Antarctica, the southernmost continent beyond 60° S, is one of Earth’s most remote and extreme environments. Its persistent sea ice, low winter light, and frigid temperatures make it challenging to study. This also provides opportunities to uncover fundamental biological and environmental processes (Vernet et al., 2019; Hutchinson et al., 2020). The Weddell Sea, located within the Southern Ocean, is significant for its biological richness and its contribution to global ocean circulation and climate (Hutchinson et al., 2020). It plays a critical role in forming water mass interactions that drive large-scale ocean currents, regulate global gas exchanges, and influence climate patterns (Foldvik et al., 2004; Vernet et al., 2019). These interactions make the area a hotspot for biological productivity, activity, and abundance (Nachtsheim et al., 2019).

The Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) has proposed the establishment of The Weddell Sea MPA (WSMPA; (Teschke, 2023). This area hosts one of the richest communities in the Antarctic region for fish (Baena et al., 2023), brittle stars (Lau et al., 2021), cephalopods (Staffer et al., 2021; Nesis et al., 1998), sponges (Brey et al., 1994; Brandt et al., 2007; Barthel and Gutt, 1992), marine birds (Reisinger et al., 2022; Teschke et al., 2021), and mammals (Reisinger et al., 2022; Teschke et al., 2020). Breeding grounds or key hunting and foraging areas for cryonotothenioids, marine mammals, and birds all occur within the proposed protection area (Hindell et al., 2020; Fretwell et al., 2012; van Franeker, 1996; Purser et al., 2022).

The suborder Notothenioidei represents a group of fish uniquely adapted to Antarctic environments (Near et al., 2015; La Mesa et al., 2021). They exhibit slow maturation, low fecundity, and large egg production, with most species spawning demersally and showing parental care during incubation, which can exceed 100 days (Everson, 1984; Gon and Heemstra, 1990; La Mesa et al., 2021; Marshall, 1953; Novillo et al., 2022).

The two species in Lindbergichthys have a benthic lifestyle as adults and reach a maximum length of 15cm and 19.5cm for L. mizops and L. nudifrons, respectively (Froese and Pauly, 2022). The latter occupies a greater depth range between 3–400 m compared to 20–220 m for the former. Parental care behaviour has been found in Antarctic icefish, with nesting and egg-guarding being the most common forms (Ferrando et al., 2014; Novillo et al., 2022; Kock et al., 2006). Species in the genus Lindbergichthys often exhibit parental care nesting behaviour, which has been well-studied (Eastman, 2013; La Mesa et al., 2021; Konecki and Targett, 1989). For L. nudifrons, sexual maturity is reached at age 4 to 5 years and length of 8 to 9cm (Hourigan and Radtke, 1989; La Mesa et al., 2017). Females spawn in a nest in late Austral autumn to winter (May to June), usually protected by crevices or rocks (Hourigan and Radtke, 1989). The male guards the nest and eggs, including chasing away egg predators. Post-hatching, larvae migrate to the pelagic zone before returning to the benthos in April (Kellermann, 1989).

Nests were generally circular and shallow, parabolic in shape, with the sediment built up on all sides of the nests. Only L. nudifrons were seen in the nests, suggesting these nests were created by this species, which aligns with previous research (Gon and Heemstra, 1990). The average nest size recorded here (12.3cm) was larger than L. nudifrons published size at maturity (9.1 to 9.5cm). The average fish size (10.3cm) was also determined to be greater than their maturation size. Other fishes identified as species other than L. nudifrons were larger (greater than the published maximum length of L. nudifrons) and are known to have larger sizes at maturity (Hourigan and Radtke, 1989). No eggs were seen within any of the nests, presumably as the WSE occurred after known hatching times (Austral Spring; (Hourigan and Radtke, 1989). Some larvae were seen within the nest circumference.

This study documents a large and widely dispersed nesting aggregation of the yellowfin notothenioid, Lindbergichthys nudifrons, in the western Weddell Sea, an area made accessible by the recent calving of the Larsen C Ice Shelf. The most significant finding is the discovery of six distinct, geometric nesting patterns, which we propose are primarily driven by biotic interactions, namely predation pressure, rather than the abiotic factors observed in other large Antarctic fish breeding colonies. These findings, summarised in Table 1, provide new insights into the complex behavioural ecology of Antarctic notothenioids and underscore the ecological significance of this recently uncovered region.

The sampling locations were chosen to represent the time since ice coverage, allowing for a comparison between the duration of ice-free conditions and nest characteristics; however, no relationships were found. A more influential factor was the local substratum. Different substrata were found between sites: visually, Site B was rockier, with a high abundance of invertebrate fauna, such as brittle stars, feather stars, and corals. In contrast, the other locations (A, C, D, and E) sediment tended to be finer and hosted more nests and teleosts. This suggests a species preference for softer sediment for nest construction.

This study provides the first documentation of complex, variable nesting patterns for L. nudifrons. The nests themselves were parabolic in structure. The depth of nest depression varied, though this was not possible to measure with precision from the ROV footage. The presence of larger rocks that remained within the nests, alongside smaller pebbles that appeared to have been moved to the nest wall, could be indicative of the amount of energy expended on nest creation. As no fish were observed actively building nests, this remains an area for future investigation. A consistent feature was that in grouped nests, the excavated sediment was typically built up on the sides away from the centre of the overall pattern, suggesting a coordinated construction effort that maintains separation and structure. In a promising sign of reproductive success, fish larvae, identified as L. nudifrons, could be seen floating within the circumference of some nests. The presence of larger rocks beside some nests, particularly evident in the ‘Sharp U’ pattern, is a behaviour that has been noted before in notothenioids (Hourigan and Radtke, 1989). We observed that smaller nests were adjacent to rocks more often than larger nests. These rocks may provide vital shelter from benthic currents or act as enhanced refugia from predation. On a seafloor that is visually rather flat and homogenous, any additional physical complexity, such as a rock, would confer a significant advantage in predation protection, corroborating earlier hypotheses by Hourigan and Radtke (1989).

We identified and categorised six different nesting patterns (Table 2). Of these, ‘Sharp U’, ‘Oval’, and ‘Crescent’ are the most similar in their geometric structure. The observable differences could be attributed to the ‘Sharp U’ and ‘Crescent’ patterns being incomplete or interrupted versions of a full oval. The most common pattern, ‘Cluster’, which accounted for over 42% of nests, lacked a specific geometric shape and was instead a dense accumulation of many nests in a small area. ‘Singular’ nests, the second most abundant pattern, were isolated and had the highest mean nest diameter. This observation could suggest that larger, older and more dominant adults are more comfortable nesting away from the group, being more capable of protecting their nests individually. The ‘Line’ pattern was very distinct and the least abundant. We hypothesise that these patterns are an anti-predator adaptation. The rarity of the ‘Line’ pattern, for example, could be associated with a reduction in its effectiveness for community anti-predator adaptation, as no single nest achieves a centrally protected position. In contrast, the ‘Cluster’ pattern strongly aligns with the selfish herd theory proposed by Hamilton (1971), where individuals reduce their domain of danger by putting other individuals between themselves and an approaching predator. This would offer a significantly reduced risk to the nests in the centre of the cluster. Such defensive patterning, while novel for this species, is a well-documented phenomenon in fish nesting colonies in shallow tropical reef scenarios, where it is also attributed to increased predation protection (Gross and MacMillan, 1981; Tyler Iii, 1995). Previous research on L. nudifrons has shown that guarding males will defend a territory up to 25cm away from their nest (Hourigan and Radtke, 1989). In the dense patterns recorded here, this defence zone would frequently encroach on their neighbour’s nests, suggesting a high level of tolerance and a communal defensive posture.

This defensive behaviour is likely a direct response to the local predators. The epifauna observed near the nests provides several candidates. It is possible that scavenging brittle stars (Ophiuroidea) positioned on the nest edges were predating on cryonotothenioid eggs, a behaviour previously noted from stomach content analysis (Volage et al., 2021; Fratt and Dearborn, 1984). The strategic location of these brittle stars also suggests they may be taking advantage of food sources carried by water currents that are channelled over the nests. A more significant threat, however, may be the predatory nemertean, Parborlasia corrugatus. This species is a voracious scavenger and predator and was observed in the general area (Gibson, 1983). We propose that the complex nesting patterns may serve to reduce predation by P. corrugatus. As this predator utilises chemotactic mechanisms to actively search for food, the aggregated nests may disguise or dilute the individual odour plumes from the eggs. This could create a confusing sensory environment for the nemertean, making it more difficult to detect and target a single nest, a crucial adaptive strategy during the long incubation period. Other fauna were present but appeared to have neutral interactions; the absence of physical contact between cnidarians like Umbellula sp. and Leptogorgia sp. and the nests could imply a neutral relationship, while the resting behaviour of holothurians near the nests may indicate a low-energy interaction with the environment.

Two key unresolved questions remain regarding the creators of these nests: the social structure and the certainty of the species identification. It is not known whether one mating pair created all the nests in a group or whether each nest was attributed to a different, single mating pair; however, it is presumed that each nest was for a single pair, given the high energetic costs and significant predation risk involved in guarding even one nest. Furthermore, a necessary limitation of this study is the certainty of species attribution for every observed nest. Our attribution to L. nudifrons is based on strong visual evidence; across 27 hours of video, it was the only species directly observed occupying or actively maintaining the nests. However, in the absence of a guarding fish or eggs at every nest, we cannot definitively exclude the possibility that a minority were created by other cryonotothenioid species. This is a particularly relevant caveat, as a recent study have shown species determination in this group is problematic (Schiavon et al., 2023), highlighting a challenge common to deep-sea visual surveys.

A striking finding of this study is the lack of a strong, localised abiotic driver for nest location, particularly when compared to other known Antarctic fish colonies. Except for Site A, our data show no significant difference between the water temperature at nesting locations and the surrounding, nest-free waters. Site B, the warmest site, exhibited slightly smaller nest diameters. Although the temperature difference across all sites is marginal (approximately 0.2 °C), it is worth considering its biological relevance. Studies on ectothermic Antarctic fish suggest that even such minor temperature variations can influence metabolic rates (Clarke, 1983; Sandersfeld et al., 2017; Enzor and Place, 2014; Johnston et al., 1991). However, the considerable overlap in temperature ranges between the sites, coupled with the lack of a consistent trend, suggests this small difference is not sufficient to induce significant biological effects such as changes in nest size or recruitment. While it is uncertain whether this temperature change affects developmental time, it requires further investigation.

This stands in stark contrast to the large Neopagetopsis ionah breeding colony reported by Purser et al. (2022) (Table 3). That site was defined by a dominant inflow of modified warm deep water (mWDW) that was up to 2 °C warmer than the surrounding bottom water, a feature hypothesised to be the key factor driving nest site selection. While it is tempting to draw parallels, the short-term nature of both temperature data, alongside the fact that the temperature change in the N. ionah colony was an order of magnitude greater, limits our ability to make a direct comparison and underscores a fundamental difference between the two sites. Since the waters of the high Antarctic shelf are known for their thermal stability, even small temperature variations may be ecologically significant, but without long-term monitoring, it is difficult to assert whether these differences represent a consistent environmental factor.

The broader oceanography of our study region is certainly a contributing factor to its overall suitability. The mWDW has been shown to reach this location, mixing with colder, denser Ice Shelf Water (ISW) and High Salinity Shelf Water (HSSW). This mixing of local shelf water masses and the mWDW is known to drive a greater level of local productivity, likely creating a favourable environment for a greater abundance of nesting sites across this area. However, our data suggest this is a regional, rather than a localised, driver. These oceanographic processes and how they influence the selection of broad nesting areas, versus the fine-scale patterns within them, are an advised focus of future research.

The physical characteristics of the nests also differed logically with species size, as did the water depths, which conformed with each species’ published depth ranges and habitat preferences. Spacing between L. nudifrons nesting sites was more stochastic, and neighbouring nests were often touching, whereas at the N. ionah site, nests were spaced ~25 cm away from each other. Furthermore, the timing of the two expeditions revealed different stages of the reproductive cycle, suggesting the wider Weddell Sea is a critical site of species-dependent nesting for over half the year.

Correctly identifying and defining these unique biological features is crucial for their conservation. A recent proposal by Teschke (2023) suggested the terms ‘nest’ and ‘potential nest’, which are established by the presence of eggs for the former, and the absence of eggs but with one or more fish present for the latter. Due to our expedition’s timing post-hatching, it was not possible to confirm the presence of eggs. However, we were able to confidently identify occupied, active nesting sites because phytodetritus was consistently ‘cleaned’ from the nests by adult fish as a form of maintenance behaviour, whereas the surrounding, unmaintained seafloor was carpeted in this detritus. This method provides a reliable proxy for identifying active nesting sites outside of the spawning season.

This discovery has immediate and significant conservation importance. Within the 43rd meeting of the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR), the Antarctic and Southern Ocean Coalition agreed that clear video evidence of fish nesting sites is required to create a Conservation Measure. Our results provide precisely this evidence for one of the few documented fish nesting sites in the entire Weddell Sea. Spawning and nesting sites play a crucial role in the wider spatial conservation landscape. They are key features used to identify Vulnerable Marine Ecosystems (VMEs) under the FAO Guidelines for Management of Deep-sea Bottom Fisheries in the High Seas and are also identified as features of Ecologically and Biologically Significant Areas (EBSAs) under the Convention on Biological Diversity (CBD). Our findings match the criteria and designations for these respective conservation measures. This work, therefore, underscores the critical importance of protecting these unique habitats and provides robust evidence for the designation of the Weddell Sea Marine Protected Area (MPA).

In conclusion, this study presents the discovery of a large, active, and well-dispersed cryonotothenioid nesting habitat in the Western Weddell Sea. The differing nesting patterns are the first described for this species and are strongly speculated to be a group behaviour for predation evasion. Crucially, the nesting sites were not shown to have any site-specificity driven by the tested abiotic variables, such as localised temperature. This suggests that biotic interactions are the primary drivers of this complex aggregation, a significant contrast to other known notothenioid colonies. This research also provides valuable ground-truthing for the cryonotothenioid nesting site suitability model from Teschke et al. (2016) and furthers the discussion into nesting site definitions as proposed by Teschke (2023). The extensive, active nesting sites documented here provide further compelling evidence to support the designation of the proposed Weddell Sea Marine Protected Area.