Drake and colleagues are concerned that our recent article (Holeck et al. 2004) might mislead policy development because of its “inference” that ship ballast water exchange (BWE) has been ineffective for the Great Lakes. Specifically, we have argued that BWE as currently practiced is insufficient to prevent ship-vectored invasions.
The best indicator of the value of any preventative management strategy is the invasion rate, which cannot be measured directly but must be estimated from the rate at which new invaders are discovered. Hence, we draw attention to data showing an increasing rate of discovery of invaders that continues several years after the implementation of BWE. Drake and colleagues misinterpret our purpose in presenting these data. Nowhere in our article do we suggest that BWE has increased the invasion rate (nor can we conceive of a process by which this should occur). Quite simply, the discovery rate shows no evidence that ship-vectored invasions have abated following BWE regulation.
We share Drake and colleagues' concern that the discovery rate is subject to biases. Although we addressed this issue (Holeck et al. 2004, pp. 923–924), we are criticized for not acknowledging a theoretical scenario in which the rate of discovery in a system increases (because of a time lag between introduction and population growth to detectable abundance) while the rate of introduction is constant or zero. However, some recent introductions cast doubt on the hypothesis that all ship-vectored invasions discovered after the implementation of BWE are attributable to extensive time lags. The predatory Eurasian waterflea, Cercopagis pengoi, was found in Lake Ontario in 1998, several years after implementation of BWE (MacIsaac et al. 1999). Genetic analysis indicates that it arrived from the Baltic Sea, which was reported invaded by Cercopagis in 1992 (Cristescu et al. 2001); therefore, it very likely invaded the Great Lakes while BWE was in effect. And even a decade after BWE was implemented, there have been multiple discoveries of Chinese mitten crab, Eriocheir sinensis, and European flounder, Platichthys flesus, which are both brackish-water species that cannot establish reproducing populations in fresh water and whose occurrences are best explained as recent ship-vectored introductions.
Drake and colleagues conclude that the discovery rate has “probably accelerated,” but that this is unrelated to BWE. We agree. We suggest that it could be due to the predominance of incoming NOBOB ships—i.e., ships reporting no ballast on board, and thus not subject to any existing regulation. As we explained in our article, these ships contain residual water and sediments that get mixed with Great Lakes water as cargo is unloaded on the inbound journey. This mixed water is subsequently discharged into the Great Lakes, both when ships pass through shallow connecting channels and when cargo is loaded for the outbound journey. Recent studies demonstrate that NOBOB ships carry diverse assemblages of nonindigenous invertebrates (as free-swimming adults and resting eggs) mainly in residual water but also in residual sediments (Bailey et al. 2003, Gray et al. 2005). Therefore, even if BWE were 100 percent effective, it would not sufficiently protect the Great Lakes from ship-vectored invasions by NOBOB vessels or by species vectored on fouled hulls of ballasted or NOBOB vessels.
For these reasons, we believe that Great Lakes invasions are best prevented by identifying and addressing shortcomings in the current program—in particular, by developing technology and policy to address the risk posed by the ship vector as a whole rather than by continuing to assume that BWE alone is sufficient.