Fish populations in Half Moon Creek, near Cairns, north Queensland were sampled in the wet season (12-16 November 1990) and in the dry season (11-15 April 1991). Replicated sampling with three gear types (multipanel gill nets, pocket seine and fish traps) was carried out.
Pelagic and large benthic fishes using the main channels were sampled with static gill nets anchored across the tidal flow, from the bank towards the centre of the creek. Gill nets were set at six locations. With the exception of the most seaward site, each net was located so as to cut off one or more side channels leading from the mangroves as well as intercepting fish moving downstream following the bank of the main channel. Two sites were located in the side branch leading to Reed Canal where most modification would occur and two sites were set in Half Moon Creek below (seaward) of this junction. These sites were selected on the basis that they would experience a different flow regime after development. Two sites were located above (landward) of this junction and were expected to remain relatively unchanged after development. Each of the five day sampling periods was chosen so that a high tide fell during the morning. All six gill nets were deployed during the last hour of the flood tide, fished for approximately five hours on the ebb, and were retrieved during the final hour of the ebb.
Four shallow locations were found where it was possible to pull a small pocket seine net to collect benthic fishes. During both of the field trips, each one of the four sites was sampled once, at which time three replicates were taken on the same day. Each replicate consisted of a 10m advance once the net was fully open in which case it should have sampled an area in front of the net of at least 50m².
To sample larger roving fishes, baited fish traps were deployed at six locations in the creek. Generally, these sites were adjacent to the gill nets although they had to be placed in microhabitats where predatory fishes might be expected to congregate such as overhanging banks or snags. These traps were deployed on the first day of sampling and thereafter checked at least daily. Baits consisting mostly of frozen Western Australian pilchards (Sardinops neopilchardus), supplemented by fresh mullet, were suspended inside the cage in a permeable canister; they were replaced as required but refreshed at least every second day.
Live fish caught in the baited fish traps were identified, measured (Standard Length, mm), tagged and released back into the creek. Later, their individual weights were estimated by extrapolation from dead fish of similar shape and size. All other fish were preserved at the point of collection in 10% formalin. Subsequently, every individual was identified, measured (Standard Length, mm) and weighed on an electronic balance (g). In the case of common species with little size variation, especially juveniles, not all individuals were measured; instead, a size range was established and their aggregate biomass determined for each sample.
This research was undertaken to characterise the fish fauna of Half Moon Creek and to evaluate options for monitoring the impact of a marina development in Reed Canal, which flows into Half Moon Creek. The planned alterations in Reed Canal involve considerable excavation to enlarge the internal space as well as dredging downstream to maintain a navigable channel to the sea.
Doherty, Peter J, Dr (Principal Investigator)
Statement: Sampling Gear:
1. Gill Nets
Each net was 30m long, with a 3m drop, and consisted of three contiguous 10m sections each of different mesh size: 19mm (ply 0.3mm), 32mm (ply 0.3mm), and 45mm (ply 0.4mm). The order of the different meshes varied among the nets which meant that care was taken to ensure that a particular combination was not confounded with orientation (position) in the stream or location (site) in the creek; i.e. it would be undesirable to sample one location consistently with the finest panel against the bank while another was always sampled with the coarsest mesh in this position. To get around this problem, the individual nets were numbered and rotated haphazardly among the sampling locations in order to achieve the most balance possible among the samples from the different sites.
While the nets were fishing, the creek was patrolled constantly to ensure that there was no interference with the gear and to ensure that each net retained the most appropriate orientation for effective fishing. The latter was important in the wet season when the extra flow down the creek was accompanied by large amounts of debris, some of which had to be removed to stop the nets from dragging their anchors and straightening out downstream.
2. Pocket Seine
This net measured 6m in length, 2m deep and was constructed of 3mm mesh throughout, with a solid cod end. It was the same one used in previous surveys of other mangrove creeks in North Queensland (Robertson and Duke 1987, 1990a,b) and the same method of deployment was followed to ensure comparable results.
Robertson, AI, Duke, NC (1987) Mangroves as nursery sites: comparisons of the abundance and species composition of fish and crustaceans in mangroves and other nearshore habitats in tropical Australia. Marine Biology 96: 193-205.
Robertson, AI, Duke, NC (1990a) Mangrove fish-communities in tropical Queensland, Australia: spatial and temporal patterns in densities, biomass and community structure. Marine Biology 104: 369-379.
Robertson, AI, Duke, NC (1990b) Recruitment, growth, and residence time of fishes in a tropical Australian mangrove system. Estuarine, Coastal and Shelf Science 31: 723-743.
3. Fish Traps
The trap design was a small version of an Antillean Z-trap which has been used with considerable success in other creeks in north Queensland (Sheaves 1992).
Sheaves. M.J. (1992) Patterns of distribution and abundance of fishes in different habitats of a mangrove-lined tropical estuary as determined by fish trapping. Australian Journal of Marine and Freshwater Research 43: 1461-1479.
All data were stored in a dBASE IV database and have been checked twice against the raw data files for recording errors.