnus). Relative predation rates by the sand bass Psammoperca waigiensis Cuvier were also studied in these three habitats and in an additional seagrass with tall, thin leaves (Syringodium isoetifolium Ashers). Burying behaviour differed with prawn size, seagrass habitat and exposure to light. Small prawns (2.5 to 3.5 mm carapace length [CL]) rarely buried, even on bare substratum, whereas large prawns (11 to 13 mm CL) buried often and in all habitat types. However, large prawns buried in the substratum less often in tall, broad-leaved seagrass than in short, thin-leaved seagrass. Small prawns remained emerged in both the light and dark, whereas large prawns spent more time above the substratum in the dark than in the light. Small prawns perched on all types of seagrass leaves, whereas large prawns perched only on the broad-leaved seagrass. Fish detected juvenile P. esculentus (4 to 5 mm CL) more quickly in narrow-leaved seagrasses than in broad-leaved seagrass, and more quickly and more often, on bare substratum than in narrow-leaved seagrasses. More juvenile prawns were detected in bare substratum and the thin-leaved seagrasses, S. isoetifolium and H. uninervis, than in the broader-leaved C. serrulata. More prawns were caught and eaten on bare substratum than in the broad-leaved seagrass. These results indicated that small juvenile tiger prawns would suffer higher rates of predation in short, thin seagrass and unvegetated habitats, and this would lead to a decrease in abundance in these habitats. In the field, observations of the abundance of prawns in seagrass habitats of different structural complexity suggests post-settlement processes such

The burying and sheltering behaviour of two different sizes of juvenile tiger prawns (shrimp) Penaeus esculentus Haswell was studied during light and dark periods, in three habitats: bare silt substratum; short, thin-leaved seagrass (Halodule uninervis Aschers.); and tall, broad-leaved seagrass (Cymodocea serrulata Ashers & Magnus). Relative predation rates by the sand bass Psammoperca waigiensis Cuvier were also studied in these three habitats and in an additional seagrass with tall, thin leaves (Syringodium isoetifolium Ashers). Burying behaviour differed with prawn size, seagrass habitat and exposure to light. Small prawns (2.5 to 3.5 mm carapace length [CL]) rarely buried, even on bare substratum, whereas large prawns (11 to 13 mm CL) buried often and in all habitat types. However, large prawns buried in the substratum less often in tall, broad-leaved seagrass than in short, thin-leaved seagrass. Small prawns remained emerged in both the light and dark, whereas large prawns spent more time above the substratum in the dark than in the light. Small prawns perched on all types of seagrass leaves, whereas large prawns perched only on the broad-leaved seagrass. Fish detected juvenile P. esculentus (4 to 5 mm CL) more quickly in narrow-leaved seagrasses than in broad-leaved seagrass, and more quickly and more often, on bare substratum than in narrow-leaved seagrasses. More juvenile prawns were detected in bare substratum and the thin-leaved seagrasses, S. isoetifolium and H. uninervis, than in the broader-leaved C. serrulata. More prawns were caught and eaten on bare substratum than in the broad-leaved seagrass. These results indicated that small juvenile tiger prawns would suffer higher rates of predation in short, thin seagrass and unvegetated habitats, and this would lead to a decrease in abundance in these habitats. In the field, observations of the abundance of prawns in seagrass habitats of different structural complexity suggests post-settlement processes such

The efficiency with which a small beam trawl (1 x 0.5 m mouth) sampled postlarvae and juveniles of tiger prawns Penaeus esculentus and P. semisulcatus at night was estimated in 3 tropical seagrass communities (dominated by Thalassia hemprichii, Syringodium isoetifolium and Enhalus acoroides, respectively) in the shallow waters of the Gulf of Carpentaria in northern Australia. An area of seagrass (40 x 3 m) was enclosed by a net and the beam trawl was repeatedly hand-hauled over the substrate. Net efficiency (q) was calculated using 4 methods: the unweighted Leslie, weighted Leslie, DeLury and Maximum-likelihood (ML) methods. The Maximum-likelihood is the preferred method for estimating efficiency because it makes the fewest assumptions and is not affected by zero catches. The major difference in net efficiencies was between postlarvae (mean ML q +/- 95% confidence limits = 0.66 +/- 0.16) and juveniles of both species (mean q for juveniles in water <=1.0 m deep = 0.47 +/- 0.05), i.e. the beam trawl was more efficient at capturing postlarvae than juveniles. There was little difference in net efficiency for P. esculentus between seagrass types (T. hemprichii versus S. isoetifolium), even though the biomass and morphologies of seagrass in these communities differed greatly (biomasses were 54 and 204 g m-2, respectively). The efficiency of the net appeared to be the same for juveniles of the 2 species in shallow water, but was lower for juvenile P. semisulcatus at high tide when the water was deeper (1.6 to 1.9 m) (0.35 +/- 0.08). The lower efficiency near the time of high tide is possibly because the prawns are more active at high than low tide, and can also escape above the net. Factors affecting net efficiency and alternative methods of estimating net efficiency are discussed.

The population dynamics of small tiger prawns (Penaeus esculentus and P. semisulcatus) were studied at three sites around north-western Groote Eylandt, Gulf of Carpentaria, Australia, between August 1983 and August 1984. Seagrasses typical of open-coastline, reef-flat and river-mouth communities were found in the shallow depths (≤2.5 m) at these sites. The temperature and salinity of the bottom waters did not differ among the shallowest depths of the three sites and mean values at night ranged from 21.9 to 32.0 °C, and from 30.1 to 37.5% S. Data from fortnightly sampling with beam trawls showed that virtually all post-larvae (∼90%) were caught in the intertidal and shallow subtidal waters (≥2.0 m deep). At one site, where the relationship between seagrass biomass, catches and depth could be studied in detail, high catches were confined to seagrass in shallow water, within 200 m of the high-water mark. This was despite the fact that seagrass beds of high biomass (>100 g m-2 between August and February) were found nearby, in only slightly deeper water (2.5 m). It is likely, therefore, that only the seagrass beds in shallow waters of the Gulf of Carpentaria act as important settlement and nursery areas for tiger prawns. In general, catches of tiger prawn postlarvae (both P. esculentus and P. semisulcatus) and juvenile P. esculentus on the seagrass in the shallowest waters at each site were higher in the tropical prewet (October–December) and wet (January–March) seasons than at other times of the year. Juvenile P. semisulcatus catches were highest in the pre-wet season. While seasonal differences accounted for the highest proportion of variation in catches of tiger prawn postlarvae and juvenile P. semisulcatus, site was the most important factor for juvenile P. esculentus. In each season, catches of juvenile P. esculentus were highest in the shallow, open-coastline seagrass, where the biomass of seagrass was highest. The fact that the type of seagrass community appears

The population dynamics of small tiger prawns (Penaeus esculentus and P. semisulcatus) were studied at three sites around north-western Groote Eylandt, Gulf of Carpentaria, Australia, between August 1983 and August 1984. Seagrasses typical of open-coastline, reef-flat and river-mouth communities were found in the shallow depths (≤2.5 m) at these sites. The temperature and salinity of the bottom waters did not differ among the shallowest depths of the three sites and mean values at night ranged from 21.9 to 32.0 °C, and from 30.1 to 37.5% S. Data from fortnightly sampling with beam trawls showed that virtually all post-larvae (∼90%) were caught in the intertidal and shallow subtidal waters (≥2.0 m deep). At one site, where the relationship between seagrass biomass, catches and depth could be studied in detail, high catches were confined to seagrass in shallow water, within 200 m of the high-water mark. This was despite the fact that seagrass beds of high biomass (>100 g m-2 between August and February) were found nearby, in only slightly deeper water (2.5 m). It is likely, therefore, that only the seagrass beds in shallow waters of the Gulf of Carpentaria act as important settlement and nursery areas for tiger prawns. In general, catches of tiger prawn postlarvae (both P. esculentus and P. semisulcatus) and juvenile P. esculentus on the seagrass in the shallowest waters at each site were higher in the tropical prewet (October–December) and wet (January–March) seasons than at other times of the year. Juvenile P. semisulcatus catches were highest in the pre-wet season. While seasonal differences accounted for the highest proportion of variation in catches of tiger prawn postlarvae and juvenile P. semisulcatus, site was the most important factor for juvenile P. esculentus. In each season, catches of juvenile P. esculentus were highest in the shallow, open-coastline seagrass, where the biomass of seagrass was highest. The fact that the type of seagrass community appears