University of Tasmania

File(s) under permanent embargo

An ecosystem approach, the acoustic assessment of the Northern Demersal Scalefish Fishery - distribution, habitat and abundance

posted on 2023-05-26, 14:59 authored by Gastauer, S

In order to progress towards a more ecosystem based approach in fisheries management (EBFM), it is important to develop holistic monitoring protocols. Fisheries acoustics is recognised as one of the most promising marine monitoring tools, to be used in conjunction with more traditional sampling methods, such as trawls or other biological sampling methods, to fulfil the data needs of EBFM. To work towards an improved understanding of the ecosystem, there is a need for large scale sampling programs on a temporal and spatial basis, monitoring the entire water column and habitat characteristics. Given the cost involved in running dedicated scientific survey, such a large scale data collection program is often impractical. As an alternative, fishing vessels, with acoustic equipment of scientific standards available or easily mounted can be used as opportunistic or chartered sampling platforms. Opportunistic data collection programs, take advantage of the time spent at sea by fishers. Such an undertaking bears the possibility to continuously collect data over a prolonged period of time, covering areas of commercial interest.

The present thesis can be seen as one of the first attempts to describe tropical, demersal habitats based on opportunistically collected acoustic data and catch information aboard a small (15 m) trap fishing vessel in the Northern Demersal Scalefish Fishery (NDSF). The NDSF is classified as a mixed fishery, targeting a multitude of demersal species, with two main indicator species, goldband snapper (Pristipomoides multidens) and red emperor (Lutjanus sebae). The main goal for this study was to derive indices, within selected fishing regions (where sufficient data was available), which could be of relevance to EBFM. Three fishing regions with simultaneously collected catch and acoustic data were identified. Optical recordings of the catch, facilitated the extraction of species group-specific composition and length frequency distributions for each recorded trap. Length estimates were based on pixel counting, referenced to the five cm steel meshes of the traps. Calibrated acoustic data was logged twenty-four hours a day at 38 and 120 kHz.

Abundance and biomass indices are the most common indices used in EBFM. In order to translate acoustic densities into biologically meaningful abundance or biomass estimates, a good understanding of the relationship between the acoustic reflection of an individual fish (TS) of a given species at a given length L (TS-L) is necessary. Given the sparsity of acoustic studies focusing on tropical demersal habitats, TS-L relationships for the observed groups had to be established. TS-L of red emperor was established through a combination of ex situ measurements and a modelling approach. Model parameters were calibrated by the ex situ measurements, using a Bayesian approach. For goldband snapper, in situ collected ˆ value could be used in combination with catch information to establish a robust TS-L relationship. For the other species groups Kirchhoff Ray mode models, based on computational tomography scans of fish samples, were used to model TS. In a following step, the distribution, abundance and biomass of key species groups were estimated. Through the application of geostatistical methods, associated estimates of variance were described.

Other indices, which can be used to describe the distribution patterns of species groups within the selected fishing groups were proposed. Acoustic density hotspots were described through geostatistical methods. Hotspots can help us better understand key areas of biological density. Bottom habitat types were detected through clustering of acoustic bottom descriptors, providing an acoustic habitat classification indicator. Indicator species associated with each habitat cluster were detected and described. Energetic, geometric and bathymetric characteristics of fish schools were described and clustered together, to form an indicator of acoustic diversity. All of these indicators have the potential, if collected over multiple years, to track changes in resource distributions and to link those to changes in environmental or anthropogenic factors. Further, they all contain information on the spatio-temporal structure of the monitored regions, providing valuable information for improved planning and timing of dedicated survey.

As such, the present thesis has shown that useful indices can be derived from acoustic and catch information collected opportunistically aboard a small trap fishing vessel during routine operations. The major outcomes of the thesis are a novel technique to estimate in situ TS-L relationships, using the distribution of length and TS measurements, rather than grouped frequencies. Robust methods for the combination of catch information and acoustic densities in situations where no directed sampling is available have been established. Variance estimates observed within three focus fishing regions were comparable with those observed in dedicated acoustic surveys conducted in other fisheries. This similarity in variance is an indicator that the quality of biomass and abundance estimates was satisfactory. Goldband snapper has been observed in single species schools and, as a result, improved variance estimates could be established. In a final step, through the use of unsupervised statistical methods, simple indicators with the potential to gain improved insights into the distribution and structure of the fish resources and associated habitats were gained and developed.

The presented methods hold the possibility to be applied to a wide range of habitats and ecosystems. The use of fishing vessels as sampling platforms not only generates useful information, but also strengthens the mutual understanding between scientists and fishers, working together towards more stakeholder involvement in data collection programs and higher acceptance of resulting policies. If data collection programs are to be continued and the presented methods applied, the true value of the latter could be explored in more detail.





Institute for Marine and Antarctic Studies


Curtin University

Repository Status

  • Restricted

Usage metrics

    Thesis collection


    Ref. manager