An archipelago within an archipelago: A checklist of liverworts and hornworts of Kepulauan Sunda Kecil (Lesser Sunda Islands), Indonesia and Timor-Leste (East Timor)

Abstract The first ever liverwort and hornwort checklist is provided for the Kepulauan Sunda Kecil (Lesser Sunda Islands) of Indonesia and Timor-Leste (East Timor). We report 129 accepted taxa, 12 doubtful taxa and three rejected taxa previously reported for the Lesser Sunda Islands. The list is based on over 130 literature references, including monographs, regional studies, and molecular investigations. It is clear that bryophytes from this region have been overlooked historically, and under collected, compared to seed plants, birds, and other organisms, forming a remarkable gap in the flora of Indonesia. Publications dealing with liverworts of Lesser Sunda Islands are few and scattered. We predict that further fieldwork, in addition to collections unveiled from regional herbaria, will uncover a number of new records that remain to be reported, especially considering that regionally widespread species have been recorded elsewhere.


Introduction
The Lesser Sunda Islands (LSI), known in Indonesia as Kepulauan Sunda Kecil, are an area covering a longitudinal distance of some 600 kilometres in the southeastern portion of Indonesia, extending between Java in the west and New Guinea in the east (Fig. 1). Lesser Sunda Islands include a multitude of islands, the major ones of which are Flores, Sumba, Sumbawa, and Timor. The region overlaps with two different countries; Indonesia, which includes four different provinces, i.e., Bali, West Nusa Tenggara, East Nusa Tenggara (including Western part of Timor island), and part of Moluccas (van Steenis-Kruseman and van Steenis 1950; Monk et al. 1997;Jepson and Whittaker 2002) and Timor-Leste (East Timor), which includes the eastern part of Timor island (Kusuma 2017). The Lesser Sunda Islands occur as two geologically distinct island chains termed the Inner and Outer Banda Arcs (Audley-Charles 2011). This archipelago also occurs at the heart of the complex crossroads of two continents, Asia and Australia, and two oceans, the Indian and Pacific (Monk et al. 1997). The Lesser Sunda Islands may act as 'stepping stones' for animals and plants dispersing between the Greater Sunda Shelf, i.e., the Malay Peninsula, Borneo, Sumatra, Java, and Bali, and the Sahul Shelf, i.e., New Guinea, Australia, and their land-bridge islands (Reilly 2016). Bordered to the west by the Greater Sunda Shelf and to the east by the Sahul Shelf, the Lesser Sunda Islands can be considered oceanic islands in the sense that they have never been connected by land to continental Asia or continental Australo-Papua.
The island chain of LSI comprises 5.2% of endemic species based on families treated in Flora Malesiana (van Welzen et al. 2005). Interestingly, the Indonesian part contains the highest percentages of endemic plants of Indonesia, i.e., 55% (Bappenas 2016). Some important and better-known endemic species in this ecoregion are the Komodo dragon (Varanus komodoensis), the largest lizard in the world and the iconic tree of East Nusa Tenggara Province, sandalwood tree (Santalum album L.), and Dracaena multiflora Warb. ex Sarasin (Monk et al. 1997).
The islands of LSI include seven ecoregions in two biomes, Tropical and Subtropical Moist Forests and Tropical and Subtropical Dry Forests ( Fig. 2; Terrestrial Ecoregions of the World; Olson et al. 200l). The "Tropical and Subtropical Moist forest" includes two ecoregions on Bali, and two ecoregions on the eastern islands. The Eastern Java-Bali rainforest (https://www.worldwildlife.org/ecoregions/ im0113) and Eastern Java-Bali Montane Rain Forest (https://www.worldwildlife. org/ecoregions/im0112) are situated on the Sunda shelf and both are classified as endangered ecoregions since a lot of forest has been cleared. It forms transitional vegetation types relative to the drier areas of the central islands situated on the Sahul shelf.
The "Tropical and Subtropical Dry forests" consist of three ecoregions from Lombok in the west to Timor and Wetar in the east. All three ecoregions are critically endangered/endangered. The largest of the ecoregions is "Lesser Sunda Deciduous Forests" (https://www.worldwildlife.org/ecoregions/aa0201) on Lombok, Sumbawa, Flores and west to Alor including smaller surrounding islands. It consists mainly of semi-evergreen dry forests. With an average annual rainfall of 1,349 mm, this is the driest area but also the most seasonal in Indonesia (Touw 1992). The "Sumba Deciduous Forest" (https://www.worldwildlife.org/ecoregions/aa0203)" was also dominated by deciduous monsoon forest, but now much of it has been replaced by savanna and grasslands (Monk et al. 1997). The "Timor and Wetar Deciduous Forest" (https:// www.worldwildlife.org/ecoregions/aa0204) is also largely deforested and replaced by grasslands and savanna.
Further east, two more ecoregions of "Tropical and Subtropical Moist Deciduous Forests" occur. The "Banda Sea Islands Moist Deciduous Forest" (https://www.worldwildlife.org/ecoregions/aa0102) includes all islands west of Timor/Wetar, except Aru Island. It includes evergreen rain forest (Kepulauan Kai), semi-evergreen rain forest, moist deciduous forest, and dry deciduous forest (Monk et al. 1997). The forests are still largely intact, but the ecoregion is still classified as vulnerable. The "Vogelkop-Aru Lowland Rain Forest" (https://www.worldwildlife.org/ecoregions/im0128) is mostly confined to New Guinea but occurs also on Aru. Large parts of the ecoregion are still intact.
Bryophytes, including mosses, liverworts and hornworts, are the second largest group of land plants after flowering plants and are pivotal in our understanding of early land plant evolution (e.g., Ligrone et al. 2012;Zhang et al. 2020). Bryophytes play a significant ecological role including CO 2 exchange (De Lucia et al. 2003), plant succession (Cremer and Mount 1965), production and phytomass (Frahm 1990), nutrient cycling (Coxson 1992) and water retention (Pócs 1980;Gradstein et al. 2001). Retnowati et al. (2019) cited 849 species of liverworts, 28 species of hornworts and 1,884 species of mosses which are scattered in the major islands of Indonesia. As with many regions in the world, it is evident that the bryophyte, especially the liverwort flora, remains very poorly known in comparison to the vascular flora. Gradstein and Culmsee (2010) noted there are few studies from Southeast Asia investigating the diversity and ecology of tropical bryophytes.
It is clear that bryophytes from this region have been overlooked historically, and under-collected, compared to seed plants, birds, and other organisms, forming a remarkable gap in the flora of Indonesia and/or Timor-Leste. Publications dealing with liverworts of LSI also are few and scattered. The first apparent report of liverworts from LSI was by Sande Lacoste (1856) in the mid-19 th century. It was not until the late 19 th century and early 20 th century that further influential works appeared, including those by Schiffner (1898Schiffner ( , 1900Schiffner ( , 1955, Stephani (1886Stephani ( , 1899Stephani ( , 1907Stephani ( , 1908Stephani ( , 1909Stephani ( , 1911Stephani ( , 1917Stephani ( , 1924, Verdoorn (1930Verdoorn ( , 1934aVerdoorn ( , 1934bVerdoorn ( , 1935Verdoorn ( , 1937, and others. Among those islands, Bali is the most explored island (Hegewald and van Zanten 1986;Eggers et al. 1998;Schäfer-Verwimp 2006Haerida et al. 2010;Alam 2012;Heinrichs et al. 2012;Girmansyah et al. 2013) with 101 species accepted here. Söderström and Séneca (2008) reported only 61 number of liverworts for Lesser Sunda Islands and considered that this low number of species was the effect of the under-explored areas. More recently, little botanical work has been done in the area; the area has occasionally been visited by students and researchers from nearby institutions, but publications are still lacking. This checklist will complement the survey of mosses of Lesser Sunda Islands by Touw (1992) who enumerated 367 species for the area, and complement other checklists of liverworts from Indonesia, including Java , Bali (Haerida 2015(Haerida , 2017 as well as Sumba (Haerida et al. 2020).
We here present the first-ever checklist of liverworts and hornworts for the Lesser Sunda Islands to serve as the baseline information in our study of the liverworts diversity of this archipelago. Furthermore, this checklist can serve to promote and encourage bryological research in the region. The significance of checklists is summarized by , including outlining the utility of checklists as powerful and important tools, and their applicability to taxonomy, systematics, and conservation.
As with many other regions in the world, given the relatively poor focus on liverworts in LSI historically, we predict that a vast number of new records are yet to be reported for the area. In this checklist we report 129 taxa, with another 12 taxa questioned and three rejected. The number of known species from individual islands varies from 101 (Bali) to 0 (Barat Daya Islands), but only Bali has more than 20 known species (Fig. 3).

Materials and methods
Nomenclature and taxonomy follows the world checklist of hornworts and liverworts (Söderström et al. 2016) with a few updates from recent taxonomic literature. Sources include over 130 publications found through the work of Early Land Plants Today (ELPT) database of liverwort taxonomy and distribution, and with some consultation with taxonomic experts. The checklist follows a similar format of previous liverwort and hornwort checklists by the authors, e.g., Java . The checklist distinguishes between records that are based on specimens seen by the author(s) (reference in bold) and second-hand reports, e.g., citations of earlier publications (reference in normal type). All names used for the references are given after the taxon name with spelling variants/errors within quotation marks. Taxa are arranged in alphabetical order. Significantly, each accepted taxon is qualified using a four level ranking system that indicates our level of knowledge about a taxon. The coding convention is outlined in detail by von Konrat et al. (2010). Briefly, ? = Problem with the taxon name; * = Serious doubts about the value of the taxon; ** = Probably a good taxon (default value); *** = Accepted, a good taxon as currently understood.
This checklist covers the Lesser Sunda Islands based on reports from the literature that have been subdivided into 10 geographical units that correspond to geopolitical units. These include from west to east ( Fig. 1) Bali, Lombok, Sumbawa, Sumba, Flores, Timor (separated in the Indonesian West Timor and the independent Timor-Leste), Barat Daya Islands, Tanimbar Islands, Kai Islands and Aru Islands.

Types from Lesser Sunda Islands
Recently, there has been some debate whether the type information by Bonner (1962Bonner ( , 1963Bonner ( , 1965, as well as the other volumes of his "Index Hepaticarum", can be accepted as lectotypes. Renner (2021) argued in favour for the volumes of Bonner (1962Bonner ( , 1963Bonner ( , 1965, including recommendations to improve typification practice, and Engel and Merrill (2019) argued that Bonner's herbarium designations do not represent lectotypifications. Here we accept the typifications made by Bonner as they pertain to the taxa treated here, but we also provide alternative typifications if ruled against. In the latter cases, we instead "validate" Bonner's typifications, yet causing no nomenclatural changes whichever view is taken.
It is not always clear if more than one collection exists for a possible type. McNeill (2014) recommends that in such a case a lectotype should be selected from known available material, but with a statement that it may be the only material, in which case a lectotypification would be superfluous. For such cases, we here follow McNeill's "best practice" advice using the suggested phrase "lectotype here designated, if not a holotype".

Taxa of unclear affinity
A couple of taxa are published from Lesser Sunda Islands as varieties of species synonymized under other names, without transferring or synonymizing the variety. We have not been able to trace any specimen that they may be based on and, thus, not been able to refer them to any valid taxon.

Taxa reported but doubtfully occurring in Lesser Sunda Islands
Marchantiophyta Bazzania ** Bazzania ceylanica (Mitt.) Steph. lesser sunda is.: Miller et al. 1983. Note: The report by Miller et al. is unclear and it may be that they meant some of the Greater Sunda Islands. It is widespread in SE Asia, so its presence is not unlikely. *** Bazzania erosa (Reinw., Blume et Nees) Trevis. lesser sunda is.: Miller et al. 1983. Note: The report by Miller et al. is unclear and it may be that they meant some of the Greater Sunda Islands. It is widespread in SE Asia so its presence is not unlikely. *** Bazzania tridens (Reinw., Blume et Nees) Trevis. lesser sunda is.: Miller et al. 1983. sumBawa: Pócs 1971. Note: We are not aware of any first hand report from Lesser Sunda Islands but it is common in SE Asia so its presence on some of the islands is not unlikely. The Biodiversity Heritage Library is acknowledged for the facility they provide that has greatly accelerated our efforts. The generous support by the National Science Foundation (Award No.'s 1145898, 1146168, 1458300, 1541506, 2001509 and 0531730) is gratefully acknowledged. We also recognize the support of the Museum Collection Spending Fund, administered by Field Museum, as well as curatorial support provided by Y. Rodriguez, L. Kawasaki, and A. Balla (Field Museum). Alfons Schäfer-Verwimp kindly provided us with specimen data from his unpublished Diplasiolejeunea cavifolia. We are very grateful for the helpful suggestions and comments made by the editor and reviewers: Matt A. Renner, S.R. Gradstein, Li Zhang, and an anonymous reviewer.