Additional floristic study of planktonic and seaweed-associated diatoms in Chuuk, Micronesia

The present study enhances knowledge of the biodiversity of diatoms in Chuuk, Micronesia following our 2018 study on the seaweed-associated diatoms. We collected planktonic samples of diatoms from five sites of reef flats using a 20 μ m mesh net, and two samples of seaweeds with epiphytes by hand from an islet on the barrier reef. In addition, the seaweed-associated diatoms from our 2018 study were analysed using scanning electron microscopy. A total of 109 diatom taxa are documented in the present study. Of these, 70 species were from net samples, and 39 species from the seaweed-associated diatoms. Thirty-one species are newly recorded from Micronesian waters. Most taxa are benthic or tychoplanktonic; euplanktonic diatoms were rare. The occurrence of benthic diatoms from the water column might be related to the Chuuk environmental conditions which include shallow water, strong light intensity and high grazing pressure, to which benthic diatoms seem to be able to better adapt than planktonic diatoms.

The coral reef diatoms of the Micronesian Islands in the Western Pacific Ocean have been intensely studied in Guam (Lobban et al., 2012;Lobban, 2015a, inter alia).The number of new species reported from Guam (Lobban et al., 2009(Lobban et al., -2019) ) is an indication of the biodiversity still to be discovered in the region, an area larger than the contiguous USA; it stretches 3700 km from Palau in the west to the Marshall Islands in the east and encompasses high islands and many atolls.Chuuk State is located at about 7°N within the Federated States of Micronesia in the western Pacific (Figure 1A).Chuuk lagoon consists of 12 major volcanic islands and 24 coral-reef islands, within a barrier reef (Figure 1B).Of 12 major volcanic islands, Weno (formerly Moen) is located in the eastern part of the lagoon.Park et al. (2018) recently began diatom floristic studies there, based on diatoms associated with seaweeds; 143 species were observed.The present collection of new records is based on net sampling of plankton samples at five sites at the same locations, to which we have added two samples from another island in Chuuk Lagoon from the GUAM diatom collection.The new analyses included scanning electron microscopy (SEM), which enabled us to identify to species some taxa previously included at the genus level.

Materials and methods
Net samples were collected from surface water (2-5 m water depth) at five sites near the coast of Weno Island on 22 May 2017 (Figure 1C, Park et al., 2018) by horizontally towing a 20 μm mesh net.The collected samples were put in 250 ml polyethylene bottles and fixed immediately with Lugol's solution to a final concentration of about 2%.Preparation of samples followed Park et al. (2018).In addition, we analysed two samples from Moch Island (Figure 1C, 7°30 ′ 50.7 ′′ N 151°57 ′ 59.8 ′′ E) at the Eastern Passage of the barrier reef, collected by CSL on 30 May 1991 by hand.These had been preserved in 4% formalin and are curated in the GUAM diatom collection.
Organic matter was removed with HCl-KMnO 4 , the remaining material was mounted in Pleurax and observed with a light microscope (LM; BX51, Olympus, Tokyo) (Park et al., 2018) or cleaned with HNO 3 and mounted in Hyrax, and observed with a Nikon i80 microscope (Lobban, 2015b).Scanning electron microscope (SEM) observations were made with a Phenom G2 Pro desktop instrument (Lobban, 2015b).

Results
A total of 109 diatom taxa were observed .Of these, 70 taxa were observed from net samples and 39 taxa were added from seaweed-associated habitats from SEM observations.Of 39 taxa, four uncertainly identified species in Park et al. (2018) were also further identified with SEM and re-listed in this study: Thalassiosira cedarkeyensis (as Thalassiosira cf.cedarkeyensis), Pleurosigma cf.elongatum (as Pleurosigma sp. 2), Amphora lunulata (as Amphora sp. 2), Nitzschia cf.amabilis (as Nitzschia sp.11).Brief diagnostics are included only for 31 species that were unrecorded from Micronesian waters (Table 1), and the figures and dimensions for all diatoms are presented because the evidence of species occurrences may be re-evaluated with further identifications in future.

Genus Tryblionella
T. apiculata • Mediterranean: Bosporus, Black Sea (Schmidt et al., 1874(Schmidt et al., -1959) ) • Indian: South Africa (Giffen, 1963) • North-western Pacific: Weno Island, Chuuk, Micronesia (This study) • South-western Pacific: Eastern Australia (Foged, 1978) Marine, benthic Comments: The distinction between Actinocyclus subtilis and A. tenuissimus is subtle.Hustedt's (1927Hustedt's ( -1930) ) key distinction between A. subtilis and A. tenuissimus is the areola density, but his drawings (figs 304, 305 respectively), show A. subtilis with dense areolae in the centre, separated by a distinct hyaline ring, and the segments not clearly separated, whereas A. tenuissimus has more scattered central areolae and distinct segments with lines of areolae stretching from the rimoportulae to the centre filled in by straight lines of areolae that end at different levels as they meet the segmental lines.The central area of A. tenuissimus is not more than 4 μm (Hustedt, 1927(Hustedt, -1930)).Al-Handal et al. (2016) mentioned Hustedt's (1927-1930) criterion to distinguish both species, namely, the areolae size in the valve: A. subtilis (18-20 in 10 μm) vs A. tenuissimus (12-15 in 10 μm).However, the areolae density of A. subtilis is more variable (12-18 in 10 μm) in Andersen et al. (1986) and Witkowski et al. (2000).We assigned the Chuuk species to A. subtilis based on the areola density and the appearance of the segments and the central area.The taxonomy of both species should be re-considered with the taxonomic history and distribution.This species is a first record from Micronesian waters including Guam.Order TOXARIALES Round, 1990Family ARDISSONEACEAE Round, 1990 Ardissonea fulgens (Greville) Grunow Samples: TK4, TK28.
Comments: Valve margins gradually sloping (in contrast to C. elongata, which is strongly spathulate), often curved.
Diagnostics: Inner setae bifurcated in valve plane, and its fused part long.
Comments: Hustedt (1927-1930: 612, figure 353) shows the very similar B. delicatulum Cleve, implying synonymy, but these species are still both considered valid.There appear to be no SEM data for B. delicatulum, and it was described from temperate waters, whereas B. furcatum is a warm-water species (Bosak et al., 2015).This species is new to Micronesian waters including Guam.Comments: Although the nominate variety of this species is cold/temperate Atlantic, var.skeleton, with extremely shallow valves and slender chaetae, 'reaches its peak in warmer seas' (Hustedt 1927(Hustedt -1930: 645): 645).
Diagnosis: Valves tangentially undulate; single fultoportula in the depressed part of valve; external tube of marginal rimoportula prominent.
Comments: With SEM examination, the morphological details of this small taxon confirm its identity as T. cedarkeyensis.This species was abundant with over 5.4 × 10 5 cells l −1 in the seaweed-associated sample at sites 1 and 2. Thalassiosira cedarkeyensis was first described from Florida benthos (Prasad et al., 1993(Prasad et al., , 2011)), and also reported from the Guangdong coast (Li et al., 2013).In addition, Risjani et al. (2021) also identified this species from Indonesian coral reef habitats.Despite the few reports of T. cedarkeyensis, the occurrences of tropical benthic diatoms in different ocean basins (Atlantic and Pacific) might be widespread along the coral reefs in the western Pacific Ocean based on its occurrence from coral reef regions.
Comments: Although the small valve size, areolae density and areolation of the Chuuk specimen match the T. catharinensis from Brazilian waters (García & Dutra, 2016), the internal structures such as number of satellite pores and slit direction of the rimoportula are needed to identify positively.
Class FRAGILARIOPHYCEAE Round, 1990Subclass FRAGILARIOPHYCIDAE Round, 1990 Order FRAGILARIALES P.  (2012) showed G. macilenta from Guam with the valve view under SEM observation.The valve in Guam specimens were heteropolar and two spines were present in the narrow pole (cf.Lobban et al., 2012, figures 3-5).In the Chuuk specimens, the valve outline was not observed, the marginal spine seems to absent from the pole, and we did not observe the apical pseudosepta.Al-Handal et al.Sample: TK28.Dimensions: 39 μm long, striae 18 in 10 μm.Diagnostics: Valve undulate, septum with a distinct wave at the apices.
Comments: We reported G. marina in our previous work (Park et al., 2018), which actually matches G. oceanica in both the septum shape and the stria density according to Sato et al. (2003).Therefore, we correctly report G. marina again.Comment: This species has been only reported from Guam, Chuuk and Marshall along the tropical coral reefs in the North-western Pacific (Ashworth et al., 2012).Comments: This species has been found in Guam and Yap but is also present in the Honduras material that Grunow used to describe L. remulus, and probably is also pantropical epiphyte on coral reef seaweeds (Lobban, 2021a), sometimes abundant, but the shape is not unique.There is a similar species, L. romuli Lobban, that may have been mistaken for L. romulus.It has decreasing vimines toward the apex.Comments: Dimensions given by Lobban et al. (2012) may confound two different species, the smaller one undescribed and also present in Chuuk samples.The size range given by Hustedt  matches the specimens shown here, but his stria density does not.Reference: Cox (1999): 140, figures 1-5, 11-14, 24, 31, 35, 40, 46, 50.Sample: CHB1-2 st1263.Dimensions: 45 μm long, 4 μm wide, striae 23 in 10 μm.Diagnostics: The length, number of striae and form of the internal central nodule serve to distinguish C. neoconstrictus, but it has hyaline extensions of the mantle beyond the stauros, which are not visible in this perpendicular image.

Gato hyalinus
Comments: The stria density distinguishes the Chuck specimen from C. paradoxa Ashworth & Lobban, recently described from Guam, which has much finer striae and a very reduced stauros (Ashworth et al., 2017).
Comments: This species, characterized by Hustedt (1931Hustedt ( -1959) ) as rare, has been reported from Java and Borneo.This is a first record from Micronesian waters.Diagnostics: Characterized by the prominent pores on the outer side of the elongate chambers.The stria density is higher than the range in Hustedt (1931Hustedt ( -1959) ) or Lobban (2015a).Diagnostics: Valves broadly elliptical.In rapheless valve, sternum elliptical in the middle and abruptly narrowing towards apices.Transapical striae radial.

Mastogloia exigua F.W. Lewis Figures 29a, b
Comments: Only two species of Anorthneis have a broad hyaline area on the sternum valve (SV), the other being A. hyalina Hustedt.The SV of A. hyalina can be distinguished from A. eurystoma by the transapically elongated areolae, and the latter has a round areolae (Pennesi et al., 2018).This is a first record from Micronesian waters.Comments: This specimen is considerably shorter than the range given in Lobban et al. (2010), 350-380 μm, but the stria density and the shapes of the areolae are similar, so there is little reason to think this is a new species.Samples: Moen_02 # 001, CHP-2 st1285.Dimensions: 261.7 μm long, 22.9 μm wide; 13 striae in 10 μm transapically, 12 areolae in 10 μm apically.
Comments: Gyrosigma sterrenburgii Stidolph (Stidolph, 1992;Sterrenburg, 1995) is a simulacrum species, differing only in some ultrastructural details and we cannot distinguish them on the basis of LM.

Navicula gregaria Donkin Figures 44 & 130
Diagnostics: Valves lanceolate with protracted apices.Transapical striae radiate in the middle and convergent at apices.Areolae visible in LM.Central area variable in size, markedly asymmetric.
Comments: The small size of Chuuk specimen is considered as a freshwater form.This is a first record from Micronesian waters.

Navicula plicatula Grunow
Comments: Park et al. (2018) reported this taxon in an unidentified status as Pleurosigma sp. 2. Valve shape and internal bisected areolae by bar are similar to P. elongatum (Sar et al. 2014), but the absence of calcar at the terminal ends in the Chuuk specimen is different from P. elongatum.We did not observe the external details such as central raphe fissure and areolae slits in the Chuuk sample.Further details are needed to identify this taxon positively.Nevertheless, there was no record on this P. elongatum-like species from Micronesian waters until now.
Comments: The stria density is at the high end of the several different ranges given in the literature, but appears to match specimens from Guam identified as this species by Lobban (2015a).Comments: This species, which has been reported as A. citronella from Guam (Lobban et al., 2012) and Tahiti (Ricard, 1977), now falls into Schizostauron (Davidovich et al., 2017;Górecka et al., 2021), but we cannot tell from the LM if the Chuuk specimen belongs to S. trachyderma as proposed or to S. kajotkei Dabek, Górecka & Witkowski.The distinction is hard to pin down because while Davidovich et al. (2017) state that their Schizostauron sp.1 is very similar to S. trachyderma except for the latter having only 9.8-12.1 striae in 10 μm on the SV, in transferring trachyderma from Achnanthes to Schizostauron, Górecka et al. (2021) give the stria density as 10-17 (vs 10-14).Even without that confusion, our cell, with 10 striae in 10 μm fits both.Schizostauron trachyderma has been reported from several Western Pacific sites as well as the Indian Ocean, S. kajotkei so far only from Madagascar.It is possible that our material fits neither but we will need good SEM to tell.Schizostauron is not related to Achnanthes, but in gene trees is sister to Astartiella and was transferred to Naviculales: Stauroneidaceae (Górecka et al., 2021).
Diagnostics: Valve linear semi-elliptic with broadly round apices.Dorsal and ventral part parallel in the middle part and smoothly curved toward ventral part.Areolae coarse and loculate.Striae slightly radiate in middle part and convergent in ends.Raphe sternum flap toward dorsal part covering adjacent areolae.
Comments: Although this species has the general characters of A. egregia, our specimen has striae all along the ventral side rather than just at the apices and along the margin, and appears to lack a longitudinal rib across the dorsal striae near the dorsal margin (Levkov, 2009).Amphora cf.helenesis Giffen Figure 143 References: Giffen (1973) Diagnostics: The variety tumidula is distinguished by an inflated centre with silica flaps at the edge of the valve face.
Comments: This group has still not been given a species name, which it lacks since Schmid (2007) restricted B. paxillifer to estuarine forms.This form is also abundant in mangrove samples from Vietnamese coasts (A.Witkowski, personal communication).This is a first record from Micronesian waters.Comments: This is a first record from Micronesian waters.
Comments: This is a brackish-water species and has been observed from the Atlantic Ocean for example the Indian River (Florida, Navarro, 1982a, 1982b), Quebec (Canada, Poulin et al., 1990).This is a first report from Micronesian waters, North-western Pacific.Comments: This species is similar to N. longissima, but N. rectilonga differs in the presence of the hyaline field, a labiate cylinder on the margin of fibulae and with respect to the size of the projection of the raphe central nodule on the inside (Shorenko et al., 2016).This is a first record from Micronesian waters.Simonsen (1987b), pl.385, figures 10-18 (as N. laevis); Witkowski et al. (2000): 387, pl. 188, figures 13-15, pl. 190, figures 1-6 (as N. laevis).
Diagnostics: Cells small, delicate.Valves broadly linearelliptical with blunt, rounded and slightly protruding ends.Sometimes middle part constricted.
Sample: Moen_03 #002.Comments: Our images of the flat sacs (palmulae) show the poration more clearly than in Paddock's images (Paddock, 1978).He describes the palmula thus: 'The stem is hollow … and the blade is composed of two laminar layers, though so far it is not known whether the envelope formed by the lamina[e] is sealed.'In the captions to his figures 5B and 5D (reproduced

Discussion
With addition of 105 diatom taxa in this study (excluding four species re-identified and as reported in Park et al., 2018), 248 diatoms are known from Chuuk, although many taxa still remain unidentified or uncertainly identified.In comparison, Lobban (2015aLobban ( , 2015b) ) reported an accumulation of 271 new records in Guam.Of the 105 species in this study, 31 species are newly recorded for the Micronesian waters (Table 1).Most diatom species were benthic around Weno Island, despite the collection of the planktonic samples using a phytoplankton net.The presence of benthic diatoms from plankton samples might be related to the wind-driven shear current in shallow water.The water depth of sampling sites did not exceed 10 m.The water column in shallow water can be vertically mixed by Langmuir circulation caused by wind in shear.Langmuir supercells can suspend sediments and transport the resuspended particles offshore by inducing turbulence along the highly sheared oscillatory wave that serves to lift sediment grains from the seafloor into the water column (Gargett et al., 2004).Langmuir circulations generally occur only for wind speeds greater than 3 m s −1 and appear within a few tens of minutes of wind onset (Talley et al., 2011).Annual mean wind speed in Chuuk was 5.14 m s −1 (Ko et al., 2015), this is sufficient to produce Langmuir circulations and caused resuspension of benthic diatoms.
were observed from the passage of the barrier reef near open water.Coral reef environments are characterized by relatively stable temperature, low nutrients and high irradiance (Lobban & Jordan, 2010).These environmental characteristics seem to have no advantages for planktonic diatoms compared with benthic ones.Strong light level can be a factor that inhibits the growth of phytoplankton (Moore & Villareal 1996), and nutrient acquisition is more favourable at the bottom than the water layer.In addition, the planktivorous zooplankton is highly abundant in coral reef systems (Nakajima et al., 2017), thus will act as a high grazing pressure for phytoplankton.
Although we collected the samples inshore, some freshwater taxa such as Gomphonema cf.lagenula, Diadesmis confervacea, Navicula gregaria, Caloneis macquariensis, Pinnularia cf.borealis and Surirella oblongoelliptica were observed.There are small streams near the sampling sites in Weno Island, and the freshwater species, observed from acid-cleaned specimens, were most likely dead.The freshwater diatoms of Micronesian islands have scarcely been studied (Zolan, 1981;Navarro & Lobban, 2009 and unpublished observations by Lobban), but we do not expect great diversity because on these small islands the streams are short, shaded and ephemeral (Lobban et al., 1991).
Many new species and some new genera have been reported from Guam in the last decade and it is not surprising that some of them are turning up elsewhere in Micronesia as exploration gets underway in Chuuk, Yap, Palau and the Marshall Islands (e.g.here reported Hanicella moenia, Astrosyne radiata, Licmophora curvata and Gato hyalinus).At the same time, those islands are also revealing new species, even from the relatively small sampling effort, some not yet found in Guam despite much more extensive sampling (e.g.Divergita macinnisii Lobban and Licmophora complanata Lobban from the Marshall Islands (Lobban, 2021a(Lobban, , 2021b)); several species from Yap (Lobban, 2021a)).While we have tended to assume a priori that all species are ubiquitous, an essentially untestable hypothesis, we are coming around to an assumption of regional endemicity as a starting point, recognizing that while there are many ubiquitous species, there are also different levels of endemicity both geographically and within genera (Lobban, in review).
It is tempting, in floristic studies such as this to compare species richness or species diversity of the list to date with the findings from other places but, apart from the question of the amount of effort put into different regions, the more important reason to avoid such comparisons is that the method in our floristic studies is to present micrographic documentation of each species as evidence for each hypothesis of identity.We have included some taxa here identified only to genus because of limited sampling, whereas Lobban's records papers (Lobban et al., 2012;Lobban, 2015a) report only taxa identified to species during the progress of a long-term study.The state of knowledge of tropical marine diatoms is still very preliminary, even in a few relatively well-studied places, and taxonomy is still in a state of flux as new studies uncover previously hidden diversity.Our studies in Guam have shown repeatedly that there are lookalike species that either cannot be distinguished in LM or that have a striking feature, thought to be unique, that leads to the similar species being overlooked.The documentation and arguments we provide for species records claims are essential for the continued utility of the work, so that findings can be reinterpreted as needed when new information is available.Lists with little or no documentation, even by very experienced authors (e.g.Giffen's (1980) list for Mahé, Seychelles), become outdated.The question of regional endemicity in benthic marine diatoms has scarcely been addressed but it is becoming clear, from our work in progress based on the approach used by Williams & Kociolek (2017), that at least some genera have strongly regional floras.Large-scale ecological studies such as those by Kryk et al. (2020) and Risjani et al. (2021) are able to make comprehensive comparisons of species diversity by including many incompletely identified species (essentially 'operational taxonomic units').Kryk et al. (2020: 161), said of their results for Madagascar, 'Amongst all 332 taxa, only 35% have been identified to the species level.Many of the taxa identified to the generic level may turn out to represent taxa new to science after further SEM examination and DNA sequencing.'The ratio is probably still lower for the studies in Micronesia but we have not yet tried to distinguish and tally the unidentified species.
There remains much work to be done on the Chuuk flora, both with the relatively few samples collected already and the need for more systematic collections.The marine flora to date continues to show strong similarities to the much better-known flora of Guam while also revealing species not yet reported from there, and for all Micronesian islands there are still few mangrove and sediment biofilm samples.

217 μm long, 140 μm wide in flattened iso
lated valves; 11 areolae in 10 μm.Diagnostics: Huge frustules elliptical in valve view, rectangular in girdle view with many copulae, numerous rimoportulae around each broad apex of the valve.Large oval central patch of randomly arranged areolae differentiated from the striae radiating from it.
Diagnostics: Valve elliptical.Striae radiate in a single row of areolae and two rows near valve margin.Raphe straight.Axial area linear, central area transapically expanded and connected with an H-shaped area.Comments: This is a first record from Micronesian waters.
Diagnostics: Valves narrowly long lanceolate with acute apices.Transapical striae parallel throughout but convergent near apices.