Pholiurus

Röser, Julia Schneider Grit Winterfeld Martin, 2012, Polyphyly of the grass tribe Hainardieae (Poaceae: Pooideae): identification of its different lineages based on molecular phylogenetics, including morphological and cytogenetic characteristics, Organisms Diversity & Evolution (New York, N. Y.) 12 (2), pp. 113-132 : 126-129

Pholiurus and allies ( lineage IV )

Based concordantly on both sequence datasets, the monotypic genus Pholiurus (only P. pannonicus ) is firmly nested

Fig. 5 a–h Karyopses. a

Hainardia cylindrica and b

Parapholis strigosa with short linear hilum. c Pholiurus pannonicus with oval hilum. d,

e Narduroides salzmannii with short linear hilum and longitudinal furrow of the karyopsis adherent to the palea.

f, g Agropyropsis lolium with embryo, short linear hilum, and large apical ovary appendage.

h Scribneria bolanderi with comparatively large embryo.

Material used: a fruits from

Göttingen Botanical Garden in

2006, no. 1850 (HAL); b

Werner 3477 (HAL 71657); c

Sadler s.n. (HAL 112064); d, e

Hernández s.n. (C); f, g Cosson

s.n. (JE); h Wilken 16163 and

Painter ( RSA 695253)

in lineage IV ( Poa lineage), and is thus rather distant to Hainardia and Parapholis . Pholiurus is close to Arctophila , Bellardiochloa , Dupontia , Hookerochloa incl. Festucella , Nephelochloa , and Ventenata (Figs. 1 and 2). Monophyly of these taxa is supported by a common insertion of 6 bp in the intron region between mat K gene and the 3′ trn K exon. Although Nephelochloa and Pholiurus have not been investigated in any previous molecular phylogenetic study, this assemblage of genera and its comparatively close relationship to the genus Phleum was found principally also in a number of previous studies ( Soreng and Davis 2000; Hunter et al. 2004; Gillespie et al. 2007, 2008, 2009, 2010; Döring et al. 2007; Davis and Soreng 2007; Soreng et al. 2007; Döring 2009; Schneider et al. 2009). Phleum resolves with support in lineage IV based on the cpDNA sequence data, whereas ITS shows it next to this lineage in a polytomy with lineage V (Figs. 1 and 2). Such an unstable position of Phleum is encountered frequently: CpDNA usually suggests a placement compatible to that seen in Fig. 1, irrespective of the particular markers used ( Döring et al. 2007; Gillespie et al. 2008, 2009; Döring 2009; Schneider et al. 2009). Nuclear ITS or combined ITS/ETS placed Phleum in more-or-less the same polytomy with taxa of lineage V as seen in Fig. 2 ( Quintanar et al. 2007; Gillespie et al. 2008, 2009, 2010;

and ploidy levels in exemplary species of the Aveneae / Poeae tribe complex

Schneider et al. 2009). A combined analyses of respective cp and nuclear ITS or cp and nuclear ITS/ETS datasets placed Phleum (and Milium ; cf. below) with moderate support even amongst other genera of the Poa lineage, where both genera were not placed by the separate analyses of the individual cp and nuclear datasets. This somewhat surprising result was obtained independently in different studies using a different set of taxa and different DNA regions (cp mat K and nuclear ITS: Schneider et al. 2009; cp trn T– trn F and nuclear ITS/ETS: Gillespie et al. 2009, 2010). The precise position of Phleum , however, differed in both studies.

Morphologically, Pholiurus —until now considered to be part of tribe Hainardieae or subtribe Parapholiinae — resembles Hainardia and Parapholis , due to its spicate inflorescences with hollowed rachis, sunk-in spikelets and the strongly thickened glumes with prominent nerves ( Fig. 3a, b, f). The inflorescence rachis, however, is tough in Pholiurus and the spikelets disarticulate below the glumes, which thus fall in their entirety ( Table 3; Fig. 3f). The latter character is not found in lineage I with Hainardia and Parapholis (Figs. 1 and 2), but occurs in some genera of the Poa lineage ( lineage IV ), in which Pholiurus is placed by the molecular phylogenetic data. These genera with spikelets falling entire were previously ascribed to either a broadly defined tribe Phleeae ( Tzvelev 1989) or to broadly or narrowly defined subtribes Phleinae and/or Alopecurinae , respectively, and a further subtribe acknowledged only in some classifications, namely Cinninae (cf. Tzvelev 1976; Clayton and Renvoize 1986; Soreng and Davis 2000; Quintanar et al. 2007; Soreng et al. 2003; 2007; Gillespie et al. 2008, 2010).

The genera Alopecurus , Beckmannia , Cornucopiae , and Limnas (subtribe Alopecurinae sensu Gillespie et al. 2010 ), Cinna and Limnodea (subtribe Cinninae sensu Soreng et al. 2007 , or included with subtribe Poinae sensu Gillespie et al. 2010 ) and additionally Rhizocephalus (subtribe Phleinae sensu Tzvelev 1976 and Gillespie et al. 2010) all have spikelets falling entire like Pholiurus and not disarticulating above the glumes or below the floret(s). The spikelets, however, are always one- (or rarely two-)flowered in Beckmannia instead of two-flowered as in Pholiurus . The hilum of the caryopses is rounded and the endosperm seems to be soft in most if not all of these genera ( Clayton and Renvoize 1986; Watson and Dallwitz 1992; 1992 onwards; Clayton et al. 2006 onwards) as newly reported here also for Pholiurus ( Table 3; Fig. 5c), in contrast to Agropyropsis , Hainardia , Narduroides , Parapholis and Scribneria with punctiform or shortly linear hilum ( Table 3; Fig. 5a, b, d–f). Fruit data are seemingly missing for Limnas .

Judging from morphology, we expect the closest extant relatives of Pholiurus among the assemblage around Alopecurus with allied genera showing the same type of spikelets falling entire and with a spikelet abscission zone below the glumes. To address this question using molecular phylogenetic data, we have started a denser sampling of potential target taxa.

Phleum , with its segregate genera Maillea and Pseudophleum (subtribe Phleinae sensu Tzvelev 1976 ; Gillespie et al. 2010), has spikelets disarticulating above the glumes ( Conert 1979 –1998; Watson and Dallwitz 1992; 1992 onwards) in contrast to Rhizocephalus ( Bor 1970; Tzvelev 1976, 1982), although this has not always been reported correctly or was overlooked (e.g. Clayton and Renvoize 1986; Gillespie et al. 2010: p 591). Altogether a maintenance of the subtribes Alopecurinae and Phleinae as recently reinstated and delineated ( Gillespie et al. 2010) needs to be examined further.

Figures

Fig. 2 Bayesian phylogenetic tree obtained from nuclear ITS DNA sequences. Branches collapsed in the strict consensus tree based on maximum parsimony analysis are in bold. Posterior probabilities ≥0.95