Evaluating the effectiveness of ballast water exchange policy in the Great Lakes.

Ballast water discharge from ships has been the most important vector for introducing nonindigenous species to the Laurentian Great Lakes, and is responsible for many of the most ecologically and economically disruptive invasions in the basin (Mills et al. 1993, Ricciardi 2006). In an attempt to limit further invasions, ships entering the Great Lakes–St. Lawrence River system with ballast water on board are obliged to perform a mid-ocean ballast water exchange (BWE). BWE was implemented as a voluntary procedure by Canada in 1989 and the United States in 1990, and then became mandatory by U.S. regulation in 1993. Given that scores of species have invaded—and continue to invade—European ports from which the Great Lakes receives the bulk of its transoceanic ship traffic, the basin is vulnerable to further, potentially costly, invasions via this vector in the absence of effective control (Ricciardi and Rasmussen 1998). Thus, C. Costello and colleagues (Costello et al. 2007) address the important question of whether BWE is an effective policy to halt ship-vectored introductions to the Great Lakes. We presume that they are referring to ballast water mediated invasions in general, and not merely those that are vectored by ships reporting pumpable ballast on board. Thus, we draw a distinction between the effectiveness of BWE as a procedure applied to a subset of ships and the effectiveness of BWE as the only existing policy (until recently) to prevent all ballast water mediated invasions. Our comments are based on the latter issue. Costello et al. (2007) argue that there is insufficient evidence to reject the possibility that the BWE policy is 100% effective and that all discoveries of invaders after 1993 could plausibly be explained by time lags between the introduction of species prior to the implementation of BWE and their detection in subsequent years. The authors assert that several more years of data are required to make a conclusive evaluation. We applaud the attempt of Costello et al. (2007) to evaluate the effectiveness of BWE policy. We agree with their assessment that time lags can plausibly account for some of the species discovered in the lakes since the policy took effect; however, it is unlikely to account for all of the recent finds. Although the BWE procedure can dramatically reduce the abundance and diversity of freshwater zooplankton in ballast tanks (Gray et al. 2007), live freshwater-tolerant zooplankton and other organisms have been found in ships that reportedly exchanged ballast and these include species not previously reported from the Great Lakes (Locke et al. 1993). Several other empirical studies have demonstrated that the BWE procedure is not 100% effective (e.g., Weathers and Reeves 1996, Harvey et al. 1999, Zhang and Dickman 1999, Wonham et al. 2001). The occurrence of freshwater species in ballast water following mid-ocean BWE signifies that the risk of invasion may have been sharply reduced but not eliminated entirely. It is currently impossible to identify an acceptable level of risk based on biological criteria, because the relationship between propagule pressure and invasion success has not been ascertained with respect to ballast water discharges. Time lags can hinder the determination of when introduced species first became established in an ecosystem. This effect is likely to be most pronounced for species that are microscopic and/or occupying areas not typically studied by invasion biologists. For example, Nicholls and MacIsaac (2004) reported three Eurasian testate rhizopod species in sand collected from beaches around the Great Lakes, which they attributed to ships’ ballast introductions. As these species are both microscopic and occupy a habitat not previously studied by invasion biologists in the region, the possibility of an extensive time lag—one that far exceeds the date of implementation of mandatory BWE on the Great Lakes—cannot be dismissed. However, some invaders are macroscopic, conspicuous, and unlikely to be confused with other taxa. One such species was discovered in Lake Ontario in 1998: the Ponto-Caspian waterflea Cercopagis pengoi, which can reproduce asexually (MacIsaac et al. 1999). Assuming (1) an introduction early in the year of a moderate inoculum, (2) minimal mortality from predators, and (3) reproductive output similar to another invasive waterflea, the confamilial Bythotrephes longimanus, Cercopagis could have achieved its observed high population abundance in the year of its discovery (H. MacIsaac, unpublished data). Even when present at low abundance, Cercopagis conspicuously fouls fishing line and is considered a nuisance to anglers. Another crustacean, the bloody-red mysid Hemimysis anomala, was discovered in both Lake Michigan and Lake Ontario in 2006