The challenges of exploring remote deep-sea abyssal environments have, thus far, insulated the deep-sea benthos from the impacts of mineral resource extraction. However, the advent of the technological means to access and exploit the considerable mineral resources found in these environments (Ghosh and Mukhopadhyay, 2000), heralds imminent, unprecedented levels of disturbance in the deep-sea (Weaver et al., 2018). Fauna that have evolved under highly stable, food-limited environmental regimes are likely to be poorly adapted to large-scale disturbances that rapidly and/or irreversibly alter their environment (Stearns, 2000). The CCZ in the tropical Northeast (NE) Pacific exemplifies this scenario, where the largest known global reserve of polymetallic nodules is located, formed over geological timescales in abyssal soft sediments characterised by very low sedimentation rates (Hein et al., 2013). Recent studies that have sought to describe the benthic fauna that typify the CCZ have also revealed extraordinarily high taxonomic diversity across representatives from all Domains of Life (e.g., Amon et al., 2016; De Smet et al., 2017; Shulse et al., 2017; Wilson, 2017; Goineau and Gooday, 2019; Hauquier et al., 2019; Brix et al., 2020; Christodoulou et al., 2020), making the CCZ of critical importance for biodiversity conservation. In many of these studies, data (e.g., species-abundance curves) strongly suggest that many – arguably most – species remain unaccounted for, with high species turnover over relatively short spatial scales even in groups that are brooders like Isopoda (e.g., Wilson, 2017; Brix et al., 2020). Remoteness of habitat and spatial variability in species composition of this sort present challenges for performing robust ecological studies, as evidenced by the scarcity of ecologically meaningful data available for even the most conspicuous “common” epifaunal species in the CCZ (Danovaro et al., 2017), including many echinoderm species (Amon et al., 2016).

Members of the phylum Echinodermata represent some of the most biomass-dominant taxa found in deep-sea abyssal plains (Gage and Tyler, 1991). Echinoderms play a significant role in global marine carbon budget (Lebrato et al., 2010) and are abundant in many soft- and hard-substrate habitats globally (Gage and Tyler, 1991). The class Ophiuroidea are known to be particularly prevalent and diverse on bathyal slopes (e.g., O’Hara et al., 2008) but data concerning species diversity and abundances at abyssal depths are scarce by comparison (Stöhr et al., 2012). In the equatorial NE Pacific, such data is limited to select historic (e.g., HMS Challenger and Albatross expeditions, Lyman, 1878, 1879, 1882; Clark, 1911, 1949) and contemporary surveys (Amon et al., 2016; Vanreusel et al., 2016; Christodoulou et al., 2020). Drivers for large-scale regional variability in abyssal ophiuroid population densities are still poorly understood; however, in the CCZ it seems that local-scale patchiness in nodule substrate availability plays a key role in determining the distribution of many species. Under certain habitat conditions, ophiuroids and the echinoderms more generally, comprise one of the largest components of mobile epifauna at nodule-rich sites (Amon et al., 2016; Vanreusel et al., 2016: up to 15 individuals per 100 m², with major contributions from ophiuroids). The presence of nodules appears particularly relevant to certain ophiuroid and echinoid species; when nodules are absent, mobile epifaunal densities fall sharply to one or two encounters in an equivalent 100 m² area, largely due to much-reduced encounters with ophiuroid species (Vanreusel et al., 2016). Two recent studies published in 2020, one based on video-transect surveys and the other on specimens collected by remotely operated vehicle (ROV) and epibenthic sledge (EBS), have also identified ophiuroids in abundance and notably, at unexpectedly high levels of diversity in the easternmost regions of CCZ (Christodoulou et al., 2020; Simon-Lledó et al., 2020) with the discovery of previously unknown ancient lineages (Christodoulou et al., 2019, 2020).

Current knowledge of reproductive biology in abyssal ophiuroid species is limited to a few papers examining gametogenic and/or size-class patterns (e.g., Ophiomusa lymani in N Atlantic & NE Pacific, Gage, 1982; Gage and Tyler, 1982; Ophiocten hastatum in NE Atlantic, Gage et al., 2004; Ophiura bathybia, Amphilepis patens, Amphiura carchara, and Ophiacantha cosmica in NE Pacific, Booth et al., 2008; Ophiuroglypha irrorata loveni, Ophiura lienosa, Salix daleus, O. cosmica, Ophiernus quadrispinus, and Ophioplexa condita in S Indian Ocean, Billett et al., 2013), with no data available for specimens collected from polymetallic-nodule habitats. Two of the more frequently encountered ophiuroids within the eastern CCZ (Christodoulou et al., 2020) are the brittle stars Ophiosphalma glabrum (Lütken and Mortensen, 1899; Ophiosphalmidae) and Ophiacantha cosmica (Lyman, 1878; Ophiacanthidae). These two species make for an interesting comparative reproductive study due to their contrasting biology. Ophiosphalma glabrum is relatively large (35–40 mm maximum disc diameters, Clark, 1911, 1913), epifaunal on soft sediments (0–2 mm burial, Amon et al., 2016; Glover et al., 2016; Christodoulou et al., 2020) and likely a generalist deposit feeder, as is the case in the closely related genus Ophiomusa (Pearson and Gage, 1984, formerly Ophiomusium). Ophiacantha cosmica, by contrast, is considerably smaller (11–12 mm maximum disc diameters, Booth et al., 2008; Billett et al., 2013) and epizoic or epifaunal on hard substrata (Billett et al., 2013), where it filters feeds with its spinous arms (Pearson and Gage, 1984). Although reproductive data are currently not available for Ophiosphalma glabrum, a few data already exist for Ophiacantha cosmica in nodule-free habitats. These indicate that O. cosmica is probably both gonochoric and lecithotrophic (oocyte 30–560 μm in diameter in specimens from the Southern Indian Ocean, Billet et al., 2013), with seasonal fluctuations in body-size structure, suggesting that specimens spawn in response to peaks in particulate organic-carbon (POC) flux (Booth et al., 2008).

As part of a wider concerted effort to address significant knowledge gaps in our ecological understanding of nodule-rich seabeds while this habitat remains relatively pristine, the current study describes the reproductive biology of the brittle stars Ophiosphalma glabrum (Lütken and Mortensen, 1899; Ophiosphalmidae) and Ophiacantha cosmica (Lyman, 1878; Ophiacanthidae) in a nodule-rich environment, based on the histological analyses of specimens collected from several mining-contract areas within the eastern CCZ.

The current study provides new ecologically relevant reproductive data for two species of ophiuroid living in an oligotrophic, highly stable low-energy environment in the NE equatorial abyss which, due to the presence of polymetallic nodules, is under future threat from deep-sea mining activities (Weaver et al., 2018). The current study documents the reproductive biology of O. glabrum for the first time. Equal sex ratios for intermediate modal disc-diameter size classes and the absence of any simultaneous hermaphrodites both point to gonochorism in O. glabrum. Unfortunately, the manner (and possibly timing) of sampling may have created biases in the size classes of each sex. Large (and ripe) females (i.e., >20-mm disc diameters) were conspicuously absent, suggesting either an unlikely upper limit on female disc diameter, or most probably, that large females were overlooked. This may simply be symptomatic of a patchy (or sex-specific) species distributions, where large communal echinoderm aggregations in the soft-sediment abyss only really occur during the episodic arrival of labile organic matter to the seafloor (Kuhnz et al., 2014; Smith et al., 2018) or during aggregative spawning events (Mercier and Hamel, 2009).

Oocyte feret diameters frequently exceeding 300 μm in the current study (max 453 μm in a mature oocyte) are indicative of a lecithotrophic (non-feeding) larval mode with substantial yolk-protein and lipid reserves (Sewell and Young, 1997; Young, 2003). Although this places an upper limit on transport time relative to planktotrophy, lecithotrophy releases larvae of nutritional constraints that likely exist in food-impoverished transport environments. In addition, if lecithotrophic dispersal occurs in deep waters, where temperatures remain relatively low, this can act to extend transport times through metabolic suppression, with capacities for dispersal equivalent to planktotrophy (Mercier et al., 2013). Planktotrophic (feeding) larvae, by contrast, would be confined to water masses with higher detrital input (Young et al., 1997). Lecithotrophy necessitates greater energy investment per oocyte than in planktotrophy. This likely places constraints on fecundity in a food-limiting environment (Ramirez Llodra, 2002), as evidenced by the inverse relationship between oocyte densities and feret diameter in the current study. Ovaries in the current study when nearly ripe, were classified as such based mainly on oogenic characteristics, since striking external indications of maturity were lacking in female specimens. Coloration was inconsistent and ovaries were not engorged, nor had they spread extensively into the bursae, as was witnessed in at least one ripe male. This indicates considerable scope remained for further reproductive investment in the females collected, under suitable environmental conditions. Without ripe females and intra- and inter-annual time-series sampling it is not possible to ascertain whether gametogenesis is seasonal (Tyler and Gage, 1980; Gage and Tyler, 1982) or aperiodic (“continuous”), though the presence of all oogenic stages in most specimens suggests aperiodic or semi-continuous spawning behaviour (Brogger et al., 2013). Oocyte size-class data also appear to suggest that slight differences exist between 2015 and 2019, with a higher proportion of females approaching maturity in 2015. This could reflect the slightly different timing of cruises, as no post-spawn recovery females were found in 2015 or site-related differences for each year since site locations were not the same. Further data are necessary to confirm this.

In an oligotrophic environment lacking strong seasonal fluctuations in POC flux, apparent aperiodic (or “continuous”) gametogenesis may arise in species that spawn (1) periodically, but histological studies fail to identify their cyclical nature (Mercier et al., 2007; Mercier and Hamel, 2008) or (2) opportunistically, occurring in rapid response to increased food supply related to episodic massive changes in surface productivity and as such, remain undetected (Mercier and Hamel, 2009). Coupled with lecithotrophy, aperiodic opportunistic spawning would be resource-driven, allowing a species to respond to fluctuating food availability by investing in gametogenesis only when suitable conditions arise (e.g., Booth et al., 2008). Responsive reproductive modes of this sort appear to be a feature of bathyal and abyssal ecosystems in the tropical NE Pacific (Kuhnz et al., 2014; Smith et al., 2018), where the influences of episodic disturbance events are of greater influence than seasonal changes in surface productivity. However, additional sampling from highly oligotrophic NW regions of the CCZ (e.g., APEI 3, Vanreusel et al., 2016; Christodoulou et al., 2020) would be needed to establish whether lower POC flux translates into reduced energetic investment in reproduction.

Maximum sizes recorded for O. glabrum vary considerably in the literature, but the largest attributed to this species is a 35 mm disc diameter (Clark, 1913), though there exists an Ophiomusium multispinum specimen – now synonymised with Ophiosphalma glabrum – with a 40-mm wide disc (Clark, 1911). In either case, the largest individual in the current study falls considerably short of these sizes, at 25.2 mm. Aside from some concerns around species assignation in historical samples and difficulties in collecting representative samples of species that occur at low densities in expansive and remote environments, these larger specimens also originate from shallower habitats (e.g., Cocos, Malpelo and Carnegie Ridge, Panama Basin, depths 800–3,200 m, Clark, 1911, 1913), where greater food availability arriving from productive coastal systems (Pennington et al., 2006) could enable larger maximum sizes. Although smaller specimens were undoubtedly underrepresented, enough immature O. glabrum were collected to establish a rough size at first maturity of <12 mm, around 30–35% maximum size. Without knowledge of growth rates, the time taken to reach these sizes remains unknown; however, if measures of growth (e.g., Gage et al., 2004) or temporal size-class frequency analyses can be compiled in future studies, this value could be easily converted to an estimate of age at first maturity, a highly relevant conservation metric for response to disturbance.

Sample sizes for the smaller species Ophiacantha cosmica limit any detailed interpretation of their reproductive biology. However, some data on size-class distributions and oocyte feret diameters already exist for this species from nodule-free habitats at Station M in the NE Pacific and around Crozet Island in the Southern Indian Ocean (Booth et al., 2008; Billett et al., 2013). Although sex ratios were significantly different in the current study, little stock can be placed in this result as the total number of mature specimens numbered less than 10 and maximum disc diameters for both sexes are markedly lower than in the literature. Evidence from other studies indicate that sex ratios follow a 1:1 ratio when sampled in larger numbers (Billett et al., 2013), with maximum disc diameters of 11–12 mm (Booth et al., 2008; Billett et al., 2013; M. Christodoulou, personal observations), although the disc diameter for this species’ holotype is unaccountably large, at 18 mm (Lyman, 1878). Maximum oocyte feret diameter in the current study (VIII oocyte at 273 μm) also appears to be underestimated when compared with those of specimens from the Crozet Plateau (Billett et al., 2013), which reached maximum feret diameters exceeding 500 μm, indicating lecithotrophy (or direct development, though no evidence of brooding has ever been recorded, Billett et al., 2013; this study). This is to be expected as females in the current study were relatively small by comparison and only in developing states, with no oocytes in the final stages of meiosis (i.e., asymmetric or later). The only individual classified as post-spawn/recovery was probably developing for the first time, particularly in light of its small disc diameter. Unlike the protandric hermaphrodite Ophiacantha fraterna (Tyler and Gage, 1982; species identification updated from Ophiacantha bidentata by Martynov and Litvinova, 2008), data from the current study – corroborated by that of Billett et al. (2013) – indicate that O. cosmica is gonochoric. However, the current study lacks sufficient data to confirm whether O. cosmica is also iteroparous, as is the case for O. fraterna (Tyler and Gage, 1982). Size at first maturity for O. cosmica was assessed in the current study but remains approximate at <4.96-mm disc diameter (in females), due to ambiguity around two relatively large specimens (6–7 μm disc diameters) that possessed tiny bud-like gonads in which no evidence for gametogenesis could be found. These gonads did possess nodes in which cell clumps with highly basophilic nuclei were located, which are believed to be gonadal precursors housing primordial germ cells (Figure 5C).

Ophiacantha spp. have been recorded from various hard substrata and vertically elevated fauna such as deep-sea corals (Tyler and Gage, 1980) and tube worms (Lauerman et al., 1996; Billett et al., 2013). In the current study, this species is recorded on stalked-sponge species Caulophacus sp. and Hyalonema spp. for the first time (Figure 9). Although alternative hard substrates are available in the CCZ (most obviously, polymetallic nodules), O. cosmica specimens here were exclusively epizoic; this species was not encountered elsewhere. Stalked sponges are some of the most elevated of fauna on the seafloor in the CCZ; colonising them at positions above the benthic boundary layer would mitigate the influences of shear on current speeds, which likely aids in a suspension-feeding lifestyle (Booth et al., 2008). The specimens in the current study either occurred as single individuals or, in certain instances, as size-stratified groups (Figure 9), suggestive of gregarious behaviour, offspring retention or fissiparity (though pentameral Ophiacantha spp. are not known to be fissiparous, Lee et al., 2019). A more detailed molecular work, beyond the scope of this study, would help to reveal common or distinct genetic origins for the members of these groups.

In contrast, Ophiosphalma glabrum examined in the current study was epibenthic as found in previous studies both within the CCZ (Amon et al., 2016; Glover et al., 2016; Christodoulou et al., 2020) and more generally in the NE Pacific (Booth et al., 2008). Stomach contents of dissected specimens also indicate that O. glabrum is a deposit feeder, like members of the closely related genus Ophiomusa (Pearson and Gage, 1984). Single individuals generally occurred in the sediment around nodules, or under habitat-forming fauna, such as non-stalked sponges and xenophyophores – giant, deep-sea foraminifera with delicate agglutinated tests (Goineau and Gooday, 2019).

The limited availability of food in the deep-sea soft-sediment benthos is a principal driver in structuring deep-sea benthic communities characterised by low population densities but high levels of diversity (Hardy et al., 2015). The CCZ is a vast, food-limited ecosystem, reliant on the arrival of particulate organic matter (POM) to the seafloor, dictated largely by productivity at the surface (Pennington et al., 2006). While nodule substrate availability plays a role in determining species distributions (Vanreusel et al., 2016), food availability – largely dependent on POM flux – will ultimately determine population densities. Having evolved under these relatively stable conditions, communities in the abyssal deep-sea may be poorly adapted, and thus highly susceptible to future mining impacts. If, for example, elevated substrata provide a nursery-type habitat for Ophiacantha cosmica, then the ramifications for reproductive success following future mining disturbance events are cause for concern. In addition, filter-feeding organisms are expected to be severely impacted by plumes and settling aggregates generated by the resuspension of sea-floor sediments during nodule collections, resulting in the long-term loss of suspension feeding organisms (Simon-Lledó et al., 2019). As with all deposit feeders, changes in surface-sediment chemical and physical characteristics following mining activity (e.g., remobilisation of formerly sequestered heavy metals and sediment compaction) could increase the toxicity of ingested sediments and impact Ophiosphalma glabrum’s ability to forage for food.

Food availability (i.e., POC flux) in soft-sediment abyssal habitats is subject to spatial variability both regionally, due to an undulating seafloor topography and at ocean-basin scales, as its dependency on surface productivity (Smith et al., 2008). Such heterogeneity in POC flux could create habitat networks of richer or poorer food supply over spatial scales at which reproductive kinetics and larval dispersal become ecologically relevant (Hardy et al., 2015). The CCZ lies along a POC flux gradient: lowest in regions that underlie central oligotrophic ocean gyres and highest beneath or adjacent to coastal and equatorial regions characterised by productive upwelling zones (Smith et al., 2008). Since population densities decrease concomitantly with POC flux (Tittensor et al., 2011), species in food impoverished areas often occur at densities below the minimum needed for reproductive success (e.g., critically low encounter rates for mating), rendering them wholly dependent on larval supply from populations in higher POC-flux habitats (the oligotrophic sink hypothesis, Hardy et al., 2015). The two species examined in the current study are relatively common in the CCZ and beyond (Booth et al., 2008; Amon et al., 2016; Christodoulou et al., 2019, 2020; Simon-Lledó et al., 2019). In fact, it is for this reason that the current reproductive study was even possible. However, the relative densities at which they occur at regional scales (such as between contract areas/APEIs) varies considerably (e.g., Christodoulou et al., 2020). Should mining activities proceed in regions of higher-POC flux that play a formative role as larval sources, this may have wide-reaching detrimental effects on the most food-limited populations of these and other species that employ planktonic larval dispersal.

Reproductive kinetics plays a fundamental role in mediating biogeographic patterns, population dynamics, metapopulation connectivity, and ultimately, species survival (Ramirez Llodra, 2002). While molecular approaches can be used to examine both historical and contemporary connectivity, the temporal resolution of these data remains too poor to discriminate between sporadic regional genetic exchange that then spreads locally over successive generations, and regular regional genetic exchange that acts to supplement local populations to a biologically meaningful degree. Reproductive ecology remains vital in bridging gaps in our understanding. The current study sought to add to the paucity of reproductive data for deep-sea species in nodule environments. It also highlights an uncomfortable truth in deep-sea ecology: these classical reproductive studies of this sort, which necessitate large numbers of individuals, may not be practicable in habitats such as the CCZ, where the most susceptible species are typically the rarest. Support for more ambitious approaches that compensate for sampling constraints, such as temporal studies or permanent monitoring networks are urgently needed in order for conservation measures to be effective and informed.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

AH conceived the reproductive study. AH, MC, and PM coordinated the sample collection, fixation, and preservation. MC and PM performed all aspects of molecular ID. SL performed the specimen imagery, dissections, sectioning, histological, graphical and statistical analyses, prepared figures, and wrote manuscript with scientific input from remaining authors. All authors contributed to the article and approved the submitted version.