|
Harvest of reef finfish from the Great Barrier Reef (GBR) is an important
commercial and recreational activity. Minimum sizes for retention are
an important instrument in managing the harvest of many species. Fishers’ compliance with minimum legal sizes, and enforcement of these by officials,
however, assumes the reliable and consistent measurement of the lengths
of landed fish.
Recently, breaches by commercial fishers retaining allegedly undersized
live coral trout have been challenged based on assertions that fish kept
on board vessels effectively ‘shrink’ after capture and that
measures of live fish are inherently variable. Prosecutions have failed
due to arguments that either of these factors may occur, but there is
currently little understanding of their magnitudes. Failed legal proceedings
are cause for concern for fishers, who object to unproved accusations
of illegal behaviour, regulators, who seek certainty in the prosecution
of regulations, and society, which ultimately bears the cost of enforcement.
In this project we sought to assess the veracity of claims that live
common coral trout Plectropomus leopardus change size following capture
and to estimate the magnitude of variations associated with repeated measurements
of coral trout, such as would occur when fishers measured fish at the
time of capture and enforcement officers did so subsequently during compliance
inspections.
We found that fish kept on board a commercial fishing vessel during normal
fishing operations varied in size both positively and negatively over
a 10 day trip. There was an overall average “shrinkage” of
1.76 ± 0.76 mm (95% CI). Fish showed a decrease in average size
from soon after they were brought aboard the vessel, with the greatest
reduction in size occurring in the first 1-2 days. Fish subsequently continued
to “shrink” by smaller amounts, reaching their smallest size
on the third last day of the trip when they were on average 3.2 ±
1.02 mm (95% CI) below their initially measured size. The greatest daily
change in mean size was a positive change, or elongation, of 1.43 ±
0.31 mm (±SE) during the final day of the trip. It was notable
that if the large changes in size over the first 1-2 days were ignored,
the significant relationship between length of the fish and time at sea
disappeared.
Large variations in length measurements were also observed between repeated
measurements by different observers (three commercial fishers and two
research officers) over short periods (<22.5 hours). These measurement
errors, or uncertainties, were generally greater than the effect of fish
shrinkage over the 10 days of the trip. Average 95% and 99% confidence
intervals on measurements for common coral trout by different observers
were 2.91mm and 4.83mm respectively. Average confidence intervals attributable
to variation amongst repeated measurements by the same observer were 1.72mm
(95% CI) and 2.35mm (99% CI).
Whilst there was a real overall uncertainty due to measurement error,
much of this could be eliminated by careful attention to the techniques
used to measure the fish. For example, the two researchers, who were using
a pre-determined and consistent technique for length estimation, had similar
measurement biases and considerably greater precision of measurements
than the fishers. The 95% confidence intervals on measurements attributable
to between- and within-observer variations for observers using similar
measurement techniques were reduced to 2.08mm and 2.16mm respectively.
The fundamental inability to determine an exact ‘true length’
of a fish means that enforcement of minimum legal sizes cannot be ‘knife-edged’
and definite. With knowledge of the magnitude of this uncertainty, however,
it is possible to determine the numbers of a given sample of fish that
would be reasonably expected to fall below the legal size on repeated
measurement simply because of measurement uncertainty. It is also possible
then to stipulate how many fish would have to appear to be undersize before
a reasonable conclusion would be made that the observations were not attributable
to measurement uncertainty and that at least some of those fish were indeed
being kept illegally. We present exemplar tables of such ‘unlikely
numbers’ to illustrate how knowledge of measurement uncertainty
might be used to construct rules for prosecution that include allowances
for ‘reasonable’ margins of error. These tables allow fishers
and enforcement officers to determine how many fish, from a given sample,
they would expect to appear undersized by given amounts due only to measurement
variation and see the points at which records of undersized fish become
sufficiently frequent to be considered improbable and so warrant prosecution.
Very many such tables can be generated, however, each associated with
a different level of risk that a prosecution might proceed in error. The
choice of what is an acceptable margin for such error ultimately resides
with legislators and / or the judiciary. Once such a criterion is set,
however, application of a standardised measurement methodology by all
fishers and enforcement officers is likely to make enforcement of minimum
legal sizes considerably more precise and robust than is currently feasible.
Conclusions and Recommendations
-
Uncertainty in measurement of live reef fish has a clear empirical
basis and we have quantified both between-observer uncertainty and within-observer
uncertainty and their effects of compliance monitoring.
-
Evidence was found for short-term (days-weeks) changes in sizes of
live coral trout held aboard a commercial fishing vessel, but changes
were generally small (<5mm).
-
We provide tables of the expected frequencies of fish measured to
be under the legal size because of measurement uncertainty in a format
designed to provide decision-making criteria for when prosecution is
justified on the balance of probabilities.
-
Enforcement, administrative, legislative and industry personnel should
be advised of the potential magnitudes both of the uncertainty in measurements
and the frequency with which given discrepancies in measurement might
be expected during compliance and enforcement monitoring.
-
Further work should be done to verify whether rudimentary instruction
in measurement procedures is effective in reducing uncertainties in
the measurement of fish and, more importantly, reducing the amount of
variation between measurements by different people (e.g. at capture
and at compliance inspection).
-
A standard measuring technique should be defined and sent to all fishers
and enforcement officers. In the interest of reducing the potential
variation between observers’ measures the authors suggest the
following: the fish’s mouth should be closed or gently and carefully
forced closed, there should be no lateral compression of the body or
caudal peduncle, and the length should be measured from the tip of the
lower jaw to the posterior margin of the dorsal lobe of the tail. If
the dorsal lobe is damaged, a measurement to the end of the ventral
lobe should be used.
Download a complete copy of the report (Adobe Acrobat file 366KB)
For a hard copy of the report contact CRC Reef on info@crcreef.com
|
