Natural Pathways
Atmospheric, oceanic, and river currents have always formed pathways or routes for plant dispersal. For instance, the Gulf Stream has carried seeds and plant propagules not only through the Caribbean but also as far as the British Isles. One consequence of this long-term dispersal has been the arrival of many subtropical species in the Scilly Isles, southwest of the main islands of Britain (Lousley 1971). Similar long-distance dispersal has facilitated the spread of mangrove species through Oceania and elsewhere (Murray 1986). Such movement can be highly directional: off the western coast of North America, the California Current carries plant flotsam in a distinctly southerly direction each summer. Alternatively, species are not as likely to reach a new range along these natural pathways, if the range lies outside the path of prevailing currents.
Ocean and air currents, of course, continue to affect plant distribution; their influence on the dynamic composition of any flora is a direct function of the frequency with which they carry plants to new ranges (Ridley 1930). Not surprisingly, strand and shoreline species commonly immigrate in this manner (Smith 1999). The sprawling morning glory, Ipomoea pescaprae, is a cosmopolitan subtropical/tropical shoreline species (Ridley 1930); its huge geographic distribution is a direct consequence of its floating seed and tolerance for sandy, salt-spray environments. Even among species not confined to shorelines, natural forces continue to transport species across impressive distances. Natural forces have apparently dispersed plants from Australia to New Zealand in modern times (Mack and Lonsdale 2001); these immigrations have produced adventives and possibly some naturalizations but no invasions. The role that the natural forces of moving air and water play today in changing plant distributions is small, compared with the role of humans. These natural forces are not feeble, just infrequent, in their global impact.
Pathways Developed by Humans
The human dispersal of plants has long followed many natural pathways, especially when human transport was substantially dependent on wind and water currents. The Gulf Stream along with the weaker Canary Current off North Africa and the westerly South Equatorial Current formed a great triangular route for sailing ships between Europe, West Africa, and the Caribbean that was in full swing by the early eighteenth century (Viola and Margolis 1991). Thus, ships augmented flotsam and other living rafts as conveyors of plants from Africa to South America and the Caribbean. The list of locations to which plants could be carried by ships was greatly expanded with the well-known advances in ship construction and navigation, beginning in the late fifteenth century and the later advent of steam-powered ships. It is hard to overestimate the increased likelihood of transoceanic plant dispersal as a result of these innovations. In effect, any two anchorages now potentially share a connecting pathway. As a result, species have been introduced to new ranges that they would not have reached by ocean currents alone; for example, species native to temperate Britain reached temperate New Zealand and vice versa (Good 1964).
With the enormous versatility of steamship travel, webs of transoceanic pathways soon became well established. These routes were shaped by the desire to speed commerce between trading partners, which were often European (and later American) nations and their overseas colonies. Even before the advent of steamships, a path for plant dispersal had developed between western Europe and the North American eastern seaboard and soon thereafter between the Netherlands and Britain and their colonies at the Cape of Good Hope, in the South China Sea, and Australia (Mack 1999 and references therein). These routes initially relied on ocean and wind currents; they multiplied as coaling stations were established. For instance, by 1889 Britain alone had established with its colonies and other trading partners 156 coaling stations that spanned the world (Porter 1991) (Fig. 1.1). France, Germany, Portugal, and Spain also maintained ports and connecting routes with their overseas colonies (Emmer and Gaastra 1996).
With steamships, landfalls that were exceptionally remote and unlikely to receive plant immigrants by natural forces (e.g., Ascension and St. Helena Islands in the southern Atlantic) became the recipients of many deliberately and accidentally introduced plants (Cronk 1989). Furthermore, along these webs of routes thousands of plant species were efficiently moved within colonial empires. By the late nineteenth century Britain had established botanical gardens at key locations worldwide (e.g., Calcutta, Cape Town, Hobart, Port of Spain, Singapore) (McCracken 1997). These facilities became bases for plant collection from which newly discovered species could be shuttled within the British Empire via Londonās Kew Gardensāan early hub-and-spoke transportation network. Never before had plant immigration operated on such a massive scale, replete with test gardens in potential new ranges (McCracken 1997). As discussed below, scores of species that were transported along pathways developed in the nineteenth century among Europe, North America, and Europeās colonies later became naturalized; some have become invasive (e.g., the woody plants, Rhododendron ponticum, Acacia ieucocephala, Leucaena nilotica, and Lantana camara). The tradition of massive worldwide plant exchanges can be traced to the development of these colonial programs, beginning as early as 1600 (Kloot 1987, Mack 1991, 1999, McCracken 1997).
The spread of plants was extended inland with the growth of canals (Mills et al. 2000) and, later, railroads (Dewey 1896, Mack 1991). Canals were seen as essential routes for eighteenth-century commerce in the United States, a view that continued until they were largely supplanted by railroads at the end of the nineteenth century. In the meantime, navigable canal networks laced the interior of the eastern United States and provided a mode for both deliberate and inadvertent spread of nonindigenous plants. Reconstructing the nineteenth-century spread for accidental introductions along canals is necessarily circumstantial and relies on the collection history of these species alongside or near old canal routes. Invasion by the aquatic herb Lythrum salicaria (purple loosestrife) has been documented in this manner. Thompson et al. (1987) contend that with the exception of the canal system in interior Pennsylvania, the early-nineteenth-century inland spread of L. salicaria in the northeastern United States was tied closely to canal traffic moving from navigable East Coast estuaries. And most sites of purple loosestrifeās pre-1880 establishment in the region were along the Erie Canal, which bisected New York and the Delaware & Raritan Canal, which bisected New Jersey (Thompson et al. 1987). Canals in the United States still provide avenues along which nonindigenous plants spread. Dispersal of highly invasive species (e.g., Elodea canadensis [waterweed], Hydrilla verticillata...