The need to optimise spawning success and survival of offspring by fishes has resulted in the evolution of a vast array of reproductive strategies, such as spawning aggregations. A spawning aggregation is defined by Domeier and Colin (1997) as, “a group of con-specific fish gathered for the purpose of spawning with fish densities or numbers significantly higher than those found in the area of aggregation during non-reproductive periods”. However, Mackie et al. (2009) noted that this is not always the case, as individuals of a species may aggregate to spawn in numbers lower than those found in schools of the same species, at the same location, outside of spawning times. A more pertinent description of an aggregation from a management perspective is that reproductively active fish are grouped together in a manner which increases their vulnerability to fishing.
Over-fishing of spawning aggregations is often associated with the collapse of the fisheries they supported. Management of fisheries that exploit such aggregations has become a crucial element in sustaining such resources. To effectively assess the biomass, distribution, behaviour and ecological importance of spawning aggregations, techniques are required which are non-invasive, incite as little behavioural bias as possible, can repetitively acquire high-resolution data for periods up to entire spawning seasons and are comparatively easy and cost effective to deploy. Conventional sampling techniques, such as video census (for example, Diver Operated Video surveys or Baited Remote Underwater Video systems), egg tows or catch related sampling often offer only a snapshot-in-time of the aggregation and in some cases suffer from sampling bias induced by the method itself. No sampling method is exempt from bias; however, the integration of a suite of techniques pertinent to the biological and behavioural characteristics of the species can provide complementary data sets that allow a better understanding of their biases.
Although not without limitations, acoustic techniques offer unique, complementary methods to other sampling techniques. Sound waves propagate efficiently through water, allowing acoustic ‘observation’ of sound-producing marine animals over considerably greater distances and ranges of conditions than visual techniques. Over 800 species of fish reportedly produce sound, many during spawning and the recording of these vocalisations is being increasingly used to acquire information on species that aggregate. The lack of human interaction during the recording of fish calls means that passive acoustics offers a non-extractive method of monitoring vocal fish. Sound pressure levels (SPLs) of fish choruses have been shown to be related to the number of calling fish present and provide important information on likely ecological (environmental or anthropogenic) correlates behind the timing, spatial distribution and relative size of aggregations and the behaviours associated with them.
The proposal for this project was developed in liaison with the Department of Fisheries, Western Australia (DoFWA); Aquaculture Development Unit (ADU) at Fremantle Challenger Institute of Technology; the Shark Bay Ecosystem Research Project (SBERP); the Department of Parks and Wildlife (DPaW) and; the Curtin Aquatic Research Laboratory (CARL), who have all provided in-kind support to various areas of the project. Recreational and commercial fishing groups including Recfishwest, Recfishing Research and Western Australia Fishing Industry Council (WAFIC) were consulted for advice and support. The Swan River Trust was also consulted to identify the benefits this project could offer towards their current objectives, and have provided in-kind support in the form of access to a substantial environmental dataset for the Swan River to compare environmental variables with sound production levels.
One of the primary objectives of this project was to quantify the spatial and temporal variations in sound production of mulloway at various aggregation sites. In addition, at the beginning of this project no reports of sound production by WA dhufish, snapper and black bream existed; thus, the project aimed to determine whether these species are vocal and, if so, whether their vocalisations are of use as a fisheries-independent monitoring tool. Deployments have been successfully completed along the WA coastline from Augusta to Shark Bay to record vocalisations by the four target species. In excess of 1.84 Tbytes of data have been collected during the course of the project, comprising a total of 9015 hrs (375 days) of recordings over a total deployment period of 818 days. Analysis from a further nine datasets, comprising 894 days deployment (287 days of recording in total), provided from affiliated projects has helped investigate a variety of fish calls and choruses off Western Australia.
Keywords: Sound production, spawning aggregation, fisheries management, propagation, mulloway, dhufish, snapper, black bream, Western Australia.
