Biogeography of nonindigenous species: From description to prediction.

摘要

Nonindigenous species are a major threat to the ecological integrity and biodiversity of marine and estuarine ecosystems. To become a successful invader, species must pass through four phases: (1) survive transport, (2) survive release, (3) establish a population, and (4) expand their range. To better understand these processes, an integrated framework was designed to capture life history characteristics, environmental preferences, dispersal mechanisms, and geographic distribution information for both native and nonindigenous marine and estuarine flora and fauna. Key aspects of this framework include: (1) consistent terminology; (2) translation of numerical habitat values and physiological requirements into classes; (3) development of classification schemas for natural history, environmental attributes, and geographic distributions; and (4) integration of biotic attributes to allow database queries on single or multiple species across spatial scales. Species data for the North Pacific were collected from the literature, local surveys, and regional databases.;Ballast water discharges have been identified as a major source of species introductions. To predict the potential rate of invasion from ballast water, a linear invasion model predicting per capita invasion probabilities (PCIP) of new invaders was developed based on historic invasion rates and ballast discharge volumes for estuaries on the west coast of the United States. While the probability of invasion is likely to vary with ballast discharge values, organism concentrations in the ballast, and invasibility of individual ports, the PCIP provides a quantitative methodology for establishing protective ballast water discharge standards based on organism concentrations, the approach being used to regulate ballast water discharges both nationally and internationally.;Habitat or niche models can be used to predict a nonindigenous species' potential distribution in invaded areas over several spatial scales. The utility of non-parametric multiplicative regression (NPMR) was evaluated for predicting habitat- and estuary-scale distributions of native and nonindigenous species. Results indicate that NPMR generally performs well at both spatial scales and that distributions of nonindigenous species are predicted as well as those of native species. Development of approaches for regulating ballast water and identifying areas at risk through predictive modeling are useful tools in the management of the nonindigenous species threat.