Phragmites australis (Common Reed, Phragmites), appears to have native populations on every continent except Antarctica (Haslam 1972). It has probably been so affected by human transport and habitat disturbance that its original range is unknown (Holm et al. 1977; Cook 1985). As a species, it is native to North America, and has been found in 3,000 year-old peat in CT, and 1,400-1,100 year old woven artifacts in CO (Marks et al. 1994). Phragmites australis is listed here because of its apparently recent appearance (1910-1950) in many parts of the Chesapeake Bay and surrounding region, and a widespread suspicion that a dramatic increase in its abundance and invasiveness on the Atlantic Coast is due to introduction of a new genotype to eastern North America (Marks et al. 1993; Metzler and Rozas 1987; Tucker 1990; Chambers et al. 1999). Recent genetic research has confirmed this suspicion (Saltonstall 2002a).
At least 11 chloroplast DNA haplotypes of P. australis are native and unique to North America. They are now recognized as P. ausralis ssp. americanus (Saltonstall et al. 2004). The form prevailing on the Gulf Coast and Atlantic coast of FL, haplotype I or Med, is shared with South America, southern Europe, Asia, and Africa, but its introduction status is uncertain (Saltonstall 2002a; Lambertini et al. 2006; Guo et al. 2013). At least 5 haplotypes are native to northeastern North America (Saltonstall 2002b), but the invasive form now predominating in this region, haplotype M, appears to be native to Eurasia, and introduced to North America in the 19th century (Saltonstall 2002a). Dry ballast or garden varieties are possible sources for supposed exotic invasive genotypes (Marks et al. 1993). Several cultivated ornamental varieties have been introduced (Metzler and Rozas 1987).
The first collections of P. australis in North America are from the Chesapeake region. An herbarium specimen of this plant from MD, collected by William Krieg before 1698, is in the British Museum (Mountford 1997), although P. australis is not listed among MD specimens collected by Krieg and others in British herbaria by Brown et al. (1987). John Clayton (Clerk of the Court for Gloucester County) collected this plant in VA in the 1730's, [as 'Arundo panicula laxa' in Gronovius' 'Flora Virginica' (Gronovius 1739; Reveal 1983)]. Pursh also found it in southeast VA when he collected there, while headquartered at a plantation in Southampton County (located inland, south of the James River) in 1806 (Pursh 1814; Harvill et al. 1992). However, there are no further reports of this plant from the Chesapeake Bay region until the early 20th century (Kearney 1901; Shreve et al. 1910; Hitchcock and Standley 1919).
The genetic identity of the early (1697-1806) specimens is not known. Two native genotypes have been found in recent collections in Chesapeake Bay, in Allens Fresh MD (along the Wicomico River, a Potomac River tributary), and along the Rappahannock River in VA (Saltonstall 2002b). Records ofP. australis in a 1901 survey of the Norfolk-Virginia Beach area ( Kearney (1901), could represent a native population:' Phragmites communis is not uncommon near the heads of bayous, but rarely makes a dense stand to the exclusion of other species' (Kearney 1901). Most of the early records of P. australis probably represent P. australis ssp. americanus
19th century floras suggest that the distribution of P. australis was centered along the northeastern coast in the early 19th century, with a range expansion southward. Some examples are listed below:
Canada-ME-VA (1814-1848)- Pursh (1814) lists P. australis as 'On the banks of rivers and in large salt-marshes, Canada to Virginia, common. Gray described its distribution as 'Edges of ponds and swamps, common northwards' (Gray 1848). Saltonstall did not examine specimens from this period, but the early populations are presumed to have consisted largely or wholly of native genotypes. In recent (1997-2001) collections from the northeast Atlantic Coast and Great Lakes regions, native haplotypes were found in Nova Scotia, New Brunswick, Ontario, and 2 Chesapeake Bay locations. Altogether 14 specimens represented native haplotypes, versus 105 for the introduced haplotype M (Saltonstall 2002b).
Boston MA and vicinity- Phragmites australis 'resembles a field of standing corn' (Bigelow 1814). Native haplotypes were collected as late as 1926 on Martha's Vineyard. The introduced haplotype M was collected in 1915 in South Boston. Haplotype M is the only form found in recent collections near Boston and on Cape Cod (Saltonstall 2002b).
CT- The latest specimen of a native haplotype was collected in 1935, on Fishers Island NY, off the CT coast in Long Island Sound. Specimens of haplotype M were collected in Madison and New Haven CT in 1875 and 1885 respectively. Haplotype M is the only form found in recent collections (33 samples) in CT (Saltonstall 2002b).
New York City area- Phragmites australis was 'common in the Newark meadows...also in Pennsylvania and Delaware' (Torrey 1823). Haplotype M was the only form found in 5 collections from Long Island and northern NJ (Saltonstall 2002b).
Philadelphia, and adjacent NJ-DE-- Phragmites australis was not mentioned in Barton (1818), but was listed as rare by Darlington (1853).
Near Wilmington DE, P. australis was found on 'Muddy shores of Christiana Creek, near the lighthouse' (Tatnall 1860). Phragmites australis was abundant on dry ballast piles at Philadelphia. The form present there thought to be a distinct and unusual variety (Martindale 1875; Burk 1878). A specimen of haplotype M was collected from ballast grounds at Camden NJ in 1877 (Saltonstall 2002b).
Baltimore MD and Washington DC - Phragmites australis was apparently absent in the 1800's, and was first recorded by Shreve et al. (1910) near Baltimore, and by Hitchcock and Stanley (1919) for DC. However, Herbarium specimens were collected within 40 km of the two cities in 1883-1902 (US National Herbarium collections). One of these early specimens, from 1905, was identified as haplotype M (Saltonstall 2002b). (Details of the Chesapeake invasion are given below).
NC-SC-GA - Phragmites australis was not mentioned by Walter (1788) , Elliott (1824) or Curtis (1867). Hitchcock's (1935) range map shows no records in the southeastern states from VA to GA. The first record from the Carolinas or GA was in 1973 by Stalter (1975) for SC. By 1991, this species had become a dominant species in portions of brackish marshes on the SC coast (Stalter and Baden 1994). Recent specimens from NC and SC were haplotype M (Saltonstall 2002b).
FL and Gulf States - Phragmites australis was present in southern LA (Langlois 1887) and Chapman (1860) gives the habitat and range as 'deep river marshes near the coast, Florida and northward.' Four herbarium (1905-1910) and 9 recent specimens examined by Saltonstall from FL, AL, LA, and TX were the cryptogenic haplotype I. Three specimens from Plaquemines Parish LA, on the Mississippi River, were the introduced haplotype M (Saltonstall 2002b).
Populations of the invasive haplotype M now appear to be scattered across North America, with some occurring on the Pacfic coast. However, occurrences are most frequent in the northeast and Great Lakes regions (Saltonstall 2002a).
Chesapeake Bay records after 1814 are summarized below:
Phragmites australis was found in a survey of the Norfolk-Virginia Beach area by Kearney (1901), not too far from Clayton and Pursh's collecting area. 'Phragmites communis is not uncommon near the heads of bayous, but rarely makes a dense stand to the exclusion of other species' (Kearney 1901).
Phragmites australis was not listed for Baltimore by Aikin (1837), Sollers (1888), or the District of Columbia by Brereton (1831), Potomac Side Naturalists Club (1876), Ward (1881), or additions to Ward's flora (1885-1901). However, herbarium specimens were collected in Anne Arundel County MD in 1883, and in adjacent Chesapeake Beach (Calvert County) in 1905, 1912 and 1919. This reed was also collected along the Bush River (Harford County) in 1902 (US National Herbarium collections). The 1905 Chesapeake Beach specimen was sampled by Saltonstall and identified as haplotype M (Saltonstall 2002b).
The first published report of Phragmites australis from MD was by Shreve et al. (1910), with two specific locations mentioned, (Patapsco River above, and the Nanticoke River, Vienna, Wicomico County, MD) and treated overall as 'frequent' in marshes. The first report from the District of Columbia was by Hitchcock and Standley (1919), along Potomac ('near the steel plant'). Phragmites australis may have been somewhat local in the early 20th century. It was absent from two local floras, one for the vicinity of Williamsburg VA (Erlanson 1924) and another for Worcester Co. MD (Redmond 1932). However, it was reported from another outlying location; Pocomoke Swamp, MD-VA (Beaven and Oosting 1939), and was considered 'Common... as far south as Sussex (DE) and Dorchester (MD) Counties' by Tatnall (1946).
Phragmites australis was reported in 9 of 17 local studies of Chesapeake marsh and shore vegetation conducted from 1966-1994. Interviews with local botanists (Silberhorne 1995, Sipple 1995; Humaira Khan 1996 ) indicate that the abundance of P. australis around Chesapeake Bay has increased drastically in the last 20-30 years. Studies of 6 Chesapeake Bay marshes (4 on the Patuxent River, 2 on the Eastern Shore) using aerial photography and GIS indicated that the area of P. australis colonies increased by 0.4-12.2% per year, with fastest spread in oligohaline marshes, and in the most recently formed colonies (Rice and Stevenson 1996). The local rate of spread of P. australis is correlated with development and shoreline hardening (McCormick et al. 2011)
Twelve of 15 recent specimens tested from the Chesapeake Bay region were the introduced haplotype M (Saltonstall 2002b). In Chesapeake Bay, based on Saltonstall's surveys (Saltonstall 2002a; Saltonstall 2002b; Meadows and Saltonstall 2007) and the historical data presented here, native populations can be assumed to be rare and localized.
History References - Aikin 1837; Barton 1818; Beaven and Oosting 1939; Bigelow 1814; Brereton 1831; Brown et al. 1987; Burk 1878; Chambers et al. 1998; Chapman 1860; Curtis 1867; Darlington 1853; Elliott 1824; Erlanson 1924; Gray 1848; Gronovius 1739; Hitchcock and Standley 1919; Hitchcock 1935; Holm et al. 1977; Langlois 1887; Marks et al. 1994; Martindale 1875; Metzler and Rozas 1987; Mountford 1997; Potomac-Side Naturalists Club 1867; Pursh 1814; Redmond 1932; Reveal 1983; Rice and Stevenson 1996; Saltonstall 2002a; Saltonstall 2002b; Shreve et al. 1910; Sollers 1888; Tatnall 1860; Torrey 1823; Tucker 1990; Walter 1788; Ward 1881.