Taxonomy of duckweeds ( Lemnaceae ) , potential new crop plants

doi: 10.9755/ejfa.2016-01-038


INTRODUCTION
The present "blooming era" of "resurgence of duckweed research and applications" (Zhao et al., 2012;Lam et al., 2014;Appenroth et al., 2015b) is based mainly on their high potential for practical application.Lemnaceae comprises one of the fastest growing angiosperms (Sree et al., 2015c;Ziegler et al., 2015).In some Asian countries, duckweeds have been used for human nutrition for generations (Bhanthumnavin and McGarry, 1971;cf. van der Spiegel et al., 2013), because of their high protein content (Appenroth et al., 1982;Cheng and Stomp, 2009).Duckweeds have the potential to replace soya bean products at least partially as a source of protein.
As demonstrated by several projects, duckweeds proved to be advantageous when fed to domesticated animals e.g., cattle, sheep, horses, rabbits (reviews in Landolt and Kandeler, 1987;Cheng and Stomp, 2009), poultry and water fowls (Muztar et al., 1979;Anderson et al., 2011), fish (Hassan and Edwards, 1992;Pipalova, 2003), and pigs (Van et al., 1997).Apart from their use as food and feed, duckweed biomass with high starch content produced under suitable cultivation conditions (Sree and Appenroth, 2014;Sree et al., 2015a, b) can be used as a raw material for biofuel and biogas plants (Jain et al., 1992;Su et al., 2014;Cui and Cheng, 2015).This would substantially replace the use of food crops that are grown in arable lands as raw materials for biofuel production.Taken together, plants of this family have the potential as new crop plants.
Elias Landolt from ETH Zurich, Switzerland has collected duckweeds from all over the world since the mid-1950s and established a duckweed stock collection (Landolt, 1957;Appenroth et al., 2015a) which originally exceeded one thousand clones.Based mainly on morphological investigations, he published a "key of determination" of duckweeds (Landolt, 1980) which was later slightly modified in the first part of his famous monograph (Landolt, 1986).Within the last three decades, quite a number of changes have been introduced to taxonomic classification of duckweeds in comparison to the published key.Moreover, molecular taxonomy and barcoding of duckweed species (Wang et al., 2010;Bog et al., 2010Bog et al., , 2013Bog et al., , 2015;;Borisjuk et al., 2015) has made huge progress and it aids in identification of duckweed species.However, it is not practically and economically viable to determine each of the collected duckweed clones by Amplified Fragment Length Polymorphism (AFLP) or by sequencing of DNA markers.Therefore, determination of duckweed species based on their morphology remains crucial for their identification.An update to the "key of determination" of all the duckweed species known to date is a prerequisite to facilitate accuracy in communication amongst researchers in academia and industry, and also entrepreneurs in this field, especially keeping in view the practical applications of duckweeds.
In this review we discuss the taxonomic position of duckweeds and the classification changes made since the publication of Landolt's monograph in 1986 (Table 2).This revision shall help students and newcomers to the field of duckweed research and applications to become quickly acquainted with this expanding field and to use the accurate terms and nomenclature.

Taxonomic position of the family Lemnaceae
Duckweed species are monophyletic with those of the Araceae (Cabrera et al., 2008;Cusimano et al., 2011).However, the branching appears to have started about 104 Ma ago in the Early Cretaceous when the breakup of Pangea was in its final stage (Nauheimer et al., 2012).This is one reason why we prefer keeping Lemnaceae as a family of its own (Appenroth et al., 2015a).The monophyly does not necessarily mean that duckweeds have to be integrated into the family of Araceae and be treated as a subfamily Lemnoideae, as preferred e.g. by Nauheimer et al. (2012).One can separate the small group of Protoaraceae together with Lemnaceae from the "true Araceae" (as these two groups are the most basal elements in this group), which results in three monophyletic plant families, i.e. true Araceae, Lemnaceae and Protoaraceae (Appenroth et al., 2015a).The following additional arguments speak in favour of designating Lemnaceae as a family on its own (Appenroth et al., 2013): (1) The notion to unify Araceae and Lemnaceae originated from a certain similarity between Pistia spp.and some duckweeds, especially Spirodela polyrhiza L. (Schleid.)(Hooker and Brown, in Smith, 1824).On the basis of more recent morphological and molecular investigations (Cusimano et al., 2011), it is now clear that this similarity is only superficial.(2) The characteristics of the members of Lemnaceae which are similar to those of Araceae at the morphological, anatomical or biochemical levels are not only restricted to Araceae but are similar in all Alismatales.(3) All species belonging to Lemnaceae together form a well-defined unit on morphological, biochemical and habitat bases.
The current systematic investigations into the evolutionary and taxonomic relationships between the five genera of Lemnaceae are represented in Fig. 1 (cf.Les et al., 2002;Tippery et al., 2015) and the classification of these five genera into 37 species has been depicted in Table 1.

Newly discovered species
The following three species were discovered after 1986: Wolffiella caudata Landolt Wolffiella caudata was first described by E. Landolt in 1992.It was found to grow in the southwestern part of the Amazon basin within the inundation area of the tributary rivers, e.g.Yacuma and Tahuamanu, in Bolivia.More information on its identification can be found in Landolt (1992).A chemotaxonomic study of the genus Wolffiella including W. caudata was performed by Crawford et al. (1997) based on the allozyme profiles.Barcoding (Wang et al., 2010;Borisjuk et al., 2015) and a detailed molecular taxonomic investigation of this species based on AFLP and plastidic marker sequences (unpublished data; Bog, Sree and Appenroth) has been carried out recently together with other species of the genus Wolffiella.These investigations further support its species identity.
Wolffia cylindracea was found in seasonal waters of winterdry subtropical to tropical regions of Africa, south of the equator, e.g.Angola (Landolt, 1994).Although this species was first described by C.F. Hegelmaier as early as 1868 (Hegelmaier, 1868), researchers of the later century did not consider it and incorporated it into the species W. globosa (Roxb.)Hartog & Plas (Daubs, 1965;Landolt, 1986).The distinctness of the African clones of W. globosa in the chemotaxonomic investigations led to the rediscovery of this group of African plants as a separate species, W. cylindracea, as already suggested by Hegelmaier (Landolt, 1994;Crawford and Landolt, 1995).The molecular taxonomic studies on this species also clearly distinguish this species from its relatives (Bog at al., 2013;Borisjuk et al., 2015).

Wolffia neglecta Landolt
Wolffia neglecta (Landolt, 1994) was found in the winter-dry subtropical to tropical regions of India (e.g.Rajasthan), Pakistan, and Sri Lanka and probably exists in seasonal waters.The classification of these plants as new species was supported by chemotaxonomic methods (Crawford and Landolt, 1995).
Detailed morphological description of W. cylindracea and W. neglecta annexed with photographs and very useful drawings are presented in Landolt (1994).

Changes in the species nohmenclature
It is worth mentioning that in plant systematics the term "synonym" has an unusual meaning.It describes the same species (which would be in agreement with the common usage of the term) but also means that this name is no Wolffia cylindracea Hegelm.
Newly defined Landolt (1994) longer valid and should not be used.We would like to cite from a letter of W. S. Hillman published by E. Landolt (1986, p. 444): "When the late Dr. Wm.S. HILLMAN (Brookhaven, NY) learned that the name L. paucicostata had to be changed again, he wrote to me in an understandable reaction: "I am filled with horror or laughter (I am not sure which) at the thought of again changing the name under which we write articles about poor old 6746 and its brethren.And the names get worse and worse for the ordinary (English-speaking anyway) person to handle: perpusilla -" paucicostata -" aequinoctialis.And please, I do beg of you, conceal the reference from me as long as possible, since people who read the physiological literature tend to be intolerant of such things"."Although this reaction is understandable, results of the investigations in the field of plant systematics should be respected by other plant scientists.
The most dramatic change in duckweed terminology took place when the molecular taxonomic investigations of Les and Crawford (1999) Ward (2011).Because the type specimen described as Lemna punctata by G. F. W. Meyer was lost, he suggested a neotypification but with Spirodela intermedia as a type.As already stressed by E. Landolt (cited in Wiersema, 2015) this would lead to considerable confusion in duckweed nomenclature (cf.for the genus Spirodela in Plant List, Version 1.1, 2013, http://www.theplantlist.org).More recently, Wiersema (2015) made it clear that the valid name of this species is Landoltia punctata.However, such a drastic change in botanical nomenclature has the unfortunate consequence that the old but crucial and basic research (e.g.Mattoo et al., 1984;Greenberg et al., 1987) published under the old name of the species (e.g.Spirodela oligorrhiza) cannot be retrieved by the internet search machines using the new nomenclature (e.g., Landoltia punctata).
Another interesting change is the spelling of the species name, Spirodela polyrhiza.Landolt (1986) considered this spelling of the species to be an orthographic mistake of the term polyrrhiza with two 'r's and used the spelling Spirodela polyrrhiza.But as pointed out by Appenroth et al. (1990), Spirodela polyrhiza with one 'r' is the valid name as written by its author von Linné (Linnaeus, 1753).
Two of the Lemna species should be mentioned in this context.The first one is Lemna aequinoctialis Welw.Its synonym, Lemna paucicostata, which was replaced many years ago is still being used and the species name is wrongly cited (e.g.Murata et al., 2014).Lemna aoukikusa (as mentioned in Yamaga et al., 2010) is also a synonym of Lemna aequinoctialis (Landolt, 1986).This decision was recently confirmed by Borisjuk et al. (2015) based on the plastidic sequence analysis.In a second case, Reveal (1990) found out that there was already an older and valid name for Lemna minuscula.As a consequence and following the rules of plant systematics, the name of the species Lemna minuscula had to be changed to Lemna minuta Kunth and the term L. minuscula became a synonym.It is worth to mention that the species of the genus Lemna are characterized by the presence of one root (Landolt, 1980).The names Lemna polyrhiza and Lemna arrhiza are therefore not valid (Landolt, 1986).The valid names are Spirodela polyrhiza and Wolffia arrhiza Horkel ex Wimm.(Iamonico and Iberite, 2014).Iamonico and Iberite (2014) also lectotypified the two Linnaean names: Lemna polyrhiza and Lemna arrhiza.
Although invalid, the first one is still being used (e.g.Lemna polyrhiza in Bairagi et al., 2002;Lemna polyrrhiza in John et al., 2008).Apart from the fact that these names are invalid, they are in evident conflict with the meaning of the terms "polyrhiza" (= many roots) or "arrhiza" (= rootless) under the genus Lemna.
As a supplementary material we have compiled all the synonyms of the 37 duckweed species, which we became aware over the years.The valid names are given in each case.

CONCLUSIONS
Use of accurate scientific nomenclature of the duckweed species needs to be stressed in view of the considerable number of changes in duckweed systematics in the recent past.Efforts should be made to disseminate this up-todate information to duckweed researchers as well as to the general readers.In general interest and convenience of both the authors and the readers, it is recommended that the synonym (invalid old name of a species) of a duckweed species be mentioned in braces at the first mention of the new name of the duckweed species in a text.With the current advancements in the field of molecular biology, barcoding and molecular taxonomy of Lemnaceae members have already reached a considerable level.Nevertheless, morphological identification of duckweeds still remains significantly important and cannot be superseded.For instance, Wolffia neglecta, a species described by E. Landolt on a morphological basis, can hardly be separated out as a species on its own using the methods of molecular taxonomy (Bog et al., 2013).On the other hand, molecular taxonomy aids in investigating some of the minute details.The term Lemna yungensis Landolt must be specially mentioned.This species was introduced mainly by eco-geographic reasons because it was discovered on wet rocks that are in permanent contact with flowing waters (Landolt, 1998).However, according to our investigations, the morphological differences between L. yungensis and L. valdiviana Phil.are not very distinct and the molecular analysis also does not provide markers for a clear differentiation between the two species (Bog et al., 2010;Borisjuk et al., 2015).In-depth molecular analysis might provide a substantial basis for further conclusions in this respect.

Fig 1 .
Fig 1. Internal structure of the family: Lemnaceae.The relation of genera and subfamilies based on Les et al. (2002) and Tippery et al. (2015).

Table 1 : Presently accepted genera and species of the plant family Lemnaceae Dumort Subfamily Genera Species Lemnoideae
Engl.