Australia: State of the Environment Second Technical Paper Series (Coasts and Oceans), Series 2
David Barratt, John Garvey and Jean Chesson
Bureau of Resource Sciences, Australia
Department of the Environment and Heritage, 2001
ISBN 0 6425 4745 9
There is a discussion section within each indicator chapter. This chapter addresses issues that transcend individual indicators.
State of environment reporting seeks to provide an easily digestible, yet meaningful, overview of the state of Australia's environment and how it is changing over time. In order to do this, information must be aggregated in various ways (across space, across time, across species, across types of activities, across fisheries). While carrying out the work described in the preceding chapters it has become increasingly obvious that the approach to aggregation is of critical importance.
Three of the four indicators are proposed as indicators of pressure. While it is reasonable to suppose that increases in these indicators (or their sub-components) could result in greater impacts on the marine environment, it is unlikely that the relationship is always linear or is identical in different places or at different times. Combining information by simple addition or averaging implicitly assumes a linear relationship and may give misleading results. For example, the death of 100 seabirds in one fishery may be inconsequential whereas the death of 1 seabird in another fishery may be of great significance. Combining these data and reporting the death of 101 seabirds obscures important information.
Environment Australia specifically requested that an attempt be made to combine Indicators 3 and 4. The result of this exercise is shown in the box below. This experience together with consideration of Indicators 1 and 2 has demonstrated that there are at least two different approaches to developing national indicators (Figure 8.1). The first approach adds or otherwise combines information from each contributing unit (for example, a fishery). To do this, data need to be compatible across units and impacts are assumed additive. The significance of the value of the indicator is then determined against some national objective. In the second approach, data are collected and their significance assessed against an objective specified at the level of the unit. The national indicator is then expressed in the form "number or proportion of units meeting their individual objectives". In the second approach, the type of information collected may differ among units.
The first approach is suitable for some indicators. The second approach is more suitable for other indicators. Applying the first approach inappropriately may require a large effort and provide little or even misleading information.
The ideal unit would be determined by a combination of physical, ecological and management criteria so that desired states, measurements of pressure and their impact and responses to those pressures could be sensibly discussed and evaluated. Catchments are a terrestrial example of such units. The IMCRA regions represent an attempt to classify the marine environment, but they are inadequate for the indicators considered here because they take no account of management arrangements. A significant amount of fishing effort and petroleum exploration and extraction takes place outside the IMCRA regions. At present, the best available unit for fishery-related issues is the fishery since this is the unit of data collection and management decisions.
Figure 8.1 Two approaches to deriving national indicators. In Approach 1, compatible data from each reporting unit is combined and then the indicator is assessed against a national objective. In Approach 2, possibly different data are assessed within each reporting unit and the national indicator provides a summary of the results.
We suggest that Approach 1 can be used for indicator 1 (bycatch) if the indicator is defined in terms of total removals. The basic data is or should be collected for every fishery. However, once attention moves to more detailed issues such as individual species, the approach should shift to Approach 2 where data collection and interpretation is specific to the circumstances of the individual fishery.
Indicator 2 (management plans) uses Approach 2 by definition.
As illustrated in the box below, it is possible to use Approach 1 with Indicators 3 and 4. Some sub-indicators, such as number of drilling wells can be handled quite satisfactorily using Approach 2. Those involving fishing effort need to be broken down further by gear type to allow sensible summation of data. Even then, the problem remains that the same effort value may represent a quite different pressure in different ecosystems. We suggest that Approach 2 may prove to be a more fruitful approach in many cases.
Results of Combining Indicators 3 and 4
The extent and intensity of operations contributing to disturbance indicator 3 and indicator 4(a) in 1998 for all of the case study fisheries combined, is shown in Figure Ind. 3 and Figure Ind. 4. When combining datasets to report against indicator 3, the number of operations, rather than actual fishing effort, was used because of the variety of gears and gear units used by droplining, longlining and meshing methods. Note that only that part of the Australian Pelagic Longline Fishery analysed in the case studies is shown. The trawling effort shown in Figure Ind. 4 also represents numbers of operations rather than area trawled, as area trawled was only calculated for the Northern Prawn Trawl Fishery. An appropriate method of determining area trawled for the South East Trawl Fishery, given the logbook data currently collected, is still being developed. The methods used to generate these gridded products are basically the same as have been described for individual fisheries case studies in this report. A neighbourhood area of 30x30 minutes was used as this reflects the spatial precision of some of the data and ensures the information is interpretable on an A4 map of the continent. To help remove geocoding errors, cells having less the 5 operations in their neighbourhood area have been masked.
The use of operations as a surrogate for fishing effort provides a means of developing a national representation of fishing area and intensity for indicators 3 and 4(a), but fishing return from this effort is still not described. Mapping the spatial relationship between change in fishing effort indicators and change in fish stocks indicators may make it possible to infer the effect of fishing and trawling pressure on marine species over time. However, this derived "change-in-stocks-per-change-in-effort" indicator is likely to be confounded by the fact that effort information is often used to calculate stock estimates in the first place. If the effort information used to calculate stocks is independent of the commercial logbook effort information, then the analysis of these two indicators together may perhaps be used to infer the presence or absence of cause and effect. This means that not only will the system response (stock abundance) be monitored over time, but also the importance of the pressure (fishing and trawling effort) in generating the observed response can be inferred, with obvious implications for management of the system. Alternatively, fishing effort data from logbooks may on their own be mapped as a surrogate of marine disturbance if different levels of fishing and trawling effort can be shown experimentally to cause quantifiable impacts on marine populations or ecosystems, and these impacts can be generalised across different ecosystems and across time.
Validation and Checking
In the course of generating results for disturbance indicators 3 and 4 it became apparent that significantly more work is required in the area of data validation and error checking. A formal, repeatable and, as much as possible, automated process of validating data and resolving spatial and other errors as data enters a management authority system is urgently needed. Subsequent queries of the databases storing this information by authorised users can then be made in confidence that the data is free of errors with respect to the checking criteria known to have been applied. Users may then wish to do further checking and validation for their own purposes, but an explicit quality standard has at least been implemented. Furthermore, at the data collection end of proceedings, management bodies should be encouraged to improve the spatial precision of information being received from fishing vessels by requiring that records be georeferenced using a GPS, or at least referenced to a 6nm2 grid. As the spatial accuracy and precision of regional environmental data improves, and spatial analytical tools become more widely adopted, the needs for accurate and precise resource use information will also rapidly increase. This will also naturally intensify concerns over fisheries logbook data confidentiality.
The issue of confidentiality of fisheries logbook data is one that requires more discussion in political, legal and scientific contexts. Indicators describing the area and intensity of fishing and trawling immediately come into conflict with the issue of confidentiality of logbook data. The States and the Commonwealth have policies of not allowing the public reporting of logbook data, unless those data are aggregated by a certain number of vessels or fishers. However, in many fisheries, a small number of boats in an area does not necessarily mean a low level of fishing effort, either absolutely, or relative to other parts of the fishery.
Prawn trawling is a widespread activity over most of the northern half of Australia. The North West Slope Trawl Fishery (NWSTF) is a demersal trawl fishery using prawn gear that operates in deep waters off the north-western coast of Australia. While this fishery has been operating since 1986, for many years in the 1990's the number of vessels active in the fishery has been lower that that allowing public presentation of the logbook data. Because of this, it is impossible to show a true picture of the area of seabed subject to prawn trawling in northern Australia.
Demersal longline operations and dropline operations among the southern non-trawl fisheries and the Danish seine sector of the South East Trawl Fishery are also good examples of the potential magnitude of this problem. For droplining, the area in which fishing operations cannot be reported makes up more than 90% of the total area of the fishery and for demersal longlining the figure is over 95%. In both cases, regions of relatively high fishing effort are included in these non-reportable or masked areas. The area in which operations cannot be reported in the Danish seine fishery represents almost two thirds of the fishery.
One possible solution to this problem may be to allow the distribution of fishing effort to be shown, irrespective of the number of boats involved, provided catch information is not described. However, most fisheries operate in a competitive environment, and fishers often seek to maintain exclusive knowledge to "secret" locations of good catches. This perception of exclusivity is one factor for non-compliance in logbook programs. If the marine disturbance indicators examined in this report are to be used in the long term, it will be critically important for data confidentiality issues to be resolved between Environment Australia and the relevant fisheries agencies. As well as undermining spatial analyses of harvesting effort and resource use, censorship of information on the extent of fishing activities may result in improper representation of areas described as "pristine", or undisturbed by fishing activities.