Spartina Introductions and Consequences in Salt Marshesments with the result of marsh elevation. The...

3

Spartina species, cordgrasses, are powerfulecosystem engineers and grow over a greatrange of elevations in the intertidal zone (seechap. 16 in this volume). Their tall, stiff stems

reduce wave energy and cause sediments car-ried in estuarine waters to precipitate (seechap. 2); the plants then grow into these sedi-ments with the result of marsh elevation. Themotivation for many early Spartina introduc-tions was to stabilize shorelines and terrestrial-ize intertidal lands. Storm and tidal defenses in

1

Spartina Introductions and Consequences in Salt Marshes

arrive, survive, thrive, and sometimes hybridize*

D. R. Strong and D. A. Ayres

Maritime Spartina species define and maintain the shoreline along broad expanses oftemperate coasts where they are native. The large Spartina species grow lower on thetidal plane than other vascular plants; tall, stiff stems reduce waves and currents toprecipitate sediments from turbid estuarine waters. With the right conditions, roots growupward through harvested sediments to elevate the marsh. This engineering can alterthe physical, hydrological, and ecological environments of salt marshes and estuaries.Where native, Spartinas are uniformly valued, mostly for defining and solidifying theshore. The potential to terrestrialize the shore was the rationale of many of the scores ofSpartina introductions. In a time of rising sea levels, these plants are valued as a barrierto the sea in native areas and in China and Europe where they have been cultivated.In contrast, in North America, Australia, Tasmania, and New Zealand, nonnativeSpartinas are seen as a bane both to ecology and to human uses of salt marshes andestuaries. Four of the seven large-scale invasions involved interspecific hybrids betweenintroduced and native Spartinas, or intraspecific hybridization between formerlyallopatric populations. Rapid evolution driven by selection of genotypes particularlyadapted for invasive behavior could be the cause of observed high spread rates of hybridcordgrass. The study of Spartina introductions is a rich mixture of social and basicsciences, with interaction of human values, ecology, and evolution.

3

*“Arrive, Survive, and Thrive” as a rubric for species intro-duction and invasion was coined by Kevin Rice, RichardMack, and Spencer Barrett.

Silliman_ch01.qxd 12/3/08 3:59 PM Page 3

Copyrighted Material

a time of rising sea levels are more recent valuesfor these plants on the low coasts of Europe andChina, as well as where the maritime Spartinaspecies are native. In other places, such asNorth America, Australia, Tasmania, and NewZealand, the potent ecosystem engineering ofnonnative and hybrid Spartinas is seen as abane both to ecology and to human uses of saltmarshes and estuaries.

Many of the scores of Spartina introductionsfor which we have records have been purpose-ful, and a few have come about as inadvertenthitchhikers on other human activities. Most in-troductions failed, some spread little, and a fewhave spread widely by dint of the ability of theseed of these plants to float on the tide. Seed-laden inflorescences can disperse great dis-tances within rafts of wrack, which form in thefall when the aboveground parts of the plantsenesce. The four most extensively introducedspecies are the maritime species S alterniflora,S. patens, S. densiflora, and the allopolyploid ofrecent origin, S. anglica.

Growing near docks, cordgrasses were anabundant and convenient source of cushioningfor ballast. The oldest known introduction, S.densiflora, from South America to the Gulf ofCadiz, Spain, in the sixteenth century (Castilloet al. 2000), may well have been ballast packing.The first records of this practice are from seven-teenth-century England (Ranwell 1967). On theEast Coast of North America, bales of S. patensand S. alterniflora were packed as cushionsamong heavy items in the holds of ships (Civilleet al. 2005). However, the large numbers ofvoyages from the Atlantic to the Pacific from thesixteenth century onward resulted in no knownintroductions of Spartina. All of the Pacific intro-ductions came with the twentieth century, andnone are known to have resulted from ballast.

The first recorded introductions of S. al-terniflora were to France (1803) and England(1816) from North America. In both countries,the invasions led to stands of the plant thatwere used by people. For example, early in thenineteenth century, residents used S. alterniflorato thatch roofs on the Itchen River, England

(Marchant 1967). These invasions initiallythrived then receded. Patches of dieback arenot unusual for S. alterniflora (Mendelssohnand McKee 1988) and S. anglica (Goodman,Braybrooks, and Lambert 1959). S. alternifloraapparently did not spread much beyond its in-troduction sites in Brittany and the southwestof France (Baumel et al. 2003), and it is nowextremely rare in the United Kingdom (Gray,Marshall, and Raybauld 1991). In both coun-tries, early ranges of S. alterniflora overlappedwith S. maritima in multiple places, givinggreat potential for hybridization. Marchant(1967) speculates that these S. alternifloraintroductions were from ships’ ballast fromAmerica. While it is possible that the Europeanintroductions were inadvertent, it is alsoreasonable that European visitors to NorthAmerica purposefully carried the plant home.S. alterniflora is a much taller and more robustplant than the European S. maritima, and theAmerican species had obviously greater poten-tial to produce fodder and fiber. It is also rea-sonable to assume that people who used theplant would encourage its growth and spread.

The contrast between salt marshes along theAtlantic Coast of North America, the center ofdiversity for Spartina, and those where cord-grasses have been introduced illustrates the in-fluences of these plants. In Atlantic NorthAmerican estuaries, there are no invasive cord-grasses, and the natives are distinctly valuablein maintaining the shoreline habitat (Warrenet al. 2002). Only at the fringes of the range ofthe genus—in Europe, the South Atlantic, andthe Pacific—do native and invading Spartinascoexist. In estuaries of the Pacific, Australia,Europe, and China, nonnative Spartinas bringlarge changes (see chap. 16 in this volume).Human introductions of Spartinas have led tohybrids, some of which spread rapidly and arepowerful ecological engineers. Four of theseven largest invasions are by hybrids betweenintroduced and native Spartinas (S. anglica inEurope, China, and Puget Sound, Washington;and S. alterniflora � S. foliosa in San FranciscoBay, California) and another (S. alterniflora to

4 i n v a s i o n s i n n o r t h a m e r i c a n s a l t m a r s h e s



China) was by intraspecific hybrids selected forvigor and fertility. The other two large-scale in-vasions are S. alterniflora in Willapa Bay,Washington, which is not hybridized, and S.densiflora in Humboldt Bay, California, whichcould bear a remnant of hybridization with S.alterniflora (Baumel et al. 2002).

Spartina is a small genus of halophyticspecies in the Chloridoideae, a monophyleticlineage of the Poaceae (Hsiao et al. 1999). Mostspecies are native to the maritime, north tem-perate Atlantic, New World (Mobberley 1956;Baumel et al. 2002). S. alterniflora is native tothe Atlantic shore of South America, and S.densiflora is native to both the Atlantic andPacific of South America. Before S. anglica,only one species was native to Western Europe(S. maritima); S. anglica is an allopolyploid hy-brid species that arose in the nineteenth cen-tury, in Europe, shortly after one of its parents,S. alterniflora, was introduced there from theNew World. S. maritima was the European,male parental species (Ferris, King, and Gray1997). North America is home to most species.Many maritime salt marshes of the Atlanticand Gulf coasts of North America are domi-nated by S. alterniflora, the smooth cordgrass,which ranges from Canada through the Gulf ofMexico and along the Atlantic coast of SouthAmerica. The second-most widespread andabundant North American species is S. patens,the salt hay grass, which ranges from Canadathrough the Gulf and into Mexico and into theCaribbean. Pacific maritime estuaries haveonly two natives, S. foliosa in California fromSan Francisco Bay south through BajaCalifornia, and S. densiflora in the Pacific ofChile and the Atlantic of South America. TheArgentine S. longispica is inferred to be a hybridof the native S. densiflora and S. alterniflorathere (Orensanz et al. 2002). While no evi-dence exists that such species as S. alternifloraand S. densiflora are hybridized, they are hexa-ploids. Research will possibly reveal evidenceof hybridization in their history, such as thathinted at in S. densiflora in Humboldt Bay,California (Baumel et al. 2002). In this review,

we focus on Spartina species that have beenintroduced beyond their native ranges.

SPARTINAS ARE ECOSYSTEM ENGINEERS

Spartinas can be particularly influential to otherspecies especially when they affect marsh eleva-tion of the marsh. In one well-studied marsh inNew England, the accumulation and loss ofmineral sediment and organic matter have main-tained equilibrium of the marsh surface with sealevel for four thousand years (Redfield 1972).Modification and regulation of the environmentthat is mediated biologically and large relative toabiotic influences is termed ecological or ecosys-tem engineering (Jones, Lawton, and Shachak1994, 1997). Although ecological engineering byorganisms is neither a simple nor a single phe-nomenon (Reichman and Seabloom 2002),wide-ranging discussion over the past decade hasverified ecological engineering as widely impor-tant in nature; the concept comprises a multifac-eted set of notions with evolutionary, ecosystem,and community implications as well as shorter-termed autecological ones (Bruno 2000).

Spartina plants are the head engineers intemperate marshes where they occur, just asare mangroves in the tropics. Spartinas epito-mize organisms with powerful reciprocalinfluences between biotic and physical envi-ronmental features and have been called“foundation species” (Pennings and Bertness2001). The ecosystem engineering prowess ofSpartinas comes from their erect, stiff stems,which create drag, dissipate hydrodynamicforces, and reduce wave height and current ve-locity (see fig. 1.1). Stiffer stems are more costlymetabolically than more flexible stems and canbe seen as an adaptation to harvest sediment,which increases plant fitness (Bouma et al.2005). Sediments that are delivered to marshesby the tides, storms, rainfall, and riverine input(Allen and Pye 1992; Neumeier and Ciavola2004) are trapped within the Spartina canopy.Roots grow up through the harvested sediment,elevating the marsh. With these influences,cordgrasses can affect the ecology of the marsh

s p a r t i n a i n t r o d u c t i o n s a n d c o n s e q u e n c e s 5



well above and below where they grow on thetidal plane.

With the appropriate conditions, the sedi-ment harvesting of S. alterniflora can maintainbroad, flat monospecific salt marsh platformsclose to the mean high tide. In a South Carolinaestuary, primary productivity of S. alterniflorawas greatest toward the lower tidal elevation tol-erance of Spartina, at a depth between 40 and60 centimeters below mean high tide (seefig. 1.2). This was where most sediment wastrapped and salinity lowest. Hypoxia limitedproductivity at depths below maximum produc-tivity, while hypersaline pore water owing toevaporation limited growth at higher elevations;stunted, short-form S. alterniflora grew at thesehigher elevations. This processes created apotentially hump-shaped function of primaryproductivity with position along the intertidalgradient (Morris et al. 2002). It is these relation-ships that allow S. alterniflora marshes to main-tain equilibrium with sea-level changes overshort and very long time scales.

The coast of the Mississippi Delta was builton sediments deposited by the river and ele-vated in this manner by S. alterniflora as sea lev-els rose during the Holocene (Redfern 1983).Decreasing sediment supply has led to loss ofcoast areas in the delta (Stockstad 2005). In theNetherlands, a declining sediment budget couldaffect many aspects of salt marshes and S.anglica dynamics in the future (Riese 2005). InSan Francisco Bay, a deficit of sediment willhamper restoration of the massive South BaySalt Ponds. Sediment supply in the south armof the estuary is less than 1 million cubic yards(MCY) per year, while the restoration projectwill require well over 100 MCY (Siegel andBachland 2002). In New Zealand, both intro-duced S. alterniflora and S. anglica accreted somuch sediment as to negatively affect man-groves and salt marshes (Lee and Partridge1983). In Ireland (Hammand and Cooper2002), the United Kingdom (Goss-Custard andMoser 1988), Tasmania (Kriwoken and Hedge2000), and San Francisco Bay (Stralberg et al.


Spartina

Zostera

Stiff strips

Flexible strips

SG

SG

100 1000 10,000 100,000

SF

SFZF

100

Dis

sip

ated

wav

e h

eig

ht,

δh

(m

m/m

)

10

1

FIGURE 1.1 Drag as a function of wave energy for Spartina anglica and Zostera noltii. The stiff culms of S. anglica give ahigher slope of drag with increasing wave energy than the flexible stems of Z. noltii. The high drag lows water movement,which results in sediment deposition. From Bouma et al. 2005, fig. 6.)



2004), nonnative Spartinas and hybrids coverthe soft sediments and exclude the invertebratefood of shorebirds.

RATIONALES, HISTORIES, AND EVOLVINGVALUES IN SPARTINA INTRODUCTIONS

And that old man asked me to think ofUnited States Marines in a Godforsakenswamp. “Their trucks and tanks and how-itzers are wallowing,” he complained, “sink-ing in stinking miasma and ooze.” Heraised a finger and winked at me. But sup-pose, young man, that one Marine had withhim a tiny capsule containing a seed of ice-nine, a new way for the atoms of water tostack and lock, to freeze. If that Marinethrew that Seed into the nearest puddle.

Kurt Vonnegut, Cat’s Cradle

Many Spartina introductions were the prod-uct of high, if vague, hopes to solidify soft mudand increase the economic value of saltmarshes. The “most obvious economic applica-tion of Spartina is to use it for the reclamationand stabilization of muddy foreshores. There isno plant in the world better fitted for this partic-ular purpose” (Oliver 1925, 84). Though muted,early concerns about Spartina introductions an-ticipated those of today: overly optimistic sce-narios, adverse effects on other economic usesof estuaries and salt marshes (e.g., oyster cul-ture and navigation), limited agricultural valueof reclaimed salt marshes, decreased recre-ational and aesthetic value of beaches invadedby the plant, and threats to the ecology and wildlife of salt marshes (Ranwell 1967). “Whetherthe result will in the end be beneficial or to the


0 10 20 30 40 50 60 70

Depth below MHT, D (cm)

80

2800

2400

2000

1600

1200

800

400

Ab

ove

gro

un

d p

rod

uct

ion

(g

. m–2

.yr–1

)

0

Stable region Unstable

FIGURE 1.2 The curved relationship between net productivity of Spartina alterniflora as a function of position on theintertidal plane (depth below mean high tide [MHT]). Highest productivity occurs near the lowest elevations where salinitiesare lowest; at slightly lower elevations, hypoxia reduces productivity so much that S. alterniflora is excluded. At elevationsabove the peak productivity, evapotranspiration increases sediment salinity, which limits growth. Open circles are data fromlow marsh; closed circles, from high marsh. From Morris et al. 2002, fig. 2.



contrary will depend greatly on local conditions.In any case it will be a change worth watchingand studying” (Stapf 1908, 34). Herbicide appli-cations to introduced Spartina were under wayin the United Kingdom by the mid–twentiethcentury (Ranwell and Downing 1960).

EUROPE

Today in Iberian and Mediterranean regions, S.maritima gives desired protection against ero-sion during storms; unlike the more robustspecies and hybrids forming the bulk of thisreview, the small, diffusely growing S. maritimadoes not accrete sediment during calm periods(Neumeier and Ciavola 2004). The distributionof S. maritima may have been influenced byhumans; it “has been known for a long time(since 1629), and is beyond doubt truly indige-nous to Europe” (Stapf 1908, 33). At the sametime, this species is found along the west coastof Africa and in South Africa (Mobberley 1956).It is possible that S. maritima is a SouthernHemisphere (“tropical”) species long ago intro-duced to Europe by early shipping (Marchant1967, fig. 3). A report of Spartina pollen in earlyHolocene sediments of South Africa (Meadowsand Baxter 2001) raises the possibility of frag-mented S. maritima populations strung alongthe Atlantic coasts of southern Europe andAfrica. Such a distribution would mirror thedisjunctive populations of S. alterniflora thatrange along the South American Atlantic coast.

S. patens, first detected in Europe in 1849, wasthought until recently to be native there(Mobberley 1956). It has long been widespreadand harvested for hay and fodder on the Atlanticand the Mediterranean coasts of Europe as well asin its native range (Ainouche, Baumel, andSalmon 2004). Reports of the invasive S. patens inSpain (Javier et al. 2005) do not mention any de-sirable features of the species, such as protectionagainst erosion. Forming dense monocultures, itis now seen as a threat to native high marsh vege-tation in natural areas of the Mediterranean andIberian coasts (Castillo et al. 2005).

S. densiflora, a native of Chile andArgentina, is believed to have been introduced,

either accidentally or purposefully, in the six-teenth century to the southwestern corner ofSpain on the Atlantic. In the Gulf of Cadiz, S.densiflora has spread to eight estuaries and anumber of shoreline habitats including dunes,high marsh, salt pans, and intertidal flats. ANorth African invasive population is presum-ably derived from the Spanish introduction(Castillo et al. 2005).

S. anglica is presumed to have invadedFrance by floating without human aid acrossthe English Channel. It was first detected inFrance in 1906, at Baie des Veys, Normandy(Baumel, Ainouche, and Levasseur 2001). Theinvasion proceeded during the remainder of thetwentieth century from Normandy southwardthrough Brittany mainly without human inter-vention. S. anglica now occupies nearly all suit-able habitats along this shore of France. Theprodigious abilities to accrete sediment of S.anglica can greatly change intertidal elevationsand ecology where it has invaded in France(Guénégou et al. 1991).

The Netherlands and Germany were in-vaded by S. anglica during the 1920s (Gray et al.1991). It grew lower than the native vegetationand trapped volumes of sediment there. Theinfluence of S. anglica is now greatest in areaswith most available sediment, and its influencedecreases to the north where colder winters hin-der the plant (Bakker et al. 2002). In the south-ern Netherlands, S. anglica was important forholding sediments and building elevated landthat is now being restored with native vegeta-tion (de Jonge and de Jong 2002). S. anglicaspread to Ireland in the 1930s and is now seenas a bane to natural areas and conservation. Theintentions are to eradicate it from Ireland, andonly a shortage of funds has prevented attain-ment of this goal (Hammond and Cooper2002). The twentieth century saw a shift fromagriculture to sea defenses as the major humanvalue of salt marshes in both the Netherlandsand the United Kingdom.

Rising sea levels and increasing stormstrength maintain the value of S. anglica inparts of Europe as a protection from coastal




flooding (Bouma et al. 2005). Productivity ofS. anglica in the United Kingdom could in-crease with rising temperatures and higher at-mospheric CO2 concentrations, but otherfactors, such as competition with other marshspecies, complicate this picture (Gray and Mogg2001). In the Netherlands, human influenceson salt marshes have long rivaled, if notexceeded, those of weather, geology, and hydrol-ogy. In the late twentieth century, terrestrializa-tion of the shore has vied for limited finesediments with ecologically and economicallyvaluable intertidal mud flats—a phenomenoncalled “the Wadden Sea squeeze” (Delafontaine,Flemming, and Mai 2000). The sedimentsfrom large rivers that would feed the marshes ofshallow coastal areas are now shunted offshoredown deep shipping channels into the NorthSea (Riese 2005).

Sediment loss via shipping channels is also athreat to coastal marshes of Louisiana in NorthAmerica (Redfern 1983). The dredged, deep-water passage of the Mississippi River carriessediments offshore into the Gulf of Mexico. Sixhundred square kilometers of coastal Louisianasalt marsh have washed into the sea in the lastdecade, and the loss is greater during hurri-canes (Stockstad 2005). Were these sedimentsto be delivered over the fresh- and saltwatermarshes of Louisiana, as before dredging of theMississippi channels, the rate of coastal losswould be much less (Committee on theRestoration and Protection of Coastal Louisiana2006). Vast monospecific stands of S. alterni-flora define this coast, and the prodigioussediment-holding and -elevating capacities ofthis plant would contribute substantially to themaintenance of the land there.

Protection from the sea is the value of S.anglica in the southeastern United Kingdom,where the Earth’s crust is subsiding while sealevel rises. However, multiple values of restora-tion, conservation, mariculture, and sea defensecombine—and even come into conflict—in saltmarsh issues in the United Kingdom. Forexample, the removal of S. anglica in marshrestoration and conservation has led to lawsuits

based on allegations that liberated sedimentharmed nearby oyster culturing (Kirby 1994).Complications for managing S. anglica inEuropean Union countries arise from needs tomeld traditional with modern uses of the shore-line differently among regions (Pethick 2002).Moreover, the human values of salt marshes areevolving rapidly. “Until the last decade, saltmarsh has often been conceived as coastalwasteland with minimal economic value, whichhas led to considerable loss through land recla-mation for use as agriculture, caravan sites, in-dustrial developments and marinas.” (King andLester 1995, 181). In recent years, sea defensesthat include S. anglica have risen to the highestlevels among these multiple values of saltmarshes (King and Lester 1995), but this canconflict with conservation. In parts of the UnitedKingdom, dense monospecific swards of S.anglica replaced mud flats and excluded nativeinvertebrates and the shorebirds that feed onthem (Goss-Custard and Moser 1988; Frid,Chandrasekara, and Davey 1999). Subsidence ofthe southeastern UK coast means a sedimentdeficit in the long run, and even S. anglica can af-ford little protection to rising sea levels underthese conditions.

NEW ZEALAND

Early in the twentieth century, there was un-abashed enthusiasm for the potential of nonna-tive Spartina in New Zealand. “For thousands ofyears tidal salt mud flats the world over havemade entrances to harbours unsightly andtreacherous and have remained as vast areas ofwaste flats. . . . In the past they have provided analmost unconquerable challenge to man. . . .Now such mud can be conquered, and . . . re-claimed to form useful and stable farmlands.This plant which has such an important roleis . . . Spartina townsendii” (Harbord 1949, 507).By 1950, the slow growth of introductions ofS. anglica and S. � townsendii on the NorthIsland of New Zealand shifted attention to S.alterniflora, albeit with a soft counterpoint ofcaution. “Extensive areas of tidal flats roundNew Zealand’s coastline, usually difficult and




costly to develop, have become the subject ofrenewed interest with the introduction of themaritime grass Spartina alterniflora, which willenable many farmers to capitalize on thesenaturally fertile soils. . . . However, farmers arewarned that the adverse influences may notalways be readily apparent” (Blick 1965, 275). Inthe subsequent decade or so, attitudes reversedfrom conquering to preserving salt marshes. “Insome places the problems caused by its spreadare virtually insurmountable. With renewed ap-preciation of estuarine wetlands in their naturalstates, planting of any species of Spartinaaround the coast of New Zealand should not beallowed to take place. Suitable control and eradi-cation measures need to be developed whereSpartina is already present” (Partridge et al.1987, 567). During the last sixteen years, newherbicides and new methods of application haveeradicated all meadows and patches of S. anglicaon the South Island (Miller 2004).

AUSTRALIA AND TASMANIA

The motivations for introduction of S. anglicainto Australia were vague and mostly the productof little more than curiosity among Europeancolonists about what might grow there. Unlikethe United Kingdom, the Netherlands, and NewZealand, where farmland from marshland was afocused objective, Australia had no general engi-neering or agricultural problems to be solved bythe plant (Boston 1981). The successful plantingswere clustered in southwestern Australia, wherethe last recorded planting was made in 1962.

In Tasmania, the objective for S. anglica wasclear: “to stabilize the mudflats so that theywould eventually be above high water level andbecome relatively useful land . . . [and] . . . forcestream flow into the central part of the river,creating a scouring effect and keeping themain channel free of mud” (Wells 1995, 12).Plantings of S. anglica were done from 1930until 1968. Values changed and the negativeresults of the plant came into focus: reductionof areas of soft sediments where shorebirdsforage, harm to native animal communities,large unwanted changes to the appearance of

the shore and beaches, and lack of access to thewater across the S. anglica sward, which hadbeen an open shore (Hedge and Kriwoken2000). S. anglica was also seen as a threat to theTasmanian oyster industry (Hedge andKriwoken 2000). During the early 1990s, S.anglica was seen as undesirable and was re-moved with herbicides and other means in bothAustralia and Tasmania (Wells 1995). In recentyears in Tasmania, opinion has shifted to retain-ing S. anglica as a habitat and food source fornative species that have lost habitat to landclearing and industrial activity (M. Sheehan,personal communication)

CHINA

Shorelines in China have been densely occu-pied for thousands of years, and shoreline dy-namics have long been affected as much byhuman activities as by geology, weather, andhydrology (Li et al. 1991). In the last half of thetwentieth century, several species of Spartinawere introduced. While future coastal sedimentloads will decrease with new dams (Chen et al.2005), recent decades have seen staggeringlylarge volumes of sediment discharged from thevast, densely populated watersheds of theYangtze, Yellow, and Pearl rivers. From two tothree square kilometers of new intertidal landsappeared on these sediments annually (Hanet al. 2000). Some botanically inclined authorshave praised introduced Spartinas in China.S. anglica, introduced in 1963, was planted inabout a hundred locations over 2,700 kilome-ters of coastline and spread to 33,000 hectaresby the 1980s. Accreting sediment at high inter-tidal levels, S. anglica was touted as protectingdikes from erosion during typhoons and con-tributing to the creation of new pastureland (aspart and parcel of polderizing: diking, freshwa-ter flooding, and drainage). Spartina was har-vested for green manure, animal fodder, fishfood, and cellulose for paper and rope (Chung,Zhuo, and Xu 1983, 1993). The taller S. alterni-flora was planted widely in China in 1975 andgrew lower on the tidal gradient than S. anglica.It accreted sediment at a prodigious rate in a




band two hundred kilometers long and ninetykilometers wide by 1997 (Chung et al. 2004;Zhang et al. 2004). Brief accounts of concernthat S. alterniflora displaces native marsh specieshave been published recently (Zhang et al.2004; Xie et al. 2001; An et al. 2004). At leastsome view nonnative Spartina to be distinctlyundesirable in China (Ding et al. 2008).

A different perspective comes from the liter-ature of physical geography, where Spartinacomprises no more than the most minor offootnotes. Coastal areas of China are extremelyvulnerable to sea level rise (Han et al. 2000).Agriculture by diking has been practiced on theshores of the Pearl River Delta since the HanDynasty, beginning around 200 BCE. Age-oldhuman effects on the shoreline accelerated withindustrial development when China opened tothe outside world. Dikes have always been themain means of reclamation of intertidal lands,and building of dikes does not need Spartina.Dikes line almost the entire Pearl River Delta,and any appearance of a tidal flat is immediatelydiked, drained, and flushed of salts. Wetlands,mangroves, and tidal flats have been completelyeliminated. The entire region is on very low,muddy, subsiding ground that is vulnerable tofreshwater floods from inland and typhoonflooding from the sea.

SAN FRANCISCO BAY, CALIFORNIA

The salt marshes of San Francisco Bay are a ca-sualty of 150 years of “reclamation,” in which thevast majority of intertidal and supertidal habitat,both brackish and salty, were diked and drainedfor agriculture, urban development, and indus-try (Williams and Faber 2001). The remaining125 square kilometers, 5 percent of those foundby Euro-Americans in this huge Pacific Coast es-tuary, are the foundation of conservation effortswith shorebirds of the Pacific flyway and of threefederally endangered species (a mammal, a bird,and a plant). These tenuous salt marshes, andbrackish and freshwater wetlands upstream,form a narrow buffer for wildlife and ecosystemservices between San Francisco Bay and sur-rounding human population and agriculture.

Introduced Spartina played virtually no roleuntil 1975, when the U.S. Army Corps ofEngineers planted S. alterniflora, which hy-bridized with the native S. foliosa soon afterward(Faber 2000).

Ironically, native S. foliosa was a poster childfor the destruction of salt marshes. “This speciesis useful in reclaiming salt marshes, and in sev-eral places about San Francisco Bay it has modi-fied the coast line and increased the acreage ofmany farms. The town of Reclamation receivedits name from the fact that in that vicinity manyacres of land have been reclaimed chiefly by theuse of this grass” (Merrill 1902, 6). Actually,Reclamation was not a town but rather a “local-ity” with few inhabitants on San Pablo Bay,upstream on the Sacramento River from SanFrancisco Bay (Durham 1998). A post office wasestablished at Reclamation in 1891 and discon-tinued in 1903 (Salley 1977). S. foliosa is amodest ecosystem engineer, small, diffuselygrowing, and shallow rooted. It is but a minorleague sediment trapper compared to S. anglica,S. alterniflora, and the massive S. alterniflora �

S. foliosa hybrids now spreading through SanFrancisco Bay. Reclamation, California, origi-nally salt- and freshwater marshes and now ex-tensive hay and oat fields, was formed by diking,dredging, and draining the wetlands. WhileS. foliosa probably made little contribution to thereclamation, it is playing a role in restoration ofa small fraction of the shore. Conservation inter-ests opened 340 hectares to the sea during the1990s, and S. foliosa will grow to form a band atthe lowest tidal elevation on at least part of erst-while Reclamation, California (Marcus 2000).

INVASIVE SPARTINAS ON THE PACIFICCOAST OF NORTH AMERICA

Perhaps the first Pacific introduction was S.densiflora to Humboldt Bay, California, specu-lated to have been introduced there as early asthe 1850s from Chile (Spicher and Josselyn1985; Kittleson and Boyd 1997) and identifiedin 1984 (Faber 2000). In 1999, it was found in329 hectares and 94 percent of the salt marshes




of Humboldt Bay (Pickart 2001). S. densiflorawas dominant over a wide tidal range, from lowSalicornia virginica marsh to the species-richhigh marsh (Eicher 1987). It is the object of con-trol by herbicides and cutting, as it threatenstwo rare plant species classified as endangeredunder State of California and federal regula-tions, Menzies’ wallflower, Erysimum menziesii,and the beach Layia, Layia carnosa (Pickart2005). S. densiflora from Humboldt Bay wasplanted at least twice in San Francisco Bay inthe 1970s and 1980s (Faber 2000) and hadspread from Marin County to about fivehectares at three sites by 1999 (Ayres et al.2004). In 2002, a few plants of unknown originwere found in Gray’s Harbor, Washington(Murphy 2004); and in 2005, it was discoveredin Vancouver, Canada (G. Williams, personalcommunication).

The first known Pacific introduction of S.alterniflora is to Willapa Bay, Washington, anestuary forty-four kilometers long and twentykilometers wide just north of the ColumbiaRiver. The plant probably arrived as a hitchhikerin live oyster shipments rather than fromshipping ballast (Civille et al. 2005). The firstEuro-American settlements on Willapa Bay ex-ploited native oysters and began export to SanFrancisco in the 1850s. The oyster industry grewrapidly, but by the 1880s, native oysters were indecline owing to overexploitation in Willapa Bay.For a few years, Atlantic oysters, Crassostrea vir-ginica, were imported from populations culti-vated in San Francisco Bay. If Californiacordgrass, S. foliosa, was introduced in theseshipments, it was never detected and has be-come extinct in Willapa Bay. In 1893, eighty bar-rels of Atlantic oysters were imported to WillapaBay directly from Atlantic marshes on the newtranscontinental railroad. The origins of the oys-ters were from areas in New York Harbor andLong Island, where native S. alterniflora flour-ished. Upon arrival at Willapa Bay, after thenine- to thirteen-day rail journey, the contents ofthe barrels were spread widely in the intertidal.More than three hundred railcars of eighty toone hundred barrels each of oysters were

imported to Willapa Bay from New York be-tween 1893 and 1919. Cordgrass seed and plantscould easily have been placed into the barrelswith the oysters. Most likely, the introductionwas inadvertent. We know of no economic valuefor cordgrasses at that time and place, and theabundant discussion in the press during theearly twentieth century about introductionsfrom the Atlantic to the Pacific Coast of oystersand other species with commercial potentiallacks any mention of Spartina. Whether thelarge colonies of smooth cordgrass that ap-peared in several places in Willapa Bay a fewdecades later were introduced and spread inad-vertently or had some intentional component isnot known. Oystermen, who were the biggestusers of the bay in the early twentieth century,viewed the plant as undesirable (Sayce 1988).

The earliest written record is an anecdotalaccount that implies but does not say that S. al-terniflora was in the bay in 1911 (Sheffer 1945).The earliest photographs of it show large plantsin several widely spaced areas in Willapa Bay(Civille et al. 2005). A Sheffer photograph of1940 (Civille et al. 2005), the earliest known ofsmooth cordgrass in Willapa Bay, shows a plantor group of plants about 42 meters in diameter,covering nearly 1,385 square meters. Aerial pho-tographs from 1945 show similarly large plantsat seven widely separated sites on the bay, sometwenty kilometers apart from one another.Present-day growth rates of S. alterniflora inWillapa Bay suggest that the plants in the pho-tos were about fifty years old. This implies thatS. alterniflora had dispersed, or was spread byhumans, widely in Willapa Bay and thrived verysoon after the earliest oyster trains dumped oys-ters there from the Atlantic, in 1893 (see fig. 1.3).

Maritime Spartina seed disperses long dis-tances quickly without human help by floatingon tides and currents. There is no evidence thatmaritime Spartina has a seed bank, and all re-cruitment comes from seeds less than a yearold. Floating seed was responsible for virtuallythe entire invasion of Willapa Bay, and rhizomefragments have made virtually no contributionto the spread of S. alterniflora there (Civille et al.


Silliman_ch01.qxd 12/4/08 11:19 AM Page 12


2005). New areas have consistently been in-vaded at densities so low that young individualplants grew separated by meters of open inter-tidal mud. These isolated recruits grew rhi-zomatously into isolated circular plants, eachcomposed of a single genet. The lack of otheremergent plants on the open mud makes newcolonies of cordgrass obvious on aerial pho-tographs, reminiscent of bacterial coloniesgrowing on agar. Over several decades, isolatedplants grew rhizomatously into continuousmeadows of S. alterniflora. These meadowscompletely cover the intertidal mud (Davis,Taylor, Civille, et al. 2004).

Isolated plants comprised a very large frac-tion of S. alterniflora in Willapa Bay through thetwentieth century and contributed little to thespread of the invasion because they set verylittle seed (Davis, Taylor, Civille, et al. 2004).

The meadows produced most of the seed andwere responsible for driving the invasion. Theseed set of meadow plants was nearly tenfoldthat of isolated plants, 20 percent compared to2 percent of florets set seed, respectively. While92 percent of meadow plants produced at leastsome seed, only 37 percent of isolated plantsproduced any seed at all. The lower reproduc-tive rate of isolated plants than of those inmeadows was an Allee effect. Because the low-est densities of isolated recruits had high sur-vival, grew rhizomatously, and did set someseed, the Allee effect was weak (Taylor et al.2004). S. alterniflora is self-incompatible, out-breeding, and therefore requires pollen fromanother plant in order to set seed (Daehler1998). The cause of the Allee effect was sparsepollen among the isolated plants. Only in themuch older meadows was the density of the


B C

D

E

F

A1945

B C

D

E

F

L

GG

K

H

M

2000

FIGURE 1.3 The relative abundance of the earliest known colonies of Spartina alterniflora inWillapa Bay, Washington, in 1945 (left) and the growth of this plant by 2000 (right). Note theproximity of the earliest known colonies of S. alterniflora to the approximate locations of oysterbeds at the beginning of the twentieth century (Jensen Spit, North Cove, Kindred Slough, PalixRiver, Nemah River, and Seal Slough).



wind-borne pollen sufficient for substantialseed set in Willapa Bay (Davis, Taylor,Lambrinos, et al. 2004).

The coverage of S. alterniflora in Willapa Baygrew at a remarkably constant 12 percent or soannually over the fifty-five-year history of aerialphotographs. In 2000, nearly a hundred yearsafter the invasion began, the invader coveredabout 1,670 hectares, or 27 percent of the 6,000hectares of the intertidal habitat of Willapa Bay.Without the Allee effect, the invasion wouldhave covered virtually the entire bay long ago(Taylor et al. 2004).

Other small infestations of S. alternifloraappeared in Washington, Oregon, and northernCalifornia during the twentieth century. Duckhunters introduced S. alterniflora to Dike Islandin Padilla Bay, Puget Sound, Washington, be-tween 1940 and 1946 (Riggs and Bulthuis1994). The seed was provided by a nursery inWisconsin, but there is no record of the mar-itime area from which the nursery obtained theseed. The Padilla Bay introduction had spreadto cover about 1.4 hectares by 1979 and about4.9 hectares by 1991. Flowering stems werefirst seen in October 1992. Some of this seedwas viable, and seedlings had appeared aroundDike Island by the late 1990s. Control effortsare greatly reducing S. alterniflora and S. anglicain Padilla Bay (Riggs 2005). In the 1990s,several small infestations were discovered: inGrays Harbor, Washington, twenty-five kilome-ters to the north of Willapa Bay; in ConnerCreek farther north; and in the Copalis River yetfarther to the north (Murphy 2004). These pre-sumably arose from seed that floated northwardin S. alterniflora wrack on currents from WillapaBay, without human help. The environmentalcommunity in Washington is aware of severalother small infestations of S. alterniflora inPuget Sound (Riggs 2005; Murphy 2004), andwe infer that the lack of reporting indicates thatthese either have been eradicated or are notspreading.

In Oregon, S. alterniflora from Georgia onthe Atlantic Coast of North America waspurposefully planted in the 1970s, in the

Siuslaw River at the entry to Coos Bay, Oregon(Frenkel and Boss 1988). The patch had grownto cover about two hectares by 1994, when itwas sprayed with herbicides and dug up. Withno visible growth, it was declared to be eradi-cated in 1997. In 2005, five culms of S. alterni-flora were found on the same site, and anotherpatch of this species was discovered down-stream in Coos Bay, presumably brought thereas rhizome material in dredge spoils from thefirst site (Howard et al. 2007). This indicatespotential for local dispersal by rhizomes. InHumboldt Bay of northern California, a singlelarge patch of S. alterniflora was presumablyeradicated by the efforts of the CaliforniaDepartment of Fish and Game during the1980s or early 1990s (Cohen and Carlton 1995).

The Pacific Coast of North America had twoknown introductions of S. anglica, both recent.That in Puget Sound was planted in 1961, per-haps with seed from England (Hacker et al.2001). During the intervening forty years or so,the infestation spread to more than seventysites over a roughly linear course of about 160kilometers to affect 3,300 hectares and to coversolidly nearly 400 hectares. The San Franciscopopulation of S. anglica was detected in the1980s and by 2004 had grown to only twenty-four individuals covering 360 square meters.The largest plant was eight meters in diameter,and what was inferred to be the oldest plant wassix meters in diameter (Ayres et al. 2004).

In August 2003, S. anglica was discoveredin southwest British Columbia, Canada, in theFraser River estuary, the largest estuary on thePacific Coast of Canada (Williams et al. 2004).Seeds presumably originated from the S. an-glica populations in Puget Sound, which ex-tend as far north as Orcas Island, about twelvemiles south of the Fraser River mouth. ByOctober, mapping had been completed, andmanual removal began. Surveys in Novemberdiscovered another invaded site in BoundaryBay. In 2004, a multiagency committee wasestablished to conduct removal, begin out-reach and education of naturalists, and enlistvolunteer support.




S. patens, salt marsh hay, has been intro-duced to the Pacific Coast of North America, aswell as to Europe and China (described earlier).It was first known in Oregon as three smallpatches in a 1939 aerial photograph of CoxIsland in the Siuslaw River. Fifty years later, ithad grown to ninety large monospecific patches(Frenkel and Boss 1988). In San Francisco Bay,two plants of S. patens were found in 1970 at themouth of Suisun Bay near the town of Benicia,in the Southampton marsh (Ayres et al. 2004).

SPARTINA HYBRIDIZATION

In 1974, the U.S. Army Corps of Engineers con-ducted test plantings of S. foliosa and Salicorniavirginica on unconfined dredge spoils depositedalong Coyote Hills Slough (also known asNew Alameda Creek) near Fremont, California(37�33�58.62� N by 122�07�44.52� W) as part ofa larger study to quantify and ameliorate theimpact of dredging and dredged material dis-posal in San Francisco Bay (U.S. Army Corps ofEngineers 1976). The immediate goal of theseplantings was to determine the feasibility ofusing native marsh species to create a salt marshon dredge spoils confined in a former salt pond(pond 3) adjacent to New Coyote Hills Slough,while the eventual goal was to evaluate whetherthis approach had potential utility as a means toboth dispose of dredge spoil and create new saltmarsh. After determining that native specieswere able to establish from sown and tidallyborne seed, the Corps of Engineers inexplicablyplanted S. alterniflora in pond 3 from seed it hadobtained from an environmental consultingfirm in Maryland (Faber 2000). According topersonal communication in 2005 between JunBando of the University of California, Davis, andL. Hunter-Cario, the firm’s nursery manager, theS. alterniflora seed came from marshes in Maineand Virginia in the 1970s. A genetic survey ofplants growing along Coyote Hills Slough in1994 found equal numbers of S. alterniflora andhybrids between S. alterniflora and S. foliosa, anda single S. foliosa (out of forty-five specimens)(Ayres et al. 1999).

In 1978, the Corps of Engineers planned touse both native and exotic smooth Spartina tocontrol shoreline erosion at Alameda Island,fifteen miles to the north of pond 3 (U.S. ArmyCorps of Engineers 1978). When this popula-tion was genetically surveyed in 1998, only hy-brid and smooth cordgrass was found; thenative species was absent (Ayres et al. 1999).Similarly, the native species was absent from anintroduced population at San Bruno marsh thatcontained equal numbers of hybrid and S.alterniflora plants in 1994.

In recent surveys, hybrid cordgrass is spread-ing rapidly in the San Francisco estuary, whilethe native and nonnative parents are now be-coming rarer (Ayres, Baye, and Strong 2003;Sloop 2005). The hybrid swarm hinders accessto shorelines, blocks flood control channels,overgrows intertidal foraging areas of shore-birds, and without control could lead to the ex-tinction of the native S. foliosa in San FranciscoBay by means of competition and pollen swamp-ing (Ayres et al. 2003). The potential spread ofthese hybrids southward through the range ofS. foliosa, in southern California and BajaCalifornia, carries the risk of global extinction ofS. foliosa (Ayres et al. 2003). A large control pro-gram is trying to eradicate hybrid cordgrassfrom San Francisco Bay (www.spartina.org).

Hybridizations resulting from human intro-ductions loom large in Spartina biogeographyand in the huge influence that nonnative cord-grasses have had in salt marshes around theworld. The history and incidence of Spartina hy-bridizations are incompletely known. For exam-ple, the S. densiflora in Humboldt Bay, California,has genetic features that suggest introgressionfrom S. alterniflora (Baumel et al. 2002). S. �

townsendii is a sterile F1 homoploid hybridspecies that formed in Southampton Water insoutheastern England in the nineteenth century(Gray et al. 1991). The parental species were S. alterniflora, introduced from America, andS. maritima, presumably native to Europe. Thefirst notice of these homoploid hybrids noted wasin Southampton Water, England, in 1870. It wasgiven the name S. � townsendii (Marchant 1967),




while other hybridizations of these parentalspecies occurred in France and are named S.neyrautii (Ainouche et al. 2003). Only inSouthampton Water did subsequent chromoso-mal doubling occur. It gave rise to the fertile,dodecaploid, allotetraploid species, S. anglica,around 1890 (Gray et al. 1991).

S. anglica was spread and dispersed on itsown to salt marshes throughout Britain(Raybould et al. 1991), France (Gray et al. 1991),and beyond. The introduced S. alterniflora isnow extremely narrowly restricted in Britain, atSouthampton Water at the hybridization site(Gray et al. 2001). S. alterniflora maintainssmall, vigorous populations at the three sites inFrance to which it was introduced in the earlytwentieth century (Baumel et al. 2002). Thelack of spread could be due to pollen scarcity asS. alterniflora is largely self-incompatible (Davis,Taylor, Civille, et al. 2004; Davis, Taylor,Lambrinos, et al. 2004).

S. � townsendii is rare today. It was not fullydiscriminated from S. anglica until the 1960s,after many years of exportation from PooleHarbor. The Spartina nursery at Arne had bothforms in the sward in the 1950s. S. � townsendiigrew together with S. anglica in the south andsoutheast of England and on the Isle of Wight(Goodman et al. 1969, fig. 3). While little infor-mation exists on export to most places exceptNew Zealand, sterile S. � townsendii would havebeen outcompeted by fertile S. anglica in mixedswards. East Anglia marshes had both formsthrough the 1930s and 1940s (Gray et al. 1991).

In a world survey, Ranwell (1967) groupedS. � townsendii with S. anglica under the rubricS. townsendii sensu lato (s.l.) and noted that by1870, one or both of these hybrids had beenspread about the United Kingdom. While seedof the allotetraploid S. anglica disperses on thetide, the sterile diploid sets no seed and could bespread only by humans or, rarely, by means offloating vegetative fragments that might breakloose from an eroding bank. Globally, both werespread widely and covered more than twenty-eight thousand hectares around the world whenRanwell (1967) recorded twenty-two successful

introductions and twenty-two failures. Specifi-cally, for 1967, Europe saw thirteen success andno failures; Australia and New Zealand, eightsuccesses and three failures; and Puget Sound,one success, no failures. The rest were apparentfailures: India and Asia, seven; South Africa,two; the Red Sea, one; the Mediterranean, one;the western Atlantic, five; Greenland, one;British Columbia, one; and Hawaii, one(Ranwell 1967, fig. 1). The higher success ratein Europe was matched by wider spread there(in maximum estimated hectares in 1967): theUnited Kingdom, 12,000; France, 8,000;Netherlands, 5,800; Germany, 800; Denmark,500; Ireland, 400; Tasmania and New Zealand,each 40; Australia, 20; and Puget Sound, lessthan 1.

The second known hybridization of S.alterniflora with a native species occurred in SanFrancisco Bay after introduction of this Atlanticnative in 1976 by the U.S. Army Corps ofEngineers (Faber 2000). Most introductions ofS. alterniflora to the Pacific were beyond theranges of native Spartina species and posed nopossibility of hybridization. The northern limitof the only native north temperate species in thePacific, S. foliosa, California cordgrass, isDrake’s Estero, forty kilometers north of SanFrancisco Bay. Introductions north of SanFrancisco Bay, to Washington, Oregon, andnorthern California, were into regions with nonative Spartina. California cordgrass, S. foliosa,was the native parent species of these hybrids.Although not known at the time, many of thehybrids were purposefully spread during the1980s from the earliest site of hybridization,which was in pond 3, adjacent to Coyote HillsSlough (New Alameda Creek) at the southeast-ern end of San Francisco Bay. Seed floating onthe tide spread the hybrids to many othermarshes over eastern and western shores of thesixty kilometers of shoreline in the southernarm of the bay. A few hybrid colonies estab-lished in salt marshes of Marin County, on thenorth side of the Golden Gate. In the earliestpublished record of this invasion (Callaway andJosselyn 1992), cordgrasses assumed to be




S. alterniflora, but probably hybrids, were foundto be competitively superior to, and to have awider tidal range than, the native S. foliosa.

Hybrids were detected in the mid-1990swhen plants were found that contained geneticmaterial unique to each parent (Daehler andStrong 1997). The highly diverse nuclear andcytoplasmic composition together with chromo-some numbers equal to or close to the parents’suggested that these plants comprised a swarmof backcrossing hybrids rather than an allopoly-ploidization, as per S. anglica (Ayres et al. 1999;Anttila et al. 2000). Much of the shoreline and

most creeks flowing into the southern arm ofSan Francisco Bay were invaded by hybrid cord-grass by 2004 (see fig. 1.4).

Native California cordgrass, S. foliosa, isshorter, with shallower roots, and grows lessdensely than vigorous hybrid genotypes (Daehlerand Strong 1997). The vigorous subset of hybridgenotypes is transgressive; it grew larger andproduced more inflorescences, pollen, and seedthan either parent species (Ayres et al. 1999,2003). Transgressively vigorous hybrids are nowfound at the two leading edges of the hybrid inva-sion. One leading edge is in the native marshes


Spartina patensSpartina anglicaSpartina alterniflora/hybridSpartina densifloraCoast line

N

0 12

km

FIGURE 1.4 Distribution of exotic cordgrasses in San Francisco Bay, 2001. From Ayres et al. 2004.



of dense S. foliosa growing above mean sea level.In native marshes, hybrids spread by vegetativecompetitive displacement of the native speciesand by swamping native stigmas with pollen,which leads to hybrid seed. Where large hybridsare growing, most S. foliosa flowers set hybridseed. The other leading edge of the invasion isbelow mean sea level on the vast, open, intertidalmud flats of the bay. While very few recruits ofnative California cordgrass appeared in openareas in the San Francisco estuary in recent years(Sloop 2005), the hybrid swarm increasedtremendously in numbers and coverage in bothnative marshes and previously open mudflats.Both vegetative expansion and multiple episodesof seedling recruitment contribute to the inva-sion by hybrids.

The area of San Francisco Bay covered by hy-brids in 1975 was about two hectares. By 1990,about 650 circular plants expanding vegeta-tively were noted on aerial photos, and we esti-mated this to amount to five hectares. In 1993,the number of such hybrids had increased toone thousand, many had coalesced to formmeadows, and we estimated the cover to benearly ten hectares. In 2001, cover of hybridswas 190 hectares. While exponential growth incover gives a semilog straight line, the semilogplot of the data for hybrid cover from theseyears gives a convex line. This indicates that therate of spread of hybrid cordgrass has increased.The transgressive traits caused by hybridizationundoubtedly contributed greatly to the veryhigh rate of spread in San Francisco Bay.

In China, intraspecific hybridization of mul-tiple S. alterniflora populations may have cre-ated genotypes with a propensity to spreadrapidly. Seeds and cutting of Spartina alterni-flora from Morehead City, North Carolina,Altamaha Estuary, Georgia, and Tampa Bay,Florida, were sent to C. H. Chung at NanjingUniversity in China in 1979 (Chung et al.2004). They were grown in the BotanicalGarden their first year, and their growth wascarefully monitored. There were striking differ-ences among the populations in phenotypiccharacters; for example, the maximum height

of Georgia plants was almost three meters,while the largest Florida plants were half thatheight (An et al. 2004). Differences were alsofound in isozyme banding patterns, indicatinggenetic as well as phenotypic variation amongthe populations. In 1981, rooted cuttings fromthe nursery population were planted into a1,300-square-meter field site at Luouyuan Bay.Ecotypic differences persisted in the field plan-tation. In 1985, a nursery was established in avillage paddy near Chengmengkou using mixedseed collected from the Luouyuan Bay field site.Plants and/or seeds (references don’t saywhich) from the nursery were outplanted intothree field sites for experimental monitoring.

The allopatric S. alterniflora provenancescultured together in China could have readily hy-bridized. They were mixed within a small 1,300-square-meter plot. The potentially hybrid seedwas sown into a common paddy, and then a mix-ture of genotypes was selected, with preferencefor Georgia provenances, in the first large field tri-als. Adding to this genetic farrago, at some point0.5 kilogram of seeds from North Carolina wasintroduced, and one source claims that it was thissource that led to most of the salt marshes incoastal China (An et al. 2004). It is not unlikelythat hybrids and the most vigorous progeny werespread widely. Intraspecific hybridization couldwell have played a role in the rapid spread of S. al-terniflora thorough Chinese marshes.

The high planting densities in the Chineseplanting could have overcome Allee effects(Davis, Taylor, Civille, et al. 2004; Davis, Taylor,Lambrinos, et al. 2004). It is also possible thatthe rapid spread is a product of evolution of self-compatibility, which occurred in the hybrids ofS. foliosa � S. alterniflora in San Francisco Bay(Sloop 2005). No interspecific hybridization isknown for Chinese S. alterniflora, but it is possi-ble that the intraspecific hybridization of thethree North American provenances producedself-compatible genotypes.

Acknowledgments. We thank Daniel Goldstein ofthe University of California, Davis, library for helpwith the history of Reclamation, California; Tjeerd




Bouma and Alan Gray for advice about S. anglicaon the coast of the Netherlands and the UnitedKingdom, respectively; Kevin Rice, Richard Mack,and Spencer Barrett for coining “Arrive, Survive,and Thrive”; and three anonymous reviewers forhelpful suggestions on the manuscript. T. B. C.Shaw provided inspiration.

REFERENCES

Ainouche, M. L., A. Baumel, and A. Salmon. 2004.Spartina anglica C. E. Hubbard: A natural modelsystem for analysing early evolutionary changesthat affect allopolyploid genomes. BiologicalJournal of the Linnean Society 82: 475–484.

Ainouche, M. L., A. Baumel, A. Salmon, and G.Yannic. 2003. Hybridization, polyploidy and spe-ciation in Spartina (Poaceae). New Phytologist 161:165–172.

Allen, J. R. L., and K. Pye. 1992. Coastal salt marshes:Their nature and importance. Pages 1–18 inJ. R. Allen and K. Pye (eds.), Salt Marshes,Morphodynamics, Conservation and EngineeringSignificance. Cambridge: Cambridge UniversityPress.

An, S., C. Zhou, Z. Wang, Z. Deng, Y. Zhi, and L.Chen. 2004. Spartina in China: Introduction, his-tory, current status, and recent research. Paper pre-sented at the Third International Conference onInvasive Spartina, San Francisco, November 8–10.

Anttila, C. K., R. A. King, C. Ferris, D. R. Ayres, andD. R. Strong. 2000. Reciprocal hybrid formationof Spartina in San Francisco Bay. MolecularEcology 9: 765–770.

Ayres, D. A., P. Baye, and D. R. Strong. 2003.Spartina foliosa—A common species on the roadto rarity? Madroño 50: 209–213.

Ayres, D. A., D. Garcia-Rossi, H. G. Davis, and D. R.Strong. 1999. Extent and degree of hybridizationbetween exotic (Spartina alterniflora) and native(S. foliosa) cordgrass (Poaceae) in California, USAdetermined by random amplified polymorphicDNA (RAPDs). Molecular Ecology 8: 1179–1186.

Ayres, D. A., D. L. Smith, K. Zaremba, S. Klohr, andD. R. Strong. 2004. Spread of exotic cordgrassesand hybrids (Spartina sp.) in the tidal marshes ofSan Francisco Bay, California, USA. BiologicalInvasions 6: 221–231.

Bakker, J. P., P. Esselink, K. S. Dijkema, W. E. vanDuin, and D. J. de Jong. 2002. Restoration of saltmarshes in the Netherlands. Hydrobiologia 478:29–51.

Baumel, A., M. L. Ainouche, R. J. Bayer, A. K.Ainouche, and M. T. Misset. 2002. Molecular phy-

logeny of hybridizing species from the genusSpartina Schreb. (Poaceae). Molecular Phylogeneticsand Evolution 22: 303–314.

Baumel, A., M. L. Ainouche, and J. E. Levasseur.2001. Molecular investigations in populations ofSpartina anglica C.E. Hubbard (Poaceae) invadingcoastal Brittany (France). Molecular Ecology 10:1689–1701.

Baumel, A., M. L. Ainouche, M. T. Misset, J. P.Gourret, and R. J. Bayer. 2003. Genetic evidencefor hybridization between the native Spartinamaritima and the introduced Spartina alterniflora(Poaceae) in south-west France: Spartina �

neyrautii re-examined. Plant Systematics andEvolution 237: 87–97.

Blick, C. M. 1965. Tidal flats can be productive. NewZealand Journal of Agriculture 110: 274–276.

Boston, K. G. 1981. The introduction of Spartinatownsendii (sl) to Australia. Unpublished manu-script, Melbourne State College, Melbourne,Victoria, Australia.

Bouma, T. J., M. B. De Vries, E. Low, G. Peralta, C.Tanczos, J. Van de Koppel, and P. M. J. Herman.2005. Trade-offs related to ecosystem engineer-ing: A case study on stiffness of emerging macro-phytes. Ecology 86: 2187–2199.

Bruno, J. F. 2000. Facilitation of cobble beach plantcommunities through habitat modification bySpartina alterniflora. Ecology 81: 1179–1192.

Callaway, J. C., and M. N. Josselyn. 1992. The intro-duction and spread of smooth cordgrass (Spartinaalterniflora) in south San Francisco Bay. Estuaries15: 218–226.

Castillo, J. M., L. Fernandez-Baco, E. M. Castellanos,C. J. Luque, M. E. Figueroa, and A. J. Davy. 2000.Lower limits of Spartina densiflora and S. mar-itima in a Mediterranean salt marsh determinedby different ecophysiological tolerances. Journal ofEcology 88: 801–812.

Castillo, J. M., A. E. Rubio-Casal, S. Redondo, A. A.Alvarez-Lopez, T. Luque, C. Luque, F. J. Nieva,E. M. Castellanos, and M. E. Figueroa. 2005.Short-term responses to salinity of an invasivecordgrass. Biological Invasions 7: 29–35.

Chen, X. Q., E. F. Zhang, H. Q. Mu, and Y. Zong.2005. A preliminary analysis of human impactson sediment discharges from the Yangtze, China,into the sea. Journal of Coastal Research 21:515–521.

Chung, C. H. 1983. Geographical distribution ofSpartina anglica Hubbard, C. E. in China. Bulletinof Marine Science 33: 753–758.

———. 1993. 30 years of ecological engineering withSpartina plantations in China. EcologicalEngineering 2: 261–289.




Chung, C. H., R. Z. Zhuo, and G. W. Xu. 2004.Creation of Spartina plantations for reclaimingDongtai, China, tidal flats and offshore sands.Ecological Engineering 23: 135–150.

Civille, J. C., K. Sayce, S. D. Smith, and D. R. Strong.2005. Reconstructing a century of Spartina alterni-flora invasion with historical records and contem-porary remote sensing. Ecoscience 12: 330–338.

Cohen, A. N., and J. T. Carlton. 1995. NonindigenousAquatic Species in a United States Estuary: A CaseStudy of the Biological Invasions of the San FranciscoBay and Delta. Washington, DC: U.S. Fish andWildlife Service.

Committee on the Restoration and Protection ofCoastal Louisiana, National Research Council.2006. Drawing Louisiana’s New Map: AddressingLand Loss in Coastal Louisiana. Washington, DC:National Academies Press.

Daehler, C. C. 1998. Variation in self-fertility and thereproductive advantage of self-fertility for an in-vading plant (Spartina alterniflora). EvolutionaryEcology 12: 553–568.

Daehler, C. C., and D. R. Strong. 1997. Hybridizationbetween introduced smooth cordgrass (Spartinaalterniflora; Poaceae) and native California cord-grass (S. foliosa) in San Francisco Bay, California,USA. American Journal of Botany 84: 607–611.

Davis, H. G., C. M. Taylor, J. C. Civille, and D. R.Strong. 2004. An Allee effect at the front of aplant invasion: Spartina in a Pacific estuary.Journal of Ecology 92: 321–327.

Davis, H. G., C. M. Taylor, J. G. Lambrinos, and D. R.Strong. 2004. Pollen limitation causes an Alleeeffect in a wind-pollinated invasive grass(Spartina alterniflora). Proceedings of the NationalAcademy of Sciences of the USA 101: 13804–13807.

de Jonge, V. N., and D. J. de Jong. 2002. Ecologicalrestoration in coastal areas in the Netherlands:Concepts, dilemmas and some examples.Hydrobiologia 478: 7–28.

Delafontaine, M. T., B. W. Flemming, and S. Mai.2000. The Wadden Sea squeeze as a cause ofdecreasing sedimentary organic loading. Pages273–286 in B. W. Flemming, M. T. Delafontaine,and G. Liebezeit (eds.), Muddy Coast Dynamicsand Resource Management. Amsterdam: Elsevier.

Ding, J., R. N. Mack, P. Lu, M. Ren, and H. Huang.2008. China’s booming economy is sparking andaccelerating biological invasions. BioScience 58,no. 4: 317–324.

Durham, D. L. 1998. California’s Geographic PlaceNames: A Gazetteer of Historical and ModernNames of the State. Clovis, CA: Word Dancer.

Eicher, A. L. 1987. Salt marsh vascular plant distribu-tion in relation to tidal elevation in Humboldt Bay,

California. Unpublished MA thesis, HumboldtState University, Arcata, CA.

Faber, P. 2000. Grass wars: Good intentions goneawry: Why would anyone bring an alien cordgrassinto San Francisco Bay? Coast and Ocean Spring2000: 9–20.

Ferris, C., R. A. King, and A. J. Gray. 1997. Molecularevidence for the maternal parentage in the hybridorigin of Spartina anglica C.E. Hubbard. MolecularEcology 6: 185–187.

Frenkel, R. E., and T. R. Boss. 1988. Introduction, estab-lishment and spread of Spartina patens on CoxIsland, Siuslaw Estuary, Oregon. Wetlands 8: 33–49.

Frid, C. L. J., W. U. Chandrasekara, and P. Davey.1999. The restoration of mud flats invaded bycommon cordgrass (Spartina anglica, CE Hubbard)using mechanical disturbance and its effects onthe macrobenthic fauna. Aquatic Conservation—Marine and Freshwater Ecosystems 9: 47–61.

Goodman, P. J., E. M. Braybrooks, and J. M. Lambert.1959. Investigations into “die-back” in Spartinatownsendii AGG: I. The present status of Spartinatownsendii in Britain. Journal of Ecology 47:651–677.

Goodman, P. J., E. M. Braybrooks, C. J. Marchant, andJ. M. Lambert. 1969. Spartina � townsendii H. & J.Groves sensu lato. Journal of Ecology 57: 298–313.

Goss-Custard, J. D., and M. E. Moser. 1988. Rates ofchange in the numbers of dunlin, Calidris alpina,wintering in British estuaries in relation to thespread of Spartina anglica. Journal of AppliedEcology 25: 95–109.

Gray, A. J., D. F. Marshall, and A. F. Raybould. 1991.A century of evolution in Spartina anglica.Advances in Ecological Research 21: 1–62.

Gray, A. J., and R. J. Mogg. 2001. Climate impacts onpioneer saltmarsh plants. Climate Research 18:105–112.

Guénégou, M. C., J. E. Levasseur, C. Bonnot-Courtis,L. R. Lafond, and J. Le Rhun, J. 1991. The ecomor-phological and botanical changes in Kernic Bay(Brittany, France): Influence on coastal manage-ment. Journal of Coastal Research 7: 331–339.

Hacker, S. D., D. Heimer, E. Helquivst, T. G. Reeder,B. Reeves, T. J. Riordan, and M. N. Dethier. 2001.A marine plant (Spartina anglica) invades widelyvarying habitats: Potential mechanisms of inva-sion and control. Biological Invasions 3: 211–217.

Hammond, M. E. R., and A. Cooper. 2002. Spartinaanglica eradication and inter-tidal recovery inNorthern Ireland estuaries. Pages 124–131 in D.Veitch and M. Clout (eds.), Turning the Tide: TheEradication of Invasive Species. Gland, Switzerland,and Cambridge: Invasive Species Specialist Group,International Union for Conservation of Nature.




Han, M. K., L. Wu, Y. F. Liu, and N. Mimura. 2000.Impacts of seal-level rise and human activities onthe evolution of the Pearl River delta, SouthChina. Pages 237–246 in B. W. Flemming, M. T.Delafontaine, and G. Liebezeit (eds.), MuddyCoast Dynamics and Resource Management.Amsterdam: Elsevier.

Harbord, W. L. 1949. Spartina townsendii: A valuablegrass on tidal mud flats. New Zealand Journal ofAgriculture 78: 507–508.

Hedge, P., and L. K. Kriwoken. 2000. Evidence foreffects of Spartina anglica invasion on benthicmacrofauna in Little Swanport estuary, Tasmania.Austral Ecology 25: 150–159.

Howard, V., M. Pfauth, M. Sytsma, and D. Isaacson.2007. Oregon Spartina Response Plan. Portland:Oregon Department of Agriculture. http://www.clr.pdx.edu/docs/OSRP.pdf.

Hsiao, C., S. W. L. Jacobs, N. J. Chaterton, andK. H. Saya. 1999. A molecular phylogeny of grassfamily (Poaceae) based on the sequences of nu-clear ribosomal DNA (ITS). Australian SystematicBotany 11: 667–668.

Javier, F., J. Nieva, E. M. Castellanos, J. M. Castillo,and M. E. Figueroa. 2005. Clonal growth and tillerdemography of the invader cordgrass Spartinadensiflora brongn. at two contrasting habitats inSW European salt marshes. Wetlands 25: 122–129.

Jones, C. G., J. H. Lawton, and M. Shachak. 1994.Organisms as ecosystem engineers. Oikos 69:373–386.

———. 1997. Positive and negative effects of organ-isms as physical ecosystem engineers. Ecology 78:1946–1957.

King, S. E., and J. N. Lester. 1995. The value of salt-marsh as a sea defense. Marine Pollution Bulletin30: 180–189.

Kirby, R. 1994. Sediments 2-oysters-0: The case-histories of 2 legal disputes involving fine sedi-ment and oysters. Journal of Coastal Research 10:466–487.

Kittelson, P. M., and M. J. Boyd. 1997. Mechanismsof expansion for an introduced species of cord-grass, Spartina densiflora, in Humboldt Bay,California. Estuaries 20: 770–778.

Kriwoken, L. K., and P. Hedge. 2000. Exotic speciesand estuaries: Managing Spartina anglica inTasmania, Australia. Ocean and CoastalManagement 43: 573–584.

Lee, W. G., and T. R. Partridge. 1983. Rates of spreadof Spartina anglica and sediment accretion in thenew river estuary Invercargill New Zealand. NewZealand Journal of Botany 21: 231–236.

Li, P. R., P. N. Quio, H. H. Zheng, G. X. Fang, andG. Q. Huang. 1991. The Environmental Evolution

of the Zhujiang (Pearl River) Delta in the Last10,000 Years. Beijing: Ocean.

Marchant, C. J. 1967. Evolution in Spartina(Gramineae). I. History and morphology of thegenus in Britain. Botanical Journal of the LinneanSociety 60: 1–24.

Marcus, L. 2000. Restoring tidal wetlands at SonomaBaylands, San Francisco Bay, California. EcologicalEngineering 15: 373–383.

Meadows, M. E., and A. J. Baxter. 2001. Holocenevegetation history and palaeoenvironments atKlaarfontein Springs, Western Cape, SouthAfrica. Holocene 11: 699–706.

Mendelssohn, I. A., and K. L. McKee. 1988. Spartinaalterniflora die-back in Louisiana: Time- course in-vestigation of soil waterlogging effects. Journal ofEcology 76: 509–521.

Merrill, E. D. 1902. The North American species ofSpartina. Bulletin of the Bureau of Plains Indians,U.S. Department of Agriculture 9: 5–15.

Miller, G. 2004. Controlling invasive Spartina spp.,the New Zealand success story. Paper presented atthe Third International Conference on InvasiveSpartina, San Francisco, November 8–10.

Mobberley, D. G. 1956. Taxonomy and distribution ofthe genus Spartina. Iowa State College Journal ofScience 30: 471–574.

Morris, J. T., P. V. Sundareshwar, C. T. Nietch,B. Kjerfve, and D. R. Cahoon. 2002. Responses ofcoastal wetlands to rising sea level. Ecology 83:2869–2877.

Murphy, K. C. 2004. Progress of the 2004 SpartinaEradication Program: Report to the Legislature. AGRPub. 850-132(N/1/05). Olympia: Washington StateDepartment of Agriculture Spartina Program.

Neumeier, U., and P. Ciavola. 2004. Flow resistanceand associated sedimentary processes in aSpartina maritima salt-marsh. Journal of CoastalResearch 20: 435–447.

Oliver, F. W. 1925. Spartina townsendii: Its mode ofestablishment, economic uses, and taxonomicstatus. Journal of Ecology 13: 74–91.

Orensanz, J. M. L., S. E., G. Pastorino, B. A., G. Casas,G. Darrigran, R. Elias, J. J. Lopez, S. O. Gappa,M. Pascual, P. Penchaszadeh, M. Luz Piriz, M.Scarabino, E. D. Spivak, and E. A. Vallarino. 2002.No longer the pristine confines of the world ocean:A survey of exotic marine species in the south-western Atlantic. Biological Invasions 4: 115–143.

Partridge, T. R. 1987. Spartina in New Zealand. NewZealand Journal of Botany 25: 567–575.

Pennings, S. C., and M. D. Bertness. 2001. Salt marshcommunities. Pages 289–316 in M. D. Bertness,S. D. Gaines, and M. E. Hay (eds.), MarineCommunity Ecology. Sunderland, MA: Sinauer.




Pethick, J. 2002. Estuarine and tidal wetland restora-tion in the United Kingdom: Policy versus prac-tice. Restoration Ecology 10: 431–437.

Pickart, A. 2001. The distribution of Spartina densi-flora and two rare plant species in Humboldt Bay,California 1998–1999. Arcata, CA: U.S. Fish andWildlife Service.

———. 2005. Control of Invasive Spartina densiflorain a High-Elevation Salt Marsh, Mad River Slough,Humboldt Bay National Wildlife Refuge. Arcata,CA: U.S. Fish and Wildlife Service.

Ranwell, D. S. 1967. World resources of Spartinatownsendii (sensu lato) and economic use ofSpartina marshland. Journal of Applied Ecology 4:239–256.

Ranwell, D. S., and B. M. Downing. 1960. The use ofDalapon and substituted urea herbicides for con-trol of seed-bearing Spartina (cord-grass) in inter-tidal zones of estuarine marsh. Weeds 8: 78–88.

Raybould, A. F., A. J. Gray, M. J. Lawrence, and D. F.Marshall. 1991. The evolution of Spartina anglicaHubbard, C.E. (Gramineae)—Genetic-variation andstatus of the parental species in Britain. BiologicalJournal of the Linnean Society 44: 369–380.

Redfern, R. 1983. The Making of a Continent. Toronto:Fitzhenry & Whiteside.

Redfield, A. C. 1972. Development of a New Englandsalt marsh. Ecological Monographs 42: 201–237.

Reichman, O. J., and E. W. Seabloom. 2002. The roleof pocket gophers as subterranean ecosystem en-gineers. Trends in Ecology and Evolution 17:44–49.

Riese, K. 2005. Coast of change: Habitat loss andtransformations in the Wadden Sea. HelgolandMarine Research 59: 9–21.

Riggs, S. 2005. Survey and Control of Spartina inPadilla Bay in 2005. Mount Vernon: WashingtonState Department of Ecology, Shorelands andEnvironmental Assistance Program.

Riggs, S., and D. A. Bulthuis. 1994. Estimated Net AerialPrimary Productivity and Monitoring of SelectedCharacteristics of Spartina alterniflora in Padilla Bay,WA, April 1992–May 1993. 94-176, Technical Reportno. 11. Mount Vernon: Washington State Depart-ment of Ecology, Shorelands and EnvironmentalAssistance Program.

Salley, H. E. 1977. History of California Post Offices,1849–1976. La Mesa, CA: Postal History Associates.

Sayce, K. 1988. Introduced Cordgrass, Spartina alterni-flora Loisel, in Saltmarshes and Tidelands of WillapaBay, WA. USFWS-87058 (TS). Ilwaco, WA:Willapa National Wildlife Refuge, U.S. Fish andWildlife Service.

Siegel, S. W., and P. A. M. Bachland, eds. 2002.Feasibility Analysis: South Bay Salt Pond RestorationSan Francisco Estuary, California. San Rafael:California Wetlands and Water Resources.

Sloop, C. M. 2005. Reproductive and recruitmentdynamics of invasive hybrid cordgrasses (S.alterniflora � S. foliosa) in San Francisco Bay tidalflats. Unpublished PhD diss., University ofCalifornia, Davis.

Spicher, D., and M. Josselyn. 1985. Spartina(Gramineae) in northern California: Distributionand taxonomic notes. Madroño 32: 158–167.

Stapf, O. 1908. Spartina townsendii. Gardener’sChronicle 43: 33–35.

Stockstad, E. 2005. Louisiana’s wetlands struggle forsurvival. Science 310: 1264–1266.

Stralberg, D., V. Toniolo, G. W. Page, and L. E.Stenzel. 2004. Potential Impacts of Non-nativeSpartina Spread on Shorebird Populations in SouthSan Francisco Bay. Stinson Beach, CA: PointReyes Bird Observatory.

Taylor, C. M., H. G. Davis, J. C. Civille, F. S. Grevstad,and A. Hastings. 2004. Consequences of an Alleeeffect in the invasion of a pacific estuary bySpartina alterniflora. Ecology 85: 3254–3266.

U.S. Army Corps of Engineers. 1976. Dredge DisposalStudy San Francisco Bay and Estuary: Appendix K,Marshland Development. San Francisco: Author.

———. 1978. Shoreline Erosion Control DemonstrationProgram, Alameda, California: PreconstructionReport. San Francisco: Author.

Warren, R. S., P. E. Fell, R. Rozsa, A. H. Brawley, A. C.Orsted, E. T. Olson, V. Swamy, and W. A. Niering.2002. Salt marsh restoration in Connecticut:20 years of science and management. RestorationEcology 10: 497–513.

Wells A. 1995. Rice grass in Tasmania: An overview.Pages 11–13 in J. E. Rash, R. C. Williamson, andS. J. Taylor (eds.), Proceedings of the AustralasianConference on Spartina Control. Melbourne:Victorian Government Publications.

Williams, G., J. Baumann, D. Buffett, and et al. 2004.Discovery and management of Spartina anglica inthe Fraser River estuary, British Colombia, Canada.Paper presented at the Third InternationalConference on Invasive Spartina, November 8–10,San Francisco.

Williams, P., and P. Faber. 2001. Salt marsh restora-tion experience in San Francisco Bay. Journal ofCoastal Research 27: 203–211.

Xie, Y., Z. Y. Li, W. P. Gregg, and L. Dianmo. 2001.Invasive species in China: An overview. Biodiversityand Conservation 10: 1317–1341.

Zhang, R. S., Y. M. Shen, L. Y. Lu, S. G. Yan, Y. H.Wang, J. L. Li, and Z. L. Zhang. 2004. Formationof Spartina alterniflora salt marshes on the coastof Jiangsu Province, China. Ecological Engineering23: 95–105.




Spartina Introductions and Consequences in Salt Marshesments with the result of marsh elevation. The...

Documents

Transcript of Spartina Introductions and Consequences in Salt Marshesments with the result of marsh elevation. The...