﻿Rhynchosporamesoatlantica (Cyperaceae), an imperiled new species of beaksedge from eastern U.S.A.

﻿Abstract Rhynchosporamesoatlanticasp. nov. (Cyperaceae) is described, illustrated, and compared with morphologically similar species. Rhynchosporamesoatlantica is known only from southern Delaware, southeastern Maryland, and southern New Jersey, all within the Mid-Atlantic region of the U.S.A. It inhabits sunny, wet margins of natural, shallow, nutrient-poor, seasonal ponds of the Coastal Plain. Narrow leaf blades; fruits obpyriform in outline; faces of mature fruits possessing a central, pale, well-demarcated disk; and fruit tubercle margins denticulate for most of their lengths indicate R.mesoatlantica is most similar to R.filifolia and R.harperi. Rhynchosporamesoatlantica is unique in its fruit dimensions, scales intermediate in length between R.filifolia and R.harperi, and relatively long fruit stipe. The NatureServe rank of Critically Imperiled and the IUCN rank of Endangered appear warranted for R.mesoatlantica because only six populations are known to be extant, most quite small and isolated; all populations occur within a small geographic area; populations have declined; and serious threats confront the survival of the species.


Introduction
Rhynchospora Vahl section Fuscae (C.B.Clarke ex Gale) Kük., as circumscribed by Gale (1944) and Kral (1996), is a group of sedges characterized by perianth bristles antrorsely barbed, fruit bodies widest in distal half and with smooth faces, and tubercle margins denticulate.Six species belong to Rhynchospora section Fuscae: R. crinipes Gale, R. curtissii Britton, R. filifolia A.Gray, R. fusca (L.) W.T.Aiton, R. harperi Small, and R. pleiantha (Kük.)Gale.These species occur from eastern North America south to northeastern South America, with R. fusca also occurring in Europe.The center of diversity is the southeastern U.S.A.
The most recent phylogenetic analysis that included members of Rhynchospora section Fuscae indicated the section was polyphyletic (Budden-hagen et al. 2017).This analysis indicated the four included members of Rhynchospora sect.Fuscae (R. curtissii, R. filifolia, R. fusca, and R. pleiantha) belonged to three clades separated from each other by multiple clades composed of species from other sections of Rhynchospora, e.g.R. ciliaris (Michx.)C. Mohr, R. fascicularis (Michx.)Vahl, and R. lindeniana Griseb.However, this analysis is based on DNA sequence data from only one marker, trnL/F.Future analyses with more ample taxon and molecular sampling will be necessary to settle questions of monophyly of Rhynchospora section Fuscae and relationships of taxa within it.
Several previous authors have included Delaware and Maryland within the range of R. harperi (Kral 1996(Kral , 2002;;LeBlond 1997;McMillan 2007;McAvoy 2013;Maryland Natural Heritage Program 2016, 2021;Knapp and Naczi 2021).In addition, McMillan (2007) included New Jersey within the range of this species.Both Kral (1996: 396) and LeBlond (1997: 278) reported northern plants (from Delaware and Maryland) resembled R. filifolia in aspect, but had fruits resembling R. harperi plants from the southeastern U.S.A.These authors contended that the northern plants fit within the concept of R. harperi and included them within this species.
Among plants previously identified as Rhynchospora harperi, we observed substantial differences between plants of the Mid-Atlantic (Delaware, Maryland, and New Jersey) and plants from farther south, including characters not noted by previous authors.Our observations led us to hypothesize that the Mid-Atlantic plants were a species distinct from R. harperi.We tested this hypothesis with field work, herbarium work, and morphometric analyses.The purpose of this paper is to report our results, which supported our hypothesis.Accordingly, we also describe the new species Rhynchospora mesoatlantica.

Materials and methods
We studied the morphology, geography, and ecology of Rhynchospora through herbarium work and field work.For herbarium work, we borrowed specimens from, or studied specimens during, visits to DOV, GA, GH, MO, NCU, NY, PH, US (abbreviations as in Thiers 2023).We directly examined all specimens cited in this paper.
The measurements we report in this paper are ones we made directly from specimens.For morphometric analyses, we selected a representative set of 68 specimens of Rhynchospora filifolia and R. harperi to measure.We chose mature, ample specimens from throughout the geographic ranges of these species that exhibited their full range of morphologic variation.Among these specimens, we measured type specimens.All measured specimens are from different populations.We considered populations to be different if their localities are at least 1 km apart and separated by unsuitable habitat.For each of these measured specimens (Appendix 1), we measured eight continuous characters and calculated one ratio of measured characters (Table 1).The characters we studied morphometrically are those suggested to be diagnostic for species and infraspecific taxa by previous authors (Gale 1944;Kükenthal 1950;Kral 2002;McMillan 2007;Naczi and Moyer 2016;Ciafré and Naczi 2022), as well as additional ones we suspected to be diagnostic based on our observations from field and herbarium work.
We plotted measurements of characters that were not highly correlated with each other (r < 0.7, thus probably not genetically redundant) in order to detect groups within the morphometric dataset.We then used ANOVA to test for differences among the groups.We performed all statistical analyses with SYSTAT version 11 (SYSTAT Software 2004).
Field work furnished geographic and ecologic data.To determine the geographic range of Rhynchospora mesoatlantica, we used specimen collection data to map all known occurrences.For each population of Rhynchospora that we studied in the field, we noted vascular plant taxa growing in close association with the target species.We considered closely associated plant taxa to be those growing within 10 meters of R. mesoatlantica.

Results
Within Rhynchospora section Fuscae, R. crinipes, R. curtissii, R. filifolia, and R. harperi comprise a group characterized by four features: cespitose habit, fruit body compressed, mature fruit body with well-demarcated pale disk on center of each face, and fruit with tubercle margins denticulate for most of their lengths (both proximally and distally).The other members of Rhynchospora sect.Fuscae, R. fusca and R. pleiantha, have a long-creeping rhizomatous habit, fruit bodies biconvex, mature fruit body uniformly brown, and tubercles denticulate only in the proximal half.
Rhynchospora crinipes is distinctive in having relatively wide leaves (widest leaf blade per plant 2.2-3.8mm wide in R. crinipes vs. 0.6-1.9mm wide in other members of Rhynchospora section Fuscae).Also, R. crinipes has relatively long fruits [fruits, including tubercles, 2.6-2.9 mm long in R. crinipes vs. 1.5-2.6(-2.8)mm long in other members of Rhynchospora sect.Fuscae] with a long stipe (stipe 0.45-0.83mm long in R. crinipes vs. 0.11-0.38mm long in other members of Rhynchospora sect.Fuscae).The remaining species of Rhynchospora sect.Fuscae, R. filifolia and R. harperi, are more similar to each other than they are to other species of the section.A plot of scale length (SCLL) vs. fruit width (FRW) for these two species reveals three clusters of points (Fig. 1).Inclusion of measurements from type specimens enables identification of these groups.These clusters correspond to R. filifolia, another to R. harperi, and a third corresponds to plants from Delaware, Maryland, and New Jersey that we propose as a new species and describe below as R. mesoatlantica.For this plot, all of the clusters of points are separate and non-overlapping.
Most characters measured are statistically significantly different among Rhynchospora filifolia, R. harperi, and R. mesoatlantica (Table 2).The characters that best differentiate the species, judging by ANOVA F scores, are scale length (SCLL), spikelet length (SPKLTL), tubercle length (TL), fruit width (FRW), and fruit length (FRL), in descending order of discriminatory power.For every one of these five characters, R. filifolia has the lowest values, R. harperi has the highest values, and R. mesoatlantica has intermediate values.For measurements of one character, stipe length (STPL), R. mesoatlantica has greater values than both R. filifolia and R. harperi (Table 2, Fig. 2).For fruit body length (FRBDYL) and longest perianth bristle length (LBRL), R. mesoatlantica and R. harperi have very similar measurements, and both have larger values than in R. filifolia.

Discussion
Comparative morphology and morphometric analyses support the hypothesis that R. mesoatlantica is distinct from all other Rhynchospora species.Multiple morphologic features clearly place Rhynchospora mesoatlantica in Rhynchospora sect.Fuscae: perianth bristles antrorsely barbed, fruit bodies widest in distal half and with smooth faces, and tubercle margins denticulate.Additional features place R. mesoatlantica as most similar to R. filifolia and R. harperi: habit cespitose, widest leaf blade < 2.0 mm wide, longest perianth bristle < 3.0 mm long, fruit body compressed and with a well-demarcated pale disk on the center of each face, and tubercle margins denticulate for most of their lengths.
Several morphologic characters distinguish Rhynchospora mesoatlantica from R. filifolia and R. harperi.A plot of SCLL vs. FRW provides complete separation of R. mesoatlantica from both R. filifolia and R. harperi (Fig. 1).In addition, ANOVA reveals R. mesoatlantica is statistically significantly different from both R. filifolia and R. harperi in six of the nine characters studied in the morphometric analysis: SPKLTL, SCLL, FRL, FRW, STPL, and TL.Two additional characters distinguish R. mesoatlantica from R. filifolia (FRBDYL, LBRL), and one other character distinguishes R. mesoatlantica from R. harperi (RTLFRL).All these diagnostic characters are from fruits, scales, and spikelets.
Syntopy of Rhynchospora filifolia and R. mesoatlantica is another line of evidence supporting species status for Rhynchospora mesoatlantica.Despite R. mesoatlantica growing in close proximity to R. filifolia, the two species maintain their morphologic distinctions at the syntopic site, as well as in all known populations.This naturally occurring syntopy provides a strong test of species distinctions for R. filifolia and R. mesoatlantica.However, the geographic ranges of Rhynchospora mesoatlantica and R. harperi do not overlap, making syntopy of these two species impossible.
Specimens of Rhynchospora mesoatlantica collected prior to our work had been determined as R. filifolia or R. harperi.Now that we have presented support for species status for R. mesoatlantica, we name and describe this species in order to clarify its status and bring attention to it as a species of conservation concern.Diagnosis.Rhynchospora mesoatlantica is similar to R. filifolia and R. harperi, but R. mesoatlantica differs by its fruit dimensions, scales intermediate in length between R. filifolia and R. harperi, and relatively long fruit stipe.In R. mesoatlantica, scales are 3.0-3.4mm long, and tubercles are 0.6-0.7 mm long and 26-32% of fruit length, in contrast to R. harperi, which has scales 3.8-5.0mm long, and tubercles 0.7-1.0mm long and (30-)33-39(-45)% of fruit length.In R. mesoatlantica, scales are 3.0-3.4mm long, and fruits are 2.1-2.3 mm long and 0.9 mm wide, in contrast to R. filifolia, which has scales 2.1-3.0 mm long, and fruits 1.5-1.9mm long and 0.6-0.8mm wide.Rhynchospora mesoatlantica has fruit stipes 0.29-0.38mm long, in contrast to R. filifolia (0.16-0.34 mm long) and R. harperi (0.20-0.35 mm long).
Etymology.We name Rhynchospora mesoatlantica for the Mid-Atlantic region of the U.S.A., the region in which all known populations occur.
Geographic distribution.Rhynchospora mesoatlantica is a narrow endemic of a portion of the Mid-Atlantic U.S.A. (Fig. 5).It is known only from southern New Jersey, southern Delaware, and southeastern Maryland, where it occurs on the Coastal Plain physiographic province.Specimens document its occurrence from a total of 12 populations, each separated by at least 1 km from other populations.Two of the populations in the vicinity of Ellendale, Delaware [E of Ellendale, Commons s.n.(PH); S of Ellendale, McAvoy 6333 (DOV) and later collections] are sufficiently close (3 km apart) that they map as one population (Fig. 5).Other populations are separated by greater distances.The greatest distance separating nearest neighbors among populations (Moyer G0272 in Cape May County, New Jersey, and Commons s.n. in Sussex County, Delaware) is 70 km.
Habitat.Rhynchospora mesoatlantica grows on the sunny, moist upper portions of natural, shallow, nutrient-poor, seasonal ponds and depressions with gently sloping shorelines and sandy-peaty soils (Fig. 6).Surrounding these wetlands are dry-mesic forests or pine plantations.Water levels are typically highest in winter and spring, which is characteristic of Coastal Plain ponds (Phillips and Shedlock 1993).By the time of fruiting, the ponds are usually devoid of standing water, and the plants grow in soils that are merely moist.At most sites we visited, natural seasonal fluctuations in water levels were disrupted by extensive ditching and draining that apparently lowered the water table.Drier soils throughout the year have provided favorable growing conditions for woody vegetation, which is slowly overgrowing and shading some of the sites.The least disturbed site had few trees and shrubs (Fig. 6).In the absence of the natural disturbance of fluctuating water levels, management appears necessary to maintain a sunny environment.Rhynchospora mesoatlantica may persist vegetatively or in the seed bank during periods of unfavorable conditions, but research is needed to understand its persistence and dormancy.
Close    Preliminary conservation assessment.Rhynchospora mesoatlantica is at a high risk of extinction due to a restricted geographic range, small number of occurrences, small population sizes, and historic and ongoing declines due to numerous threats.All historic and current populations total 12. Six of the populations have not been seen for over 20 years, despite repeated, more recent surveys at most of the sites.Three of these populations had been documented in the 1990s, yet we could not relocate them.Thus, declines are apparent in number of populations and number of plants.We are sufficiently familiar with some of these sites to identify likely causes for extirpations: habitat destruction for some and, for others, habitat degradation, including changes to hydrology.
Only six populations are known to be extant.Populations are typically small, ranging from 25 to a maximum of 200-300 plants at the population northwest of Belleplain (R. Moyer, pers.comm.).Only three populations contain more than 100 plants.Our estimate of the total number of mature plants present in extant populations is 700.
Five of the six populations known to be extant are in protected areas.Most of these protected areas are state forests that allow resource extraction and consequent habitat alteration.
Most extant and historic occurrences are/were in Coastal Plain ponds in Delaware and Maryland, one of the most threatened habitats on the Delmarva Peninsula and host to many rare species (McAvoy and Bowman 2002).Most of these ponds and surrounding forests are highly degraded due to direct and indirect anthropogenic impacts.Land-use changes resulting in habitat fragmen- tation, conversion of forest to pine plantations, destructive forestry practices like clear-cutting, and hydrologic alterations due to extensive ditching and draining are among the threats contributing to past and ongoing declines (McAvoy and Bowman 2002).Quantifying declines in Rhynchospora mesoatlantica is challenging; the historic record is sparse, with only four populations documented prior to 1990.Landscape changes are evident throughout the Delmarva Peninsula, including the extent of ditching and draining.In Delaware alone, there are over 2,000 miles of ditches intended to redirect normal water flows across the land and sustain productive agricultural lands (DE DNREC 2023).Unfortunately, these ditches negatively impact natural plant communities hosting R. mesoatlantica by interrupting seasonal water-level fluctuations that suppress woody vegetation.Habitat restoration with ongoing maintenance, especially for natural hydrologic cycles, appears to be warranted at most sites, including those on public lands.
Also noteworthy is the fact that R. mesoatlantica plants usually occupy only a portion, and often a small portion, of the Coastal Plain ponds that host this species.For example, the area of one pond is 0.008 km 2 (8,000 m 2 ), yet plants of R. mesoatlantica occupy only 0.004 km 2 (4,000 m 2 ) of the pond.Our estimate of the area occupied by all known R. mesoatlantica populations, historic and extant, is 0.031 km 2 (31,000 m 2 ).For R. mesoatlantica populations known to be extant, our estimate of area occupied is 0.017 km 2 (17,000 m 2 ).
Due to decades-long recognition of Coastal Plain ponds as centers of rare plant diversity (e.g., Hirst 1983;Boone et al. 1984;McAvoy and Bowman 2002) and our own extensive field efforts to rediscover formerly documented populations of Rhynchospora mesoatlantica and discover new ones, we regard the likelihood of discovery of new populations as low.Simply, most Coastal Plain ponds within the geographic range of R. mesoatlantica have been botanically explored, many very extensively during multiple years and multiple seasons.
We recommend a NatureServe Global Rank of Critically Imperiled (G1, Faber-Langendoen et al. 2012) for Rhynchospora mesoatlantica, based on considerations of rarity, threats, and trends (Master et al. 2012).There are 12 known occurrences (6 historic and 6 extant), a Range Extent (Extent of Occur- rence, EOO) of 4,495 km 2 , and an Area of Occupancy (AOO) of 44 km 2 .Threat impact is estimated at very high, and short-term trends and long-term trends are estimated to be at least 10% and 40%, respectively, based on declines in AOO, population size, and number of occurrences.
As a preliminary assessment, we consider the IUCN category Endangered ( IUCN Standards and Petitions Committee 2022) to apply to Rhynchospora mesoatlantica for the following reasons: EOO of 4,495 km 2 is < the 5,000 km 2 threshold (B1); AOO of 44 km 2 is < the 500 km 2 threshold (B2); and we have observed continuing decline in AOO, habitat quality, and number of populations (Bb).Tentatively, we assess the metapopulation as severely fragmented since at least 50% of the populations are isolated and small (< 50 plants) and occurring in a very rare and localized habitat surrounded by unsuitable habitats and with limited capacity for dispersal between distant extant populations 11-70 km apart (Ba).
Due to the severity of conservation threats, few known extant populations, small population sizes, and apparent necessity of human-mediated intervention to maintain habitats, we recommend Rhynchospora mesoatlantica for protection under the U.S.A. Endangered Species Act.

Identification key to Rhynchospora section Fuscae
This key is for specimens bearing mature fruits.Measurements of fruit length include the tubercle, but not perianth bristles.Scale length is for scales from middle of spikelets.

Figure 5 .
Figure 5. Known geographic distribution of Rhynchospora mesoatlantica.Based on all known collections.

Table 1 .
Morphologic characters, with their abbreviations, measured on herbarium specimens of Rhynchospora.

Table 2 .
Means ± 1 SD and ranges for morphologic characters measured for Rhynchospora.Character abbreviations correspond to those in Table1.All measurements are in millimeters.N = sample size.Within a row, means with different superscripts differ significantly (ANOVA, P < 0.01).