Allium burdickii in the southern Appalachians:  

Does it exist and how rare is it?

Wild ramps or leeks, Allium tricoccum, has been recognized since the late 1700’s (Aiton).  Dr. J. H. Burdick first noted a variation in Allium tricoccum in the 1870’s from some specimens he sent to Asa Gray at the Harvard herbarium (Jones 1979, Swink & Wilhelm 1994).  This separation was not formally recognized as two varieties, Allium tricoccum Ait. var. tricoccum and var. burdickii Hanes, until the mid 1900’s  (Hanes 1953).  By the late 70’s, Jones recognized two discrete species.   

Allium tricoccum is distinguished from A. burdickii by the presence of wider leaves, bigger bulbs and more flowers per individual umbel (see Table 1).  Most A. tricoccum individuals have a red-pigmented scape and petiole. A. tricoccum tends to  form more extensive colonies then A. burdickii populations, which occur in scattered small clumps of 3-12 individuals.   Jones found that these species bloom at different times, which serve as a phenological barrier to hybridization.  Jones suggests that both species are essentially self-incompatible.  

Table 1: Characters separating Allium tricoccum and Allium burdickii.

Electrophetic work completed on six populations near the northern limit of wild ramp distribution was unable to detect a difference between the two species  (Vasseur et al. 1990).  In the northern region, both species are autogamous and flower in synchrony (Vasseur et al. 1990), which contrasts with Jones (1979) findings.  Differences between these species in the Southern Appalachians are not well known.

A.  tricoccum and A . burdickii have similar ranges, occurring throughout the eastern United States with population concentrations in the Midwest.  Both are considered secure globally.  A.  tricoccum has a G5 rank, while  A. burdickii has a G4G5 rank.  Range maps follow for each plus a state rank map for A. burdickii.   It should be noted that the state rank map by the Association for Biodiversity Information (offshoot of The Nature Conservancy) indicates a more tenuous and unknown distribution for A. burdickii compared to Kartesz’s map.     

Figure 1:  Allium tricoccum state Range Map (yellow highlight denotes formally tracked by the state, light green denotes presence in the state) Kartesz  1999. 

        Figure 2: Allium burdickii state Range Map (yellow highlight denotes formally tracked by the state, light green denotes presence in the state, red denotes extirpated occurrences within the state) Kartesz  1999.   

Figure 3: Allium burdickii state range map by S rank (red is S1—critically imperiled, orange is S2—imperiled, yellow is S3--vulnerable, grey is SU--unrankable, dark blue is S?--unranked, light blue is SR—reported) Natureserve 2001. 

 

Distinction between the two ramp species has been accepted in most recent floras or checklists  (Swink & Wilhelm 1994, Weakley 2000, Kartesz 1999, Wofford 1989, Yatskievych & Turner 1990, Voss 1985, Owenby & Morley 1991, Brown & Athey 1992, Medley 1993, Mohlenbrock 1986), although not all (Gleason & Cronquist 1991, Hanes & Vining 1998).   Of course, even the variety separation was not universally accepted as evidenced by the lack of recognition in some local  (Braun 1969, Weishaupt 1960, Strasbaugh & Core 1978, Clarkson 1966) and regional floras (Seymour 1969, Britton & Brown 1952).  It appears A. burdickii individuals in the Midwest are easier to separate from A. tricoccum individuals while at the edge of their range to the east and south, a clear distinction is indeed nebulous.   For instance, a recent flora completed in Maine does not indicate the presence of A. burdickii (Haines & Vining 1998).   Jones, on the other hand, placed A. burdickii within Maine based on herbarium specimens she examined from there (1979).   

 Both ramp species occur within the southern Appalachians at the southern edge of their range.  A.  tricoccum has a slightly more southern extent with occurrences in northern Georgia and northern Alabama (Duncan & Kartesz 1981, Mellinger 1984, Jones 1979).  A. tricoccum has been documented for centuries and is not formally tracked by any heritage program in the southern Appalachians.   The status of A. burdickii is less certain.    In Tennessee, A. burdickii has a S3 status, apparently secure, although it is tracked as a threatened and commercially exploited plant.  The Tennessee map below documents the odd county distribution with apparent occurrences in at least 3 physiographic regions (Chester et. al. 1993).  It has been documented in 3 counties in the Blue Ridge (Sullivan, Carter, and Sevier), in 4 counties in north and south Middle Tennessee (Sumner, Trousdale in north plus Lincoln & Coffee to the south), and 1 county in west Tennessee (Hardeman).  

Figure 4:  Allium burdickii distribution in Tennessee (shaded county indicates presence) Chester et. al.  1993.

The Virginia Natural Heritage Program does not formally track A. burdickii as a rare species, but does keep it on a watch list and ranks it as SU, denoting a species with an unknown state range.  Fred Huber, botanist with the George Washington and Jefferson National Forest, indicates little is known on the species distribution in USFS lands in Virginia (pers. comm.)   Almut Jones also indicates an uncertainty on the distribution in Virginia (Jones 1979).   Her geographical distribution map has open circles for the 2 site locations in the state.  These open circles indicate sites with collections doubtfully identified as A. burdickii.  

The Kentucky Natural Heritage Commission does not formally track A. burdickii on any rare list (Kartesz 1999).   The following figure indicates the distribution for this species in Kentucky (USDA Plant Source 2001).  A. burdickii is primarily distributed in the Bluegrass region and scattered in the Cumberland Plateau (Brown & Athey 1992).   According to both David Taylor, Daniel Boone National Forest botanist, and Julian Campbell, Kentucky Nature Conservancy botanist, it is uncertain how abundant it is in the state and if it is worthy of conservation status. 

Figure 5:  Allium burdickii distribution in Virginia (shaded county indicates presence). USDA PLANTS national database 2001.   

Little information is available on the distribution of A. burdickii in West Virginia.   Its presence is not listed either at the species or the variety level in the vascular flora of the Monongahela National Forest (Clarkson 1966) nor in the state (Strasbaugh & Core 1978).  The taxon is not tracked by the WVa Natural Heritage Program (Jan Garrett, Monongahela National Forest botanist, pers. comm.).  

The North Carolina Natural Heritage Program tracks A. burdickii. (Amoroso 1999).  It has a S1 rank with a significantly rare status.  There are only 4 documented occurrences of A.  burdickii within the NC Biological Conservation Database (BCD).  All four of these sites occur in either Ashe or Mitchell Counties.  In addition, 14 occurrences have been recorded during the past six years at sites with permanent vegetation plots measured across western North Carolina (Newell & Peet 1996 &1997, Ulrey 1999).   All of these plots were originally visited during anthesis with species determination exclusively based on flower or fruit numbers per umbel.   Five of these sites were visited this past year during the spring ramp season.  Based on leaf shape, color and size, all five of the populations showed stronger characteristics indicative of A. tricoccum (pers. obs.).   It should be noted that the 4 documented populations within the NC BCD all where determined during the flowering season.   Thus it is uncertain how extensive this species is within North Carolina. 

Except for the Nantahala and Pisgah National Forests, no USFS office officially tracks this as a rare species within the southeast.  It has a forest concern status within the USFS lands in NC (USFS 1998).  This status requires the National Forests in NC to maintain the viability of this species across USFS lands in NC.   Field surveys are completed when a proposed project occurs in likely habitat.  The species has been located twice in the Pisgah National Forest during these surveys.  These 2 determinations were made at the onset of flowering. 

Clearly, a status review is warranted for A. burdickii in the southern Appalachians and possibly throughout the eastern edge of its range.   Most of the current range extension for this species was extrapolated from the extensive herbarium study by Jones.   The current range map does not reflect on the rarity of the species throughout a large portion of its range.  A complete status review for this species has probably only been completed for Illinois (Dr. Almut Jones, pers. comm.).   

Ecological separation based on Southern Appalachian Vegetation Survey plots

The Southern Appalachian dataset is the culmination of over 20 years of fieldwork by various ecologists and includes detailed information from over 2000 locations (Ulrey 1999, Patterson et. al 1999, Newel & Peet 1995,1996, 1997).  It is a compilation of 18 datasets collected almost exclusively from western North Carolina.   A. tricoccum was recorded in 54 plots and A. burdickii in 17 plots from the Southern Appalachian dataset.  These ramp subsets were analyzed to determine any ecological separation between the two species within the southern Appalachians.   It should be mentioned that since this data is compiled from various datasets (9 separate studies) the distinction between the 2 Allium species might have not been apparent to all the researchers. 

337 different species were recorded in the Allium tricoccum plots.    The vast majority of these species, 241, occurred in less than 20% of the plots.  Ninety-two of these species only occurred in one plot.   These plots averaged 52 species.   Species diversity was highly variable among the 54 plots, varying from a high of 92 total species to a low of 24 species.   Aesculus flava, Betula alleghaniensis, Acer saccharum, Fagus grandifolia, and Tilia americana were the species with the highest frequency occupying the canopy layer (Table 2).   Of these common species, all have 10-25%  average cover.  No one individual species tended to dominate within a single plant community.  Acer pensylvanicum is commonly located in the midstory canopy.  Hydrangea arborescens (51%) is the most frequently encountered shrub.   Those frequently associated herbs include Laportea canadensis, Caulophyllum thalictroides, Dryopteris intermedia, Actaea racemosa (=Cimicifuga racemosa), Polystichum acrostichoides, Trillium erectum, Impatiens pallida, Polygonatum pubescens, Osmorhiza claytonii, Maianthemum racemosum, Stellaria pubera, Aster divaricatus, Arisaema triphyllum, and Hydrophyllum canadense.    

 222 species were located in the 17 plots recorded with Allium burdickii.   Most of these species, 150, only occurred in four or fewer plots.  Seventy-five of these individual species only occurred in one plot.  An average of 53 species occurred in the Allium burdickii plots.  Species numbers were highly variable among these 1/10-hectare plots ranging from moderately diverse with 80 individual species to sparse with 36 total species.   Canopy dominants included Fagus grandifolia, Aesculus flava, Betula alleghaniensis, Quercus rubra, Acer saccharum, and Tilia americana (Table 2). As with the A. tricoccum plots, no single canopy species tended to dominate.  Common subcanopy tree species include Acer pensylvanicum and Acer spicatum, plus Cornus alternifolia.  The most common shrub in this open understory habitat is Hydrangea arborescens (47%).  As was determined in the A.  tricoccum plots; Euonymus obovatus (42%) is almost as common.  The herb species occurring with greater than 50% frequency, in order of frequency, include Trillium erectum, Laportea canadensis, Maianthemum racemosum, Galium triflorum, Aster chlorolepsis, Polystichum acrostichoides, Caulophyllum thalictroides, Athyrium asplenioides, Carex pensylvanica, Angelica triquinata, Ageratina altissima var. roanensis, Rudbeckia laciniata, Viola rotundifolia, Dryopteris intermedia, Arisaema triphyllum, and Osmorhiza claytonii. 

Few species differences were seen across the 71 Allium plots.  Seven herb species occurred in at least 50% of the A.  tricoccum plots but not in as a high a frequency in the A. burdickii plots.  However most were almost as common in the A. burdickii plots.  These seven follow with the percent frequency (A. burdickii plots) in parentheses:  Actaea racemosa (=Cimicifuga racemosa) (47%), Impatiens pallida (41%), Polygonatum pubescens (41%), Stellaria pubera (41%), Aster divaricatus (29%), Viola canadensis (29%), and Hydrophyllum canadense (41%).    Nine herb species occurred in at least 50% of A. burdickii plots but less frequently in the A. tricoccum plots.  These 9 species showed a variable association with A. tricoccum.     The 9 herbaceous species follow with the percent frequency (A. tricoccum plots) in parentheses:  Galium triflorum (40%), Aster chlorolepsis (28%), Athyrium asplenioides (33%), Carex pensylvanica (39%), Angelica triquinata (35%), Ageratina altissima var. roanensis (37%), Rudbeckia laciniata (13%), Viola rotundifolia (35%), and Solidago curtisii (35%).   

It has been suggested that A. burdickii is more commonly encountered at higher elevations (Rock 1996).  One could infer this distinction from the higher frequency of occurrence for Abies fraseri, Picea rubens, Stellaria corei, Aster chlorolepsis, Brachyelytrum septentrionale and Ageratina altissima var. roanensis within the A. burdickii plots.   All of these species are more frequently encountered at higher elevations (Weakley 2000, Cronquist 1980, Lamboy 1992).  Conversely, Liriodendron tulipifera is absent from the A. burdickii plots.  It is a minor associate (17%) in the Allium tricoccum plots. This species is uncommon above 3500 feet and is not known above 4000 feet elevation.   However, this pattern does not hold true for all the species encountered in these plots.  For instance, Viburnum lantanoides, Ribes rotundifolium, Acer spicatum and Sambucus racemosa var. pubens are all commonly associated with higher elevation forests (Weakley 2000).  Viburnum lantanoides occurs more frequently in the A. tricoccum plots (44% vs. 30%).   There is little separation in the frequency of Sambucus racemosa var. pubens (28% in the A. tricoccum plots vs. 24% in the A. burdickii plots) or Acer spicatum (46% in the A. tricoccum plots vs. 53% in the A. burdickii plots).   Ribes rotundifolium does not occur in the A. burdickii plots, while occurring in 28% of the A. tricoccum plots.   

It has also been suggested that A. burdickii occurs in drier microhabitats compared to A. tricoccum (Jones 1979, Hanes 1953).  Few species within the data sets are indicative of drier coves.  Two species more commonly encountered on drier slopes are Solidago curtisii and Carex pennsylvanica.   Solidago curtisii occurs in 59% of the A. burdickii plots compared to 35% of the A. tricoccum plots.   Carex pensylvanica was more frequently located in A. burdickii plots (65% vs. 39% in A. tricoccum plots).   

There is a difference in the total number of species encountered in the A. tricoccum plots (337) compared to the Allium burdickii plots (222).  This is undoubtedly an artifact of the greater number of plots located with Allium tricoccum (54) compared to Allium burdickii (17).   Over 160 of the species encountered in the Allium tricoccum plots each occurred in 3 or fewer plots.   196 species occur within both Allium species datasets.   141 exclusively occur in the A. tricoccum dataset.   Most of these are quite infrequent.  Only 15 species occur in more than 10% of the plots.   All 15 of these species occur in fewer than 25% of all the A. tricoccum plots.  These 15 species in order of greatest occurrence are Ribes rotundifolium, Polypodium virginianum, Campanula divaricata, Sedum ternatum, Prenanthes serpentaria, Dicentra canadensis, Dicentra cucullaria, Castanea dentata, Liriodendron tulipifera, Parthenocissus quinquefolia, Hepatica acutiloba, Arnoglossum muhlenbergii, Goodyera pubescens, Prosartes maculatum and Viola pubescens var. leiocarpon.    Twenty-six species occur exclusively within the A.  burdickii dataset, however only two of these occur in more than one plot.  The 2 species are Brachyelytrum septentrionale and Viburnum nudum var cassinoides. 

 Click here to view the REFERENCES

Go back to Other Research links:

 More related links of interest...