POACEAE

POACEAE View in CoL View at ENA

The Andropogon elliottii complex: refined concepts and description of a cryptic species Primary authors: Brett J. Budach and Edwin L. Bridges

Vorontsova et al. (2023) reestablished the genus Anatherum P. Beauv. for the former sect. Leptopogon Stapf within the broader genus Andropogon L. This required numerous new combinations for almost all North American Andropogon . While many of these combinations were made in Vorontsova et al. (2023), several recently described or elevated Andropogon species were left in need of new combinations within Anatherum P. Beauv. This need prompted a closer investigation of members of Campbell’s (1983) concept of Andropogon gyrans Ashe and provided an opportunity to redefine concepts therein.

From Andropogon subtenuis Nash to Anatherum gyrans (Ashe) Voronts. & E.A. Kellogg

Without much discussion, Campbell (1983) lumped Andropogon subtenuis Nash (1912) into his concept of Andropogon gyrans Ashe. It was not until recently that the name reappeared in Weakley’s Flora of the Southeastern US ( Weakley & Southeastern Flora Team 2023). There have been variable applications of the concept for A. subtenuis Nash , which lack obvious patterns and often do not match the type material. At one end of this concept are plants that do not match the type of A. subtenuis Nash and more closely resemble typical A. gyrans , with widened upper stem sheaths obscuring at least some racemes, but with extremely narrow basal leaves, glabrous basal leaf sheaths, and at least some exerted racemes. This application, as noted in Weakley and Southeastern Flora Team (2023), is assumed to include Campbell’s “tenuous variant” of his concept for A. gyrans . However, we have found that Campbell’s use of that designation was inconsistent with our contemporary applications of it. At the other end of the A. subtenuis Nash concept are plants that more closely match the A. subtenuis Nash type, lacking the wider stem sheaths and more obscured racemes of typical A. gyrans , but with similar dispersal unit characters and appressed-pubescent basal leaf sheaths. We believe that the former application of this concept is indeed patternable and descriptive of a true taxonomic entity while the latter only applies to a disturbance-induced growth form unique to cleistogamous grasses but is not taxonomically significant in and of itself.

The type specimen for Andropogon subtenuis Nash is clearly an example of a disturbance-induced anomalous growth form and not a taxonomically significant entity. With its long awns, long dispersal unit pubescence, wider leaf blades, and appressed-pubescent basal leaf sheaths, the type resembles typical A. gyrans material in every way aside from the extremely reduced stem sheaths and exerted racemes. This behavior of a typically cleistogamous grass shifting to chasmogamy in the presence of disturbance is not novel, indeed Campbell was aware of it ( Campbell 1982). Hitchcock and Chase (1950) also mention the possibility of Andropogon subtenuis Nash being a reduced, chasmogamous form of Andropogon elliottii Chapm. (= A. gyrans Ashe ). This is a phenomenon the authors have observed in other partially cleistogamous grasses, including Sporobolus clandestinus (Biehler) Hitchc. in north Florida and Sporobolus compositus (Poir.) Merr. in southwest Missouri, even though it is not well represented in the literature.

Following our examination of the type and interpretation of the potential for disturbance-induced anomalous growth forms, we believe that Campbell’s choice to lump Andropogon subtenuis Nash into his concept for Andropogon gyrans Ashe was correct. As such, our concept of Anatherum gyrans (Ashe) Voronts. & E.A. Kellogg includes the former Andropogon subtenuis Nash. Material recently identified as Andropogon subtenuis Nash using Weakley and Southeastern Flora Team (2023) could be either the anomalous, disturbance-induced growth form of Anatherum gyrans or the new species described below ( Anatherum subtile Budach & E.L.Bridges ).

From Andropogon gyrans Ashe ‘tenuous variant’ to Anatherum subtile Budach & E. Bridges

The concept of Andropogon subtenuis Nash as currently applied in Weakley and Southeastern Flora Team (2023) includes the named but not formally published concept of Campbell’s “tenuous variant” of Andropogon gyrans Ashe. However , upon close study of Campbell’s own annotations and descriptions, we found confusing applications. Campbell lumped A. subtenuis Nash into his concept of A. gyrans Ashe while maintaining the separate distinction of the “tenuous variant” of the same species in contrast to the “common variant.” Clearly these concepts were not meant to be applied to the same material. To further complicate this issue, a close study of the material that Campbell annotated as the “tenuous variant” of A. gyrans Ashe shows a very wide range of habitats and morphological expressions, relying only on what we believe to be wildly variable and unreliable characters related to the visibility and exertion of the racemes and raceme sheaths. There is certainly overlap between entities annotated by Campbell as the “tenuous variant” and “common variant.” There are also several Campbell-annotated sheets of what he called his “tenuous variant” that are clearly more like his “common variant” of A. gyrans Ashe and consistent with our updated concept of Anatherum gyrans (Ashe) Voronts. & E.A. Kellogg. Peduncle length and whether an individual plant is fully cleistogamous (all racemes hidden within the raceme sheaths; raceme sheaths hidden within the stem sheaths) or mostly chasmogamous (many racemes at least partially exerted from their sheaths; raceme sheaths well-exerted from the stem sheaths) are highly variable and respond dramatically to extreme forms of disturbance. Intensive field experience in both burned and degraded habitats throughout the state of Florida has shown that a severe late-season burn or recent roller-chopping or mowing can induce extreme chasmogamy in material that corresponds to either Campbell’s “common variant” or his “tenuous variant” of A. gyrans Ashe , and as such, it is not possible to perfectly and directly apply his “tenuous variant” concept to any one entity.

With that chaos of concepts and applications as a backdrop, we studied Floridian populations through meticulous field study of numerous individuals, with close attention paid to ecology and morphological responses to disturbances over time. An ongoing project studying growth and phenology responses to seasons of burning required the authors to locate and identify thousands of sterile vegetative individuals of over a dozen Anatherum species. This challenge allowed us to determine patterns in habitat and morphology that were initially unbiased by reproductive material or the use of existing keys and names. As the project proceeded, we were able to match these unbiased concepts with their reproductive characters and published names, and some striking patterns emerged. These patterns were repeatedly found in similar populations across numerous sites from the Florida panhandle to the Florida Everglades , helping construct more refined

species concepts built on the marriage of morphology to ecology and biogeography. The result is, in part, the following description of a cryptic species:

Anatherum subtile Budach & E.L. Bridges , sp. nov. ( Fig17 View FIG ). TYPE. U.S. A. FLORIDA. Highlands Co.: recently burned dry /

dry-mesic flatwoods ecotone below scrubby sandhill savanna, Silver Lake Sandhill, Lake Wales Ridge Wildlife and Environmental

Area, (27.5758324, –81.5263988), 20 Oct 2021, Budach 652 (HOLOTYPE: USF; ISOTYPES: FLAS, FSU, MO, NCU, NY, US) GoogleMaps .

Diagnosis.— Similar to Anatherum gyrans (Ashe) Voronts. & E.A. Kellogg , differing in having glabrous basal leaf sheath faces, narrower basal leaf sheaths, narrower basal leaf blades, slightly narrower culms, slightly narrower upper stem sheaths, shorter rachis internodes, and shorter pedicels ( Fig. 17 View FIG ).

Plants densely cespitose with very short, stout rhizomes. Culms (50–)60–80(–90)cm tall, (1.0–)1.3–1.7 (–2.0) mm in diameter at the midpoint, internodes glabrous and often glaucous at and just below the nodes, the branches slender, short, erect to ascending and often obscured by the widened upper stem sheaths. Upper stem sheaths widened, the largest (6.0–)7.2–9.0 (–10) cm long and (4.0–)4.7–6.3(–7.0) mm wide when flattened, often overlapping and obscuring at least some of the inflorescence units. Basal leaf sheaths smooth, glabrous, pale stramineous to dull yellow-green, narrow, (0.5–1.4–2.6(–3.0) mm wide. Basal leaf blades erect to ascending, appearing yellow-green in situ , dull when pressed, 30–50 cm long, (0.6–)1.0–2.2(–2.8) mm wide (unfolded), difficult to unfold due to nearly complete folding of lamina, glabrous on the blade surfaces but with sparse to dense, ascending, stout marginal cilia concentrated near the base. Ligule thin-membranaceous, 0.6–0.8 mm long, very finely ciliate on the margin and dorsal surface, the cilia usually <0.2 mm long. Inflorescences variable, mostly congested in the upper 1/3 of the fertile culm, with slender, erect to slightly flexuous branches concealed within and/or exserted beyond the widened upper stem sheaths, maximum branch order 1(–2), inflorescence units (3–)6–12(–15) per culm; inflorescence units often of two kinds: some with peduncles longer (up to 7.0 cm) and borne on long terminal stem branches such as to be long-exserted beyond both the widened stem sheaths and raceme sheaths, some with peduncles shorter (as short as 1 mm) and borne on short terminal stem branches such as to be completely obscured by both the widened stem sheaths and subtending raceme sheaths; the longer, exserted inflorescence form often absent but the shorter, obscured form always present; summit of terminal stem branches often with a dense callus of ascending whitish hairs. Raceme sheaths (3.2–)3.6–6.4(–7.8) cm long, some borne on longer stem branches and visible beyond the widened stem sheaths, others borne on shorter branches and obscured; the base sometimes partially obscured by the callus hairs on the summit of the subtending stem branch. Peduncles 0.2–5.0(7.0) cm long, ranging from fully included within the raceme sheath to slightly exerted beyond the raceme sheath summit. Rames in groups of 2–3(–4), terminal on each peduncle, (5.0–)5.3–5.7(–6.0) mm long, some remaining fully enclosed within the raceme sheaths while others fully exerted beyond the sheath summit; some plants with no exerted rames but all plants with at least some fully enclosed rames. Rachis Internodes (2.5–)3.0–3.8(–4.0) mm long, the pubescence 7.5–9.5(–10.4) mm long, the distal internode pubescence slightly longer than the proximal. Sessile spikelets (4.0–)4.2–4.8(–5.0) mm long, with 3.5—4.5 mm long callus hairs at the base, keels of lower glume scabrous from the midpoint, fertile lemma, including its awn, 19–24 mm long. Pedicellate spikelets absent or vestigial, with a pedicel (4.5–)4.6–5.6(–5.8) mm long, the pubescence (6.0–)6.2–8.8(–9.5) mm long, the distal pedicel pubescence slightly longer than the proximal.

In the flatwoods, pine savanna, and subtropical grassland landscapes of the extreme southeastern coastal plain, typically on hyperseasonal and highly acidic spodosols, we frequently found an entity that resembled a diminutive, narrow-leaved expression of Anatherum gyrans (Ashe) Voronts. & E.A. Kellogg. It has very narrow basal leaf sheaths and leaf blades, with the faces of those basal leaf sheaths always smooth and glabrous ( Figs. 17–19 View FIG View FIG View FIG ). This immediately caught the authors’ attention, as typical A.gyrans of more well-drained uplands of the region always has appressed-pubescent basal leaf sheaths. Typical A.gyrans material from as far away as southwest Missouri maintains this basal leaf sheath pubescence character.

Within the genus Anatherum (= Andropogon sect. Leptopogon ), basal leaf sheath pubescence is not always a conserved or reliable character and can be impossible to discern from digital specimen images, so an effort was made to observe and document this character in situ wherever and whenever possible. After over three years of field work throughout the state of Florida spanning the full range of potential habitat, we were able to consistently use this basal leaf sheath pubescence character in conjunction with a habitat separation to differentiate Anatherum subtile from typical Anatherum gyrans . Along with the basal leaf sheath pubescence character, a set of morphological measurements was made from Floridian specimens of both entities, and we found the following characters to be reliable in separating these species: fertile culm diameter at the midpoint, width of the basal leaf sheaths, width of longest basal leaf blade, width of the widest upper stem sheath, length of the shortest raceme sheath, length of the rachis internode, and length of the pedicel of pedicellate/vestigial spikelet. Given the restricted range and size of the sample, we suspect some of these characters to become more significant and others to become less significant as other researchers put them to the test over a wider geographical range (see below).

Characters that separate A. subtile from A. gyrans are based on 9 sheets of A. subtile and 8 sheets of A. gyrans from Florida. The ranges given represent the mean ±1 standard deviation.

Culm Diameter: 1.3–1.7 mm ( A. subtile ) vs. 1.8–2.2 mm ( A. gyrans )

As measured near the middle of a mature or nearly mature reproductive culm, typically below the widened summit of the internode and above the preceding leaf sheath summit.This character may become less informative if more material across the species’ ranges is included.

Sterile Leaf Sheath Width: 1.4–2.6 mm vs. 2.4–3.8 mm

The combined width of the tightly overlapping basal leaf sheaths of a single sterile, non-reproductive ramet (or tiller); only including leaves from that growing season (dense layers of long-dead material from past growing seasons are excluded).This character was expected to be more informative, but A. gyrans in extremely dry or sterile situations will be stunted and approach the dimensions more typical of A. subtile .

Sterile Leaf Sheath Pubescence: glabrous vs. appressed-pubescent

This applies to the faces (lamina), not the margins of the sheaths of the basal leaves of a sterile, non-reproductive ramet; pubescence can be severely worn on very old material of A. gyrans ( Fig. 19 View FIG ).This is the easiest and most consistent field character that seems to hold across the ranges of both entities.

Width of Longest Sterile Leaf: 1.0– 2.2 mm vs. 2.0–3.0 mm

It can be difficult to get a leaf width measurement from pressed/dried A. subtile due to the extremely tightly folded nature of the lamina, whereas A. gyrans leaf blades are much more easily flattened for measurement ( Fig.19 View FIG ).A larger sample size from across the range of A.gyrans may pull the mean and standard deviation even further away from that of A. subtile .

Widest Upper Steam Sheath: 4.7–6.3 mm vs. 6.6–8.6 mm

The upper stem sheaths are the diagnostic, wider, bract-like structures that often obscure all the inflorescences within, and are the structures that make both entities so immediately recognizable in the field. The widest stem sheath can be chosen subjectively, with the width then carefully measured by rolling out the relatively fragile sheath blade. A much larger sample size from across the range of A. gyrans may pull the mean and standard deviation even further away from A. subtile .

Shortest Raceme Sheath: 3.3–4.5 mm vs 4.9–5.7 mm

The shortest raceme sheath is often,but not always, associated with the most strongly exerted inflorescences. It is sometimes associated with the greatly reduced inflorescence units hidden within the congested and obscured layer of stem and raceme sheaths. This character is not likely to become more informative with larger sampling across the range of these two entities but is included here due to the wide separation in our regionally specific measurements.

Rachis Internode Length: 3.0– 3.8 mm vs 3.8–5.0 mm

This was an unexpected character difference between the two species.A greater sample of this character across the species’ ranges would be very interesting.This character includes the full length of the rachis internode,from the base where it joins the base of the adjacent sessile (fertile) spikelet to the tip where it disarticulates from the next dispersal unit.

Pedicel of Pedicellate Spikelet Length: 4.6–5.6 mm vs 5.6–8.0 mm

This was an unexpected character difference between the two entities. A greater sample of this character across the species’ ranges would be very interesting.This character is measured from the base where it joins the base of the adjacent sessile spikelet to the tip of the pedicel but NOT including the vestigial pedicellate spikelet if present.

It is possible to find nearly unrecognizable or unkeyable examples of both Anatherum gyrans and Anatherum subtile , made so by extreme forms of disturbance (intense or late-season fires, roller chopping, extreme soil disturbance, nutrient flushes, etc.). These anomalous examples will often have extremely exerted inflorescence units and reduced stem and raceme sheaths, while retaining the long inflorescence pubescence, twisted awns, at least some very reduced and fully cleistogamous rames, and diagnostic sheath pubescence (or lack thereof). In the past, these troubling individuals were often, in part, assigned to Andropogon subtenuis Nash , but the application of that name has been inconsistent across its range. A closer study of the type specimen for A. subtenuis Nash reveals it to be a burned or mowed example of what would otherwise be considered typical Anatherum gyrans , confirmed by its inflorescence structure and basal leaf sheath pubescence, along with awareness of this novel, disturbance-induced growth form. This behavior of a typically cleistogamous grass with hidden inflorescence units reverting to chasmogamy and strongly exerted inflorescence units in response to extreme disturbances is known from other grass species but not well documented in the literature.

Anatherum subtile is restricted to highly acidic and often hyperseasonal spodosols of the extreme southeastern coastal plain, with most records from Florida and only a few confirmed records from Alabama, Georgia, and Louisiana. The pubescence of the basal leaf sheaths, or lack thereof, is often not interpretable from specimen images, so it is difficult to confirm its identification without close examination of the physical specimens. Closer study at herbaria and in the field across the Coastal Plain is likely to expand the range of this new, cryptic species. It is most frequent in high quality mesic to dry pine flatwoods, savannas, and subtropical grasslands. It is rarely found as a waif on drier hummocks within seepage slopes and wetter grasslands and savannas and in slightly lower, wetter zones within xeric or scrubby flatwoods and savannas. In comparison, Anatherum gyrans is a common and occasionally weedy denizen of well-drained upland habitats throughout the southeast and is characteristic of Pinus palustris sandhill-savannas on deep sands. Both species are far easier to detect in areas that have been recently burned or otherwise disturbed, as both appear to produce much larger and more numerous fertile culms in those situations. Ongoing research suggests that A. subtile is far more dependent on fire than A. gyrans , producing far fewer and, in some cases, no fertile culms without recent fire, making it even more difficult to detect in unburned habitats. A. gyrans , while certainly stimulated by burning, does not require it and as a result can still be easily found in unburned habitats (Budach, Bridges, and Orzell, unpublished data).

In peninsular Florida specifically, A. gyrans is restricted to the deeper, well-drained yellow sand entisols of the uplands and ridgetops. On the Lake Wales Ridge, where narrow sandhill ridges are often separated from adjacent flatwoods and depressions by extremely sharp and narrow ecotones, it is possible to find typical A. gyrans at the top of the ridge in the well-drained yellow sands with A. subtile just a stone’s throw away downslope where the sands are white or gray and occasionally to rarely saturated. Even in these narrow zones of close contact, the authors have not seen signs of hybridization or clinal patterns of intermediacy between the two entities. Indeed, a quick glance at the basal leaf sheath pubescence of material from central Florida is enough to estimate the landscape position and habitat of the specimen.