Humans as Global Plant Dispersers: Getting More Than We Bargained For

Humans have surpassed natural forces as the principal global disperser of vascular plants. Some of the means of dispersal are accidental: Seeds and other plant disseminules and vegetative propagules are transported inadvertently in clothing; cling to or are ingested by our domesticated animals; and are found within and attached to all manner of commerce, particularly as contaminants in seed lots (Muenscher 1955). Plants are also deliberately transported. Almost all human societies have long been dependent on the deliberate transport of plants as a means to satisfy basic human needs (Mack 1999). Many ancient accounts of plant transport are probably apocryphal, such as the importation of incense trees by Queen Hatshepsut to Egypt in 1500 BC from the Land of Punt (Hodge and Erlanson 1956), but there is nevertheless a verifiable fossil record documenting the cultivation of plants far from their native ranges for thousands of years (Godwin 1975). Our actions as global plant dispersers can be beneficial, neutral, or detrimental. Establishing plants beyond their native ranges has been not only beneficial but also essential to agriculture (Hodge and Erlanson 1956). Few agricultural economies today operate exclusively with native crops, and none of these support an industrialized society. Many plants that humans transport long distances die en route or soon thereafter, unless carefully protected; the consequence of these plants' dispersal is nil. But some of these immigrants prosper in the new range, even without cultivation, and a few of these wreak much environmental and even economic damage (Vitousek et al. 1996). These species are variously termed, depending on whether they form permanent but nonspreading populations (naturalized species) or prolific, permanent populations that usually spread over large new ranges (invaders) (Mack 1997). The admittedly anthropocentric but familiar term, weed, is often applied to species in both these ecologic categories as well as to species that are destructive in their own native range. Probably no definition of weeds is universally agreed upon, but Baker's (1974) definition has at least wide acceptance: species that not only have no detected human value but actually interfere with human activities. We deal here exclusively with those introduced species that meet this definition in a new range. These species cause economic losses (e.g., weeds in pastures and crops), adverse effects on human health (such as allergies and toxic reactions), and ecologic losses (reduction of biodiversity and ecosystem services, such as the supply of water ) (Vitousek et al. 1996). This damage has an enormous aggregate cost. For example, each year introduced weeds cost the Australian economy perhaps A$3 billion (ANWS 1997) and the US economy more than $26 billion (Pimentel et al. 2000). In many, perhaps most, regions of the world, the emerging assessment on the origin of weeds comes to two conclusions. First, most weeds in a region are not native but were introduced by humans. Second, the largest single group of these unwelcome intruders was originally introduced deliberately (Panetta 1993). For example, 60% of the more than 600 naturalized taxa listed in Fernald (1950) for the northeastern United States were deliberately introduced; the true percentage is undoubtedly higher because the original use under which many naturalized species were introduced is obscure (Sturtevant 1919). Both conclusions hold major implications for the modes by which a nation's current weeds might be controlled as well as warnings on the origin of future weed introductions. In this article we explore some of the apparent motivations that lead to the introduction of vascular plants, both transoceanically and throughout a potential new range. We also explore the biological basis for factors that foster the spread of some introduced species in a new range. We also make some predictions about the characteristics of future weeds and the steps that could prevent such new weeds from arising, and we suggest areas for research in the epidemiology of potentially weedy species. We begin by exploring briefly the role that natural forces play in moving potentially weedy species around the globe. Whether natural forces of plant dispersal, such as wind and ocean currents, introduce plants that become weeds in new ranges has been little studied. The much broader issue of dispersal of all plants by these agents, however, has long fascinated biologists, as illustrated by the seminal work of Ridley (1930). More recently, Smith (1999) has followed the arrival of seeds that have drifted onto Australian beaches, identifying them to species or genus and testing them for viability. He notes that dispersal patterns are generally consistent with prevailing winds and currents (e.g., there is a westward drift of propagules toward Australia at all times of year; Smith 1999). He found few drift species that were not native to Australia (e.g., Excoecaria indica, Nypa fruticans, Lithocarpus spp., Inocarpus fagifer). However, given the small numbers of seeds and species involved, and the disparity between the number of species that arrive on a beach and the number of species that form viable populations, plant movement by sea currents appears insignificant compared with human-aided movement in fostering the spread of species. Plants can also be transported aerially great distances between land masses. New Zealand has served as the recipient for seeds of Australian native species that apparently arrived by being blown across the Tasman Sea, a minimum distance of approximately 1800 km. Species presumed to have arrived in modern times in this manner include Gratiola pedunculata, Macothelypteris torresiana, Mazus pumilia, and Wilsonia backhousei (Ewen Cameron [Auckland Museum], personal communication, 1999). As with the seaborne immigrants, none has proven weedy. Furthermore, the list of species known to spread routinely by natural agents across great ocean distances includes no species that have become detrimental (i.e., weedy) anywhere. For example, Ipomoea pes-caprae, a broad-leafed leguminous vine, has a floating seed that is eminently suited for long sea voyages. Not surprisingly, I. pes-caprae is a common sight on beaches throughout the tropics (Ridley 1930). Yet we are unaware of Ipomoea having become weedy or detrimental anywhere. Natural forces thus represent a vanishingly small threat as vectors for the transoceanic movement of weedy species. However, they can facilitate the subsequent spread of nonindigenous species once these are established in a new range (Ridley 1930). The global movement of weed species by humans has three phases: accidental, utilitarian, and aesthetic. In terms of the number of examples and their subsequent effects, these phases of weed movement equate largely to European colonial and postcolonial history. Very broadly, the three phases form a chronological sequence beginning around AD 1500, but there is much overlap between them because they began at different times with the colonization of different continents. As Europeans immigrated to the New World, they brought their agricultural weeds and ru derals with them. These species are short-lived, mostly annuals, and capable of colonizing bare ground and capturing the water, nutrients, and light before the crop emerges. They arrived remarkably early in the settlement of new colonies. By 1672, John Josselyn could tally 22 European weed species that had become common around Massachusetts Bay—only about 50 years after the first European settlers arrived in the region. These included Taraxacum officinale, Stellaria spp., Urtica dioica, and a mullein (probably Verbascum thapsus). Plantago major was so ubiquitous wherever Europeans trod that Native Americans called it “Englishman's foot” (Cronon 1983). Although some of these plant hitchhikers to New England were used as fodder, most had few or no redeeming features and were very likely accidentally, rather than intentionally, introduced to the new agricultural lands of the Americas and later to Australia (Kloot 1985) and elsewhere. Thus, introduction of nonindigenous weeds was accidental in an early (if not the distinct first) phase, in the manner that the general public still believes is the predominant mode today. Although accidental transoceanic introductions of weeds still occur, they are uncommon because seed cleaning techniques and quarantine inspection services ensure that seed lots are only minimally contaminated with extraneous seeds. Among those now naturalized species that arrived in Australia between 1971 and 1995, only 2% are known to have been introduced as contaminants. Admittedly, another 20% arrived through an undetected mode of introduction that could certainly include contaminants (Groves 1998). But even if contaminants have contributed as much as one-quarter of all recently naturalized species, such accidental introductions are no longer the major factor in plant entry. As a result, there are few modern examples of accidental first introductions of weedy species to a new range (among introductions in Australia, Chondrilla juncea arrived in vinestocks and Parthenium hysterophorus probably arrived attached to aircraft [Parsons and Cuthbertson 1992]). As colonists began to introduce more and more useful species in an attempt to provide reliable sources of food and other plant commodities, new weeds began to emerge from among the deliberately introduced species themselves. These often included longer-lived species that were capable of escaping from cultivation and invading natural and seminatural ecosystems, such as grazing lands (Harlan 1975). This utilitarian phase of weed introduction is still occurring today. Interestingly, it began quite early in the story of European colonization: José de Acosta, touring South America at the end of the 16th century, noted thick forests of introduced orange trees that had sprung from seed spread by water from citrus groves (Crosby 1986). Such deliberately introduced species arrived with the first wave of colonists, and their importation continued to grow in taxonomic breadth and frequency (Kloot 1985, Mack 1999). The colonists' zeal for importing plants was partly a consequence of the strikingly uneven geographic distribution of plant taxa that occurs even at the family level. Good (1964) considered about 120 plant families (e.g., Oliniaceae, Columelliaceae, Cephalotaceae) to be “endemic” (confined to one continent or equivalent area), compared with a few families, such as the Poaceae, Asteraceae, and Primulaceae, that have some native representatives on all continents except Antarctica. This general pattern of geographical restriction becomes abundantly apparent with the distribution of genera. There are few genera that are nearly cosmopolitan (e.g., Poa, Festuca, Astragalus, Carex, Drosera); most genera are quite restricted. Predictably, there are only a handful of nearly cosmopolitan species that are distributed by natural means (e.g., Phragmites communis, Deschampsia caespitosa, Lemna minor) (Good 1964). The probability that any species would be carried everywhere by natural agents alone (and persist) is clearly slight. This restriction in species' ranges bears on the probability that any native flora contains plants adequate to support humans, including people in a hunting–gathering society. Despite the possibility that convergent evolution could yield species with remarkably similar traits, properties, and attributes on different continents, only a tiny fraction of the approximately 250,000 vascular plants have traits that are deemed so valuable by humans that they have been widely domesticated. For example, although many grasses have seeds that can be milled, only a handful are ever ground for flour (Simmonds 1976). And although many species have fleshy fruit, few are by to the of toxic in and The of Australia and America are particularly in native species that have become widely domesticated as For example, modern with in America make a and species (Harlan 1976). Australia has contributed only the (Simmonds 1976). Yet European colonists in America and Australia (and everywhere that they for all plant needs were by from native species. For example, the of native to Australia the of many plant species. among the plant European colonists could have an if including and to this plant probably to the human to unless familiar food is (Mack 1999). in Australia, and early colonists to native species Mack although the of these was For example, in the first known European of native plant food in Australia, the de that in and they so that was any between and to de the seeds are but only after a and Thus, an much by from colonists for plants from their These were by no means to Plants for and were Such for plants from a European could be (Mack 1999). For example, in alone at least for another for the Although a usually only one some the a year Thus, even in the century, transoceanic was to ensure within a early of plant introductions on that would be and South and immigrants a that a of natural both (e.g., the in and regions (e.g., the in Although these regions the native grasses (such as and were soon deemed than as By the century, grasses native to and (e.g., and and both had had been deliberately introduced to South America because these grasses had proven much more as began the of in South of native to by naturalized and As a result, an but undoubtedly in and and South America is now by these grasses The role of in the and introduction of potentially useful plants in and in the 1997). in plant importation in the United States began by at least By John was to and facilitate the importation to the United States of any of whether useful as food for or the or for with or any of the useful This early in of the US of of and in (Hodge and Erlanson 1956), which and of nonindigenous species. species have among US including and (Hodge and Erlanson 1956). Not all introductions of grasses in the United States can be to but the of has been with and for example, (Hodge and Erlanson and et al. 1986). This pattern of is similar in of in Australian of the deliberate introduction of species as Among the common were grasses from to and from the to As these introductions have more economic and damage than the nonindigenous species introduced between and 1985, only have been as useful as in however, have become naturalized including of the species. In Australia (and probably the potential for plant was by to even species as (e.g., or by without the plants introductions also have to the The colonists to the of Good soon the of native plants, even for sources of with from and was an early but (Mack 1999). was deemed by the in the to a by nonindigenous trees in South Some of these species (e.g., in water in South has some to or even is at the water supply by the of nonindigenous trees in these and the native that is much of water and 1999). in plant dispersal more than the of species long in to colonies. The and soon for plant dispersal and established under not only the introduction of potentially useful plants to the but also served as for recently taxa for to and in new ranges 1997). For example, the had established a in by the end of the and colonial were established in the in and the These early were in the by including in Australia, New and By had established a of on in that the transport of thousands of plant taxa new ranges 1997). Such an not only the for species to be in new regions but also the of species that could in a new range. As with so much of plant we the to which weedy species were introduced in this although there are such as This list of plants introduced both species that from the and those that were deliberately distributed in the new range 1997). As colonists more in their new they began to species from their and in for both familiar and These plant from or of human by people in and Among colonists, the of a supply of and was not Many to be with familiar plants from (Mack and they also had to plants. These the of such plants as spp., and the United States (Mack This of plant from the utilitarian to include the is clearly in the plant in seed plants have a long both in and European Australia, In the these were to the seeds and propagules of essential species for a new crops, and plants. But by the in the United States and the of species Both and seed their in recently and in their were nonindigenous species. Thus, in the United and even nonindigenous species to in areas of the United States (Mack This the taxonomic breadth of species that were given the to become weeds in a new range. The phase and in the introduction of weedy species The in nonindigenous plants that are as is both in economic terms and in the of species in the one of has both and seeds of many nonindigenous species from plant in the United Many of these species are now naturalized in the United and some have clearly weedy traits in their new This list includes and Verbascum Among these species are particularly to in their new including and represent human health such as and Urtica This pattern of of potentially weeds is not to the United for the continued of such species is given the damage they wreak in many regions (Mack human most of the transoceanic dispersal of plants, including new they play a similar role in these species' in the new forces facilitate the spread of weedy nonindigenous species within a new humans play a but natural forces are the of the in this it is the common between and The epidemiology of weed similar to that of all other with the spread of (Mack et al. 2000). in of the the minimum viable the and of and the of species in the new range their and the effects of the new have in the of the and spread of weed populations, of the species (Mack plants are not among the and to be plants spread than would be from seed dispersal For example, the range of trees in to has been to be on the of the seed dispersal distance of the trees is than This known as has been by dispersal and Such play an in the of plant and that they to an of spread as an et al. 1998). of the in the of Australia is an The populations spread at between and the of seeds the of Seeds of far beyond the because they can and the that this weed are by the 1993). Natural such as can be in a of a new range that was originally through human For example, is an European that was introduced deliberately the United for or as an or both and Mack new range in the United States is to widely in the et al. 1998). Although it arrived through human subsequent spread has been by which seeds in their dispersal, the would probably still have a very distribution among it is one of the most and potentially of in the of this region and Mack Humans also play both accidental and deliberate in moving weeds around a and areas are certainly not to this have often at the for to weed seeds on their or But this has been with little Although seeds are often carried on or the of this movement for the and spread of weed in is still was by and They that the number of to and South was with the number of species. (1999) their to for the that which most likely more environmental and thus could support more nonindigenous species, also would to have more after for the of the the by and However, this still not that we are a of seeds the could be that the of of within the in which weeds can 1999). In any such have little for and that most in Australia had or one that any attempt to quarantine of would in largely they would be at for and weed in the human in plants within a potential new range in the potential for these species' As with transoceanic plant the role of humans here is in plants over natural thus new has also the of the number of to a species' of a the of range is much if the spread from many that have the aggregate as a single large (Mack There are for this including the that many of the seeds by a large not to range the at the In the few seeds by populations be much more likely to on The of that species known to be weeds arrive in new This also transport an of species to ranges in which they become The potential of these deliberate introductions are because on the by some species is often and their around the a about the potential spread of plant species. By no would the of humans or their domesticated be The in one region are to cause of quarantine these a new range this only the deliberate spread of other Species such as and are still and distributed in many This equivalent of with has been The enormous economic to species new to and likely many species with weedy to become More than taxa are for cultivation in New Zealand at least that many are in the United States 1996). Despite the of this of species, most vascular plant species about 250,000 have probably been Among these species are some that could and without cultivation once introduced a new range. As valuable as of the spread of weeds can we also a much of the epidemiology of plant through of at all including the of plant immigrants of cultivation, much become weeds (Mack et al. 2000). Thus, future weeds largely on the to and a tiny number of species within species) that could become weedy before or soon after their a Our of the for these the recipient and their (Mack et al. 2000). But this only through in by from accounts of weed spread (Mack 1996). from the here on dispersal can (and also on the deliberate in plants. and which usually arrive by most detrimental plants are deliberately their new range. This deliberate plant importation the for research in the and of introductions et al. 1999). be to carefully the introduction of species for their that have no but the of Although it is to for the introduction of species that have become weeds in new much can be in the future by that the deliberate dispersal of plants can have for spread of or from sources et al. and and 1995, et al. 1995, and 1997). The are on the with the the the the the are the and The within the the and are as over each are the number of each contains