Accelerating Invasion Rate in a Highly Invaded Estuary

See allHide authors and affiliations

Science  23 Jan 1998:
Vol. 279, Issue 5350, pp. 555-558
DOI: 10.1126/science.279.5350.555


Biological invasions are a major global environmental and economic problem. Analysis of the San Francisco Bay and Delta ecosystem revealed a large number of exotic species that dominate many habitats in terms of number of species, number of individuals and biomass, and a high and accelerating rate of invasion. These factors suggest that this may be the most invaded estuary in the world. Possible causes include a large number and variety of transport vectors, a depauperate native biota, and extensive natural and anthropogenic disturbance.

Over the past few centuries, thousands of species of freshwater, estuarine, and marine organisms have dispersed outward from their native regions through human-mediated transport and have established sustaining populations in distant parts of the globe (1, 2). Many of these organisms have profoundly affected the abundance and diversity of native biota in the regions they have invaded (3, 4), and in some cases they have had substantial economic impacts (5). Despite these many invasions, data sets suitable for analysis of spatial or temporal patterns of aquatic invasions are rare. Here, we analyzed a synthesis of such data for one of the largest estuarine and freshwater ecosystems in North America, the San Francisco Bay and Delta. This ecosystem comprises 1500 km2 of aquatic habitat within the range of normal tidal influence. It receives runoff from a 163,000-km2 watershed with an extensive system of dams, water diversions, and flood channels that partially control river flows, which nevertheless exhibit considerable seasonal and annual variation. Water salinities also vary widely, generally ranging from fresh in the Delta to coastal salinities near the mouth of the Bay, and sometimes hypersaline conditions during droughts in southern parts of the Bay (6).

We assembled data on introduced aquatic organisms from published and unpublished data, augmented by our field work. Exotic species were defined as those not present in the eastern North Pacific bioregion before the entry of Europeans in the 16th century, or present in distant parts of that region and later introduced to the Bay/Delta ecosystem by human-mediated mechanisms (7). Cryptogenic species are species that are neither clearly native nor exotic (8). In earlier reports we described the evidence used to distinguish between native and exotic species and to determine establishment (9).

We identified a total of 234 exotic species established in the ecosystem, including plants, protists, invertebrates, and vertebrates (Table 1). At least 125 additional species are cryptogenic. We sorted species according to their native regions, counting purely freshwater organisms as continental organisms and the rest as marine. Most continental organisms derived from eastern North America and Europe, most marine organisms from the North Atlantic and the western North Pacific (Table 2). A review of sampling data and species lists revealed that exotic species dominate many of the ecosystem's biotic associations, including infaunal and epifaunal soft-bottom benthos (organisms living within or on the bottom sediments), fouling communities, brackish-water zooplankton, and freshwater fish. In these communities, exotic organisms typically account for 40 to 100% of the common species, up to 97% of the total number of organisms, and up to 99% of the biomass (10).

Table 1

Exotic species established in the San Francisco Bay and Delta. Organisms that reproduce in both fresh and salt or brackish waters, or that move between them as a regular part of their life cycle (anadromous and catadromous species), were counted in both environments. Other organisms were counted in the environment in which they reproduce.

View this table:
Table 2

Native regions of exotic species that have become established in the San Francisco Bay and Delta. Where native regions are imprecisely known (for example, the organism could be native to either the eastern or western North Atlantic), the count was split, so figures for some regions include “half” species.

View this table:

In our analysis of changes in the rate of invasion, we used raw data as well as a data set modified by the exclusion of each record for which we could not determine the year of planting, observation, or collection, or that we judged to result from special expertise, an extraordinary collection effort, or the chance discovery of a highly localized species. The modified data set excluded about one-third of the raw data. The shapes of the cumulative invasions curves (Fig.1) are similar for the raw and modified data. The raw data show that 55.2% of the total number of invasions were recorded after 1960, versus 49.7% for the modified data set. Thus, about half of all invasions in the 145-year record were reported in the last 35 years. On the basis of the raw data, the rate of invasions has increased from an average of one new species established every 55 weeks from 1851 to 1960, to an average of one new species every 14 weeks from 1961 to 1995.

Figure 1

Cumulative number of exotic species established in the San Francisco Estuary: (A) raw data; (B) modified data. A time series analysis conducted by C. Parmesan [Box-Jenkins methodology (19) on RATS (Regression Analysis of Time Series) software, version 4.0 (Estima); final models used only terms that contributed significantly atP < 0.05 and had uncorrelated residuals by Durban-Watson and Lung-Box Q statistics] showed a significant increase in the number of invasions over time (linear regression on raw and modified data separately, trend terms significant at P < 0.001). Trend = (time)2 models explained 5% and 2% more of the variation than did trend = time models for raw and modified data sets, respectively (final models:R 2 raw = 0.34,R 2 modified = 0.21). Additional lag terms on time were unnecessary (and when forced into the models were not significant at P > 0.50 for both data sets), indicating that invasion events are independent between years. A better fit of trend = (time)2 models indicates that the rate of invasions, as well as the absolute number, increases with time.

Taken together, the large number of exotic species, their dominance in many habitats, and the rapid and accelerating rate of invasion suggest that the San Francisco Bay and Delta may be the most invaded estuary and possibly the most invaded aquatic ecosystem in the world. Although some other aquatic systems are known to host many exotic species (3, 11), few report as many and none report the extensive dominance by exotic organisms that is shown here (12). Several factors could be contributing to the extent of invasion in this system:

1) Transport vectors. The Bay/Delta system may have been inoculated by a greater number or diversity of potential invaders than have other estuaries. Many mechanisms, connecting at different times to different regions of the world and favoring different components of the biota, have transported organisms into this ecosystem, including transport with ships; transport with oysters imported for cultivation; fish stocking; releases or escapes from commercial and government breeding and rearing facilities, ornamental ponds, and aquariums; introductions for biological control; plantings of exotic vegetation for marsh “restoration” and erosion control; and importation with shipments of live seafood or bait. The scale and diversity of such mechanisms may be increasing with the expansion in international commerce.

2) Depauperate biota. It has been argued that the relative youth of northeast Pacific estuaries, or their island-like isolation, prevented the development of a species-rich native biota that could resist invasions (13). However, theories that island, young, or species-poor communities are more vulnerable to invasion than are other communities (1, 14) have been tested only occasionally and have been challenged (15). Nor can such theories explain the observed increase in invasion rate as more species became established.

3) Disturbance. It has been argued both that extreme natural disturbance events facilitated the establishment of some exotic organisms in the Bay/Delta system (16) and that human alterations of habitat have made the system vulnerable to invasions (17), consistent with the view that disturbed environments are more easily invaded (1, 18). Although greater numbers or greater dominance of exotic species in disturbed areas relative to undisturbed areas has often been observed, it is unclear whether this occurs because these areas are more easily invaded, because they are more heavily inoculated with exotic organisms, or because a greater number of the inoculated organisms are adapted to the disturbed environments from which they were imported. It is also unclear whether the Bay/Delta system is any more disturbed than other estuaries.

The relative contribution of these factors to the extent of invasion might be illuminated by comparative studies with other estuaries, but these must await the development of comprehensive regional data sets on invasions similar to the one discussed here. Proceeding with such work would be of great value. Our study shows the San Francisco Bay and Delta to be extensively invaded, seemingly far more so than other large aquatic ecosystems. Given the potential impact of such invasions on both native biological diversity and human economic activities, it is a matter of some urgency to learn why.

  • * To whom correspondence should be addressed. E-mail: acohen{at}


View Abstract

Navigate This Article