HORTSCIENCE 40(5):1168–1174. 2005. mans, most of the earth’s approximately 240,000 angiosperms have no immigration history. If the collective performance of those Predicting the Identity of that have been already transported beyond their native range is any indication (and we currently Invaders: Future Contributions from have no alternative index), this large pool of potential immigrants contains taxa capable of Horticulture naturalization and even invasion. The likelihood of immigration for these species obviously varies Richard N. Mack as a function of the diverse incentives among School of Biological Sciences, Washington State University, Pullman, WA 99164 humans for transporting to new ranges. Although the emphasis from antiquity through the European colonial era was on transporting Horticulturists have joined ecologists, posed by alien species released in new ranges species for food, medicine, forage, forestry or conservationists, land managers and others in are becoming increasingly apparent. For a com- soil stabilization (Mack, 1999, 2001; Warming- seeking to identity future plant invaders from bination of reasons, including their release from ton, 1974), the record of plants distributed for amongst the vast array of species introduced native predators, parasites and competitors, a their perceived ornamental value has accelerated into the U.S. The routine practice in horticulture minority of alien species can not only become (Mack, 1991), particularly in the last 200 years of evaluating plants in fi eld trials coupled with naturalized, i.e., persistent, in a new range but (Campbell-Culver, 2001). careful observation will serve this effort well. can also proliferate, spread and consequently Species prized as ornamentals refl ect a bewil- Most plant introductions into any new range are wreak enormous damage. Those naturalized dering array of human perceptions of desirable deliberate, and the bulk of these are motivated species that display this epidemiology are termed plant traits and features. These perceptions vary by horticultural pursuits. Unlike most species in- biological invaders (Mack et al., 2000). They among and within societies, as well as within troduced for soil reclamation, pastures or range- represent an exceedingly small minority of the the same society over time. Some species (e.g., lands, horticultural introductions are intensively alien species that arrive, but their numbers belie Aegilops cylindrica Host [bearded goatgrass]) cultivated (e.g., irrigation, pesticide application, their negative effects (Williamson, 1996). now naturalized in the U.S. were introduced in tillage). Cultivation simultaneously minimizes For the vast majority of species, including the 19th century for dried fl ower arrangements. stochastic, i.e., randomly appearing, and chronic plants, dispersal in the past 500 years has been Goatgrass is an exceptionally damaging weed in hazards for an immigrant population. Chronic through human agencies (Ruiz and Carlton, cereal fi elds in part through its hybridization with hazards, such as desertic conditions for an in- 2003). Unlike the post-Columbian dispersal wheat (Mack, 1991). Other naturalized species troduced species with high water demand, will of microorganisms and insects by humans, were introduced much earlier for medicinal pur- inevitably destroy the population once irrigation which has been almost entirely accidental, poses or as seasonings, e.g., Digitalis purpurea is withdrawn. Introduced species with such plant species that have subsequently become L., Taraxacum offi cinale Weber, and Verbascum chronic limitations pose no threat of invasion. naturalized were introduced largely through thapsus L. These species were deliberately Alien species that lack a chronic limitation in a deliberate introductions (Groves, 1998; Mack carried far and wide across the U.S. by early new range can nevertheless undergo extinction and Erneberg, 2002). As a result, a growing settlers (Haughton, 1978). Once introduced by random detrimental events (e.g., brief bouts fraction of the earth’s fl ora has both a native into temperate North America, a species often of low temperatures or drought coincident with range and an increasing introduced range. In spread locally through accidental dispersal. For the population’s introduction). The power of many cases, these introduced ranges could not example, during the westward expansion of random events to infl uence a population’s fate have been produced through natural dispersal European settlers, Native Americans reputedly is infl uenced strongly by the inverse relationship (e.g., the mass introduction of European species termed Plantago major L. white man’s foot; P. between population size and the probability of to New Zealand, the introduction of Central major was to them a harbinger of approaching local extinction. Cultivation may protect small, Asian species to interior North America). The European settlers (Darlington, 1859). introduced populations through a hazardous number of species that have been Strong incentive remains to introduce spe- period; the population survives and increases in introduced to North America since 1500 AD cies heretofore unknown in the U.S., despite the size, thereby lowering the risk of its extinction. probably cannot be determined but a conserva- thousands of alien species currently available Propagules from this protected population may tive estimate would be approximately 40,000 from commercial nurseries in the U.S. The then spread to other sites and can eventually taxa, based on the number of species and their public, who collectively display an enormous give rise to new, persistent populations. Hor- varieties now commercially available in the U.S. diversity of interest in plants, welcomes these ticulturists can provide invaluable information (Isaacson, 1996). This fi gure dwarfs the native importations. Some plant collectors specialize toward predicting invasions by reporting cases in North American vascular fl ora of approx. 18,000 in morphologic groups, such as succulents or which an introduced species require little, if any, species (Kartesz and Meacham, 1999). epiphytes; the desire among others is to maintain cultivation to reproduce. Cases in which plants under cultivation as many species of a taxonomic derived from an introduced population appear Deliberate Plant Introductions group as possible (e.g., bamboos or orchids) frequently well outside the cultivated area will (Mack, 2001). Much plant exploration continues be especially informative. Extermination and The overall result of these plant introductions a centuries old enterprise that has scoured the withdrawal from the market may be warranted to North America has been decidedly positive globe repeatedly for ornamental species that are for introduced species that display this second from the standpoint of human welfare. For ex- new to the marketplace (Musgrave et al., 1999; category of behavior. ample, the current population of the U.S. could Spongberg, 1990). The threats to national economies (Pimentel not have been attained, much less maintained, This search can inadvertently select species et al., 2000), the environment (Vitousek et al., with reliance solely on edible native U.S. plants with features that enhance their risk to become 1996) and even human health (Tyler, 2004) (Mack, 1999). The fate of most alien plant spe- naturalized and even invasive. For example, cies in the U.S. is either residence only within priority is placed on species with a long fl ower- Received for publication 30 Oct. 2004. Accepted for cultivation or extinction soon after introduction. ing time or repeated bouts of fl owering, or both. publication 24 Feb. 2005. I thank Neil Anderson, The small remainder, 2500 to 3000 species, have Most gardeners seek species that are attractive: Sarah H. Reichard, Judy Warnement and Peter Whar- become persistent (Kartesz and Meacham, 1999) ton for valuable references or advice in preparing if not the fl ower, then the foliage or the fruits this paper. Rich Scott once again provided skillful and of these species perhaps 300 species are truly should have an attractive display (Mack, 2001). assistance in illustrations. I especially thank Robert invasive in the conterminous U.S. (Randall and Attractive fl owers and foliage probably impart Stamps and the American Society of Horticultural Marinelli, 1996) little advantage to the likelihood of persistence Science for inviting me to the 2004 ASHS colloquium Despite the array of alien species that have in a new range, although a showy fl ower could on plant invasiveness. been introduced outside their native range by hu- be an advantage in attracting pollinators to the

1168 HORTSCIENCE VOL. 40(5) AUGUST 2005

AAugustBook.indbugustBook.indb 11168168 66/14/05/14/05 112:07:292:07:29 PPMM FEATURES

new species (Pickering, 2001). Choosing spe- goal of biologists even before the establishment 2000; Veltman et al., 1996). Such cases suggest cies and cultivars with large numbers of showy, of plant quarantine lists (Gray, 1879). As a fi rst that other factors contribute to immigrants’ fertile fl owers or seeds could also increase the approximation to prediction, biologists often fate beyond their attributes and tolerance of propagule pressure of the species. Species with search for correlation between invasive species the local climate. Certainly such outcomes attractive fruit (e.g., Hedychium gardnerianum) and traits or attributes of their biology (e.g., may refl ect low genetic variation among the may have a distinct advantage in persistence no seed dormancy, long fl owering periods, immigrants (Newman and Pilson, 1997). But because frugivorous birds are attracted to fl eshy evergreen leaves) (Reichard and Hamilton, even the fates of different genotypes do not fully fruits, which they can disperse far from the 1997). This approach is eminently reasonable: explain the apparent randomness with which source plant (Cuddihy and Stone, 1990). a species with attributes that enhance its ability some species eventually become naturalized in Ease in propagation is also sought among to survive in a new environment could serve a new range. Other important factors, such as new introductions. To this end, plant explorers as a guide to the performance of other species the size of the immigrant population (Berggren, sometimes peruse lists of native weeds wherever with the same attribute(s). 2001; Veltman et al., 1996) and especially the they are collecting, searching for ruderal species The record of these investigations as predic- character of the randomness or stochasticity (low maintenance) (Anderson, 2004) that fi t tors of potentially invasive species is however of the environment in the new range, could their other criteria (showy fl oral display, fl eshy mixed. In some cases, the link appears strong. also explain these varied outcomes (Drake and fruit). For instance, amabile Stapf. Rejmanek and Richardson (1996) fi nd a strong Lodge, 2004). I contend that prediction of the & J.R. Drumm., a native of southern China, is correlation between the persistence of some response of immigrant populations to environ- common along roadsides in its native range pines in South Africa and these same pines’ mental stochasticity and especially the role we and has a attractive display of intense, purple animal-dispersed seeds. Similar successes have humans play in buffering this stochasticity are fl owers (RNM, personal observation). It has been reported for larger, unrelated groups of essential parts of the predictive framework been introduced into the U.S. and has already species, including invasive species in Hawaii for identifying which species, amongst so become naturalized locally (Randall, 2002 and (Daehler et al., 2004) and woody species in many, will not only become naturalized but references therein). the U.S. (Reichard and Hamilton, 1997). The even invasive. Plant introductions to the U.S. should and problems with sole reliance on attribute-based will continue; the horticulture of any nation, systems are evident in their inability, so far, The size of Immigrant Populations and including the U.S., could benefi t from taxa that to satisfactorily screen all introductions. For Environmental Stochasticity have yet to be introduced. The continuance of instance, the high level of false positives (i.e., this free trade is protected both in U.S. trade species declared invasive when in fact they are It is a most optimistic gardener who laws governing the importation of living plant not) (Lonsdale and Smith, 2001; Reichard and plants a single seed in a new location with material as well as the much more recent agree- Hamilton, 1997) runs afoul of international the expectation that this sole individual will ments of the World Trade Organization (WTO), rules minimizing restrictions on free trade survive and multiply. Sowing multiple seeds of which the U.S. is a signatory (FAO, 2000). (Shine et al., 2000). These limitations further or planting multiple cuttings of a species is However, both U.S. law and WTO agreements suggest that information beyond a species’ attri- implicit recognition that a plant’s fate can recognize that some alien species can be damag- butes is needed to predict its fate in a new range. vary enormously from point to point, even ing and that deliberate steps must be taken to Species’ attributes are nevertheless part of the in the same small locale (Harper, 1977, p. identify these species and to prevent or diminish equation needed to develop a comprehensive 123). Abundant theory in population biol- their spread. Furthermore, the WTO requires system for predicting invasive species. ogy supports any gardener’s experience and all member nations to erect a national invasive Another common sense approach for pre- practice. Even for a population of genetically species policy that can identify and restrict the dicting the identity of future invasive species identical individuals, the larger the population entry of harmful plant species without imposing employs comparisons between the climate of (Belovsky et al., 1999) and the more patches excessive restrictions on free trade (Shine et al., the immigrant species’ native (or other donor) in which they occur (Bascompte et al., 2002), 2000). Although the U.S. bans the importation range with the climate in a potential recipient the more likely some individuals will produce of some alien plant species under the Plant locale (Kriticos and Randall, 2001). Climate is progeny. This correlation between survival and Protection Act of 2000, Public Law 106-224 often the strongest single deterrent of a species’ population size is important here because an (http://www.aphis.usda.gov/ppq/weeds/PPA- range (Woodward, 1987). In fact, so strong is immigrant populations may be small, even if Text.PDF), these prohibitions do not address the link between climate and plant distribu- deliberately introduced (Mack, 1995). For any the potential risk from a vast number of species tion that climatologists long ago adopted the species in an environment within its ecological that can still be imported. The need now is to names of some widespread vegetation types amplitude there is a minimum viable population build a procedure that can rapidly screen new (e.g., steppe, tundra) for the corresponding (MVP) that must be maintained for persistence introductions to the U.S. and determine upon a climatic regions (Trewartha, 1968, p. 241). (Simberloff, 1988). scientifi c basis, a WTO requirement (Shine et In the last 20 years computer software, such The random expression of most facets of al., 2000), which species constitute a risk to the as CLIMEX (Sutherst et al., 1998), has been the environment further works against any in- importing nation. Can we erect such a fl exible, developed to predict locales where an alien dividual plant’s survival. For example, weather rapid, science-basis screening system? The an- species could survive, based on similarities is the random expression of the amplitude in swer is a conditional “yes” (National Research in the climates of donor and recipient ranges climate. Accurately predicting the weather at Council, 2002), and I contend for the reasons (Holt and Boose, 2000; Kriticos et al., 2003). a site is exceedingly diffi cult, even for short that follow that the horticulture industry is in a No comparable system has been developed periods (Allaby, 2002), and weather is but one particularly good position to provide essential however to encapsulate and compare the biotic aspect of environmental stochasticity. The information needed to devise this procedure. aspects (e.g., predation, competition and para- magnitude of this stochasticity, which trans- sitism) among environments. This biological lates into uncertainty about the conditions for Predicting the Identity of Invasive information is needed because a native species plant growth or even survival, encompasses Species: Attributes and Climate Matching in the potential new rage can single-handedly not just the physical environment but also prevent naturalization (Mack, 1996). biotic forces, such as grazers, predators, and Predicting which species could become in- If a species’ naturalization and eventual parasites (Mack and Pyke, 1984). vasive amongst potentially tens of thousands of invasion depended strictly on a list of its attri- The hazards of environmental stochasticity introductions into any new range is a daunting butes or its tolerance of the climate in the new are acute for small, alien populations because task and includes statistical hurdles that may range, or both, the outcome of any immigration these populations may be reduced rapidly and not be immediately apparent (Lonsdale and would be predictable. To an unknown extent may be unlikely to receive new immigrants. Smith, 2001). Comparatively few species are however a species may go through recurring As a result, such populations may not recover legitimate concerns and should be barred entry. immigrations into a new locale before eventu- from deleterious events in environmental Identifying this distinct minority has been a ally persisting (Mack, 1995; Sax and Brown, stochasticity, except under subsequent (and

HORTSCIENCE VOL. 40(5) AUGUST 2005 1169

AAugustBook.indbugustBook.indb 11169169 66/14/05/14/05 112:07:322:07:32 PPMM extraordinarily) favorable conditions (Lande, insulates plant populations, including popula- of these new foci (Fig. 1). In this manner a 1993; Simberloff, 1988). Although the odds tions of ornamental plants, from vagaries of species can become naturalized (Mack, 1995, for persistence often appear very low, they are the environment that they might numerically 2000). Thus, the likelihood of naturalization still fi nite. Populations can become established withstand if the populations were suffi ciently is substantially enhanced by the maintenance strictly through chance alone, i.e., these popu- large. In contrast, protecting a species in lo- of a deliberately introduced founder popula- lations beat the long odds imposed through cales where it faces a chronic hazard, e.g., the tion under cultivation. This scenario contrasts random events in the environment (Ridley, maintenance of cotton under irrigation in the with the very low likelihood of naturalizations 1930). Are naturalized fl oras then simply the American Southwest (Erie et al., 1981), is also arising from repeated accidental immigrations, product of these long odds or is another fac- cultivation but any environmental stochasticity which may be infrequent and for which the tor affecting and even increasing the odds in is irrelevant for plant survival. In these cases, immigrants themselves are at high risk of favor, even temporarily, of plant survival and all members of the alien population die as soon extinction without cultivation. population persistence? as cultivation is withdrawn. The link between cultivation and natu- Cultivation applied during periods of high- ralization as described here helps explain The Role of Cultivation risk for an immigrant population provides an the strong correlation between deliberately essential service. It can indefi nitely maintain introduced species and naturalizations (Mack, Cultivation has multiple defi nitions to a seed or propagule source in the new range. 2000). Two-thirds of the naturalized fl ora in which I will add one more: a buffer against the As a result, there is an opportunity that some the U.S. are likely comprised of species that hazards imposed by environmental stochastic- seeds, bulbils or plantlets produced under were deliberately introduced (Mack and Er- ity. Cultivation can take many forms (tillage, cultivation will establish a new population neberg, 2002). Another much smaller fraction irrigation, insecticide application), and many of outside cultivation. In time, multiple foci of are seed contaminants in crop seeds that have these tasks illustrate the role cultivation plays the immigrant population may arise in the undergone strong selection, such that their in protecting an immigrant population from new range, all either direct descendants of biology strongly mimics the crop’s biology stochastic events. In these cases, cultivation plants still under cultivation or descendants (e.g., seed size, color, and dormancy as well as

Fig. 1. An alien population under cultivation is buffered indefi nitely from much environmental stochasticity that would otherwise destroy it. As a result, the population routinely produces seeds that are dispersed. Some of these seeds eventually give rise to new populations that foster further spread; most however fail to produce new populations (black circles).

1170 HORTSCIENCE VOL. 40(5) AUGUST 2005

6632-Feature.indd32-Feature.indd 11170170 66/21/05/21/05 44:51:32:51:32 PPMM plant phenology, even height and appearance) (Mack, 1999, 2000). In summary, the likelihood den, are widely observed (Clement and Foster, (Barrett, 1983; Mack and Erneberg, 2002). Few of naturalization appears to increase greatly 1994). Botanical gardens have provided some species that are deliberately introduced are with deliberate transport and cultivation. The of the best-documented examples in which an simply sown without accompanying cultiva- following examples provide evidence that alien species has escaped, spread and given tion (e.g., species introduced in arid rangelands these forces can facilitate the emergence of rise to an invasion. Botanical gardens often and soil stabilization), and these species are ornamental species as invaders. maintain alien plants that are not in general likely sown in huge founder populations that commerce, thus greatly narrowing the list increase their likelihood of persistence. In Ornamental Species that became of likely introduction points for any newly contrast, cultivation for ornamental species is Invaders detected alien species found nearby. Further- often intensive. As benefi ciaries of cultivation, more, the garden’s conservators are likely to the founder populations of these species have Garden escapes, i.e., plants that arose from detect escapes in the course of monitoring the subsequently given rise to many naturalizations seeds or propagules that dispersed from a gar- garden’s boundary for pests.

Fig. 2. Senecio squalidus spread from cultivation in the Oxford University Botanic Garden before 1800 and is now found commonly across much of Great Britain (Perring and Walters, 1976).

HORTSCIENCE VOL. 40(5) AUGUST 2005 1171

6632-Feature.indd32-Feature.indd 11171171 66/21/05/21/05 44:51:42:51:42 PPMM Mimosa pigra L. (mimosa) is a tall shrub to next 60 years, so that by 1940 the species was Vancouver was the shrub Rubus simplex Focke, small tree native to tropical America. It has long naturalized along many rail lines throughout which produces a dense mat-like stoloniferous been deemed an attractive species for tropical southern England and adjacent Wales (Kent, layer that retards the growth of other species. gardens and was planted widely throughout 1956; 1960). It had not however penetrated into More worrisome are its fl eshy yellow fruits tropical and subtropical Africa (Bentham, the countryside away from rail lines, except that are spread locally by birds. By the early 1875), a practice that continues sporadically where deliberately introduced. 1990s, the shrub was appearing with increasing (B & T World Seeds, Aigues-Vives, France. The distribution pattern of S. squalidus frequency on the grounds of the garden. For http://www.b-and-t-world-seeds.com). In changed during World War II, as the plant was the last 10 years the garden staff has consci- tropical Asia and Oceania it is an aggressive carried in the rubble from war-devastated walls entiously destroyed this perennial within the invader, which has resisted most attempts at and buildings and dumped into gravel pits. garden. Although the Botanical Garden is not control. Mimosa’s greatest damage so far has This rubble-strewn habitat proved ideal for its entirely rid of R. simplex, this goal will likely be probably been infl icted within fl oodplains in further establishment (Kent, 1964a, 1964b). achieved in the next few years. Fortunately, the the Northern Territory of Australia where it has By the late 1950s, the alien had reached the staff noted its spread and took prompt action to largely transformed a native treeless fl oodplain Tyne Valley about 300 km north of its initial remove the source plants and eradicate all the vegetation into a mimosa monoculture (Braith- point of escape from cultivation (Kent, 1964c). escapes. Nevertheless, “it almost got away” (P. waite et al., 1989 and references therein). Spread accelerated after 1960, such that it oc- Wharton, personal communication). The circumstances of its introduction curred throughout much of southern England into the Northern Territory at Darwin are not by 1976 (Perring and Walters, 1976) (Fig. 2). Horticulture’s Contribution to a altogether clear, but in the most plausible Today Oxford ragwort is found throughout Predictive Framework explanation it fi rst appeared at the Darwin the southern two-thirds of England, plus many Botanical Garden late in the 19th century. By locales in Scotland, Wales and Ireland (Preston Predicting the identity of future invasive logically eliminating other possibilities, Miller et al., 2002). species will require integrating our knowledge and Lonsdale (1987) deduced that M. pigra was Several episodes in the early history of S. of each species’ attributes that contribute to its likely the noxious mimosa being combated in squalidus in southern England reveal the role persistence within a climatic region as well as the garden by 1914. A photograph taken on the that cultivation can contribute to eventual gauging the environmental stochasticity in the garden grounds in 1936 shows the shrub grow- naturalization. The taxon had apparently been new range. Progress in identifying relevant at- ing unattended in a watercourse (Miller and maintained in the Oxford Botanic Garden for tributes will continue, even if these attributes Lonsdale, 1987). The strongest circumstantial about 100 years before escapes were detected. are detected retrospectively (Daehler et al., evidence indicates that a cultivated founder At least two nonmutually exclusive explana- 2004). Climate matching will likely become population of the tree resided in Darwin and tions could account for this prolonged delay. more discriminating among locales (Pauchard was the source for its accidental spread into the First, S. squalidus arose through the hybridiza- et al., 2004). The biggest challenge will be Adelaide River. From this point of entry, the tion of two ancestral species in Sicily. Once in assembling a comprehensive understanding shrub spread to dominate about 45,000 ha in Oxford, the immigrant population was isolated of regional environmental stochasticity and the riparian landscape south of Darwin. from its parents, and its genotype became the cultivation needed to buffer alien species As the epidemiology of Mimosa pigra il- fi xed through segregation, recombination from these forces (Mack, 2000; Minton, 2003). lustrates, direct observation of an alien species’ and possible mutation to yield the genotype To meet this challenge, risk assessment will initial escape(s) and subsequent spread from that escaped cultivation (Abbott et al., 2000). need to go beyond reliance on our current cultivation is usually not noted. The fi rst few Admittedly, earlier escapes may have been understanding of the traits of a species and occurrences of a species outside cultivation overlooked. This explanation seems unlikely its performance elsewhere as illustrated in the would usually not attract attention, except in however, given the intense collecting activity Voluntary Code of Conduct for Nursery Profes- areas where collection intensity is extraordi- in Oxford, even in the 18th century (Clokie, sionals (http://www.centerforplantconserva- narily intense. The classic account assembled 1964). An alternative explanation holds that tion.org/invasives/nurseryN.html) and include by Kent (1956, 1960, 1964a, 1964b, 1964c) of seeds of S. squalidus had been dispersed from experimental evaluations, i.e., fi eld trials, of the escape and subsequent spread of Senecio the garden throughout the 18th century but species in their potential new ranges. squalidus L. (Oxford ragwort) from the Oxford had consistently failed to produce a persistent Horticulture, especially ornamental hor- University Botanic Garden in Great Britain population. The random appearance of circum- ticulture, can provide much of this needed further substantiates concerns that cultivated stances conducive for establishment of new information through its emphasis on fi eld alien species can indeed radiate far from a populations did not occur until about 1800. trials for recently introduced alien species. cultivated founder population and eventually Further evidence that small populations of S. As new species are introduced to the U.S. establish self-sustaining foci of invasion. squalidus are vulnerable to stochastic events for commercial evaluation, I recommend that Oxford ragwort is a native annual/peren- is provided by Turrill (1948), who failed to information on plant performance under two nial of Sicily (Abbott et al., 2000) and was establish the species through transplants and categories be systematically collected and growing in the university’s garden by 1690. fruits on walls at Woodstock (about 10 km widely shared. The fi rst record of its appearance outside the northwest of Oxford). These immigrants were Two categories of information are needed garden was in 1794 by Smith, and by 1800 the extirpated, even when deliberately transported from fi eld trials. same collector reported it “as very plentiful on to an apparently suitable habitat. Each new Identifying species that require little or no almost every wall in Oxford,” which suggests focus of S. squalidus appears to have been cultivation. Attention should be paid to the that it had escaped much earlier (Smith, 1799 dependent on the population reaching suffi cient mode (irrigation, tillage, pesticide application, as cited in Kent, 1956). The post-1800 collec- size that it could withstand local environmental frost protection, minimum nutrient require- tion records for S. squalidus are extensive and stochasticity without cultivation. ments), frequency and duration of cultivation permit construction of a detailed chronology Escapes of alien species continue. The needed for recent plant introductions to persist of the species’ spread away from Oxford, es- 1980 Sino-American Botanical Expedition through sexual reproduction or vegetative pecially along the developing railroad system to the Shennongjia Forest District in Hubei propagation in fi eld trials. Special attention (Kent, 1956). The pappus of the plant’s dried Province, China, collected a trove of species should be placed on those species that require fruit greatly aided in its aerial spread. Much new to science as well as many species of hor- little or no cultivation, as these species could greater dispersal was achieved as the species ticultural interest (Dosmann and Del Tredici, conceivably become established in the new spread along rail lines and in railcars. Druce 2003). Many of these species were distributed range. I recommend that these species continue (1927) reports seeing the fruit remaining to botanical gardens in North America, includ- to be grown in trials, while simultaneously suspended in the air in a rail carriage from ing the Botanical Garden of the University of adding them to a watch list among nurseries Oxford to Tilehurst (about 40 km away) in British Columbia in Vancouver. Among the and other horticulturists in the region. 1879. Accidental transport continued for the Chinese accessions cultivated in the garden in Identifying and reporting species that

1172 HORTSCIENCE VOL. 40(5) AUGUST 2005

AAugustBook.indbugustBook.indb 11172172 66/14/05/14/05 112:07:362:07:36 PPMM routinely escape cultivation. Heightened 2002)—the erection of a rapid, comprehensive, Haughton, C.S. 1978. Green immigrants. The plants attention should be placed on species that predictive framework for invasive species that that transformed America. Harcourt Brace Jova- routinely escape from garden trials and give is transparent, effective and imposes minimum novich, New York. rise to reproducing or vegetatively propagating constraint on the importation and commercial Holt, J.S. and A.B. Boose. 2000. Potential for spread plants at sites beyond cultivation. Although distribution of plants. of Abutilon theophrasti in California. Weed Sci. 48 (1):43–52. most of these species will reside only as tran- Isaacson, R.T. (ed.). 1996. Andersen horticultural sients in the new range, some could become Literature Cited library’s source list of plants and seeds. 4th naturalized. These species should be reported Abbott, R.J., J.K. James, J.A. Irwin, and H.P. Comes. ed. Andersen Hort. Library, Univ. Minn., Min- to the state’s plant pest council or equivalent 2000. Hybrid origin of the Oxford ragwort, Sene- neapolis. (e.g., www.caleppc.org; www.se-eppc.org; cio squalidus L. Watsonia 23:123–138. Kartesz, J.T. and C.A. Meacham. 1999. Synthesis invasives.eeb.uconn.edu/ipane). These species Allaby, M. 2002 (ed.). Encyclopedia of weather and of the North American fl ora. North Carolina could continue to be on a watch list, although climate. Facts on File, New York. Botanical Garden, University of North Carolina, the more conservative approach would be to Anderson, N. 2004. Invasive horticultural crops. Chapel Hill. destroy the founder population as well as the Part 2: Where do they come from? Minn. Land Kent, D.H. 1956. Senecio squalidus in the British Assn. News 28(4):28–33. Isles-1, early records (to 1877). Bot. Soc. Brit. garden escapes and discourage further trials Isles Proc. 2:115–118. with the species in the region. This course of Barrett, S.C.H. 1983. Crop mimicry in weeds. Econ. Bot. 37:255–282. Kent, D.H. 1960. Senecio squalidus in the British action should be reserved for cases in which Bascompte, J., H. Possingham, and J. Roughgarden. Isles-2, The spread from Oxford (1879–1939). the species is routinely appearing as an escape 2002. Patchy populations in stochastic environ- Bot. Soc. Brit. Isles Proc. 4:375–379. over the course of several consecutive years and ments: critical number of patches for persistence. Kent, D.H. 1964a. Senecio squalidus in the British establishing outside the garden. Withdrawing Amer. Natur. 159:128–137. Isles-4, Southern England (1940→). Bot. Soc. such a species from further commercial evalua- Belovsky, G.E., C. Mellison, C. Larson, and P.A. Van Brit. Isles Proc. 5:210–213. tion should be voluntary and on a case-by-case Zandt. 1999. Experimental studies of extinction Kent, D.H. 1964b. Senecio squalidus in the British → basis but will be justifi ed in some cases. dynamics. Science 286:1175–1177. Isles-5, The Midlands (1940 ). Bot. Soc. Brit. Bentham, G. 1875. VII. Revision of the subor- Isles Proc. 5:214–216. Horticultural escapes have provided a Kent, D.H. 1964c. Senecio squalidus in the British majority of the naturalized fl ora of the U.S. der Mimoseae. Trans. Linn. Soc. London 30:335–664. Isles-6, Northern England (1940→). Bot. Soc. (Mack and Erneberg, 2002) from which many Berggren, A. 2001. Colonization success in Roesel’s Brit. Isles Proc. 5:217–219. invasive species have subsequently emerged bush-cricket Metrioptera roeseli: The effects of Kriticos, D. J. and R. P. Randall. 2001. A comparison (Randall and Marinelli, 1996). This legacy propagule size. Ecology 82:274–280. of systems to analyze potential weed distribu- need not however serve as prophecy for future Braithwaite, R.W., W.M. Lonsdale and J.A. Estbergs. tions, p. 61–79. In: R.H. Groves, F.D. Panetta, plant invasions. Gardens can become instead 1989. Alien vegetation and native biota in tropi- and J.G. Virtue (eds.). Weed risk assessment. an invaluable tool in the early detection of cal Australia: The spread and impact of Mimosa CSIRO, Collingwood, Australia. those species that need careful evaluation pigra. Biol. Conser. 48:189–210. Kriticos, D.J., R.W. Sutherst, J.R. Brown, S.W. Campbell-Culver, M. 2001. The origin of plants: Adkins, G.F. Maywald. 2003. Climate change for their potential to invade. In reality, most and the potential distribution of an invasive alien species evaluated in garden trials will prove the people and plants that have shaped Britain’s garden history since the year 1000. Headline, plant: Acacia nilotica ssp. indica in Australia. J. harmless in their new locales; they will either London. Appl. Ecol. 40:111–124. not survive outside the bounds of cultivation Clement, E.J. and M.C. Foster. 1994. Alien plants of Lande, R. 1993. Risks of population extinction or will produce only the occasional escapee the British Isles. Bot. Soc. Brit. Isles, London. from demographic and environmental stochas- that soon perishes without producing progeny. Clokie, H.N. 1964. An account of the herbaria of ticity and random catastrophes. Amer. Nat. The need here is to identify the small minority the Department of Botany in the University of 142(6):911–927. that requires little or no cultivation and has Oxford. Oxford Univ. Press, London. Lonsdale, W.M. and C.S. Smith. 2001. Evaluating a propensity to proliferate and spread in the Cuddihy, L.W. and C.P. Stone. 1990. Alteration of pest-screening systems—Insights from epidemi- native Hawaiian vegetation. Effects of humans, ology and ecology. p. 52–60. In: R.H. Groves, new range. F.D. Panetta, and J.G. Virtue (eds.). Weed risk I envision horticulturists in general and the their activities and introductions. Coop. Natl.Park Resour. Studies Unit, Univ. Hawaii, Manoa. assessment. CSIRO, Collingwood, Australia. plant nursery industry in particular providing Daehler, C., J.S. Denslow, S. Ansari, and H.C. Kuo. Mack, R.N. 1991. The commercial seed trade: an essential service in the future as the early 2004. A risk-assessment system for screening out an early disperser of weeds. Econ. Bot. detectors of potentially invasive species. Many invasive pest plants from Hawaii and other Pacifi c 45:257–273. commercial growers have established test gar- Islands. Conserv. Biol. 18:360–368. Mack, R.N. 1995. Understanding the processes of dens in radically different climates across the Darlington, W. 1859. American weeds and useful weed invasions: the infl uence of environmental U.S. For example, Monrovia Nursery (Azusa, plants. 2nd ed. (revised by G. Thurber). A.O. stochasticity, p. 65–73. In: C.H. Stirton (ed.). Calif.) maintains six widespread nurseries in Moore, New York. Weeds in a changing world. Brighton, U.K. Dosmann, M. and P. Del Tredici. 2003. Plant intro- Mack, R.N. 1996. Biotic barriers to plant naturaliza- the U.S. to evaluate and produce plants for tion, p. 19–26. In: V.C. Moran and J.H. Hoff- different climatic zones (http://www.monrovia. duction, distribution, and survival: A case study of the 1980 Sino-American botanical expedition. man (eds.). Proceedings of the IX International com). Across the U.S. thousands of commer- BioScience 53:588–597 Symposium on Biological Control of Weeds. cial and private nurseries collectively span Drake, J.M. and D.M. Lodge. 2004. Effects of Univ. Cape Town, South Africa. almost all climatic and soil regions. Whether environmental variation on extinction and es- Mack, R.N. 1999. The motivation for importing large integrated networks of nurseries, such as tablishment. Ecology Letters 7:26–30. potentially invasive plant species: a primal urge? those owned by Monrovia Nursery, or much Druce, G.C. 1927. The fl ora of Oxfordshire. 2nd p. 557–562. In: D. Eldridge and D. Freuden- more modest operations, these gardens are ed. Oxford. berger (eds.). People and rangelands: building in effect a basis already in place to evaluate Erie, L. J., O. F. French, D. A. Bucks, and K. Harris. the future. Proc. VI Intl. Rangeland Congress, 1981. Consumptive water use of water by major Townsville, Australia. alien species for the likelihood of escape from Mack, R.N. 2000. Cultivation fosters plant natural- cultivation. crops in the southwestern United States. USDA Conserv. Res. Rpt. 29. ization by reducing environmental stochasticity. The type of voluntary system I envision Food and Agriculture Organization of the United Biol. Invasions 2:111–122. here contains strong incentives. No responsible Nations. 2000. Multilateral trade negotiations Mack, R.N. 2001. Motivations and consequences of horticulturist or amateur gardener wants to on agriculture: A resource manual. vol. 3. SPS the human dispersal of plants, p. 23–34. In: J.A. contribute to the nation’s invasive fl ora (Reich- and TBT agreements. FAO, Rome. McNeely (ed.). The great reshuffl ing: human ard and White, 2001). Whether informally as Gray, A. 1879. The pertinacity and predominance of dimensions in invasive alien species. IUCN, private parties or more formally in horticultural weeds. Amer. J. Sci. Arts 18:161–167. Gland, Switzerland, and Cambridge, U.K. and growers’ associations, those who evaluate Groves, R.H. 1998. Recent incursions of weeds to Mack, R.N. and D. A. Pyke. 1984. The demography Australia. Coop. Res. Ctr. Weed Mgt. Systems of Bromus tectorum: The role of microclimate, alien plants for commerce have a great oppor- predation and disease. J. Ecol. 72:731–748. tunity. They can contribute in what is already a Tech. Ser. 3. Adelaide, Australia. Harper, J.L. 1977. Population biology of plants. Mack, R.N. and M. Erneberg. 2002. The United national mandate (National Research Council, Academic Press, London. States naturalized fl ora: Largely the product of

HORTSCIENCE VOL. 40(5) AUGUST 2005 1173

AAugustBook.indbugustBook.indb 11173173 66/14/05/14/05 112:07:392:07:39 PPMM deliberate introductions. Mo. Bot. Gard. Ann. 2000. Environmental and economic costs of guide to designing legal and institutional frame- 89:176–189. nonindigenous species in the United States. works on alien invasive species. IUCN, Gland, Mack, R.N., D. Simberloff, W.M. Lonsdale, H. BioSci. 50 (1):53–65. Switzerland, and Cambridge, U.K. Evans, M. Clout, and F.A. Bazzaz. 2000. Biotic Preston, C.D., D. Pearman, and T.D. Dines. 2002. Simberloff, D. 1988. The contribution of population invasions: causes, epidemiology, global conse- New atlas of the British & Irish fl ora: an atlas and community biology to conservation science. quences and control. Ecol. Appl. 10:689–710. of the vascular plants of Britain, Ireland, the Isle Annu. Rev. Ecol. Syst. 19:473–511. Miller, I.L. and W.M. Lonsdale. 1987. Early records of Man and the Channel Islands. Oxford Univ. Spongberg, S.A. 1990. A reunion of trees: The of Mimosa pigra in the Northern Territory. Plant Press, Oxford. discovery of exotic plants and their introduction Prot. Q. 2(3):140–142. Randall, J.M. and J. Marinelli. (eds.). 1996. Invasive into North American and European landscapes. Minton, M.S. 2003. Plant naturalization: surviving plants: Weeds of the global garden. Brooklyn Harvard Univ. Press, Cambridge. the gauntlet of environmental stochasticity. PhD Bot. Garden, Brooklyn. Sutherst, R.W., G.F. Maywald, T. Yonow, and P.M. diss. Wash. State Univ., Pullman. Randall, R.P. 2002. A global compendium of Stevens. 1998. CLIMEX. Predicting the effects Musgrave, T., C. Gardner, and W. Musgrave. weeds. R.G. and F.J. Richardson, Melbourne, of climate on plants and animals. User guide. 1999. The plant hunters. Seven Dials, Cassell, Australia. CSIRO, Melbourne, Australia. London. Reichard, S.H. and C.W. Hamilton. 1997. Predicting Trewartha 1968. An introduction to climate. 4th ed. National Research Council. 2002. Predicting inva- invasions of woody plants introduced into North Mc-Graw-Hill, New York. sions by nonindigenous plants and plant pests. America. Conser. Biol. 11:193–203. Turrill, W.B. 1948. British plant life. London, Natl. Acad. Sci., Wash., D.C. Reichard, S.H. and P. White. 2001. Horticulture as Collins. Newman, D. and D. Pilson. 1997. Increased prob- a pathway of invasive plant introductions in the Tyler, K.L. 2004. West Nile virus infection in the ability of extinction due to decreased genetic ef- United States. BioSci. 51:103–113. United States. Arch. Neurol. 61(8):1190–1195. fective population size: experimental populations Rejmanek, M. and D.M. Richardson. 1996. What Veltman, C.J., S. Nee, and M.J. Crawley. 1996. of Clarkia pulchella. Evol. 51:354–362. makes some plant species more invasive? Ecol- Correlates of introduction success in exotic New Pauchard, A., L.A. Cavieres, and R.O. Bustamante. ogy 77:1655–1661. Zealand birds. Amer. Nat. 147 (4):542–557. 2004. Comparing alien plant invasions among Ridley, H. N. 1930. The dispersal of plants throughout Vitousek, P.M., C.M. D’Antonio, L.L. Loope, and regions with similar climates: Where to from the world. L. Reeve, Ashford, Kent, U.K. R.G. Westbrooks. 1996. Biological invasions here? Diversity & Distributions 10:371–375. Ruiz, G.M. and J.T. Carlton. 2003. Invasion vectors: as global environmental change. Amer. Sci. Perring, F.H. and S.M. Walters. (eds.). 1976. Atlas A conceptual framework for management, p. 84:468–478. of the British fl ora. 2nd ed. EP Publ. Bot. Soc. 459–504. In: G.M. Ruiz and J.T. Carlton (eds.). Warmington, E.H. 1974. The commerce between the Brit. Isles, Wakefi eld, U.K. Invasive species. Vectors and management strate- Roman empire and India. Curzon, London. Pickering, C.M. 2001 Size and sex of fl oral displays gies. Island Press, Wash., D.C. Williamson, M. 1996. Biological invasions. Chap- affect insect visitation rates in the dioecious Sax, D.F. and J.H. Brown. 2000. The paradox of inva- man & Hall, London. Australian alpine herb, Aciphylla glacialis. Nord. sion. Global Ecol. Biogeog. 9 (5):363–371. Woodward, F.I. 1987. Climate and plant distribution. J. Bot. 21 (4):401–409. Shine, C., N. Williams, and L. Gündling, L. 2000. A Cambridge Univ. Press, Cambridge, U.K. Pimentel, D., L. Lach, R. Zuniga, and D. Morrison.

1174 HORTSCIENCE VOL. 40(5) AUGUST 2005

AAugustBook.indbugustBook.indb 11174174 66/14/05/14/05 112:07:432:07:43 PPMM