Commercial ships have spread many species around the world1, 2, 3, but little is known of the extent and potential significance of ship-mediated transfer of microorganisms3, 4. Here we show that the global movement of ballast water by ships creates a long-distance dispersal mechanism for human pathogens and may be important in the worldwide distribution of microorganisms, as well as for the epidemiology of waterborne diseases affecting plants and animals.
Ships have used ballast water for stability since the nineteenth century, discharging water at ports of call and en route1. Ports can receive relatively large volumes of ballast water — for example, the United States receives more than 79 million tonnes of ballast water from overseas each year5. Ballast tanks carry a diverse community of organisms, resulting in many biological invasions2, 3. Pathogens, including those affecting humans, are common in coastal waters6 and can also be transferred in ballast water7, 8.
We measured the concentrations of total bacteria, virus-like particles (VLPs) and the bacteria Vibrio cholerae O1 and O139, which cause human epidemic cholera, in the ballast water of vessels arriving to Chesapeake Bay (Fig. 1) from foreign ports. We collected water samples to estimate the abundance of total bacteria and VLPs, and also took both water and plankton samples to measure the concentration of V. cholerae, as the bacterium forms associations with plankton9.
Figure 1: Chesapeake Bay, on the US East Coast, receives some ten billion litres of foreign ballast water each year.
Each litre typically contains about a billion bacteria and seven billion virus-like particles.
High resolution image and legend (85K)Our samples contained an average of 8.3
108 bacteria
(s.e., 1.7108; n=11) and 7.410
9 VLPs (s.e., 2.3109; n=7) per litre.
Given that Chesapeake Bay received an estimated 1.21010
litres of foreign ballast water in 1991 alone5, our measures
indicate that ballast water probably delivers large numbers of microbial species
and potential pathogens to this estuary.
Vibrio cholerae was found in plankton samples from all ships, and
both serotypes were detected in 93% of the ships (Fig. 2a).
The concentration of V. cholerae O1 was significantly greater than
that of O139 (Fig. 2a; Student's t-test,
t=2.296; d.f., 27; P<0.05). Furthermore, there were 100 times
more V. cholerae O1 and O139 in water samples than in plankton samples
from the same ships (Fig. 2b; paired t-tests
for serotype O1 and O139 respectively, t=8.10; d.f., 6; P<0.001
and t=10.05; d.f., 6; P<0.001), indicating that V. cholerae
was not concentrated on zooplankton larger than 80 
m in diameter.
Figure 2: Prevalence and concentration of Vibrio cholerae serotype O1 and O139 in ships' ballast water.
a, Prevalence and concentration associated with plankton samples. Prevalence shows the percentage of ships in which the respective serotypes were detected. Sample size differed between serotype O1 (n = 15 ships) and O139 (n = 14 ships).b, Comparison of concentration of each serotype between paired plankton and water samples from the same ballast tanks (n = 7 ships). Concentrations are shown as mean and standard error for respective serotypes and sample types as detected by direct count methods using fluorescent antibodies; further details are available from the authors.
High resolution image and legend (52K)Vibrio cholerae is a useful model to examine the possible significance of ballast-mediated dispersal in transmission of pathogens. Our data indicate that V. cholerae can be delivered frequently by ships to estuaries with commercial ports, and our observations of dividing cells revealed that some bacteria are viable upon arrival. Although it remains difficult to estimate the concentration of viable cells10, the transfer and release of V. cholerae by ships creates an opportunity for the colonization of coastal ecosystems. V. cholerae is a common component of freshwater and marine habitats, including Chesapeake Bay, where it persists without human contact9, 10. Thus, should a novel genotype arrive in ballast water, local conditions may favour its establishment. The extent to which this has occurred, and the degree of geographic differences in the genetic structure of V. cholerae populations, is unknown.
We predict that coastal ecosystems are frequently invaded by microorganisms from ballast water. First, concentrations of bacteria and viruses exceed those reported for other taxonomic groups in ballast water by 6–8 orders of magnitude2, and the probability of successful invasion should increase with inoculation concentration11. Second, the biology of many microorganisms may facilitate invasion, combining a high capacity for increase, asexual reproduction, and the ability to form dormant resting stages3, 12. Such flexibility in life history can broaden the opportunity for successful colonization, allowing rapid population growth when suitable environmental conditions occur. Third, many microorganisms can tolerate a broad range of environmental conditions, such as in salinity or temperature, so many sites may be suitable for colonization6, 12. This suite of factors may yield a high rate of invasion for microorganisms compared to invertebrates, which are already known to invade coastal habitats from ballast water2, 3.
Despite growing concern about biological invasions7, 11 and emergent diseases9, 13, 14, the extent and effects of the transfer of microorganisms in ballast water are virtually unexplored3. We know of no published estimates of microbial genetic diversity in ballast water, and the fate of microorganisms discharged from ballast tanks remains unknown3. Given the magnitude of ongoing transfer and its potential consequences for ecological and disease processes, large-scale movement of microorganisms by ships merits attention from both invasion biologists and epidemiologists.
