Taxonomy of the Genus Halophila Thouars (Hydocharitaceae): A Review

The seagrass genus Halophila Thouars has more than twenty described species and is predominately distributed over a wide geographic range along the tropical and the warm temperate coastlines in the Indo-West Pacific Oceans. A brief history of the Halophila taxonomic development is presented. Based on reproductive and vegetative morphology, the genus is divided into eight sections including three new sections: section Australes, section Stipulaceae and section Decipientes. A rewritten taxonomic description of the type species for the genus Halophila, H. madagascariensis Steudel ex Doty et B.C. Stone, is provided. The lectotype of H. engelmannii Asch. as well as neotypes of H. hawaiiana Doty et B.C. Stone and H. spinulosa (Br.) Asch. are designated. Furthermore, H. ovalis ssp. bullosa, ssp. ramamurthiana and ssp. linearis together with H. balforurii have been recognised as distinct species. Nomenclature, typification, morphological description and botanical illustrations are presented for each taxon. Recent molecular phylogenetic surveys on certain Halophila taxa are also discussed. Field surveys for the deep water Halophila in West Pacific regions are suggested. Morphological studies combined with molecular investigations for the Halophila on the east coast of Africa and the West Indian Ocean are urgently needed and highly recommended.


Introduction
The seagrass Halophila is the smallest in size among the entire known seagrasses, but it is the most diverse group and represents more than one quarter of all recognised seagrass species [1]. Halophila plants are minute and fragile, without stripe-like long leaf blades, and virtually unlike any other seagrasses. As a result, their taxonomic position in the monocotyledonous classification has long been unsettled. Ascherson [2] placed Halophila under the tribe Halophileae in the family Najadaceae and then Ascherson and Gürke assigned it as the tribe Halophiloideae within the family Hydrocharitaceae [3]. Rydberg [4] later included Halophila in the family Elodeaceae instead of the family Hydrocharitaceae. Den Hartog [5] considered Halophila as one of nine genera in the Hydrocharitaceae; however, in his monograph [6], he followed Nikai's interpretation [7] and placed Halophila under the subfamily Halophiloideae (one genus only) in the family Hydrocharitaceae. Den Hartog and Kuo [1] stressed that the subfamily Halophiloideae status should be maintained for the Halophila.
Ascherson [8] divided the Halophila into two sections: Barkina (containing H. ovalis) and Microhalophila (consisting of H. beccarii). However, none of his subsequent publications used the sections in the genus Halophila. Ostenfeld [9] divided Halophila (8 species) into four sections: the section Typicae-H. ovalis, H. stipulacea, H. baillonis and H. decipiens; the section Americanae-H. engelmannii and H. aschersonii; the section Spinulosae-H. spinulosa; and the section Pusillae-H. beccarii. The first three sections possess stipules and the fourth section Pusillae has none. Ostenfeld and Meeresgraser [10] made minor alterations in the species compositions for these four sections. The section Typicae included  [46]. A series of molecular analyses on the Halophila species from South-East Asian countries (i.e., Vietnam, Myanmar, Thailand, Malaysia, India) and the Red Sea by Nguyen and Papenbock's team [47][48][49][50][51] concluded that Halophila ovalis, H. major, H. minor (they referred as H. ovata), H. decipiens and H. beccarii were distinct species and H. ovalis ssp. ramamurthiana was genetically different from H. ovalis in South-Eastern India. Furthermore, they found that H. beccarii was closer to the ancestor of Halophila species than H. ovalis or H. decipiens, as indicated previously by Waycott et al. [43]. Nguyen et al. [50] suggested it that the Thai-Malay Peninsula was a geographic barrier between H. ovalis populations in the Western Pacific and the Eastern Indian Ocean is worthwhile to mention that all Nguyen and Papenbock's molecular studies were accompanied by morphological data on the studied species.
Kuo and den Hartog [39] provided a key and brief description for all seagrasses globally including fifteen Halophila species. Some field guides of seagrasses including Halophila have also been compiled for the local regions, e.g., [52][53][54][55][56][57][58]. However, updated comprehensive taxonomic treatment and the key for the identification of the Halophila species globally have not been published since 2001.
For the flowering plants, the reproductive structures such as petals, sepals, stamens, fruits and seeds are main sources for species identification. However, flowers and fruits of most Halophila are not often collected. Only a small fraction of 2000 Halophila specimens studied from forty major herbaria (Supplementary A) had reproductive material. Therefore, the identification of Halophila species and sections is more or less dependent on vegetative characteristics, such as plant appearance, blade length, width, blade tips, cross vein numbers, branching, etc. Sometimes, some of these vegetative characteristics may show considerable variation or overlapping. Nevertheless, if used in conjunction with other features such as geographic locations and habitats, the vegetative characteristics are of great value taxonomically. Therefore, vegetative characteristics and reproductive structures when available will be used for Halophila section classification, and vegetative characteristics alone will be employed in the key presented in this paper for species identification. The following key is designed and arranged to identify all described (known) Halophila sections and then the species. Under each taxon, the type locality, synonyms and morphological descriptions are given. The locality, collector and herbarium information of botanical drawings for the Halophila species are listed in Supplementary B1 while Botanical drawings for each species will be placed in Supplementary B2. Marine to estuarine, submerged, sometimes intertidal, monoecious or dioecious, perennial, rare annual herbs. Rhizomes creeping, with 1 lateral shoot and 1 unbranched root at each node. Scales 2, one covering root base, the other covering lateral shoot base. Leaves in pair or in pseudo-whorl, sessile or petiolate; blades linear, lanceolate to ovate, entire or serrulate, glabrous or hairy, with a mid-vein and 2 intramarginal veins connected by cross veins; stomata absent; tannin cells absent. Spathes with 2 imbricate, keeled bracts; 1 flower or with a male and a female flower. Male flower pedicellate; tepals 3; stamens 3; anthers 2, oblong, sessile, dehiscing longitudinally; pollen grains ellipsoid, in long chains. Female flower sessile; hypanthium persistent, bearing 3 reduced tepals; ovary ovoid to ellipsoid, unilocular; styles 3-6, filiform. Fruits ellipsoid to globose, beaked. Seeds numerous, globose, testa mostly reticulate.

Genus Halophila Thouars
The genus contains at least 24 described species, arranged within eight sections; mostly distributed in tropical and warm temperate waters. Some species are well defined and restricted in distribution, but a few species, such as H. ovalis, H. major and H. minor, are widely sympatric, distributed in the Indo-West Pacific region. Halophila decipiens is the only truly pantropic species which occurs in the Indian Ocean and tropical region of Pacific and W Atlantic Oceans, recently also in the Mediterranean Ocean. The Halophila has not been found in east coasts of both the Atlantic and the Pacific Oceans.  This species usually co-exists with H. ovalis and/or H. minor and is widely distributed in the Indo-Pacific Oceans to the Red Sea. H. major grows in shallow water but also grows down to 6-7 m in the water depth. The flowering and fruiting of this species usually occur during two-to three-month periods between June and December.  Note: The type specimen Setchell 114 had smaller leaf blades (7-10 × 4-4.5 mm with 7-8 cross veins) and without a predominated bullolate appearance.
The species is widely distributed in the tropical Indo-Western Pacific Ocean. The species usually grows in shallow water with H. ovalis and H. major. Flowering and fruiting of this species take place April-June in the Philippines; August in Kenya; September to February in tropical Australia. Halophila gaudichardii has been collected from Saipan, Guam and Yap in the Mariana Islands and in the Philippines. However, the species is not found in Caroline Islands, which are adjacent to the Mariana Islands in the Pacific Ocean. This species in the Mariana Islands and the Philippines grows in the shallow water, while it has been collected from 24 m in the Okinawa Is. Flowering and fruiting occur from January to April. Rhizome internode up to 55 mm long, petiole up to 30 mm long. Leaf blade obovate to spatulate, to 30 mm long and 7 mm wide, L:W ratio 3-4:1; apex obtuse; cross veins 14-17, some branched, distance between adjacent veins 1 mm wide; distance between intramarginal veins and blade margin 0.2 mm wide; 1 2 blade width:distance between intramarginal veins and blade margin ratio ca. 15:1. Female flower with ovary 1.5 mm long; hypanthium 8 mm long. Male flowers and fruits not seen.
Halophila mikii is only known from Tanegashima and Yakushima in Southern Japan. The species grows as small patches at 2 to 10 m on mud substratum. Flowering and fruiting may take place from August to October.  Marine, rather robust species with large elliptic or obovate transparent scales, 12-17 mm long, 6-10 mm wide, folded at the rhizome nodes covering the short petioles. Blades linear to oblong, elliptic, membranes; surfaces glabrous or with minute hairs, occasionally bullate, not papillous; up to 60 mm long, 10 mm wide; cross veins 10-40, branched. Male flower with tepals 3-4 mm long; anthers 2-3.5 mm long. Female flowers with 3 styles 20-25 mm long. Fruits ellipsoid 3-4 mm long. Seeds testa reticulate in squares appearance.
Halophila stipulacea is the most successful migrating species among the Halophila species. The species migrated from the Red Sea to the Mediterranean Sea after the opening of Suez Canal and recently spread to the Caribbean Sea. Flowering and fruiting in May to October.  Female flowers with ovary ovoid to ellipsoid, 1.5-2 mm long, 1-1.5 mm wide; styles 3, 15-25 mm long. Fruits ellipsoid (2-) 3.5-4 mm long, (2-) 3 mm wide. Seeds (12-) 20-30, globular, 0.75 × 0.5 mm, at both ends contracted; testa reticulate with a rectangular appearance.
Halophila balfourii is confined to Rodriguez and Mauritius Islands. The species has been collected from less than 1 m to 6 m in the depth. Both male and female flowers occur in October and fruits are produced from October to December in Mauritius. Marine to estuarine, submerged, annual or perennial herb. Rhizome fleshy, c. 1 mm diam.; scales usually hairy. Leaves with petiole distinctly shorter than blades. Blades oblong-elliptic, 10-25 mm long, 2.5-6.5 mm wide, finely serrulate, obtuse to acute, membranous, hairy or sometimes glabrous; cross veins 5-9, unbranched. Spathal bract hairy, enclosing a male and a female flower. Male flowers with pedicel to 25 mm long; tepals ovate to elliptic, 2 mm long; anthers linear-oblong, 1.5-2 mm long. Female flowers with hypanthium 1-2 mm; tepals minute; ovary ovoid, 1 mm long; styles 3, 1.5-2.5 mm long. Fruits ellipsoid, 2-5 mm diam. Seeds c. 30, ovoid, 0.2-0.5 mm diam.; testa reticulate.

Sect. 4. Decipientes
Halophila decipiens is the most widely distributed among the Halophila species, occurring in tropic to temperate regions of all major oceans. It grows from shallow water to 58 m and also associates with mangroves and estuaries. Flowers in summer, fruits in summer to autumn. Halophila capricorni occurs on the Great Barrier Reef and also in New Caledonia, a herbarium (K) record from Indonesia. Grows on coral sand, on the lee side of the reef at depths of 7 to 42 m. Flowers from October; fruits from January to March. Thin rhizomes with 2 scales covering the base of the erect stem bearing a group of 4-10 leaves at the top. Blades lanceolate, up to 13 mm long, 1-2 mm wide, with no cross veins; apex pointed. Male flowers up to 1 cm long; spathes ca. 4 by 2 mm, pedicel up to 8 mm long; tepals ca. 2.5 by 1.5 mm; anthers linear-oblong, 2 by 0.25 mm. Female flowers up to 2 cm long; hypanthium 2 mm long; styles 2-3, 10-15 mm long. Fruits ellipsoid to ovoid, 4-5 mm by 2 mm. Seed 1-4, globose, 1.2 by 1 mm, testa reticulate.
Halophila beccarii is widely distributed in the Bay of Bengal and the South China Sea. Taiwan is the northern limit of this species. The species is usually associated with mangrove communities and often exposed at low tide. The lack of cross veins is a unique characteristic among Halophila species. Flowering and fruiting occur from January to March and also from August to October.
Perennial, dioecious plants with 2 scales at the base and 2 scales around halfway up the erect lateral shoots, bearing a pseudo-whorl 4-8 leaves at the top. Blades ovate to oblong, cross veins present, glabrous, margins serrulate.
Two species are widely distributed in the Caribbean, the Gulf of Mexico and along the Pacific coast of Central America. Rhizome thin, fragile, with one root at each node, internodes 2-4 cm long. Scales at the base of the lateral shoots broadly obovate, 3-6 mm long; apex obtuse. Lateral shoots 2-4 cm long, bearing around halfway to the top 2 lanceolate to obovate scales, and at the top 2-4 pairs of leaves placed in a pseudo-whorl. Petioles to 2 mm long. Leaf blades oblong or linear-oblong, 10-30 mm long, 3-6 mm in width; apex obtuse, sometimes apiculate; margin finely serrate; cross veins 6-8, ascending at angles of 30-45 • . Spathal leaves lanceolate, acuminate sessile in the axiles of a leaf, containing only a flower. Mae flowers with tepals 4 mm long; anthers 4 mm long. Female flowers with a minute perianth; ovary ovoid, 3-4 mm long; hypanthium 3-5 mm long; styles 3, 30 mm long. Fruits globose to subglobose, 3-4 mm diam. Seeds subspherical, 1 mm in diam. testa reticulate.

Halophila baillonis
This species is widely distributed in the Caribbean Sea extending south to Brazil. The plants grow from low water down to 30 m on mud and sands. Dioecious. Lateral shoots erect, wiry, with 2 basal nodal scales and numerous nodes. Leaves up to 20 or more pairs on nodes per shoot, sessile, distichously arranged near top of shoot; blade with perpendicular cross veins. Flowers form on apical nodes of erect lateral shoots.
Halophila spinulosa is widely distributed in Malesia and Northern Australia. The species has been found from the level of mean low water spring down to around 45 m on coral platforms, coral sand and muddy sand substrates. Flowers March, October to December, fruits January to March.
The plants grow in NE Australia, from Prince Charlotte Bay to Gladstone, also in E Gulf of Carpentaria, between 1.4 and 54 m deep on muddy substrate, near mangrove and on coral sand. The plants also reported to grow in the Philippines. Flowers from September to October, fruits November to January in Australia.

New Sections in the Genus Halophila
The present study retains all four sections used by den Hartog [6] and one section from Greenway [9]. Based on both reproductive and vegetative characteristics, three new sections, section Decipientes, section Stipulaceae and section Australes, are created. All three sections were originally under the section Halophila of den Hartog [6]; thus, the genus Halophila contains a total of eight sections in this revision. The section Australes contains a single species, H. australis, and its female flowers, with six styles, which are uniquely formed on the extended floral shoots but not on the rhizome nodes. The section Stipulaceae has two species: H. stipulacea and H. balfourii. They are dioecious and possess unusually large, persisted scales and serrulated leaf blade margins. The section Decipientes contains three species, H. decipiens, H. capricorni and H. sulawesii; all have monoecious flowers with three styles, and leaf blade margins are serrulated. The section Halophila contains the remaining H. ovalis with its closely related species. They are dioecious plants with three styles and have long leaf petioles with entire blade margins. The section Microhalophila is monoecious, but on the other hand, the sections Spinulosae, Americanae and Tricostatae all are dioecious. Table 1 shows the species compositions in these eight sections.

Typifications in Halophila
Den Hartog and Kuo [1] listed H. madagascariensis (Steudel) Doty & B.C. Stone as the type species of the genus Halophila and declared that it should not be treated as a synonym of H ovalis. A complete re-description of this type species is therefore provided in this review.
The earliest collection of Halophila engelmannii from Florida is located at NY. In the herbarium sheet, the following was stated: "Deep water, coast of Florida, near Apalachicola, Dr. Chapman, received in a letter from Dr. Gray Sept 1865". This collection was made prior to Ascherson's species description in 1875. There are two fragments of plants in the herbarium sheet, each with four internodes with perfect vegetative material. It is determined here that this collection should be designated as lectotypes. Among sixteen paratypes of H. hawaiiana cited by Doty and Stone [31], only one specimen (Doty 13009, L) was collected from Maui Island, the original type locality. Therefore, this specimen has been selected as neotypes of H. hawaiiana. An exhaustive search for the original material of Brown's H. spinulosa has failed to locate any extant specimen. It should be noted that Bentham [59] had mentioned that "the specimen of C. spinulosa in Brown's herbarium have no fructification, but Kliner's specimens in F. Mueller's Herbarium in fruit is in other respects precisely similar". In fact, Mueller [60] had already provided a detailed morphological description of vegetative, female flowering and fruiting information using Kliner's material. Therefore, it is logical to consider Kliner's collection as neotypes. Currently, there are four mounted sheets of Kliner's collection at MEL. The specimen MEL 3807 is a complete specimen that matched the descriptions in the prologue and was also included in the subsequent descriptions of the species by Meuller [60]. Therefore, the specimen MEL 3807 is selected as neotype and MEL 3008, 3009 and 3012 as iso-neotypes of H. spinulosa.

New Combinations in Halophila
Taxonomically, H. stipulacea contains two synonyms: H. bullata Délile and H. balfourii Solereder [6]. The microscopic study carried out in this laboratory reveals that both H. stipulacea and H. bullata have normal flat blade epidermis, while H. balfourii has a unusual papillae epidermis, as described by Balfour [61] previously, which is a unique characteristic among Halophila species. Furthermore, H. balfourii is restricted to the Mauritius Islands, while H. stipulacea occurs in the Red Sea. Therefore, H. balfourii has been treated as an independent species, not as a synonym of H. stipulacea [62].
Currently, there are three subspecies in H. ovalis, vis.: H. ovalis ssp. bullosa, H. ovalis ssp. linearis and H. ovalis ssp. ramamurthiana. Based on a thorough morphological examination, the present review recommends that they should be treated as independent taxa, although it is open to more vigorous molecular investigation to confirm the status of these little-known taxa. The microscopic studies have shown that the leaf blade of H. bullosa has distinct ascending cross veins at 70-80 • , which is rather unique among Halophila species. Furthermore, most of the leaf blades possess numerous small bullations. Through an isozyme investigation, McMillan and Bridge [63] recommended that this species should be treated as an independent species. However, recent research preferred to treat this species as a subspecies of H. ovalis [64]. For H. ramamurthiana, unlike a typical H. ovalis, it has distinctive oblong membranous leaf blades and the plants were only found in estuarine environments. Nguyen et al. [49] distinguished this species from H. ovalis and H. ovata in Tamil Nadu, India by using DNA AFLP fingerprint gene marker, but they could not separate these three species using ITS, rbcL, matK gene markers. Therefore, based on the strong morphological differences and partial molecular support, as well as its unique habitat, it can be confidently concluded that H. ramamurthiana is an independent species. On the other hand, as no molecular or ecological study on the species H. linearis has been reported since they were named in 1959, it has been treated as H. ovalis [65,66]. Since this species has rather unique elongated slender leaf blades and only occurs in the restricted type location of Mozambique, it is suggested to treat this species as an independent taxon until future molecular study evidence becomes available to prove otherwise. On the other hand, as there is no sufficient biological, ecological or molecular information on the Halophila type species H. madagascariensis, it is recommended that further Halophila study should focus on the Southern Indian Ocean. In addition, it can be anticipated that more potential new interesting taxa may be found there.  [69] because H. sulawesii and H. ovalis both have ovate leaf blades, except that the former species was collected from 20 m deep water. Furthermore, the former species is monoecious, while the latter is dioecious. The above information suggests that these deep-water Halophila species are possibly more widely distributed. Therefore, it is recommended that future deep-water seagrass surveys should be conducted in the West Pacific Region in addition to those carried out in NE Australia [70,71].

Species Identification in Halophila
Species identification in most Halophila sections is more or less straightforward. On the other hand, some species in section Halophila may have to rely on the combination of several vegetative characteristics such as leaf blade appearance, shape, size and cross vein number, as well as space between blade margin and intra-marginal veins and the space relating to blade width, etc., to present a clear picture of different species' unique characteristics. Therefore, it is highly recommended to use light microscopes to measure these small structures in detail carefully.
Currently, section Halophila contains around thirteen described species, indicating that there is a high possibility of morphological plasticity. Some of these species, such as H. hawaiiana, H. johnsonii and H. bullosa, have been considered as conspecific species with H. ovalis because there is no genetic diversification among them [43,72,73]. Regardless of the fact that these species are morphologically different and have restricted regional distributions, it has been interpreted that they are different ecotypes of a single species with recent speciation, incomplete lineage sorting or ongoing gene flow through hybridization [74]. Indeed, H. x tanabensis as a hybrid between H. nipponica and H. major was described [56]. More recently, a hybrid between H. ovalis and H. major has been demonstrated molecularly from Sri Lanka [75]. However, using new sequencing methods recommended by Yu et al. [76] may resolve these conspecific species' complexity and assist in future seagrass taxonomical research.
It is true that some species have a vast geographic distribution and occur at different habitats, so when studied more closely, they may turn out to be separate species. For example, H. major and H. ovalis often grow adjacent to each together and their gross morphology is also very similar. As such, H. major has been treated as H. ovalis for 150 years. Since H. major (synonym as H. euphlebia) was identified molecularly [77] and morphologically [40] in Japan as an independent taxon, this species has been recorded from several countries in Southeast Asia and the Red Sea [47,50,51,78,79].
The species name used by previous molecular studies could be misrepresented. For example, H. minor used by Waycott et al.'s study [43] could in fact belong to H. gaudichardii, and H. ovata from India in Nguyen's molecular study [49] could belong to H. minor or even represent a new taxon because the species name of H. ovata has been replaced by H. gaudichardii and this species only exists in the tropical W Pacific region, not India [40]. H. minor normally coexists with H. ovalis and H. major in the shallow waters in W Indo-Pacific regions but does not occur in the Mariana region. Furthermore, a report on the H. ovalis from the West Atlantic Ocean [80] should belong to H. johnsonii, morphologically. Contrary to Shimada's molecular conclusion, Kim et al. [46] recognised that both tropical H. okinawensis and H. gaudichardii are independent species and evolved later than temperate H. nipponica.
Currently, the identification of the closely related species in the section Halophila is mainly restricted to the use of vegetative characteristics such as leaf blade appearance and shape and, more importantly, their relationship to cross veins. Due to the high plasticity in this section, some structure characteristics used in key and species descriptions may overlap among closely related species. However, it is strongly recommended to document the reproductive biology of the studied species from local communities, e.g., Muta Harah for the annual H. beccarii [ [64]. In addition, it is suggested that the seed coat testa could serve as additional Halophila species identification criteria [82]; see also Figure 1. These additional data would improve and strengthen the Halophila taxonomy further.

Concluding Remarks
It is anticipated that some of the described Halophila species in this review may become synonyms after more studies are conducted in the future. On the other hand, there are potentially more new species in the Halophila yet to be described [75,83]. It has been suggested that the new species should be accepted if it is supported by genetic data [84]. However, it is essential to carefully and thoroughly check through the samples used in the molecular studies to make sure that they correctly represent the species to be compared. It is interesting to note that Nguyen et al. [51] reported that the genetics of H. ovalis in the Red Sea are different from those of other parts of the world and the Red Sea H. ovalis is potentially a new taxon despite the two populations having identical plant morphology. Therefore, it is also very important that these new taxa should have the reproductive information to strengthen the claim.
The present review on the Halophila taxonomy is by no means perfect. We should acknowledge that since the plants are so small and useful morphological characteristics for the species identification are so limited, it is highly recommended that all future molecular studies should conduct analyses using the newly developed sequencing methods or invent better markers to analyse multiple samples across the species range, particularly to include specimens from the species type locality. In addition, these studies should check whether the morphological data of the studied taxa are accurate. Therefore, it is encouraged to apply microscopic technologies to provide detailed information on both vegetative and generative structures to ensure that the correct identity of each species studied is verified. It is particularly important that close collaboration or consultation with morphological experts is conducted and the stringent comparison of the specimens to be studied with the type specimen from the herbarium is carried out before commencing the molecular investigation. The future generation of taxonomists could apply more molecular genetic findings and should be encouraged to collaborate with molecular scientists to produce more solid and easy to use key and species descriptions for Halophila taxonomy.
It is anticipated that the major aims of achieving a clear differentiation of species and the reduction of difficult to determine species may be accomplished in the future under close collaboration between molecular scientists and taxonomists.

Materials and Methods
Halophila specimens from more than forty herbaria (Supplementary A) and numerous freshly collected Halophila specimens from Australia, Hawaii, Florida, Thailand, Singapore and Japan were studied. Up to 50% of herbarium specimens were imaged.
The specimens were examined using light microscopes (dissecting microscope or low magnification optical microscope) to measure rhizome internodes (length and diameter), petiole length, blade length and width, the number of cross veins, the distance between adjacent cross veins as well as the distance between the intramarginal vein and the blade margin. These measurements were made on the mature leaves (the third leaves from the shoot apex). Numbers of cross veins were counted including those branches (see Figure 2). Some of the Halophila species which belong to sections other than section Halophila possess certain associated features such as hairs and seta on the blade surface and were required to be examined using a Philips Scanning Electron Microscope (SEM) operated at 15 kV to show their special 3D structures. Selected samples of seed surfaces were also imaged using the Scanning Electron Microscope.
Reproductive structures such as flowers and fruits either from freshly collected or herbarium specimens were measured and imaged with light microscopes when they were available.
Botanical drawings of each Halophila species were made either from herbarium sheets or the freshly collected samples. The drawings of some species were made from more than one specimen. Sources of drawing samples for each species are listed in the Supplementary B1.