Assessing appropriate limits to by-catch of small cetaceans in commercial fisheries in coastal British Columbia waters

Rob Williams
Raincoast Conservation Society
$19,385

Incidental catches of cetaceans in commercial fishing nets have been described as the “most serious threat to cetaceans today” (Perrin et al. 1994). Dolphins and porpoises in the transboundary waters of British Columbia and Washington State are not immune.

In the US, cetaceans are protected among other ways by the Marine Mammal Protection Act (MMPA), which triggers conservation and management action when by-catch exceeds a certain fraction of the population: an objective, conservative threshold called Potential Biological Removal (PBR; Wade 1998). This approach was designed to be conservative, so as abundance estimate uncertainty increases, the level of allowable take decreases (Wade 1998).

A different approach to the issue of by-catch is taken on the Canadian side of the border, where management objectives are generally less quantitative. Some work has been done to assess current levels of by-catch. Using a questionnaire survey and shipboard observation, Hall et al. (2002) calculated that up to 100 harbour porpoise, Dall’s porpoise and Pacific white-sided dolphins are caught in commercial salmon gear annually in British Columbia coastal waters. Of the three species, the harbour porpoise were most frequently killed. In order to effect conservation measures, the small cetacean fisheries related injury and mortality must be framed within the context of population abundance. Though incidental mortality is known to occur in BC fisheries, reduction measures have not been initiated due to the absence of an analytical evaluation of the population level significance.

In the absence of quantitative management objectives or risk-averse decision-making frameworks, the Scientific Committee of the International Whaling Commission has issued a useful rule of thumb. They recommend that attention be paid to any anthropogenic source of mortality that annually exceeds 1% of the population. This is not an automatic call for action, but rather a generic “red flag” indicating that closer attention should be paid. In order to determine whether by-catch warrants management intervention in BC, the level of by-catch must be evaluated in terms of abundance. However, no province-wide estimates of small cetacean abundance have been available until recently.

One of us (RW, with Raincoast Conservation Society) recently designed and conducted a systematic survey of British Columbia’s Inside Passage to estimate abundance and distribution for six cetacean species in summers 2004-6. The results of the Raincoast surveys will allow us to estimate the PBR for harbour porpoise, Dall’s porpoise and Pacific white-sided dolphins in British Columbia.1 If the resulting abundance estimates from the Raincoast surveys were perfect, then calculating PBR would be quite straightforward. An easier calculation than PBR would be to assess whether the highest estimate of by-catch (100 animals/year, Hall et al. 2002) is greater than 1% of the target population (i.e., 100x100=10,000).

Unfortunately, animal abundance is rarely known with certainty, and abundance estimates must cope with associated bias and uncertainty. Preliminary estimates of abundance for small cetaceans in BC (Williams & Thomas, in review) assume: representative coverage of the study area; no animals on the trackline missed; and animals were detected before responding to the ship. We are reasonably sure that the Raincoast survey design was unbiased, in that it performed well in tens of thousands of simulations (Thomas et al., in review). The assumptions of certain trackline detection (the so-called g(0) problem) and no responsive movement, however, were no doubt violated, and varied among species. Fortunately, methods exist for accounting for these sources of bias and uncertainty. We propose to incorporate a range of plausible values for these correction factors in a simulation framework, and model their effect on PBR.

If our best estimate of by-catch falls either well above or well below PBR, then either scenario has important implications for conservation and management and is worth reporting. If, however, we are near PBR, it would be interesting to know whether this reflects true threat, or rather large uncertainty associated with abundance and by-catch estimates. We believe that this problem lends itself to simulation, wherein we simulate how PBR would be affected by incorporating main sources of uncertainty, and simulating how likely one would be to meet a range of management objectives, given that uncertainty.

We propose to: (1) improve existing estimates of abundance for harbour and Dall’s porpoise and Pacific white-sided dolphins, incorporating correction factors for detection function uncertainty, responsive movement and uncertain trackline detection; (2) calculate PBR for harbour and Dall’s porpoise and Pacific white-sided dolphins, using best estimates of abundance; (3) assess whether BC’s gillnet fishery is likely to be exceeding PBR; and (4) simulate how best to address uncertainty (either by improving accuracy and precision of abundance estimates, or accuracy and precision of by-catch estimates).