In the footsteps of Michel Allioni: Liverworts and hornworts from the surroundings of Gualaquiza (Ecuador)

Abstract Gradstein, S.R. & Á.J. Pérez (2021). In the footsteps of Michel Allioni: Liverworts and hornworts from the surroundings of Gualaquiza (Ecuador). Candollea 76: 41–54. In English, English and Spanish abstracts. The liverwort and hornwort flora of Ecuador is still very incompletely known. We report three species new to Ecuador (Lejeunea rionegrensis Steph., Notothylas javanica (Sande Lac.) Gottsche, Riccia vitalii Jov.-Ast) and 38 new to the province of Morona Santiago based on our fieldwork in the surroundings of Gualaquiza, Morona Santiago, in 2019. The bryophyte flora of this area was explored in the early 20th century by the Italian missionary and naturalist Michel Allioni. A brief eulogy of the life and work of Michel Allioni is presented together with a revision of the species described by Stephani based on Allioni's Ecuadorian collections. Almost all of them are synonyms. Sixteen new synonyms and seven new lectotypifications are proposed.


Introduction
The liverwort and hornwort flora of Ecuador has been much inventoried since the 19th century, yet is still very incompletely known (León-Yánez et al., 2006;Schäfer-Verwimp et al., 2013;Gradstein & Benitez, 2017). Rather well-explored areas include the province of Pichincha, the Galápagos Islands, the upper Pastaza river valley and readily accessible sites in Parque National Podocarpus and adjacent areas in the south of Ecuador. Other parts of the country are underexplored (Gradstein, in press).
The province of Morona Santiago, located in the southeast of country, counts among the areas that have been poorly explored. Even though including a large part of Amazonian lowland territory and Andean cordillera, only 140 species of liverworts and 4 hornworts have been recorded from this province or about three times less than are known from the adjacent province of Zamora Chinchipe (León-Yánez et al., 2006;Gradstein, in press). A remarkable recent addition to the flora of Morona Santiago was the robust dendroid liverwort Riccardia verticillata Gradst. & Reeb from Reserva El Quimi in the Cordillera del Cóndor (Gradstein et al., 2019).
During the fall of 2019, the authors of this paper had the opportunity to collect liverworts and hornworts in the surroundings of Gualaquiza, a town in the south of Morona Santiago (canton Gualaquiza) at c. 850 m elevation, 3°24'13"S, 78°34'54"W ( Fig. 1). Surprisingly, our collections yielded numerous new species records for the province. The area of Gualaquiza is characterized by a humid tropical climate with high rainfall. The natural vegetation is dominated by evergreen tropical forests with abundance of palms (Iriartea deltoidea Ruiz & Pav., Mauritia flexuosa L.f., Wettinia maynensis Spruce) as well as Lauraceae, Melastomataceae, Moraceae and Annonaceae. Bryophytes abound as epiphytes on trunks and branches, on soil and on living leaves. Currently, the natural vegetation of Gualaquiza is threatened by mining, livestock and selective extraction of fine woods (Roy et al., 2018;Mazabanda et al., 2020;Pérez et al., 2020). However, forest remnants protected by private and state initiatives still remain in the area and are in need of study.
The bryophyte flora of the surroundings of Gualaquiza has been inventoried in the early 20 th century by the Salesian missionary Michel Allioni. Many new species and the new genus Allioniella Broth. (= Alloniellopsis Ochyra) have been described based on his collections (Brotherus, 1911(Brotherus, , 1920Stephani, 1898Stephani, -1924. Virtually no bryophyte collecting has been done in the area since Allioni's explorations and very little is known about this collector. A brief eulogy of Allioni's (short) life and work is therefore given here.
Michel ("Miguel") Allioni (1880Allioni ( -1912 was an Italian naturalist and missionary of the Salesian priests of Don Bosco who lived and worked during about three years in Gualaquiza, Ecuador (Prov. Morona Santiago), as head of the local Salesian mission. Allioni had studied natural sciences, especially botany, at the University of Torino and arrived in Gualaquiza on 28 January 1909 (Allioni, 1909). The mission of Gualaquiza, established in March 1894, was the first Salesian mission in the Ecuadorian Amazon and included the vallies of the Gualaquiza and Bomboiza rivers in the south of Morona Santiago (García, 1999). During his stay in Ecuador, Allioni did extensive ethnographic and bryological work. His ethnographic studies resulted in a detailed account of the life of the indigenous people of the Gualaquiza region, "La Vida del Pueblo Shuar". This book, which was published posthumously in 1978 and again in 1993 (Allioni, 1978(Allioni, , 1993, remains an important source of information on the Shuar indigenous people of Ecuador (Bottasso, 1993). Allioni's stay in Ecuador did not last very long; tragically, he died of yellow fever on 23 May 1912 during a brief stay in Guayaquil (Creamer, 2011). Some sources (e.g., Steere, 1948;Renner, 1993) erroneously cited 1911 as the year of Allioni's death, which was probably a misrendering of Brotherus's brief text on Allioni (Brotherus, 1920: 1).
In the framework of his bryological explorations, Michel Allioni made hundreds of collections, mostly in the surroundings of Gualaquiza and Bomboiza but also in neighbouring areas; for a listing of Allioni's collecting localities see Appendix 2, and Brotherus (1920) and Steere (1948). Although being mostly from the province of Morona Santiago, Allioni's specimens are frequently cited as having being collected in the province of Azuay. The reason for this is that Morona Santiago was long part of the province of Azuay (together with the provinces of Napo and Pastaza), having been created as a separate province only in 1953. Allioni's labels therefore frequently mention Azuay as province of origin of the specimen, instead of Morona Santiago. The only collections made by Allioni in Azuay, as currently defined, are from Chunchi and Cuenca, gathered in January 1909 while travelling from Guayaquil to Gualaquiza, and from the páramo of Matanga near the provincial border with Morona Santiago, visited by Allioni in October 1909 andagain in November 1910. Allioni's bryophytes include hundreds of collections. A notable feature of his specimens are the detailed labels, providing information (in Latin) on location, habitat, substrate, elevation and date of collecting (Fig. 2, 3). As mentioned by Bottasso (1993: 23), Allioni's approach stood out by its "exactitud, minuciosidad, gusto por la clasificación, descripciones sintéticas, curiosidad insaciable". In spite of his dedication to the collecting of bryophytes, Allioni's contribution to bryology has been limited to gathering specimens. His early death at the age of 32 may well have prevented him to develop a career as a bryologist.
Allioni's moss collections were worked up by Victor F. Brotherus (1849Brotherus ( -1929 in Finland, who published two papers on these materials, an annotated species list (Brotherus, 1920) and the description of a new genus, Allioniella (Sematophyllaceae). The latter genus -based on a specimen gathered by Allioni along the Río Bomboiza (Brotherus, 1911) -is still accepted as a good taxon but the name is illegitimate, having been earlier used for a genus of Nyctaginaceae, Allioniella Rydberg 1902, dedicated to the Italian botanist Carlo Allioni. As a consequence, the name Allioniella Broth. has been replaced by Allioniellopsis (Ochyra, 1982). Allioni's mosses counted 160 species including 36 species new to science, about half of which are still accepted (Churchill et al., 2000). The specimens are kept in H and correspondence between Allioni and Brotherus is preserved in the Brotherus archives in the University of Helsinki Library (Koponen & Piippo, 2002). While Allioni's moss collections have been well documented, less is known about his collections of liverworts and hornworts. The latter were worked up by Franz Stephani (1842Stephani ( -1927 in Leipzig, who probably received the material as a gift for identification from Emil Levier (1838 -1911), the Swiss born medical doctor and botanist in Florence who maintained a large private herbarium. The collections are kept in G and are labeled "Bryotheca Ecuadorensis M. Allioni" with Allioni's collecting number (Fig. 2) or "Bryotheca E. Levier" [Bryo. Levier] with Levier's herbarium number and without collecting number (Fig. 3). Specimens serving as types of new species are documented in the online Catalogue of the Geneva Herbarium (CHG, 2020). A list of the liverwort and hornwort species collected by Michel Allioni has not been published, however. Also, there is no correspondence between Allioni and Stephani in the Stephani archives kept in the library of herbarium G. All species identifications were incorporated in Species Hepaticarum (Stephani, 1898(Stephani, -1924 without citation of specimens, except for the types of new taxa which are briefly referred to in this work as "Ecuador, Allioni". Sixty-two new species were described by Stephani based on Allioni's collections, including 58 of liverworts and four of hornworts (CHG, 2020). As shown in the present paper, however, almost all new species described by Stephani were ill-defined and are synonyms. Stephani's identifications also included species that were already known (a random check of the taxonomic literature revealed five species; see below), but their total number is unknown due to the lack of a species list.
The purpose of the present paper is to present the results of our 2019 fieldwork together with a revision of the liverwort and hornwort collections made by Michel Allioni, especially those described as new species.

Material and methods
We examined about 150 collections of liverworts and hornworts from the surroundings of Gualaquiza: c. 90 collections (including 85 type specimens) made by M. Allioni during 1909Allioni during -1910 and kept in G, and 60 collections kept in QCA and PC made by the authors during 30 September -1 October 2019 (Appendix 1, 2). All collections were identified or revised using the manual of the liverworts and hornworts of Colombia and Ecuador (Gradstein, in press). Candollea 76, 2021 Liverworts and hornworts from the surroundings of Gualaquiza (Ecuador) -43

Collections made by Michel Allioni
We recognized 48 species among the collections made by Allioni in the surroundings of Gualaquiza (Appendix 2). Among these are four species that were described by Stephani as new taxa based on Allioni's material: Isotachis allionii Steph.
( Neesioscyphus allionii (Steph.) Grolle), Isotachis obtusa Steph., Metzgeria allionii Steph. and the hornwort Dendroceros allionii Steph. The status of the latter species requires verification, as the type (Allioni 319 [G00060907]) was on loan and not available for study. The remaining new taxa described by Stephani based on Allioni's collections, including 56 species of liverworts and two of hornworts, proved to be synonyms of earlier described taxa. Sixteen synonyms are newly proposed in this paper; the remaining synonyms were already established in previous studies and are cited in Appendix 2 with their current accepted name (except for a few names from the provinces of Azuay and Guayas; for these see Grolle, 1971;Heinrichs, 2002;Burghardt & Gradstein, 2008;Uribe Meléndez, 2008). The high synonymy rate (93.5 %) agrees with findings of other studies on Stephani types (e.g., Verdoorn, 1934;Gradstein, 1975;Heinrichs, 2002;Dauphin, 2003;Ilkiu-Borges, 2016) and reflects the reckless approach of Franz Stephani in describing new species. Approximately 4000 new species of liverworts and hornworts were described by Stephani, the majority in Species Hepaticarum (Stephani, 1898(Stephani, -1924; it is estimated that more than 85 % of these are synonyms (Gradstein, 2006). The sixteen species described by Stephani based on the collections made by Michel Allioni in Ecuador and newly proposed here as synonyms are listed below. Notes. -Diplasiolejeunea brunnea is a widespread neotropical species that is distinguished by ocellate leaves without hyaline margin, large lobules with a conspicuous, erect, fingerlike first tooth and an inconspicuous second tooth, and by deeply bifid underleaves with widely diverging, lanceolate lobes with acute tips (Reyes, 1982). Diplasiolejeunea bidentiloba, known only from the type collected by Allioni, is identical to D. brunnea. The new synonymy was already noted by R. Grolle (in sched.).

= Frullania ericoides (Nees) Nees
Notes. -Frullania allionii is similar in all respects to the common, pantropical F. ericoides, a species standing out by brittle, squarrose leaves, lobules usually explanate (rarely helmet-shaped), and perianths with numerous small scales, lacinia or tubercles on the surface (Schuster, 1992). The specimen in Bryo. Levier 6443 is chosen here as the lectotype of F. allionii as it was illustrated in Stephani (1985: tab. 3403). Notes. -Frullania kunzei is a widespread neotropical species that is distinguished by its very small size (0.5 -1 mm wide), prostrate growth, ovate-orbicular leaves with a rounded apex and weakly auriculate dorsal base (sub)isodiametrical apical leaf cells without intermediate thickenings, lobules 1.3 -1.7 × longer than wide, small and distant underleaves, heteroiocus sexuality, and broadly 3-keeled perianths with a subentire involucre (Gradstein, in press). Frullania ecuadorensis is a dioicous phenotype of F. kunzei with lobules varying from upright and close to the stem to oblique and at some distance to the stem. The close similarity of F. ecuadorensis to F. kunzei (= F. neesii Lindenb.) was already noted by Stotler (1969).

Lepidozia tenuicuspis
Steph., Sp. Hepat. 6: 342. 1922 [nom. illeg., non Spruce 1890], syn. nov. Notes. -Lepidozia cupressina is a common and widespread Afro-American species that has been described under many different names. Lepidozia tenuicuspis Steph. is a phenotype of L. cupressina with longly uniseriate underleaf apices made up of quadrate to subrectangular cells (1 -2 × longer than wide). The specimen in Bryo. Levier 6560 is chosen here as the lectotype of L. tenuicuspis as it was illustrated in Stephani (1985: tab. 5071 Notes. -Bazzania gracilis is a widespread neotropical species characterized by the small plants (1 -2 mm wide) with strongly caducous, entire to shallowly 2-lobed leaves (Gradstein, 2017). Mastigobryum heterophyllum Steph. is identical to B. gracilis. M. H. Fulford (in sched.) already noted this synonymy, but subsequently published M. heterophyllum as a synonym of B. phyllobola Spruce (Fulford, 1963). Notes. -Plagiochila cristata is a robust neotropical species recognized by narrowly elongate-triangular leaves with 1 -2 large, apical teeth and a strongly ampliate ventral base forming a high crest (Heinrichs, 2002). Plagiochila allionii is identical to P. cristata and a synonym. Notes. -Plagiochila rutilans is a common neotropical species that is distinguished by horizontally spreading, narrowly oblong-rectangular leaves with a shortly toothed apex, nonampliate ventral bases and subisodiametrical cells (Heinrichs et al., 2001). Plagiochila cobana is identical to P. rutilans and a synonym. Notes. -Plagiochila deflexa is an uncommon neotropical-Hawaiian species that is recognized by elongate-triangular leaves with a large vitta-like area near the base, a narrowly truncate apex, margins with rather few teeth (frequently two larger teeth at apex), and longly decurrent, ampliate ventral bases . The species was described from Ecuador as P. chimborazensis Spruce and P. ecuadorensis Steph., which are synonyms (Gradstein, in press Notes. -Plagiochila longispina is a neotropical-montane species that is distinguished by more or less bluish stems, and leaves with numerous long-linear teeth all around the margins . Plagiochila matanga is identical to P. longispina and a synonym.

= Plagiochila montagnei Nees
Notes. -Plagiochila montagnei is a common neotropical rainforest species that is recognized by terminal branching and ovate-lingulate leaves with a longly decurrent dorsal base, toothed ventral and apical margins, and strongly ampliate ventral bases with triangular teeth . The species is common in lowland forests and is only occasionally seen in montane forests. The specimen Allioni 709 is chosen here as the lectotype of P. nova as it was illustrated in Stephani (1985: tab. 11214). A further syntype (Bomboiza, "frequens ad truncos in silva M. Sapotilla prope domum bonbori", 1000 m, 21.X.1910, Allioni 651, G [G00113038]!) belongs to P. macrostachya Lindenb.

Plagiochila paludosa
Steph., Sp. Hepat. 6: 195. 1921 Notes. -Plagiochila heteromalla is widespread in the Andes and is distinguished by the (sub)opposite and usually strongly secund leaves with a border of thick-walled cells, a vitta-like basal area, shortly toothed margins, shortly decurrent leaf bases and homogeneous oil bodies (Gradstein, 2016). Plagiochila paludosa is identical to P. heteromalla and a synonym. The new lectotypification of P. paludosa was already proposed by J. Heinrichs on an annotation label.

= Scapania portoricensis Hampe & Gottsche
Notes. -Scapania portoricensis is the only neotropical member of the large, holarctic genus Scapania (Dumort.) Dumort. and characteristic of upper montane cloud forests (Frahm & Gradstein, 1991). The species is readily recognized by its reddish color, large size (to 20 cm long and 1 cm wide), toothed leaf margins, very short keel and stellate gemmae. Scapania minutidens is a robust phenotype of S. portoricensis. Notes. -Lejeunea rotundifolia is a robust Andean species with orbicular leaves, very large and suborbicular underleaves with a short-bifid apex and strongly auriculate bases, and stems with 4 -6 cell wide ventral merophytes (Reiner-Drehwald, 2005). Taxilejeunea auriculata is identical to L. rotundifolia and a synonym. The specimen collected by Allioni is chosen here as the lectotype of T. auriculata as it was illustrated in Stephani (1985: tab. 9997). Notes. -Lejeunea flava is widely distributed in tropical and warm-temperate regions of the world and distinguished by ovate-oblong leaves with rounded tips, leaf cells with conspicuous trigones, ± longer than wide underleaves with a cordate base, and gynoecia frequently two in a row on short branches (e.g., Schuster, 1980). Stephani's description of Taxilejeunea tenuiplica adds another name to the long list of synonyms of L. flava. Notes. -This liverwort is characteristic of the Chocó biogeographic region of Colombia and Ecuador, and additionally occurs in the eastern foothills of the northern Andes (Shi & Zhu, 2005). Allioni collected this subspecies twice along the Río Sapotillo; one collection was described by Stephani as a new species of Leucolejeunea, the other one as a new species of Marchesinia.

Dendroceros allionii Steph.
Notes. -Only known from the type collected by Allioni along the Río Gualaquiza. The status of this species needs further study; the type has been on loan and was not available for examination.

Frullania confertiloba Steph.
Notes. -A rare neotropical species, in Ecuador only known from Gualaquiza where it was found by Allioni along the Río Yumana (Yuzawa, 1991).

Frullania mirabilis Jack & Steph.
Notes. -An uncommon northern Andean species, ranging northwards to Costa Rica. In Ecuador only known from Gualaquiza and Parque National Podocarpus and surrounding areas (Schäfer-Verwimp et al., 2013).

Isotachis obtusa Steph.
Notes. -This rare aquatic high Andean species was first detected by Allioni in the páramo of Matanga, in the province of Azuay. The species has usually been called I. lacustris Herzog, which is a synonym of I. obtusa (Gradstein, in press).

Lepidozia pseudocupressina Schiffn.
Notes. -This rare species is only known from southern Brazil and from Ecuador, where Allioni found it in Morona Santiago (Gualaquiza) and Azuay. The species is very close to the common L. cupressina and may be a variety of the latter (Gradstein, in press).

Metzgeria allionii Steph.
Notes. -Only known from the type from Gualaquiza, where it was collected by Allioni along the Río Curiapo. The species is very similar to the widespread but uncommon M. crassipilis (Lindb.) A. Evans -found by the authors along the Río Bomboiza and Río Quimi in 2019 -but differs in gemmae and younger parts of the thallus with falcate hairs (Kuwahara, 1986;Costa, 2008).

Plagiochila deflexa Mont. & Gottsche
Notes. -A species with an unusual distribution, being common in Hawaii and additionally found in Central America  and in the northern Andes, where it is rare (Gradstein, in press). The collection made by Allioni, near Rosario along the road from Gualaquiza to Azuay, is the southernmost record of the species.
Notes. -An uncommon neotropical species occurring in southeastern Brazil and the Ecuadorian Andes, where it is known from Reserva Biológica San Francisco, Zamora Chinchipe (Parolly et al., 2004) and Gualaquiza, where Allioni found it along the Río San Francisco.

Collections made by the authors
Our collections from the surroundings of Gualaquiza yielded 66 species, including 64 of liverworts and two of hornworts (Annexe 2). Of these, three species, Lejeunea rionegrensis Spruce, Notothylas javanica (Sande Lac.) Gottsche and Riccia vitalii Jov.-Ast, are new to Ecuador and 37 (c. 60 %) are new to the prov. of Morona Santiago. Lejeunea rionegrensis is a rather widespread, but uncommon Amazonian species that was known from Venezuela, Brazil, Ecuador, Peru and Bolivia (Reiner-Drehwald & Schäfer-Verwimp, 2008). Its occurrence in Ecuador was therefore to be expected. Riccia vitalii is a characteristic species of dry caatinga of eastern Brazil that has rarely been found elsewhere (Paraguay, northern Colombia, Costa Rica; Gradstein et al., 2016). Notothylas javanica, finally, is a pantropical species with a limited distribution in the Neotropics, where it is known from central and eastern Brazil and northern Colombia. The species was known in the Neotropics as N. vitalii Udar & D.K. Sing, which is a synonym (Leandro de Almeida & Peralta, 2020). The latter species had already been reported from Ecuador, but the Ecuadorian specimens were misidentified .
Three further species in our collections, Jensenia difformis (Nees) Grolle, Myriocoleopsis gymnocolea (Spruce) M.E. Reiner & Gradst. and Riccia crassifrons Spruce, are very rare in Ecuador. Jensenia difformis is a southern temperate species that is in the Neotropics only known from southeastern Brazil and the northern Andes, where it is uncommon. In Ecuador the species had been recorded once (Parolly et al., 2004). Riccia crassifrons is an Amazonian species that had been found once in Ecuador along the Río Nangaritza (Benitez & Gradstein, 2011) and Myriocoleopsis gymnocolea is a rheophytic species from southeastern Brazil and northern Argentina, that is known from three localities in Ecuador (Reiner-Drehwald & Gradstein, 1997;Gradstein et al., 2011). A further uncommon species, Phaeoceros tenuis (Spruce) Hässel, was collected by the authors on 2 October 2019 along the road Macas -Puyo, c. 1 km before the bridge across the Río Pastaza (P11614, QCA). This is the second record from Ecuador of this rare neotropical hornwort.
The number of species of liverworts and hornworts recorded from Morona Santiago has been raised to c. 190 based on the collections made by the authors. In comparison, the country of Ecuador counts about 700 species and several provinces have more recorded species than Morona Santiago (Léon-Yánez et al., 2006;Gradstein, in press). The results indicate that Morona Santiago, in spite of being one of the largest provinces of Ecuador, is still very incompletely collected for liverworts and hornworts. This conclusion is supported by a comparison of the species collected by Michel Allioni and by the authors (Table 1). Of 48 species gathered by Allioni, only eight were found by us; 40 species or almost 80 % of those found by Allioni were not seen during our visit. An explanation may be the greater number of localities, habitats and different elevations inventoried by Allioni. On the other hand, 58 species were collected during our brief, 2-days visit to the Gualaquiza area that were lacking in the collections of Allioni examined in this study. These figures are suggestive of the incompleteness of our knowledge of the liverworts and hornwort flora of the Gualaquiza area. However, our data are biased by the rather limited set of Allioni specimens examined in this study, which was mainly restricted to type specimens. A search of the non-type collections of liverworts and hornworts in herbarium G might yield additional species of liverworts and hornworts collected by Michel Allioni during his pioneering exploration of the bryophyte flora of Gualaquiza.