Introduction

The genus Festuca L. (Loliinae, Poaceae) is represented by 140 taxa in South America (Stančík 2003a; Darbyshire et al. 2003; Stančík and Peterson 2007). The greatest taxonomic diversity of this genus is found in the Andean mountains region, with only two native species, F. fimbriata (Parodi) Türpe and F. ulochaeta Nees ex Steud., confined to the lowland forests of south-eastern Brazil and north-eastern Argentina, respectively (Stančík 2003a; Stančík and Peterson 2007; Catalán and Müller 2012; Ospina et al. 2015). However, very little information is available about species that occur in small mountain ranges associated with the Andes (e.g. Sylvester et al. 2017) or places with a significant degree of endemism, as is the case of the Eastern Sierras Pampeanas in Argentina.

A large number of new Festuca species are described from South America based on morphological evidence (e.g. Renvoize 1998; Stančík 2001, 2003a,b, 2004; Müller 2006; Stančík and Peterson 2002, 2007). Phylogenetic studies (ITS, trnT-L and trnL-F) have revealed that Festuca, as traditionally circumscribed, is not a natural genus, but a large paraphyletic assemblage of distinctly related lineages with Lolium L., Vulpia C. C. Gmel. and several other genera falling within it (Catalán 2006; Catalán et al. 2007; Inda et al. 2008; Minaya et al. 2017). In this sense, one practical strategy has been to apply evidence from diverse sources to support the recognition of species (Wiens and Servedio 2000).

Morphometric analyses have recently been applied to clarify the taxonomy within Festuca (e.g. Foggi et al. 2006; Torrecilla et al. 2013; Ospina et al. 2013, 2016). The major problem linked to the delimitation of Festuca species in South America has been to distinguish between the species concept and the species criteria (as proposed by Wiens and Servedio 2000). Even the taxonomic identity and the geographic distribution of some Festuca species are controversial between taxonomic revisions and floristic catalogues, the reason being that they have been carried out within political boundaries (e.g. Parodi 1953; Türpe 1969; Tovar 1972; Nicora 1978; Matthei 1982; Renvoize 1998; Stančík 2003a; Zuloaga et al. 1994, 2008; Catalán and Müller 2012; Ospina et al. 2015).

One such taxon is Festuca lilloi Hack., which was described by Hackel (1914) as an endemic species from the Sierras Calchaquíes of the Eastern Sierras Pampeanas (northern Argentina). This taxon was considered a variety of F. setifolia Steud. ex Lechl., F. setifolia var. lilloi (Hack.) St.-Yves (Saint-Yves 1927), a native species from the Peruvian Andes. Later, Festuca lilloi was cited for Argentina and Peru, but the variety F. setifolia var. lilloi was synonymized under F. lilloi by Türpe (1969). Türpe also proposed ‘Festuca lilloi Hack. var. breviaristata’ as a new variety from Tucumán (northern Argentina), and this author associated this new taxon with the species Festuca setifolia. Alexeev in 1984 raised Türpe’s variety to the species level using the name Festuca tucumanica (as a replaced name). On the other hand, Festuca lilloi and Festuca tucumanica were recognized as two endemic species of northern Argentina (Zuloaga et al. 1994, 2008; Darbyshire et al. 2003), but the first taxon was also cited for the Peruvian and Bolivian Andes (Tovar 1972; Darbyshire et al. 2003; Renvoize 1998). Furthermore, Müller (2006) suggests that Festuca setifolia (from Peru) does not belong to the ‘F. lilloi group’ (F. lilloi, F. samensis Joch. Müll., F. tucumanica and other specimens without an identity assigned). However, the geographic distribution of F. lilloi was expanded to the Córdoba province (central Argentina, Eastern Sierras Pampeanas) because Catalán and Müller (2012) proposed that F. tucumanica is a synonym of F. lilloi.

Both species were recognized as dominant elements in plant associations of the highland grasslands and Polylepis forests of central Argentina (Cabido et al. 1981, 1989, 1998, 2010; Cabido and Acosta 1985a,b, 1986a,b). Many herbarium collections by Kurtz in 1885, Burkart in 1910–1940, and recent collections from the Sierras de Córdoba were assigned to F. lilloi, which are not in agreement with the diagnostic characters cited by Hackel (1914) for this last taxon, but recent fieldworks in the Sierras Grandes of Córdoba (Argentina) confirm the discovery of a new species of Festuca.

In this paper we describe the new species and clarify the identity and taxonomic concept of F. lilloi. Our study is based on herbarium specimens and new collections. Traditional and new relevant morphoanatomical characters were explored, including a phenetic analysis.

Material and methods

All data for morphoanatomical analyses were based on herbarium specimens from B, BAA, BM, CORD, G, GOET, HSB, K, LIL, LP, LPB, MVFA, NY, P, S, SI, US and W (Thiers 2018). Protologues and types specimens (or digital scan images from Global Plants on JSTOR, http://plants.jstor.org) were checked for all names. All accepted species names and their synonyms were represented in our phenetic analyses. A total of 64 specimens were selected for the phenetic analyses and listed in Table 1 (54 general collections and 10 type specimens); each specimen was recognized as an OTU – operational taxonomic unit (Sokal and Sneath 1963; Sneath and Sokal 1973). Only well-preserved specimens bearing both well-developed leaves and inflorescences were considered in our analyses. The geographic distributions of all species are presented in the Fig. 1. Complementarily, recent collections of the new species were obtained by the second author within the framework of the project ‘Altitudinal variation of plant species richness at Sierras Grandes (Córdoba, Argentina)’ and Dr Leopoldo Iannone during a field trip focused on the study of the ‘Diversity of endophytes belonging to the genus Epichloë (Clavicipitaceae, Ascomycetes) in native grasses of Argentina’.

Table 1 List of the plant materials used in this study
Full size table
Fig. 1
figure 1

Distribution of Festuca hieronymi, F. fiebrigii, F. lilloi, and F. giraldoi J.C. Ospina and P.I. Picca, according to the results of the present study. Dots indicate specimens from Table 1

Full size image

We observed macromorphological characters in all specimens under a Wild Heerbrugg M5-26799 stereo microscope. Twelve characters (seven vegetative and five reproductive) were selected for our phenetic cluster analysis (Table 2). The morphological characteristics were described following the terminology used by Lawrence (1955) and Aranda and Forceck (1990). Foliar anatomy attributes were observed in cross sections from the middle area of the blade. Transversal cuts were obtained following the methodology used by Metcalfe (1960). The transversal cuts were fixed using a semi-permanent mounting medium (Ruzin 1999). A total of four informative characters were retained (Table 2). The anatomical observation and photos were taken under a Nikon SMZ 800 microscope with a mounted camera controlled by Nikon NIS elements software (Fig. 2). The anatomical description characteristics of the new species were described following the terminology used by Ellis (1976). Foliar epidermis micromorphology attributes were observed on both the adaxial and the abaxial surface. A total of five characters were selected (Table 2). Lemma epidermis micromorphology was observed on the abaxial surface from the basal anthecium of a well-developed spikelet. A total of three characters were retained (Table 2). All samples were coated with gold (40%) and palladium (60%) for scanning electron microscopy (SEM) and images were obtained under a TEM Philips EM 301 instrument (Figs. 3, 4). Epidermal features were described following the terminology of Ellis (1979).

Table 2 Morphoanatomical characters used in the phenetic cluster analysis
Full size table
Fig. 2
figure 2

Dendrogram derived from the Ward’s cluster analysis of specimens of Festuca hieronymi, F. fiebrigii, F. lilloi and F. giraldoi. Shading differentiates the three main clusters (A, B, C). Grey shade indicates the specimens of the new species. Representative characteristics of leaf-blade anatomy characteristics are shown for each of the clusters

Full size image
Fig. 3
figure 3

SEM images of leaf blade epidermis (all pictures orientated to distal apex). a – Adaxial leaf-blade in F. hieronymi; b – abaxial leaf-blade in F. hieronymi; c – adaxial leaf-blade in F. fiebrigii; d – abaxial leaf-blade in F. fiebrigii; e – adaxial leaf-blade in F. lilloi; f – abaxial leaf-blade in F. lilloi; g – adaxial leaf-blade in F. giraldoi; h – abaxial leaf-blade in F. giraldoi [a–b From Burkart 7228 (SI), c–d from Morrone et al. 4359 (SI), e–f from Burkart 5298 (LIL), g–h from P. Piccaand M. Apellaniz 2051 (SI)]. Scale bar = 1 mm in a, c, e and g; scale bar = 1 μm in b, d, f and h

Full size image
Fig. 4
figure 4

SEM images of abaxial lemma epidermis and details of the medial zone of basal anthecium (all pictures orientated to distal apex). a–f – Festuca hieronymi; b–g – Festuca fiebrigii; c–h – Festuca lilloi; d–i and e–j – Festuca giraldoi [a–f from Burkart 7228 (SI), b–g from Morrone et al. 4359 (SI), c–h from Burkart 5298 (LIL), d–i from Burkart 10173 (SI), e–j from P. Picca and M. Apellaniz 2051 (SI)]. Scale bar = 1 mm for all images

Full size image

Dataset characteristics were obtained and a phenetic analysis was carried out based on 24 characters (Table 2). The dataset contained both morphological and anatomical characters for the 64 OUTs. Correlation coefficients among characters were computed to reveal highly correlated characters (Escobar et al. 2011). The dataset had no missing data. Variables describing quantitative characters were standardized by subtracting their mean and dividing the result by the standard derivation; semi-quantitative characters were log-transformed (Sneath and Sokal 1973; Marhold 2011). The phenetic analyses used Ward’s cluster method (minimization of the increase of the error sum of squares) with the Gower distance coefficient (Gower 1971). The co-phenetic correlation coefficient (r) was estimated to measure the distortion between the original matrix and the obtained phenogram (Sneath and Sokal 1973; Sebola and Balkwill 2009; Ospina et al. 2013, 2016). Statistical analyses were carried out using NTSYS-pc v. 2.21c (Rohlf 2009).

The species under study were all delimited on the basis of non-overlapping patterns of morphological variation as the primary criterion for inferring species boundaries. Phenetic clustering analyses were employed for this purpose, and the OTUs formed hierarchical non-overlapping groups strictly according to the degree of similarity (Sokal and Corvello 1970; de Queiroz and Good 1997; de Queiroz 2007).

The characterization of habitat and ecological preferences follows the ecoregions proposed by Pereyra (2003) based on geological, edaphic, climatic, landscape and vegetation aspects. Additionally, we used the work of Köppen (1923) and Drobe et al. (2011), who studied the geodynamic evolution of the Pampean Sierras, as well as information available in phytogeographical works (Cabrera 1957; Cabrera and Willink 1973) or catalogues specialized in grasses (Nicora and Rúgolo de Agrasar 1987).

Results

We used 30 characters in the initial cluster analyses; however, six of the characters showed correlations (ligules extended in auricles laterally, shape of the lobes of the auricles, presence or absence of ribs or furrows on the abaxial surface, presence or absence of prickles on the abaxial surface of anthecium, shape of the abaxial surface of the anthecium, and shape of the anticlinal horizontal walls on the abaxial surface of the anthecium). Therefore, 24 variables Ward’s cluster analyses aimed to test the validity of the current species delimitation (Table 2). The distribution of all the 24 variables departed from the normal distribution.

Four characters with taxonomic value were identified and explored for the first time in the species under study. These characters are marked with asterisks (*) in Table 2. These potential diagnostic characters were used in the morphoanatomical comparison between Festuca lilloi and its closest related species from the Sierras Pampeanas of Argentina (Table 3).

Table 3 Morphoanatomical comparison of Festuca lilloi with its closest relatives species
Full size table

Ward’s phenetic cluster analyses resulted in a dendrogram (Fig. 2) that divided all OTUs into three main groups: cluster I comprising only specimens of the new species, cluster II comprising all specimens of F. lilloi, and cluster III comprising two subgroups referred to as specimens of F. hieronymi and F. fiebrigii. Our phenetic Ward’s cluster analyses showed low values of distortion (r = 0.98), indicating a complete success of separation without any misidentified specimens.

All specimens of Festuca fiebrigii and F. hieronymi were similar in leaf anatomy (Fig. 2, Table 3), having an expanded to rolled and flattened shape outline, differentiated costal and intercostal zones on the abaxial surface, and 5–6(–7) vascular bundles present between the midrib and the margin of the blade, situated close to the abaxial surface and at the same level. Both these species were differentiated from F. lilloi and all specimens of the new species in our leaf anatomy analysis (Fig. 2, Table 3) by: outline ‘U-shaped’ and rounded with angled arms, the costal and intercostal zones not differentiated on the abaxial surface, and 2–3 vascular bundles present between the midrib and the margin of the blade, all located in the centre of the blade at the same level. The observed anatomical differences confirm that these taxa are not closely related (as cited by Türpe 1969). The SEM observation revealed that attributes of the epidermal lemma were important diagnostic characters for separating F. lilloi and determining the new species proposed herein (Figs. 3, 4).

Specimens from northern Argentina (Western Sierras Pampeanas of Tucumán) can be considered a separate species from F. giraldoi of central Argentina (Eastern Sierras Pampeanas of Córdoba) because the two groups are geographically separate (Fig. 1) and a result of unlinked geodynamic evolution (Drobe et al. 2011). In addition, the Eastern Sierras Pampeanas have a more humid climate than the Western ones. The dominant climatic type is steppe semiarid (Bs) with transitions between the wet mesothermal (Cf) and wet temperate (Cw), see Köppen 1923 and Pereyra 2003. Likewise, the characteristic vegetation of these mountains consists of xerophilous forests in the lower zones and open grasslands and patches of Polylepis australis above 1,500 m a.s.l. (Pereyra 2003).

New taxon

Festuca giraldoi J.C. Ospina and P.I. Picca, sp. nov. (Fig. 5)

Fig. 5
figure 5

Festuca giraldoi J.C. Ospina and P.I. Picca. a – Habit and inflorescence; b – ligular zone; c – apex of blades; d – spikelet; e – glumes; f – floret dorsal view; g – floret ventral view; h – anther; i – ovary and lodicules [from P. Picca and M. Apellaniz 2051 (SI)]

Full size image

Type: ARGENTINA. Córdoba Province, San Alberto Department, Sierras Grandes-Pampa de Achala, 1,967 m a.s.l., 31° 36′ 38.196″ S – 64°54′40.608″ W, 15 February 2011, P. Picca and M. Apellaniz 2051 (holotype: SI 202281!; isotype: BAB!).

Plants

(50–)60–90 cm tall (infertile culms), caespitose. Floriferous culms (95–)115–130 cm tall, erect, 2-nodes. Leaf sheaths scabrous at apex, not shiny, papyraceous, closed along their entire length, ligular zone not differentiated. Auricles of the leaf sheaths absent. Pulvinus of the leaf sheaths absent. Ligules 3–3.5 mm long, apex obtuse, margin membranaceous-ciliate, extended in auricles laterally. Blades 35–65 cm × 1.5–2 mm, erects to decumbent, convoluted, not rigid, apex shortly awned, not pungent; abaxial surface sparsely pilose; adaxial surface pilose. Panicles 9–16 cm long, linear, contracted, greenish to pale; branches ascendant, scabrous, with rudiments of aborted branches in axils of branches arranged in a rachis. Spikelets 19–23 mm long, with 5–6 florets, greenish to pale, rachilla scabrous, glabrous. Glumes subequal, scarcely scabrous dorsally and at apex, glabrous, nerves inconspicuous; lower glume 5–5.5 × 0.3–0.5(–0.7) mm, linear, apex shortly acuminate, 1-nerved; upper glume 5.5–6.3 × 0.8–1.2 mm, oblong-elliptic, apex acute and shortly acuminate, 3-nerved. Lemma of the basal florets 8–11 mm long (without awn), narrowly lanceolate, densely scabrous, with spherical papillae, apex awned, awn 2–3 mm long, nerves inconspicuous, 5-nervous. Paleas equal or shorter than lemmas, scarcely and shortly pilose at apex on keels. Anthers 4.5–5 mm long. Lodicules 1.5–2 mm long, lobed, margin fimbriated, glabrous. Ovaries densely hairs at apex. Caryopsis 2.5–3.5 mm long.

Distribution and ecology

Known only from the Pampa de Achála and surrounding mountains, Córdoba province, central Argentina. Festuca giraldoi grows as a dominant element in the grass matrix characteristic of the ecosystems that develop at elevations around 2,000 m a.s.l., together with Plantago myosuros Lam. (Plantaginaceae); Aristida laevis (Nees) Kunth, Jarava juncoides (Speg.) Pañail., Deyeuxia colorata Beetle, Schizachyrium spicatum (Spreng.) Herter, Paspalum quadrifarium Lam., Koeleria kurtzii Hack. ex Kurtz, Bulbostylis juncoides (Vahl) Kük. exOsten (Poaceae); Carex boliviensisVan Heurck and Müll. Arg., Carex distenta Hack. ex Kurtz (Cyperaceae); Juncus pallescens var. achalensis (Barros) Novara (Juncaeae), Neobartsia crenoloba (Wedd.) Uribe-Convers and Tank (Orobanchaeae); Berberis sp. (Berberidaceae); Blechnum laevigatumCav. (Blechnaceae); Botrychium australe R.Br. (Ophioglossaceae); Dichondra aff. Microcalix (Convolvulaceae); Conyza primulifolia (Lam.) Cuatrec. and Lourteig., Stevia satureifolia (Lam.) Cav., Acmella decumbens (Sm.) R.K. Jansen var. affinis (Asteraceae); Pfaffia gnaphaloides (L. f.) Mart. (Amaranthaceae) and Lachemilla pinnata (Ruiz and Pav.) Rothm (Rosaceae).

Etymology

We are pleased to dedicate this new species to our colleague and friend Diego Giraldo-Cañas, Professor of the Universidad Nacional de Colombia, an outstanding botanist and agrostologist, and respected explorer of the flora of Argentina.

Paratypes

ARGENTINA. Córdoba: San Alberto, Sierras Grandes-Pampa de Achala, 1967 m, 12 Feb 2011, P.I. Picca and M. Apellániz 2040 (type, SI); Sierras Grandes-Pampa de Achala, 1967 m, 13 Feb 2011, P.I. Picca and M. Apellániz 2045 (paratype, SI). Calamuchita, Pampa de Achala, 15 Jan 1910, A. Burkart 10168, 10173 (paratype, SI). Punilla, Pampa de Achala, Jan 1940, A. Burkart 10202 (paratype, SI); entre Tanti y Los Gigantes 1,600–1,700 m a.s.l., 11 Jan 2013, L. Iannoneand and M.C. Cargo 113 (paratype, SI); entre Los Gigantes y Taninga 1,550–1,750 m a.s.l., 11 January 2013, L. Iannoneand and M.C. Cargo 115 (paratype, SI). Santa Maria, entre estancia San Miguel y Cuesta del Corral, 15 Dec 1885, F. Kurtz 2960 (paratype, CORD).

Comments

The new species is morphologically similar to F. lilloi, but the latter is easily recognized by its ligular zone not differentiated and its distinctly lanceolate-acute to subulate glume shape (vs ligular zone differentiated and glumes linear in the new species); the ligular apex of F. lilloi is denticulate or emarginate with lobes acute (vs entire and obtuse in the new species). The main morphological features separating the new species from F. fiebrigii are its floriferous culms with 1–2 nodes (vs 4–5, in F. fiebrigii) and its lemma apex with an awn 2–3 mm long (vs without an awn). Finally, the ligular features distinguishing the new species from F. hieronymi are its blade apex obtuse (vs truncate) and longer ligules of 3–3.5 mm (vs 0.5–1 mm). We therefore found these taxa to be clearly distinguishable, contrary to Türpe (1969), Tovar (1972), Renvoize (1998), Müller (2006), and Catalán and Müller (2012).

The foliar anatomy of Festuca giraldoi is characterized by a ‘U-shaped’ outline that is rounded with angled arms, with 2–3 vascular bundles present between the midrib and margin of the blade, all located in the centre of the blade at the same level, all located in the centre of the blade at the same level, the central rib is located in the central area of the blade, with associated sclerenchyma layers on both surfaces. Abaxial surface with straight edges forming angles associated with the vascular bundles, ribs angular, composed of a sclerenchyma block and found opposite to all vascular bundles, prickles present in the costal zone, macro-hairs absent. Adaxial surface markedly irregular, with rounded ribs situated opposite to all vascular bundles, lacking a sclerenchyma block, prickles present and scarce in the intercostal and margin zones, macro-hairs present and densely covering all the surface (Fig. 3).

The leaf epidermal micromorphology in Festuca giraldoi is characterized by an abaxial surface with intercostal and costal zone having rectangular long cells, arranged in rows with one short cell between each successive long cell, anticlinal horizontal walls deeply undulating in a ‘Ω-shape’, anticlinal vertical walls very thick at right angles to the horizontal walls containing silicified short cells that have a rounded shape, and prickles and stomata absent. Adaxial surface intercostal and costal zone with rectangular long cells, arranged in rows with one macro-hair, anticlinal horizontal walls smooth, anticlinal vertical walls angled, short cells absent, stomata flat-topped with horizontal walls of subsidiary cells rounded, stomata present only in the intercostal zone, prickles present only on the blade margin, and macro-hairs curved and abundant in both zones (Fig. 3e–f).

The epidermal micromorphology of Festuca giraldoi is characterized by a lemma with an intercostal zone with rectangular cells on the adaxial surface, arranged in rows with one short cell between each successive long cell, anticlinal horizontal walls deeply undulating in a ‘Ω-shape’ and very thick anticlinal vertical walls at right angles to the horizontal walls containing silicified short cells that have a rounded shape.

Discussion

Our present study found that some species of Festuca from the Eastern Sierras Pampeanas of central Argentina (Córdoba province) are quite similar in their morphology to species from the Western Sierras Pampeanas of Argentina (Tucumán province). In our opinion, this situation has led to confusion in the identification of herbarium or field material analysed by Cabido et al. (1981, 1989, 1998, 2010) and Cabido and Acosta (1985a, b, 1986a, b). In our phenetic cluster analysis based on morpho-anatomy and epidermal traits (Figs. 2, 3, 4; Table 3), 64 OTUs were clearly separated into four groups corresponding to F. fiebrigii, F. hieronymi, F. lilloi and a hitherto unknown taxon, described here as Festuca giraldoi. Our phenetic analysis demonstrates that at least two distinct systematic units may be recognized (F. lilloi and F. giraldoi) and that Festuca lilloi is not related to other taxa (F. fiebrigii, F. hieronymi) as stated in previous taxonomic works.

The first classification of the genus Festuca distinguished two artificial groups called ‘fine-leaved’ (or setaceous-leaved) and ‘broad-leaved’ (or flat-leaved) on the basis of leaf morphoanatomical characteristics (Hackel 1882). This classification was used by numerous other authors (Saint-Yves 1927; Türpe 1969; Tovar 1972; Matthei 1982; Clayton and Renvoize 1986; Renvoize 1998; Müller and Catalán 2006). Although the existence of these two artificial groups was confirmed by recent molecular phylogenies (Torrecilla and Catalán 2002; Catalán et al. 2004; Catalán 2006; Inda et al. 2008), the leaf morpho-anatomical patterns of numerous Festuca taxa form an intermediate phylogenetic clade between the two characteristics, and some broad-leaved taxa are located within the fine-leaved group. Analogous morphological similarity was found between F. lilloi and F. giraldoi (fine-leaf) and F. fiebrigii and F. hieronymi.

More recent studies, by contrast, have used epidermal evidence on the leaf and lemma to address specific questions surrounding Festuca species (Hackel 1882; Türpe 1969; Tovar 1972; Zarinkamar and Jouyandeh 2011; Ortúñez and Cano-Ruiz 2013). Our results corroborate the utility of foliar epidermal characters in the taxonomic determination of groups of specific specimens, since these have a constancy in epidermal characters which permits the identification of each of the four species recognized in our phenetic cluster analyses (Figs. 2, 3, 4; Table 3). These characters are very useful for the separation of F. lilloi from the homogeneous and well-defined species such as F. giraldoi (described herein; Figs. 2, 3, 4), which allows to classify these species as two different taxonomic units. From our results (Figs. 2, 3, 4; Table 3). Evidently, the individuals of Festuca lilloi sampled exhibited homogeneous morphology (Figs. 2, 3, 4, Table 3), rejecting their splitting into numerous taxa in previous taxonomic works.

The specimens of F. setifolia from the Western Sierras Pampeanas (Tucumán) can be considered a separate species from F. giraldoi of central Argentina (Eastern Sierras Pampeanas of Córdoba) because the two groups are geographically separate (Fig. 1) and morphoanatomically distinct (Table 3). In addition, the geodynamic evolution differs between the Western Sierras Pampeanas, Tucumán and the Eastern Sierras Pampeanas, Córdoba (Drobe et al. 2011). Compared to the Western Sierras Pampeanas, the Eastern Sierras Pampeanas are an older (Precambrian–Lower Cambrian vs Cambrian–Ordovician), lower-altitude mountain system (up to 3,000 m a.s.l. vs 5,000 m a.s.l.) with a humid climate and more developed, nutrient-rich soils (Pereyra 2003).

Our phenetic analyses of morphoanatomical (including epidermal) characters and habitat preferences of four species of Festuca occurring in central Argentina allow us to conclude that (1) every Festuca species considered in this study presents unique leaf anatomy attributes and that (2) our investigation of morphoanatomical characters in different specimens of F. giraldoi (collected at different times) shows that its variability of its species-specific traits is significantly different from that of other, related species (see Türpe 1969; Tovar 1972; Renvoize 1998; Müller 2006; Catalán and Müller 2012).