A new dinoflagellate cyst genus and species are described here as Hiddenocysta gen. nov. and Hiddenocysta matsuokae sp. nov. from Holocene sediments in a core from the west coast of Vancouver Island (British Columbia, Canada). The genus Hiddenocysta encompasses spherical to ovoid skolochorate cysts, characterized by a gonyaulacoid plate pattern and a 2P precingular archeopyle. The species H. matsuokae is characterized by a granular wall and slender trifurcate processes with heavily perforated process bases. Two end members are described here based on process morphology and number of processes (formas 1 and 2). Cyst wall chemistry is analyzed using micro-Fourier transform infrared (FTIR) spectroscopy and reveals a unique dinosporin composition consistent with a gonyaulacoid autotrophic feeding strategy.
Around 360 of the >2000 known species of dinoflagellates are known to produce resting cysts as part of their life cycle (e.g. Head 1996; Gómez 2012a; Penaud et al. 2018). The majority of the resting cysts are organic-walled; they can be found in nearly all aquatic sediments and have been reported from surface sediment and sediment trap studies in coastal and estuarine waters of Western Canada (e.g. Dobell 1978; Kumar and Patterson 2002; Mudie et al. 2002; Radi et al. 2007; Krepakevich and Pospelova 2010; Pospelova et al. 2010; Price and Pospelova 2011; Gurdebeke et al. 2018b; see Figure 1). Several new species have been described from this region in recent years (e.g. Mertens et al. 2012, 2013; Gurdebeke et al. 2019a; Gurdebeke et al. 2019b). Knowledge gained on the taxonomy and ecology of dinoflagellate cysts improves the quality of their application as a proxy for past environmental change in coastal British Columbia (e.g. Bringué et al. 2016), and in biological studies that, for example, try to identify seed banks of harmful species (e.g. Gurdebeke et al. 2018b). Furthermore, in recent years, two workshops on the complex genus Spiniferites renewed the interest in spiniferate cysts (e.g. Gurdebeke et al. 2018a; Limoges et al. 2018; Londeix et al. 2018; Mertens et al. 2018; Mertens and Carbonell-Moore 2018; Van Nieuwenhove et al. 2018) and chorate and proximochorate cysts produced by autotrophic dinoflagellates in general. However, despite these improvements, some dinoflagellate cyst types from this region remain poorly understood and have been left undescribed; these can now be re-evaluated (e.g. Dobell and Taylor 1981).
The macromolecular chemical composition of dinoflagellate cyst walls, classically described as ‘dinosporin’ (e.g. Fensome et al. 1993), has been studied most extensively on whole cysts using micro-Fourier transform infrared (micro-FTIR) spectroscopy. Dinosporin was found to consist of a diverse suite of complex carbohydrates (e.g. Kokinos et al. 1998; Versteegh et al. 2012; Bogus et al. 2014; Mertens et al. 2015, 2017; Gurdebeke et al. 2018a; Luo et al. 2018) with suggestions that there are differences in composition related to trophic preference (e.g. Bogus et al. 2014) and phylogeny (e.g. Gurdebeke et al. 2018a; Gurdebeke et al. 2019a). As features such as ribosomal DNA and in vivo and culture experiments are not available for dinoflagellate cysts in the fossil record, information from cyst wall chemistry is a welcome addition to the morphological data. However, study of more cyst taxa is desirable to further explore the utility of micro-FTIR for chemotaxonomy (i.e. a classification based on similarities in chemical structure and composition) and for an increased understanding of dinoflagellate (paleo)ecology and the possibly differential preservation of their cysts (Zonneveld et al. 1997).
Dobell and Taylor (1981) documented an unusual dinoflagellate cyst type from Hidden Basin (British Columbia, Canada), which was not reported since (e.g. Mertens et al. 2018). Two intergrading morphologies were described as Spiniferites sp. 1 and Spiniferites sp. 2, and the hatched motile stages were identified as Gonyaulax sp. 1 and Gonyaulax sp. 2 (Dobell and Taylor 1981). Here, morphologically identical cysts are studied from the Holocene sediments of Kyuquot Sound (Vancouver Island, British Columbia, Canada), as Hiddenocysta matsuokae gen. et sp. nov. and its cyst wall chemistry is documented using micro-FTIR.
2. Material and methods
A total of 103 samples (UVic 15-583 to UVic 15-684; Supplementary Information Table S1) were subsampled from a 9.9 m long piston core (2012002PGC133), retrieved in 2012 by the Geological Survey of Canada 8 km from the head of Tahsish Inlet in Kyuquot Sound (50°5′27.80″N, 127°9′2.75″W; 147 m water depth; Figure 1). The samples were taken at regular intervals (every 10 cm) in the core, with the exception of the upper 50 cm, where a resolution of 5 cm was used. A plastic syringe was pushed in the core sediment to sub-sample ∼2 cm3 of material. The sediment is mainly olive-green silty mud. The subsamples were kept in 15 cc polypropylene test tubes. The chronology of core 2012002PGC133 was established from five accelerator mass spectrometry (AMS) 14C measurements on plant fragments or bulk sediment (Gurdebeke et al. 2019a; Gurdebeke et al. 2019b; see Figure 2).
Sediment subsamples were treated following the palynological preparation method used at the Paleoenvironmental Laboratory, School of Earth and Ocean Sciences, University of Victoria, Canada (e.g. Pospelova et al. 2010; Price et al. 2016). After drying (∼40 °C) and weighing the sediment sample, a fixed number of exotic marker spores (Lycopodium clavatum) was added (one tablet from batch #177,745; 18,584 spores/tablet) for quantitative abundance estimations ( Table S1). To remove carbonates, samples were treated with hydrochloric acid (room temperature 10% HCl). After rinsing with distilled water, samples were sieved through a 120 µm mesh and captured on 10 µm mesh nylon Nitex screens to separate coarse and fine fractions. Subsequently, the samples were treated for 4–5 days with hydrofluoric acid (room temperature 50% HF) to remove silicates. Possible fluorosilicates were removed by a second round of room-temperature 10% HCl treatment. After rinsing with distilled water twice, the residue was sonicated for up to 30 s and collected on a 10 µm mesh screen. One or two drops of the residue were mounted in glycerine jelly, covered with a cover slip and sealed with transparent nail polish.
Slides were systematically scanned in non-overlapping traverses at 400× magnification using a transmitted light microscope (LM) (Zeiss Axioskop 2 plus). A targeted count of 300 dinoflagellate cysts were systematically identified per slide ( Table S1). LM photomicrographs were produced from the same microscope using an MRc 5 (Zeiss) camera. Cysts were picked with a micropipette from the residue of sample 15-609 and photomicrographed on an Olympus BX51 microscope with a digital sight DS-1L 1 module (Nikon, Tokyo, Japan).
Field-emission scanning electron microscope (FE-SEM) images were taken from cysts picked from the residue of sample UVic 15-609 (7.2 m core depth, ∼9.8 ka). Specimens were first isolated from the samples with a micropipette under an IX51 (Olympus) or IM35 (Zeiss) inverted microscope and transferred onto polycarbonate membrane filters (Millipore, Billerica, MA, USA; GTTP Isopore, 0.22 µm pore size). After 12 to 24 h of air-drying, the filters were affixed to aluminum stubs with adhesive tabs. Other specimens were subjected to dehydration and critical point drying, as outlined in Chomérat and Couté (2008), to prevent them from collapsing during drying. Subsequently, the mounted filters were sputter coated with gold using a Cressington 108 Auto Sputter Coater. The samples were observed with a Sigma 300 (Zeiss) field-emission SEM equipped with conventional Everhart–Thornley and in-lens detectors of secondary electrons at 1.5 and 5 kV, located at the Station of Marine Biology of Concarneau.
For micro-FTIR analysis, cysts were manually picked from palynological residue from the core-top sediment sample (UVic 15-684). The residue was ultrasonicated for 30 s and rinsed 3 times with organic solvents (methanol and dichloromethane, 1:1 by volume) and MilliQ water over a 20 µm mesh-size filter to remove polar and apolar compounds that might have adhered to the cyst walls. Visually clean and empty cysts were manually isolated, placed on a gold-coated mirror and air-dried. Specimens were analyzed with a Bruker Hyperion 2000 microscope coupled to a Bruker Vertex 80v FTIR spectrometer in the Department of Solid State Sciences (Ghent University, Belgium) over an infrared range of ∼4000–650 cm–1. The objective magnification of the microscope was set at 15× and the aperture at 100 × 100 µm. All presented spectra were recorded in reflection mode, at a resolution of 2 cm–1, and averaged over 100 scans. Data were analyzed with the OPUS© software (Bruker 2014). The presented absorbance spectra were obtained after background subtraction, atmospheric correction and baseline correction with a rubberband method using polynomes. Functional groups were identified based on comparison with Colthup et al. (1990) and published data (e.g. Cárdenas et al. 2004; Bogus et al. 2014; Gurdebeke et al. 2018a). Data reduction was performed as a principal component analysis (PCA) using PAST 3.2 (Hammer et al. 2001).
3.1. Systematic paleontology
Division DINOFLAGELLATA (Bütschli 1855) Fensome et al.
Class DINOPHYCEAE Pascher 1914
Order GONYAULACALES Taylor 1980
Family GONYAULACACEAE Lindemann 1928
Hiddenocysta. P Gurdebeke, V Pospelova, KN Mertens and S Louwye gen. nov.
Etymology. In reference to Hidden Basin (Nelson Island, British Columbia, Canada), where similar cysts were reported by Dobell and Taylor (1981).
Type species. Hiddenocysta matsuokae sp. nov.
Diagnosis. Skolochorate cyst with spherical to ovoid central body bearing gonal and intergonal processes reflecting a gonyaulacacean tabulation (?3–4′, 6″, 5‴, 1p, 1′‴). Processes connected or not by crests. Compound archeopyle consisting of 3″ and 4″, operculum free.
Remarks. This cyst was first reported by Dobell and Taylor (1981) in surface sediments from Hidden Basin (British Columbia, Canada), who considered it to belong to the genus Spiniferites (Spiniferites sp. 1 and S. sp. 2).
Comparison. Hiddenocysta's 2P precingular archeopyle distinguishes the genus from Spiniferites and Achomosphaera, which have a single (3″) plate archeopyle. A further difference from Spiniferites is that the latter genus always has pronounced crests connecting the process bases, which are often absent in Hiddenocysta. The genus Achomosphaera, on the other hand, never has crests (Mertens et al. 2018). The extant genus Bitectatodinium has a similar 2P precingular archeopyle but is ‘without obvious surface projections’ (Wilson 1973, p. 351). A number of Mesozoic cyst genera have a similar general appearance and a two-plate archeopyle, but each can be distinguished from Hiddenocysta: Kiokansium (Duxbury 1983) has a smooth wall, simpler processes and a complete absence of tabulation; Taleisphaera (Masure 1986) has diagnostic penitabular crests; Pulchrasphaera (Schiøler et al. 1997) has low non-tabular ornamentation; Nexosispinum (Davey 1979) has anastomosing processes; Corradinium (Masure 1986) mainly has intratabular processes; Gongylodinium (Fenton et al. 1980) is non-tabulate and does not bear processes; and Occisucysta (Gitmez 1970) is characterized by a pronounced apical boss and denticulate or spiny sutural crests. The genera Durotrigia, Dissiliodinium and Cavatodissiliodinium have a multiplate precingular archeopyle but one that most often involves more than two plates (Feist-Burkhardt and Monteil 2001). Furthermore, species of Dissiliodinium are characterized by negative structures (i.e. grooves), Cavatodissiliodinium is cavate, and Durotrigia species have ridges and septae, but no processes (Feist-Burkhardt and Monteil 2001).
Hiddenocysta matsuokae. P Gurdebeke, V Pospelova, KN Mertens and S Louwye sp. nov.
Spiniferites sp. 1 Dobell and Taylor (1981), pl. 1, figs 4, 11.
Spiniferites sp. 2 Dobell and Taylor (1981), pl. 1, figs 9, 10, 12, 13, 16.
Spiniferites sp. Dobell and Taylor (1981), pl. 1, figs 14, 15, 19, 20.
Etymology. In honor of Dr. Kazumi Matsuoka, who made fundamental contributions to the taxonomy of dinoflagellates, mainly from the Pacific domain, specifically about cyst stages.
Repository. Royal Belgian Institute of Natural Sciences (RBINS, Brussels), accession no. IRSNB b7162.
Diagnosis. Cysts relatively large, spherical or slightly elongated, skolochorate. Wall granular and ornamented with numerous processes. Generally, gonal processes trifurcate and intergonal processes bifurcate. The distribution of the processes reflects a gonyaulacoid plate pattern. Process bases are hollow and periphragm is heavily fenestrated. Neighboring processes can be connected by crests, mostly by a discrete lineation reflecting a septum or by an expanded periphragm that merges between two processes. One type of process is long, slender and trifurcate, with each of the three ends bifurcate and recurved. Other processes are shorter and broader, with more expanded bases, and are also often trifurcate with bifurcate terminations. Distally from the pericoel, the processes are solid, but occasionally with one or more bubble-like structures along the length of the process shaft, above, below or at the trifurcation. Archeopyle compound; operculum involves two precingular plates (3″ and 4″) that release separately, and is free.
Dimensions. Holotype: central body 62 × 51 µm; process length 14.8 µm (N = 4). Forma 1 has a smaller central body and longer processes, whereas Forma 2 has a larger central body and shorter processes. The average sum of the central body diameter and process length is ∼71.9 µm for Forma 1 and ∼71.2 µm for Forma 2. The central body wall is ∼1.1 µm thick (N = 3). For Forma 1, the height of the trifurcation above the process base is ∼7.7 µm (N = 16) (see Table 1).
Description. The cysts are spherical to ovoidal. The archeopyle margins follow the suture lines, and a projection from the apical margin of the archeopyle suggests 2P (Plate 3, figures 1–3). These plates are of similar size; plate 3″ is pentagonal and 4″ is quadrangular (Plate 3, figures 1–2; see also Dobell and Taylor 1981). The cyst wall is granular on the outside and smooth on the inside (Plate 2, figure 5), and an endophragm and periphragm can be distinguished (cf. Evitt 1985; Figure 3). Near the processes, the phragma separate, and a pericoel forms the hollow process base. The periphragm lining the pericoel is heavily fenestrated (Plate 1–Plate 3). Two morphological end members can be defined within this species, and morphologies ranging between these end members were observed. The first end member (Forma 1; Figure 3; Plate 1, figures 1–6; Plate 2, figures 1–2) has long and slender trifurcate processes terminating in recurved bifurcations resembling tendrils (cf. a Greek Ionian pillar). A single bubble-like structure is present below, at or above the trifurcation (Plate 2, figure 1). The processes in Forma 1 are never connected by crests but by very low lineations only. The other end member (Forma 2; Figure 3; Plate 2, figures 2–9) has no clear single bubble-like structures as in Forma 1 but sometimes has numerous distributed small bubble-like structures in the processes (Plate 2, figure 7); process bases are more expanded, more heavily fenestrated and more often connected by the merging of the expanded periphragm at the respective process bases. There are more intergonal processes in Forma 2. There is no apparent heterogeneity in processes with respect to position on the cyst body. Distally from the pericoel, the process is solid, but in high-magnification SEM micrographs of broken processes of Forma 1 (see Plate 3, figure 6, indicated by an arrow), eight holes are visible, spread more or less evenly along the width of the process. It is unclear whether these are small bubble-like structures or canals running along the length of the processes. An apical boss is occasionally observed (Plate 2, figures 2, 8), but appears not to be diagnostic. The cysts are often collapsed, folded and torn in palynologically prepared slides, which can also be observed when manipulating cysts (Plate 2; Plate 3, figures 4, 9, 10).
Remarks. Two end members of this species were distinguished based on the process type. Specimens were observed bearing both types of processes, suggesting intraspecific variation. The species described here is identical to the specimens depicted by Dobell and Taylor (1981) based on size, tabulation, wall structure and the structure and organization of the processes.
Comparison. As it is the only species described in the genus Hiddenocysta, the comparison is as for the genus. The processes of H. matsuokae Forma 1 are often more slender than those of most Spiniferites species (Limoges et al. 2018; Mertens et al. 2018). Spiniferites solidago of de Verteuil and Norris (1996) has a comparable shape and the same characteristic vacuoles in the processes. However, this species differs from Hiddenocysta matsuokae by a 1 P archeopyle, its wall ornamentation, its characteristic alveoles in the cyst wall and process terminations that are more simply bifurcate and do not resemble tendrils. Spiniferites lazus has similar perforations at the process base, but has a more elongated central body and a 1 P archeopyle (Reid 1974; Gurdebeke et al. 2018a).
Distribution and biostratigraphy. The species is described here from estuarine waters of southwestern British Columbia and Vancouver Island (Canada), specifically from surface sediments of Hidden Basin (∼6 m water depth, sea surface temperature (SST) 7–16 °C, sea surface salinity (SSS) 24–32 PSU; e.g. Gurdebeke et al. 2018b) and as a present to rare species (up to 1.6% of the assemblage; terminology of Pospelova et al. 2004) throughout Holocene sediments of Kyuquot Sound (∼150 m water depth; Figure 2; Table S1). A specimen with endospore/cell content (Plate 1, figures 1–6) was found in the early Holocene, ∼11 ka (UVic 2015-594). This excludes reworking of the cysts, which is also improbable because of the local geology and hydrology of the fjord system (e.g. Muller et al. 1974) and because Dobell and Taylor (1981) successfully hatched similar cysts. The species was also observed in Ise Bay, Central Japan (34°57.24′N, 136°43.84′E; 25 m water depth; SST 10–23 °C and SSS 33–34 PSU; e.g. Narita et al. 2006; K. Matsuoka, pers. comm.). From these limited observations, H. matsuokae occurs in ice-free estuaries of up to 150 m water depth in the temperate North Pacific. At present it is not clear whether formas 1 and 2 represent ecotypes, and if so, what the driving environmental parameter is. It is observed, however, that at the majority of the stratigraphic levels at which H. matsuokae is recorded in the Kyuquot Sound core, it is dominated by one forma (Figure 2).
Motile stage equivalent. Dobell and Taylor (1981) incubated cysts from surface sediments in Hidden Basin, which yielded dinoflagellates assignable to Gonyaulax, one of which they considered perhaps conspecific with Gonyaulax alaskensis, originally described by Kofoid (1911) and later restudied by Borgese (1987). Dobell and Taylor (1981) did note significant differences between their Gonyaulax sp. 1 and Gonyaulax alaskensis, particularly in size and cingular displacement. A relationship to G. alaskensis is interesting and would support a separate classification of this group of species from other Gonyaulax, initially proposed by Kofoid (1911). Kofoid (1911) placed G. alaskensis with G. fragilis in the subgenus Steiniella, separate from the typical Gonyaulax that have been related to Spiniferites. These latter were placed in the spinifera group of the subgenus Gonyaulax. Molecular phylogenies support such a separation: G. fragilis from the subgenus Steiniella (there is no sequence for G. alaskensis yet) is placed separately from a clade containing species from the spinifera group of the subgenus Gonyaulax, G. spinifera (sensu stricto) and G. digitale (Escalera et al. 2018; however see also Carbonell-Moore and Mertens 2019). However, the situation is more complicated since G. polygramma is in the same clade as G. fragilis in the phylogenies; and Kofoid (1911) placed G. polygramma in the polygramma group of the subgenus Gonyaulax. It is also of interest to note that Ataxiodinium choanophorum can be found in the same clade as G. fragilis and G. polygramma (Mertens et al. 2017). New cyst-theca experiments with SEM observations and molecular phylogenetics are needed to establish the correct cyst-theca relationship of this and related species.
Morphometrics of Hiddenocysta matsuokae sp. nov.
Micro-Fourier transform infrared assignments of functional groups present in the spectra of Hiddenocysta matsuokae sp. nov.
3.2. Chemical composition
Micro-FTIR spectra were retrieved from three specimens from the core-top sediment sample in Kyuquot Sound (UVic 2015-684) (Figure 4). The absorption peak assignments are given in Table 2. The spectra show a broad absorption peak around 3275cm–1, characteristic of O-H stretching vibrations, and two peaks at 2920 and 2860cm–1 (C-H stretching). Within the fingerprint region, the main absorptions are centered at ∼1040cm–1, with a second maximum around 1640cm–1. The absorption spectra most closely resemble those of Spiniferites species (Gurdebeke, et al. 2018a; the spectrum of S. elongatus is shown in Figure 4 for comparison) but differ in having a pronounced shoulder at 1540cm–1 and in the region between 1500 and 1185cm–1.
In all, the spectra reveal a cyst wall composition similar to those recorded for autotrophic taxa (e.g. Mertens et al. 2015, 2017; Gurdebeke, et al. 2018a; Luo et al. 2018), which suggests a similar trophic preference for the motile stage that produces Hiddenocysta matsuokae sp. nov. This is in line with the morphology-based placement of this species with the Gonyaulacaceae, in which the majority of the species are autotrophic (e.g. Gaines and Elbrächter 1987; Gómez 2012b). In a PCA analysis comparing data from earlier studies (Gurdebeke, et al. 2018a; Luo et al. 2018; Gurdebeke et al. 2019a; Gurdebeke et al. 2019b), the spectra for H. matsuokae form a distinct cluster confirming quantitatively the chemical distinctiveness of this species (Figure 5). Also, the three spectra of H. matsuokae line up with other spectra of cysts of autotrophic taxa (Spiniferites, Bysmatrum, Prorocentrum) and are well separated from spectra of typical brown cysts of heterotrophic taxa (Quinquecuspis, Votadinium, Trinovantedinium pallidifulvum) (Figure 5). The spectra for Trinovantedinium applanatum cysts are positioned intermediately (Figure 5). These transparent cysts are produced by a heterotrophic species and were shown to have an aberrant wall composition compared with other cysts of heterotrophic dinoflagellates (Gurdebeke et al. 2019a). In all, the PCA analysis supports autotrophy for Hiddenocysta matsuokae.
A new gonyaulacoid dinoflagellate cyst genus and species, Hiddenocysta matsuokae gen. nov. et sp. nov., is described from Holocene sediments from Kyuquot Sound (Vancouver Island, British Columbia, Canada). The species is presently known only from a few reports from estuarine sites in the northern Pacific Ocean. Cyst wall chemistry reveals a distinct macromolecular composition and supports the placement of H. matsuokae as an autotrophic species within the Gonyaulacaceae. These results once more show the utility of micro-FTIR in the study of dinoflagellate cyst taxonomy and ecology. New incubation experiments and sequencing (e.g. ribosomal DNA sequences) are required to further evaluate the relationship with other gonyaulacoid species.
We are grateful to Dr. Audrey Dallimore for the sediment core collection and for sharing this material with us. Victor Pospelov and Alison Thomas are thanked for their help with sediment sampling and palynological preparation at the University of Victoria. VP acknowledges funding support from the Natural Science and Engineering Research Council of Canada (NSERC) Discovery Grant. The Regional Council of Brittany, the General Council of Finistère and the urban community of Concarneau-Cornouaille-Agglomération are acknowledged for the funding of the Sigma 300 FE-SEM of the station of Marine Biology in Concarneau that was used in this study. The Research Foundation Flanders (Fonds Wetenschappelijk Onderzoek, Hercules Foundation) is thanked for funding the micro-FTIR facility (FT-IMAGER, project number AUGE/13/16), and Henk Vrielinck is thanked for technical assistance and discussion of the micro-FTIR data. We thank Dr. Kazumi Matsuoka for sharing his observations from Japanese coastal waters. Stijn De Schepper and two anonymous referees are gratefully acknowledged for reviewing the manuscript.