54
Biological Pollution: l he lun i ruu
Distribution and Spread of the Invasive Biennial Alliaria petiolata (Garlic Mustard) in North America
Victoria Nuzzo'
INTRODUCTION
Alliaria petiolata (Bieb. [Cavara & Grande]) is a cool-season, shade-tolerant, obligate biennial herb that invades forested natural communities in the Midwestern and northeastern United States and adjacent Canada (Fig. 1). Alliaria petiolata is often referred to as Alliaria officinalis Andrz; earlier names included Alliaria alliaria L. (Britton), Sisymbrium alliaria Scop., Sisymbrium officinalis DC and Erysimum alliaria L. Alliaria petiolata (hereafter referred to as Alliariä) is commonly known as garlic mustard, in reference to the strong garlic fragrance produced when the plant is crushed. Other common names include hedge garlic, jack-by-the-hedge, and sauce-alone (Georgia 1920). The plant is native to northern Europe, south of 68°N (Tutin et al. 1964), ranging from England to Sweden to the western region of the former USSR, and south to Italy. From this native range, Alliaria has spread to North Africa, India, Sri Lanka (Rai et al. 1972 and Fernaldo 1971, in Cavers et al. 1979) as well as Canada (Cavers et al. 1979) and the United States (Gray et al. 1889; Fernald 1970). The North American range of Alliaria extends from British Columbia (Cavers et al. 1979) to New England (Gleason and Cronquist 1963) and south to Missouri (Yatskievych and Turner 1990). Regional distribution has been mapped for Wisconsin (Patman and litis 1961), North Carolina (Radford et al. 1965), Kansas, Minnesota, North Dakota (Barkley 1977), Canada (Cavers et al. 1979), southern New York (Brooks 1983), Michigan (Voss 1985), Illinois (Nuzzo 1991c), and Ohio (Furlow 1991).
Life history and biology of this European mustard have been investigated by Murley (1951), Trimbur (1973), Lhostka (1975), Cavers et al. (1979), Roberts and Boddrell (1983), Byers and Quinn (1987, 1988), Babonjo et al. (1990) and Kelley et al. (1991) and summarized in Cavers et al. (1979), and Nuzzo (1991a). Alliaria germinates in early spring, overwinters as a basal rosette, flowers the following spring, and produces seed in summer, some 15 months after germinating. Plants produce an average of 350 seeds, and individual plants may produce up to 7,900 seeds (Nuzzo, unpubl.). Seeds have a variable dormancy period, germinating in 8 months in Kentucky (Baskin and Baskin 1992) and in 20 months in Ontario (Cavers et al. 1979).
Alliaria invades shaded communities, habitat essentially unoccupied by other invasive alien herbs. The plant most frequently invades wet to dry-mesic deciduous forests but also occurs in the partial shade characteristic of oak savannas, forest edges, shaded roadsides, in urban areas, and occasionally in full sun, particularly in areas exposed to periodic disturbance.
Alliaria threatens the floristic structure of invaded natural communities (Schwegman 1989), particularly the herbaceous layer. First recognized as a prob-
'er and Lea 1972; : chameleon goby
0 Bay in the same-ely in the form of \igas). This goby, h, doubtless arriv-
to be established ip ballast (Hoese is found in ballast e recent introduc-zx japonicus), also
ed from waters of lished, apparently Teleki 1978). Re-
1 to be established year classes pres-
in coasts (Selgeby an percid, several lallast water as the ankters is a large in Lake Huron in pme an important heavily oh native n native cladocer-; history stages of
(Neogobius mela-ecame established Huron, the proba-ne or both species ilymorplxd) (G. R. efore zebra mussel
ed by ballast water ports before.trans-gly brought aboard ^sterns. Such transit waters, with one ppakoski 1984).
nvasive fishes ant g rate. Those wh<-p claim that ther ted from such n introductions ar >r which there hi traductions; whc 1, even when hab s are found to I bably a correct a
1. Native Landscapes, 1947 Madron Dr., Rockford, IL 61107
137
138
Biological Pollution: The Control and Impact of Invasive Exotic Species
FIGURE 1.—Garlic mustard in Cook County Forest Preserve District, Illinois, (photo by Bill Glass)
lem species in Ontario (Cavers et al. 1979), this European mustard was not considered a threat in the United States until the late 1980s. In a 1986 survey of 25 natural areas botanists throughout Illinois, which requested a rank-ordered list of the 10 most problematic exotic plants in the state, Alliaria was not listed by any of those surveyed (B. N. McKnight pers. comm.)- By 1991 many Midwest states had identified Alliaria as a species of concern, and methods to control the plant in natural communities were under study (Nuzzo 1991a, 1991b).
DISTRIBUTION AND SPREAD
Entry, spread, and distribution of Alliaria were determined using > 1,150 collection records obtained from 77 North American herbaria, supplemented by sight observations made between 1989 and 1991 by field biologists, site stewards, and the author. The resulting distribution map (Fig. 2) depicts within 20-year periods the first record or observation of Alliaria within 15-minute topographic quadrangles (approximately 590 km2 in northern Illinois) in the northcentral
54
Biological Pollution: The Control and Impact of Invasive Exotic Species
Distribution and Spread of Garlic Mustard in North America
139
o
Q.
o
E <
ra
er and Lea 1972; : chameleon goby
0 Bay in the same ely in the form of >igas). This goby, ii, doubtless arriv-
1 to be established lip ballast (Hoese is found in ballast e recent introduc-ixjaponicus), also
;ed from waters of lished, apparently Teleki 1978). Re-1 to be established [ year classes pres-in coasts (Selgeby an percid, several >allast water as the ankters is a large in Lake Huron in pine an important sheavily on native n native cladocer-p history stages of
i (Neogobius mela-Secame established Huron, the proba-ne or both species olymorpha) (G. R. sfore zebra mussel ■
ed by ballast water ports before.trans-gly brought aboard ystems. Such transit waters, with one :ppakoski 1984).
nvasive fishes and ig rate. Those who claim that there ,'ted from such re-| introductions are :Dr which there has [traductions; when ■d, even when habi-|s are found to be bably a correct as-
140 Biological Pollution: The Control and Impact of Invasive Exotic Si'ECiES
U.S. and adjacent Canada. Infrequent collections from western regions were not mapped.
This map is subject to certain limitations. First, distribution maps based on herbarium records represent only the minimum presence of a species and presumably underestimate actual rate of spread and distribution. Second, mapping the first record of an invasive species within equi-sized blocks provides a reliable measure of the rate of invasion through space but not of the extent of invasion within a block (Mack 1985).
Alliaria was first recorded in North America in 1868 on Long Island, New York (Leggett s.n. NYS). Spread was initially gradual, with presence reported in just six quadrangles in the next 21 years (Table 1). Multiple collections made within these quadrangles indicate that the species was locally abundant. Lack of collection outside the six quadrangles implies that the species had not spread far from the initial invasion site and/or that collectors worked only within a limited area.
TABLE 1. First record ot Alliaria petiolata occurrence within 15-minute topographic quadrangles (approximately 590 km2) in the United States and Canada.
	1868-	1890-	1910-	1930-	1950-	1970-	1990-	
	1889	1909	1929	1949	1969	1989	1991	TOTAL
United States	5	16	13	58	158	233	57	540
Canada	1	3	0	8	21	21	0	54
Total	6	19	13	66	179	254	57	594
Cumulative	6	25	38	104	283	537	594	
In the next 20 years (1890-1909) Alliaria was collected from an additional 19 locations, including the St. Lawrence Valley in Canada as well as Idaho, Iowa, and Ohio. By 1929, 61 years after the first collection, Alliaria had been recorded from a total of 38 quadrangles in 11 states and 2 provinces. The rate of spread, or of collection, increased in the following decades. By 1989 Alliaria was recorded from 537 topographic blocks. An additional 57 blocks were recorded in 1990-1991, all but two based on sight observations. The paucity of herbarium collections since 1989 may reflect the tendency for an invasive species to be well-collected in the early stages of invasion and under-collected in the later stages, once it is established in a locale and considered to be an 'undesirable' or 'weedy' species.
By the end of 1991 Alliaria occurred in a minimum of 594 topographic quadrangles in 30 states and 3 provinces. What is not apparent is the extent of infestation within each quadrangle. In Illinois virtually all available habitat in the Chicago region supports Alliaria, while southern counties have very localized infestations. Throughout the state 30 percent of state parks and 31 percent of dedicated nature preserves support populations of this species (Nuzzo 1991c).
Graphing the quadrangle occurrences cumulatively by 30-year periods produces a distinct j curve (Fig. 3), reflecting the exponential spread of this plant. Alliaria spread gradually during the first 60 years (1868-1929), at a rate of approximately 0.6 quadrangle or 366 knr/year. The rate of spread increased substantially beginning in 1930 (3.3 quadrangles or 1,950 knr/year between 1930 and 1949) and jumped dramatically again after 1950 (8.9 quadrangles or 5,280 knr/year). As of 1991 there was no indication of leveling oft. On the basis of the 433 quadrangle occurrences between 1950 and 1989, Alliaria is spreading, or being recorded, at the rate of some 10.8 quadrangles or 6,400 knr/year. It must
Distribution and Spread of Garlic Mustard in North America T-T
141
1850 1880
FIGURE 3.—Alliaria petiolata presence in North America,
1910        1940 1970 YEAR
2000
be emphasized that this figure reflects only presence, and not abundance, of plants within each topographic quadrangle.
The rate of spread of Alliaria is greater than that recorded for purple loosestrife {Lythram salicaria; 645 kmVyear since 1940 [Thompson 1991]) but considerably less than documented for cheat grass (Bromus tectorum; all available habitat in 30 years [Mack 1981]). The rapid increase in Alliaria spread is consistent with increases recorded for other alien species, beginning approximately 40 to 50 years after initial invasion (Lacey 1957; Thompson et al. 1987).
Spread progressed from the northeastern seaboard westward, a pattern typical of many invasive European species. Regional spread of Alliaria has two components; establishment of multiple satellite populations (sensu Auld et al. 1978; Auld and Coote 1980) often separated by great distances, and spread as an advancing front from population centers. Eventually the satellite populations coalesce. Both modes are noticeable in Fig. 2, which depicts the minimum presence of Alliaria as of 1991. Infrequent collections from western North America imply that the species may be a sporadic rather than established component of the regional flora.
Within individual communities Alliaria population size may fluctuate widely from year to year, reflecting both the biennial nature of this plant and the 20-month seed dormancy of northern plants. Across a region Alliaria presence consistently increases through time. In seven northern Illinois forests Alliaria occurred at an average frequency of 24 percent in 1989, 34 percent in 1990, and 46 percent in 1991 (Nuzzo, unpubl.). Alliaria abundance in these same communities, as measured by percent cover, showed a modest but non-significant increase during the same time period. This implies an invasion strategy whereby Alliaria initially
|er and Lea 1972; : chameleon goby o Bay in the same ely in the form of figas). This goby, , doubtless arriv-to be established ip ballast (Hoese is found in ballast e recent introduc-zx japonicus), also
:ed from waters of lished, apparently Teleki-1978). Re-'i to be established
year classes pres-in coasts (Selgeby an percid, several >allast water as the ankters is a large
in Lake Huron in 3me an important heavily on native n native cladocer-
history stages of
(Neogobius mela-ecame established Huron, the proba-ne or both species ilymorpha) (G. R.
fore zebra mussel •
ed by ballast water ports before, trans-*ly brought aboard ^sterns. Such transit waters, with one ippakoski 1984).
nvasive fishes and g rate. Those who n claim that there ted from such re-/ introductions are Dr which there has itroductions; when rj, even when habi-;s are found to be bably a correct as-
142 Biological Pollution: The Control and Impact of Invasive Exotic Species
spreads relatively rapidly through a site at low density and subsequently establishes higher density populations. On a generational basis Alliaria's presence within a community also consistently increases, by an average of 250 percent (Nuzzo, unpubl). The increase is considerably greater at sites subjected to disturbance, where presence of Alliaria increased more than 14-fold between generations.
Alliaria, like many invasive alien herbs, is disturbance-adapted, and both natural and anthropogenic disturbance factors are associated with the rapid invasion rate in natural communities (Nuzzo 1991c). Naturally disturbed habitats such as floodplains and riverbanks and anthropogenically disturbed habitats such as roadsides, heavily used preserves, and urban areas are the primary dispersal corridors for Alliaria. Disturbances create habitat suitable for initial entry, and continued disturbances maintain habitat suitable for expansion. However, once established in a locale, additional disturbance may not be necessary for continued spread of Alliaria {sensu Mack 1985). Exponential expansion of other weed species has been correlated with various anthropogenic disturbances (Lacey 1957; Mack 1981; Forcella and Harvey 1983; Thompson et al. 1987).
HABITAT
Habitat data were compiled from the 705 U.S. collection records that indicated habitat. Alliaria was most frequently collected in forests, along roads, and near rivers; > 75 percent of collections were made near one of these habitats. Eight percent were made in urban areas, 4 percent in arboreta or on campuses, and 3 percent along railroads.
Roads (21%) and rivers (23%) were the primary collection locations indicated on 296 records from northeastern states (Conn., Del., Mass., Md., Maine, N.H., N.J., N.Y., Pa., Vt., and the District of Columbia), followed by forested riverbanks (14%), other forested areas (13%), and urban areas (9%) (Fig. 4). In 146 records from southeastern states (Ky., N.C., Va., and W.Va.) slightly more collections were made on forested riverbanks (21%) than along roads (18%), rivers (16%), or forests (16%). In contrast in the Midwestern states (Iowa, 111., Ind., Kans., Mich., Minn., Mo., Ohio, and Wis.), 30 percent of the 260 collections were made in forests, 14 percent along roads, and 10 percent along rivers or other wet areas. Interestingly, 5 percent of the Midwestern collections were made along railroads.
Regional differences in collection locations have many causes, and cannot be attributed solely to habitat preference of the collected species. However, the decline in river-associated habitat from Northeast to Southeast to Midwest and concomitant increase in non-riverine forest habitat may indicate that Alliaria preferentially invades drier forest communities in the Midwest than in the Northeast. This is supported by the higher presence along railroads, which are generally indicative of drier habitats.
CONTROL
Active management to eliminate Alliaria from natural areas and to limit invasion into new locales may slow the spread of this plant. Generally effective methods include removal of the flowerstalk prior to seed production, dormant-season prescribed fire, and dormant-season herbicide application (Nuzzo 1991a, 1991b). Such management is feasible in isolated forested communities and in regions where Alliaria has low presence. Alliaria is self-compatible (Cavers et al. 1979; Babonjo et al. 1990), and a single plant is sufficient to populate or repopulate a site.
Distribution and Spread of Garlic
Mustard in North Amer.ca
Railroad
Open
Campus, Arboreta
Waste Ground
Urban
Woods River Road
Road and River
Woods and Road
Woods and River
Road
River
Woods
20%
30%
FIGURE 4.—Alliaria petiolata habitat by geographic region, derived from 705 herbarium specimens of Alliaria petiolata collected in the United States between 1868 and 1989. Northeastern states are Conn., Del., tylass., Md., Maine, N.H., N.J., N.Y., Pa., Vt., and the District of Columbia; southeastern states are Ky., N.C., Va., and W.Va.; Midwestern states are III., Ind., Iowa, Kans., Mich., Minn., Mo., Ohio, and Wis. Percentages based on total number of collections within each region.
Annual monitoring for and immediate removal of Alliaria will prevent establishment in individual natural areas. Once the species is well-established in a site, successful removal is unlikely without considerable expenditure of labor and
er and Lea 1972; '.chameleon goby j Bay in the same
ly in the form of \igas). This goby, (i, doubtless arriv-
to be established ip ballast (Hoese s found in ballast e recent introduc-. zx japonicus), also
ed from waters of lished, apparently Teleki 1978). Re-1 to be established
year classes pres-\n coasts (Selgeby an percid, several ballast water as the ankters is a~ large -iffiLak&Hurqn in p^^:a^;ijnpo'rtaht heavily on native n native* cladocer-
histdry stages of
{Neogobius mela-ecame established Huron, the proba-ne or both species jlymorpha) (G. R. efore zebra mussel
ed by ballast water ports before.trans-jgly brought aboard ystems: Such transit waters, with one ppakoski 1984).
nvasive fishes and g rate. Those who in claim that there ted from such re-introductions are br which there has troductions; when (d, even when habi-s are found to be fbably a correct as-
144 Biological Pollution: The Control and Impact of Invasive Exotic Species
money [sensu MacDonald et al. 1989; Coblentz 1990) over an extended period of time (sensu Usher 1988). In regions with multiple infestations, the frequency and abundance of Alliaria in unmanaged habitats limit the effectiveness of single-site management, as seeds are continually imported into the managed site.
Long-term effective control will require significant reduction or elimination of Alliaria populations from both public and private lands throughout the infested region. Auld et al. (1978) theorize that 20 percent of a weed's populations must be eliminated annually over a 20-year period in order to eradicate the species; elimination of 5 percent of the populations will not slow the rate of spread, and at least 15 percent of all infested sites must be managed to effect an immediate decrease in the total numbers of the weed species. Successful control of Alliaria at this level is unlikely with current management techniques and budgets. Development and implementation of biological control agents provide the most likely means to effectively reduce presence of Alliaria in the northcentral U.S. and adjacent Canada.
ACKNOWLEDGMENTS
I would like to thank the curatorial staffs of the 77 contributing herbaria in the United States and Canada for making collection information available and the staffs of the more than 40 additional herbaria who searched for Alliaria collections. Many curators provided additional reference materials for which I am most grateful. I especially appreciate the considerable effort expended by field biologists who reported presence of Alliaria. I would also like to thank Bill Mc-Knight and an anonymous reviewer for their helpful comments on an earlier version of this paper, and Matt Paulson for his meticulous preparation of the distribution map.
REFERENCES
Auld, B. A. and B. G. Coote, 1980. A model of a spreading plant population. Oikos 34:287-92.
Auld, B. A., K. M. Menz and N. M. Monaghan. 1978. Dynamics of weed spread: implications for policies of public control. Protection Ecol. 1:141-48.
Baronjo, F., S. S. Dhillioh and R. C. Anderson. 1990. Floral biology and breeding system of garlic mustard (Alliaria petiolata). Trans. Illinois State Acad. Sei., suppl. to vol. 83:32. (Abstract).
Barkley, T. M. 1977. Atlas of the flora of the Great Plains. Iowa State Univ., Ames.
Baskin, J. M. and C. C. Baskin. 1992. Seed germination biology of the weedy biennial Alliaria petiolata. (Draft manuscript).
Brooks, K. L. 1983. A Catskill flora and economic botany IV. (Part I.) Polypetalae, Chenopodiaceae
through Capparidaceae. Univ. of the State of New York, Albany. Byers, D. L. and J. A. Quinn. 1987. The effect of habitat variation in Alliaria petiolata on life history
characteristics. Amer. J. Bol. 74:647. (Abstract). -. 1988. Plant size as a factor in determining flowering time and reproductive output in Alliaria
petiolata. Amer. J. Bot. 75:71. (Abstract). Cavers, P. B., M. I. Heagy and R. F. Kokron. 1979. The biology of Canadian weeds. 35. Alliaria
petiolata (M. Bieb.) Cavara and Grande. Can. J. PI. Sei. 59:217-29. Coblentz, B. E. 1990. Exotic organisms: a dilemma for conservation biology. Cons. Biol. 4:261-65. Fernald, M. L. 1970. Gray's manual of botany. Van Nostrand Co., New York, N.Y. Fernaldo, L. V. S. 1971. Selection and utilization of different food plants by Pieris brassicae (L.).
Spolia Zeylon Bull. Natl. Mus. Ceylon 32:1 15-27. Forcella, F. and S. J. Harvey. 1983. Relative abundance in an alien weed flora. Oecologia 59:292-
95.
Furlow, J. 1991. unpublished distribution map.
Georgia, A. E. 1920. A manual of weeds. Macmillan Co., New York, N.Y.
Gleason, H. A. and A. Cronquist. 1963. Manual of vascular plants of northeastern United States and adjacent Canada. Van Nostrand Reinhold Co., New York, N.Y.
54
Biological Pollution: The Control and Impact of Invasive Exotic Species
Distribution and Spread of Garlic Mustard in North America
145
Gray, A., S. Watson and J. M. Coulter. 1889. Manual of the botany of the northern United States, including the district cast of the Mississippi and north of North Carolina and Tennessee. American Book Co., New York, N.Y.
Kelley, T., S. Diiillion and R. Anderson. 1990. Aspects of the seed biology of garlic mustard (.4/-liaria petiolata). Trans. Illinois State Acad. Sei., suppl. to vol. 84:33. (Abstract).
Lacey, W. S. 1957. A comparison of the spread of Galinsoga parviflora and G. ciliata in Britain. Pages 109-15 in Progress in the study of the British Flora. J. E. Lousley (ed.). The Botanical Society of the British Isles Conference Report 5, London.
Liiotska, M. 1975. Notes on the ecology of germination of Alliaria petiolata. Folia Geobot. Phytotax. (Praha) 10:179-83.
MacDonald, I. A. W., L. L. Loope, M. B. Usher, et al. 1989. Wildlife conservation and the invasion of nature reserves by introduced species; a global perspective. Pages 215-55 in Biological invasions; a global perspective. J. White (ed.). SCOPE. John Wiley and Sons LTD, New York, N.Y.
Mack, R. N. 1981. Invasion of Bromus tectorum L. into western North America: an ecological chronicle. Agro-Ecosystems 7:145-65.
-. 1985. Invading plants: their potential contribution to population biology. Pages 127-42 in
Studies on plant demography. J. White (ed.). Academic Press, London.
Murley, M. R. 1951. Seeds of the Cruciferae of Northeastern North America. Amer. Midi. Nat. 46:1-65.
Nuzzo, V. A. 1991a. Experimental control of garlic mustard Alliaria petiolata [(Bicb.) Cavara &
Grande] in northern Illinois using fire, herbicide and cutting. Nat. Areas J. 1 1:158-67. -. 1991b. Experimental control of garlic mustard (Alliaria petiolata [Bieb.] Cavara and Grande)
in four natural communities in Illinois. Interim report to the Illinois Department of Energy and
Natural Resources. Native Landscapes. -. 1991c. Current and historic distribution of garlic mustard {Alliaria petiolata) in Illinois. Report to the Illinois Department of Conservation. Native Landscapes. Patman, J. P. and H. H. Iltis. 1961. Preliminary reports on the flora of Wisconsin. No. 44
Cruciferae—mustard family. Wisconsin Acad. Sei., Arts and Letters 50:17-72. Radford, A. E., H. E. Ahles and C. R. Bell. 1965. Atlas of the vascular flora of the Carolinas.
Technical Bulletin No. 165. Univ. of North Carolina, Chapel Hill. Rai, J. N., V. C. Saxena and J. P. Tewari. 1972. New leaf and stem spot diseases of Indian wild
crucifers. Ind. Phytopathol. 25:253-56. Roberts. H. A. and J. E. Boddrell. 1983. Seed survival and periodicity of seedling emergence in
eight species of Cruciferae. Ann. Appl. Biol. 103:301-04. Schwegman, J. 1989. Illinois garlic mustard alert. Illinois Department of Conservation, Division of
Natural Heritage, Springfield. Thompson, D. Q., R. L. Stuckey and E. B. Thompson. 1987. Spread, impact, and control of purple
loosestrife (I.ythrum salicarid) in North American wetlands. U.S. Fish and Wildlife Service, Fish
and Wildlife Research 2. Washington, D.C. Thompson, D. Q. 1991. History of purple loosestrife (Lylhrum salicaria L.) biological control efforts.
Nat. Areas J. 11:148-50.
Trimbur, T. J. 1973. An ecological life history of Alliaria officinalis, a deciduous forest "weed." M.S. Thesis. Ohio State Univ., Columbus.
Tutin, T. G., V. H. Heywood, N. A. Burges, et al. 1964. Flora Europaea, vol. 1. Cambridge University Press.
Usher, M. B. 1988. Biological invasions of nature reserves: a search for generalisations. Biol. Con-serv. 44:119-35.
VoSj. E. G. 1985. Michigan flora part II. Cranbrook Institute of Science and Univ. Michigan Herbarium. Bulletin 55. Ann Arbor.
Yatskievych, G. and J. Turner. 1990. Catalogue of the flora of Missouri. Missouri Bot. Gard., St. Louis.
er and Lea 1972; : chameleon goby :o Bay in the same :ely in the form of %igas). This goby, h, doubtless arriv-i to be established lip ballast (Hoese as found in ballast :e recent introduc-ax japonicus), also
ted from waters of dished, apparently Teleki 1978). Re-i to be established 1 year classes pres-in coasts (Selgeby :an percid, several ballast water as the ilankters is a large i in Lake Huron in [ome ah important I heavily on native in native cladocer-fe history stages of
' (Neogobius mela-ecame established Huron, the probate or both species olymorphd) (G. R. efore zebra mussel
:ed by ballast water ports before, trans-gly brought aboard iystems. Such trans-bt waters, with one eppakoski 1984).
invasive fishes and ig rate. Those who \n claim that there ited from such re-v introductions are or which there has ^traductions; when ;d, even when habi-bs are found to be bably a correct as-