Typha lepechinii Mavrodiev et Kapit . sp . nov . ( Typhaceae Juss . ) — A New Endangered Endemic Cattail in the Outmost East of European Russia †

Typha lepechinii, a new species from European Russia, is described. In terms of the external morphology, it is similar to T. shuttleworthii but differs from the latter by the shape of female inflorescence, wider leaf blades and geographical distribution. Like T. shuttleworthii, T. lepechinii is clearly distinct from all other cattails of T. sect. Ebracteolatae due to its short male inflorescence that is 2–4 times shorter than the contiguous female inflorescence. The spatial disjunction of this new species from the main distribution range of predominantly European T. shuttleworthii is significant: T. lepechinii is an endemic of the outmost East of the Russian Plain, in particular, the regions of Middle and Southern Cis-Ural region (basin of the Middle and Lower Kama). In terms of the political administrative borders, this new cattail was found in the Udmurt Republic, the Republics of Tatarstan and Bashkortostan, the parts of the Russian Federation. From the elementary comparative molecular standpoint, T. lepechinii is different from its sister T. shuttleworthii due to four single nucleotide positions and two indels of the rpl32 gene and rpl32-trnL intergenic spacer (cpDNA). The new cattail is named after I. I. Lepechin (1737–1802)—a Russian scientist-encyclopedist whose primary botanical interest focused on the regions of Ural and Siberia.


Introduction
At least seven species and hybrids from Typha L. sect. Ebracteolatae have been listed as the flora of European Russia [1]. One of the most interesting taxa from this section is Typha shuttleworthii WDJ Koch and Sond., which has been described from Northern Switzerland and, as is estimated today, is distributed from Western, Southern and Central Europe to the Mediterranean regions, European Russia, Poland and Belarus [1][2][3][4][5]. This species was also reported in Northwestern Iran and Eastern Turkey [6]. However, because T. shuttleworthii is a predominantly Western European plant [1,7], here we tend to agree with Ghahreman and Sanei Chariat-Panahi [8] that the proper name for the Northwestern Iranian populations of Shuttleworth's cattail is actually T. persica Ghahr and Sanei Chariat-Panahi. The findings of T. shuttleworthii previously documented from the regions of Russian Krasnodar Territory, Crimea and some areas of Caucasus [1] must also be clarified from a taxonomic standpoint.
Despite its relatively wide distribution range, T. shuttleworthii (incl. T. transsilvanica) is not a common species. This cattail is considered extinct (EX) in Hungary, critically endangered (CR) in Slovakia, Bulgaria and the Czech Republic, endangered (EN) in Serbia, strongly threatened (ST) in Germany, and vulnerable (VU) in Switzerland and Western Ukraine [3].
This taxon is extremely rare in Belarus [2,9]. Current, well-documented records of T. shuttleworthii in Russia include only the populations of its outmost Eastern European parts [10]. The age of the youngest of three known herbarium specimens of this cattail that were collected in Central European Russia (Moscow and Kaluga regions) has now exceeded 70 years [2]. Thus, both the current distribution as well as the conservation status of T. shuttleworthii in Russia is unclear now. However, it is fairly likely that this cattail has gone extinct in Central Russia, as has happened in Hungary [3].
The populations of Shuttleworth's cattail from the outmost European East Russia ("Pre-Ural" or "Cis-Ural" region) are of our particular interest not only because of their high vulnerability [10,11] but also due to significant disjunction from the main range of T. shuttleworthii in Western and Central Europe [1,3,10]. It is also remarkable that the Eastern European plants of this cattail are different from the Western European plants from both morphological [10,11] and molecular standpoints. We would like to officially establish that, in our current opinion, the Shuttleworth's cattail from the outmost regions of Eastern Europe represents a separate species. The goals of this study include the formal description of this previously unrecognized species, as well as its justification within the framework of the most recent plastid (cpDNA) sequence data, currently available for T. sect. Ebracteolatae [12,13].

Materials and Methods
Methods of the field work are already given in Kapitonova et al. [10]. Specimens were collected following standard techniques and deposited in the IBIW, TOB, UDU, and UFA herbaria.
First, the simple summaries of the single nucleotide polymorphic positions (SNPs) and indels were established as separate figures (see Results). Of the 28 sequences examined, 22 sequences represent broadly defined T. latifolia L. Due to the clear patterns in sequence variation and the fact that many sequences were identical, not all of these 28 sequences were included in future cladistic analyses, and the placeholders of the identical sequences have been used instead. We also determined that within the obtained alignment there are five patterns of variation within the rpl32 gene and rpl32-trnL intergenic spacer sequences, and included from one to a few sequences of each of these five patterns as placeholders in future cladistic analyses.
The three-taxon statement (3TS) permutations of reduced molecular alignment were conducted with TAXODIUM ver. 1.2 [19]. Typha laxmanni Lepechin was a priori defined as an outgroup taxon for maximum parsimony analysis, as well as an operational outgroup for the 3TS permutations of the molecular data (Williams-Siebert (WS) representation of unordered multistate characters; reviewed and implemented in Mavrodiev and Madorsky [19]).
Upon obtaining the aligned molecular matrix, the summary of indels was constructed in Mesquite [17] and manually recorded as a reduced binary ("presence-absence") matrix following the assumption that the lack of indel is a plesiomorphic character-state. This binary matrix was analyzed using Wagner parsimony [20], the 3TA following fractional weights of the 3TSs [19,20], and the modified AC method [24,25].

Diagnosis
Typha lepechinii sp. nov. is similar to T. shuttleworthii but differs from the latter by the morphology of female inflorescence, wider leaf blades and geographical distribution.
Mature female inflorescence (9) 12-18 (29) cm long, slightly ashy at the base. Leaf  Like T. shuttleworthii, T. lepechinii sp. nov. clearly differs from all other cattails of sect. Ebracteolatae due to its short male inflorescence that is 2-4 times shorter than the contiguous female inflorescence.

Etymology
The species is named after a member of St. Petersburg's Academy of Science Ivan Ivanovich Lepechin (1737-1802)-a famous Russian scientist-encyclopedist, traveler, and naturalist, whose primary botanical interest focused on the regions of Ural and Siberia.

Distribution and Ecology
Typha lepechinii sp. nov. is distributed in the East of the Russian (Eastern European) Plain, in the regions of Middle and Southern Cis-Urals (basin of the Middle and Lower Kama). In terms of the political administrative borders, the new cattail was found in the Udmurt Republic, the Republics of Tatarstan and Bashkortostan (Figure 7).

Distribution and Ecology
Typha lepechinii sp. nov. is distributed in the East of the Russian (Eastern European) Plain, in the regions of Middle and Southern Cis-Urals (basin of the Middle and Lower Kama). In terms of the political administrative borders, the new cattail was found in the Udmurt Republic, the Republics of Tatarstan and Bashkortostan (Figure 7). Typha lepechinii sp. nov. is a continental species [10,11]. It grows in shallow waters, damp and dry shores of small oxbows, ponds and reclamation canals, and along riversides and streams. Within the Udmurt Republic, T. lepechinii sp. nov. is now existing within secondary habitats, such as artificial ponds and drainage channels [10,11].

Conservation Status
Typha lepechinii sp. nov. was initially considered a highly vulnerable species (VU) [10] but under binomial 'T. shuttleworthii' is currently listed in the Red List of the Republic of Bashkortostan with a first category of rarity (EN) [28].

Cladistic Analyses of cpDNA Sequence Data
The results of cladistic analyses of the cpDNA sequence data of Typha lepechinii sp. nov. and related species are summarized in Figures 8-11. Typha lepechinii sp. nov. is a continental species [10,11]. It grows in shallow waters, damp and dry shores of small oxbows, ponds and reclamation canals, and along riversides and streams. Within the Udmurt Republic, T. lepechinii sp. nov. is now existing within secondary habitats, such as artificial ponds and drainage channels [10,11].

Conservation Status
Typha lepechinii sp. nov. was initially considered a highly vulnerable species (VU) [10] but under binomial 'T. shuttleworthii' is currently listed in the Red List of the Republic of Bashkortostan with a first category of rarity (EN) [28].

Cladistic Analyses of cpDNA Sequence Data
The results of cladistic analyses of the cpDNA sequence data of Typha lepechinii sp. nov. and related species are summarized in Figures 8-11.

Standard cpDNA Matrix (No Indels Included)
The total number of characters in the final trimmed molecular alignment of T. sect. Ebracteolatae with no Outgroup added/with included Outgroup (Typha laxmannii) was 745/745, consisting of 728/720 invariable characters, and 17/25 variable characters (all mutually congruent), 13/13 of each were informative for standard maximum parsimony analysis. After the number of taxa was reduced to 10, this matrix implies 816 3TSs (again, all mutually congruent and informative) if permutated relatively T. laxmannii.
Manually analyzing the obtained cpDNA matrix, we found five distinct patterns in sequence variation that differed from each other by 1-12 SNPs (Figure 8).
While Patterns I (T. lepechinii sp. nov.) and II (T. shttleworthii) each contain a single sequence, the other patterns contain more than one sequence. Pattern III contains 17 sequences (one sequence of T. elata, T. incana, T. yakutii and 14 sequences of T. latifolia), Pattern IV is of five sequences (one of T. caspica and four of T. latifolia), Pattern V is of four sequences of T. latifolia.
Each pattern of variation is clearly congruent to the geography of the sequenced samples ( Figure 8). Pattern I is represented by T. lepechinii sp. nov., a narrow endemic of the outmost East of the European Russia (see above).        Figure 9, but simplified and established as a binary ("presence-absence") matrix under the assumption that the presence of an indel is an apomorphic character state.
Taxonomy 2022, 2, FOR PEER REVIEW 12 Figure 10. Similar to Figure 9, but simplified and established as a binary ("presence-absence") matrix under the assumption that the presence of an indel is an apomorphic character state.  Pattern II is represented by T. shttleworthii, a predominantly Western and Central European taxon. The exact sequence data of T. shttleworthii that was included in the comparisons and cladistic analyses was obtained from a sample that had been collected in Bavaria [12], the closest neighboring German region of Switzerland.
We can continue this logic, even while keeping in mind the high invasiveness of the broad-leaf cattail.
Pattern III includes species that are broadly distributed across Eurasia and North America, but are predominantly found in northern and temperate areas.
The predominant distribution of Pattern IV includes Europe (mostly Southern Russia), Caucasus, Southern Asia, Central Asia and its nearest neighboring areas of Western Siberia, as well as Africa (Kenya). Contrary to Pattern III, Pattern IV clearly gravitates towards the South, not the North.
Finally, the remarkably distributed Pattern V includes the sequence data obtained from its plants, collected solely in the Far East of China (province Heilongjiang), in the outmost South of Sakhalin Island (Russia) as well as in the Russian Kurils (islands Yurii and Kunashir). Fitch parsimony analysis (the single most parsimonious tree of length 25, CI = RI = 1.000), 3TA (the single most parsimonious tree of length 816, RI = 1.000) and 3TS-ACA (strict consensus of 15 trees of the score 0.17160) yielded the similar results ( Figure 11A): a. Typha sect. Ebracteolatae appeared as monophyletic [12]; b. Typha lepechinii sp. nov. (Pattern I) is sister of T. shuttleworthii (Pattern II); c. Clade T. lepechinii sp. nov. plus T. shuttleworthii defined as a sister of all the remaining analyzed taxa of T. sect. Ebracteolatae; d. The clade that corresponds to Pattern III, is a sister of the clade that includes sequences of Patterns IV and V, but Pattern IV is defined as a polytomy by Fitch parsimony as well as by 3TA. e. In all the analyses T. latifolia appears as clearly polyphyletic.

Indel-Based Binary Matrix
The obtained cpDNA matrix contains seven indels (Figure 9). From these seven, five are parsimony-informative.
Wagner parsimony analysis of the indel-based binary matrix with all-zeros Outgroup included ( Figure 10) (the single most parsimonious tree of length eight, CI = 0.8750; RI = 0.900), the 3TA of the same matrix (the single most parsimonious tree of length 79.384, RI = 0.9553), as well as the 3TS-ACA of this matrix (strict consensus of 225 trees of the score 0.41648), again yield similar results ( Figure 11B), which are, however, slightly different from the results of the cladistic analyses of the molecular dataset with no indels included (see above): a. Typha shuttleworthii, not the clade (T. shuttleworthii + T. lepechinii sp. nov.), is defined as a sister of all the remaining taxa of T. sect. Ebracteolatae; b. Typha lepechinii sp. nov. is a sister of all the remaining analyzed taxa of sect. Ebracteolatae, excluding T. shuttleworthii; c. The clade that corresponds to Pattern III, is a sister of the clade that includes most sequences of Patterns IV (but not T. capsica) and V, but both Patterns are again defined as polytomies in all analyses; d. Clade (Pattern IV + Pattern V) is defined as a polytomy by Fitch parsimony as well as by 3TA; e. Pattern IV appears as paraphyletic due to the sisterhood of T. capsica and the clade (Patterns III + IV + V); f.
Pattern IV is also the most variable in terms of the presence/absence of indels (see the indel-based patterns A-F (Figures 9 and 11B), which are not all necessary informative from a cladistic standpoint; g. In all the analyses of the indel-based binary matrix, T. latifolia appears as clearly polyphyletic ( Figure 11B).
A simple visual comparison of the cpDNA sequence data shows that T. lepechinii sp. nov. is different from the sister T. shuttleworthii by four SNPs [13] and two indels (Figures 8 and 9).

Discussion
William Whewell (1794-1866) pointed out that the essence of Natural Classification can be understood from this simple proposition: " . . . natural arrangements obtained from different sets of characters, must coincide with each other" (summarized in Williams and Ebach [29]. This is what we see in the case of Typha lepechinii sp. nov.-from a Cladistic point of view, this species is nothing but the original relationship, obtained from cladistic analyses of the cpDNA sequence data. However, simultaneously, it also differs from all other cattails of T. sect. Ebracteolatae in external shape and geographical distribution. In other words, the recognition of new species is not solely the result of the cladistic analyses of cpDNA sequence data. The congruence between different sources of evidence (including molecular data) is the main argument in the taxonomic recognition of T. lepechinii sp. nov.
Even if the power of the rpl32 gene and rpl32-trnL intergenic spacer is not sufficient to discriminate against T. incana and T. elata [13], this molecular marker is useful for the molecular studies at low taxonomic levels [30], in particular for the cpDNA-based systematic studies of the cattails of T. sect. Ebracteolatae. For example, the obtained trees based on the rpl32 gene and rpl32-trnL spacer cpDNA sequence data ( Figure 11) are not only sufficient to discern T. lepechinii sp. nov. from T. shuttleworthioi, but also to show the polyphyly of the genetically heterogeneous T. latifolia, and also to clearly differentiate between morphologically related T. elata and T. caspica [1].
Again, the patterns in variation of this molecular marker are congruent with predominant geographical distributions of individual plants whose cpDNA was sequenced, and thus the predictive power of the marker is essential, even if it is still not the most variable region of the plastid genome of Typha [14]. Keeping in mind the results of the current comparative analyses of cpDNA sequence data (Figure 11), it is easy to predict, for example, that the populations of T. latifolia from the outmost South of Sakhalin, Russian Kuril Islands and the Far East of China (Pattern V), actually represent species from T. sect. Ebracteolatae, previously unknown for floras of China and Russia.
Unrooted phylogenetic analyses, especially if combined with a priori selected lumping style of treating diversity within Typha [13], as well within any other group of flowering plants, may have dangerous environmental protection connotations. In this paper it is enough to stress that just the taxon (e.g., species), not the "haplotype" or any semantically similar formal construct from the language of biosystematics or the phylogeography, is an actual object of protection in any Red List of Plants., for example, in the European Red List of Vascular Plants [31]. Due to this, the status of vulnerable or endangered plants must be accurately established within a clear taxonomic framework.
The communities with a dominance of T. lepechinii are normally moderate in size and typically have low taxonomic diversity [10]. General projective cover of such habitats varies between 50-100%; the overabundance of T. lepechinii itself changes from two to four if estimated following the Braun-Blanquet scale, with the ratio of the number of "generative" to "vegetative" shoots close to 2:3 [10]; the competitive potential of populations of T. lepechinii relative to T. latifolia is fairly low. As we already mentioned above, the new species was found either in natural areas and frequently, in secondary habitats [10]. Thus, T. lepechinii may be viewed as an illustration of the phenomenon of "extinction debt" [32]. From that, the secondary landscapes may be considered as a kind of temporary refuge, as they are in a way still helping this endangered cattail to survive within the current biota [32,33].