The bryozoan biogenic reef (BBR) area is situated in the Eastern Arm of Western Port (WP), VIC, Australia, in an area termed the “Rhyll Segment” (Harris et al., 1979; Jenkins et al., 2013) between French Island, Corinella and Rhyll (Figure 1). Potential reef areas were initially identified based on the seabed textures in a 2009 bay-wide multibeam bathymetry product (2.5 m horizontal resolution) generated by the Port of Hastings Development Authority. Image-based groundtruthing was achieved by inspection of linear seabed texture features detected in the multibeam bathymetry using a BlueROV2 Remote Operated Vehicle (ROV) recording oblique high-definition video (Figure 2). Approximately 1.5 h of ROV footage was acquired in the northern cell. In addition, a drop-camera platform was used to collect imagery while station-keeping or slow drifting over seafloor features of interest and surrounding habitats. The drop camera platform comprised a Splashcam^® high-definition video camera recording in an oblique orientation and a GoPro Hero4 with separate lighting collecting still images in planar orientation (see Figure 2). Further, in situ observations supported by underwater photography were made over 10 SCUBA dives in the BBR area (see Figure 2) during various sample collection events between December 2017 and February 2020. Collectively, the results of these groundtruthing efforts were used to injunction with the low-resolution bay-wide multibeam bathymetry to define the survey extents for high-resolution multibeam echosounding (MBES). BBRs were found to be comprised of three species on species and representative samples were collected for taxonomic confirmation by a bryozoan taxonomy expert at the National Museum of Victoria (Australia). Still images and video footage were also shared with bryozoan taxonomy experts at CSIRO (Australia) and National Institute of Water and Atmospheric Research (New Zealand) to confirm species identifications. For this study, the imagery was used to make qualitative assessments of the presence of BBR, its composition and characteristics and there was no quantitative analysis. In a partner study, epifauna species co-inhabiting the reefs were scored from the imagery and these data will be reported in future publication.

The vessel-based MBES surveys were completed in the northern, central and southern cells, for a total area of 6.4 km² at a resolution of 0.5 × 0.5 m (see Figure 2). MBES surveys were completed over two periods: 9–10 October 2019 and 11–14 February 2020. Transect direction was dictated by the shape and position of the survey area extents but was principally in east-west direction. The MBES was completed using an R2Sonic^® Dual Head 2020 multibeam system (340, 400 kHz frequency). Positioning was achieved with an Applanix^® POS MV Wavemaster system with the real-time kinematic (RTK) satellite position validated against the Victorian Permanent Survey Mark “Bittern PM 259”. The MBES were corrected to Lowest Astronomical Tide (LAT). Validation of the vertical data was achieved by deploying a cage to the seafloor at the survey site which had a RTK Global Navigation Satellite System (GNSS) receiver target installed at a known distance above the seabed. The MBES system captured data over the subsea target, resulting in a Total Horizontal Uncertainty (THU) of ± 0.1 m and Total Vertical Uncertainty (TVU) of ± 0.05 m. Data processing used the Hypack^® 2020 system and digital elevation models for the survey areas were produced at a horizontal resolution of 0.5 m × 0.5 m.

Bryozoan reefs were visible in the digital elevation models (DEMs) in each of the three cells. However, fully automated terrain analysis and habitat modelling routines failed to reliably distinguish BBR due to interference patterns introduced by seams in the DEM and textures associated with Zostera nigricaulis and seasonal Caulerpa cactoides beds. Therefore, a supervised approach that relied on the circularity of reef structures was applied as follows:

Detections were checked by overlapping resultant polygons on a vertically exaggerated DEM and confirming against actual ground-truthing locations.

The LBRs morphology was observed only in the northern cell (see 3D reconstruction, Figure 3) and consisted of continuous reef mounds that are linearly aligned in approximately north-south orientated rows (see Figure 7). The reef rows ranged from ∼ 3 m to some 30 m in width and the vertical relief of LBR rows ranges from ∼ 50–120 cm above the surrounding sediment. Water depths in the LBR zones ranges from approximately 4.8–6 m. The sediment shoal separating the linear reef zones has a depth of approximately 4 m.

In the northern cell there were three discrete areas of LBR, interrupted by a sediment shoal (see Figure 2). This sediment shoal was present in the 2009 bathymetry in addition to this new bathymetry and so does not appear to be a recent feature or to have changed significantly since 2009. The northern edge of the contiguous reef rows terminates with a distinctive gully feature and a steep slope leading to a raised bank (see Figure 3).

In the northern cell, linear reef bands were separated by tracts of very fine silty muds with depauperate infauna. We hypothesise that the baffled colony structure and morphology of the reefs running perpendicular to the prevailing current flow has the effect of slowing currents causing the settlement and trapping of fine particles that are otherwise transported in tidal flows. The multibeam bathymetry showed signs of sediment processes, with potential sediment accretion on the eastern sides of reef rows. Due to the high density of habitat-forming bryozoans in the linear reefs, these are speculated to harbour the highest matrix-associated biodiversity of co-occurring fauna.

The PBR were observed in the northern, central and southern cells. Patch reefs occurred in 4.5–8.5 m water depth (Figure 4B). In some areas, the PBR were locally dense but these patches were of a different form to the LBRs. Ground-truthing of PBR has not been as thorough as that for the strikingly unusual LBR and field visual observations have been hampered by poor visibility. However, field observations suggest that individual reef patches represent sites of enhanced localised biodiversity of matrix-associated fauna compared to the surrounding sediment beds.

The discovery of PBR in the southern cell significantly expanded the initial expected distribution of bryozoan reefs. Furthermore, discovery of PBR in the far north and east portions of the northern cell represent new findings. Some PBR in the northern cell are associated with apparent remnant, or newly forming, LBR features, which requires future investigation.

The Western Port BBRs occur in seafloor depressions at the terminus of sub-channels. Harris et al. (1979) noted that due to the proximity of the Rhyll Segment to the Eastern Entrance of to bay, currents in this area do not display the directional symmetry that is typical of other sites but rather are generally counter-clockwise circulation pattern. During our surveys, underwater and surface observations were that tidal currents in the BBR areas are lower than those in the neighbouring main channels and there is significant eddying. This reduced current flow is likely to be required for the establishment and colonies with fine fenestrate lamellae that may be prohibited by the stronger currents in the main axis of nearby channels.

Linear bryozoan “swards” in Foveaux Strait, New Zealand are orientated parallel to the peak tidal current direction (Cranfield et al., 2003). The approximately north-south alignment of the LBR may be associated with the pattern of northerly movement of flood tide currents from eastern entrance, along the eastern coast toward Corinella reported by Harris et al. (1979). The Western Port LBRs may further baffle current flow and eddying owing to their perpendicular orientation to the current flow in the area, potentially creating conditions that are conducive to additional density-dependent growth and recruitment (Cranfield et al., 2003; James et al., 2008).

Subtle depth controls on bryozoan distribution were observed and these are exemplified in the central cell. Bryozoan reefs in the central cell were generally constrained to the depth range ∼4.5–8.5 m. Bryozoan reefs generally did not encroach onto the banks to the east and south of the central cell shallower than 4.5 m where Zostera nigricaulis seagrass and Caulerpa cactoides algae beds occur. Few bryozoans were detected deeper than 8.5 m in the central cell. Interestingly, this ecophysical model of bryozoan reefs bordering the offshore extent seagrass beds is reflected in the historical accounts from the oyster dredging fishery, in which fishers reported highest oyster catches at the offshore extent of seagrass (Hannan and Bennett, 2010).

On the Lacepede Shelf, South Australia, Hageman et al. (1995) identified factors of nutrients, temperature, water chemistry, environmental stability, substrate, competition, and secondary factors of water depth and currents as the potential abiotic drivers of bryozoan distributions. Additionally, on the Tasmanian continental shelf, Amini et al. (2004) noted bryozoan morphotype associations with water depth, energy regimes, sedimentation rate and substrate type, with fenestrate and foliose forms occupying deeper water where energy and sedimentation rates were low and where hard substrate is more abundant. The presence of shallow, dense growth of fenestrate bryozoan forms in the WP reefs is therefore unique and may be related to the relatively sheltered conditions, low wave action and current energy at this location. Across the northern, central, and southern cells in our study area, nutrient supply, water chemistry, temperature and water circulation patterns are presumed to be constant. Therefore, the depth-related patterns observed may be related to competition with seagrass and Caulerpa cactoides at the shallow end of their distribution. Alternatively, historic differential parent substrate types at the time of reef formation may explain today’s distribution. Limitations to distribution at the deep end of the distribution may be related to increased current exposure and substrate differences. Furthermore, given the generally turbid nature of this embayment, UV penetration would be typically reduced and hence closely mimic deeper habitats in which bryozoan reefs are more commonly found. Furthermore, the turbid conditions in the BBR area may be limiting competition with algae.

There appears to be an association between the mud oysters (Ostrea angasi) and BBRs. Field surveys identified mud oyster clumps intermingled with bryozoan colonies, particularly Celleporaria foliata. This association is supported by the fact that the BBR area was an important mud oyster fishery between 1850–1900 (Hannan and Bennett, 2010). Our field surveys which involved diver hand-cores, also identified the presence of a dense subsurface bivalve shell layer consisting of O. angasi and Anadara trapezia shell. It is speculated that these observations are indicative of previously bivalve beds that may have served as a settlement substrate for bryozoans and that is markedly different to the surface sediments today. Biogenic bivalve reefs have been severely impacted in Victoria (Ford and Hamer, 2016) and are now the subject of significant restoration investment. The apparent BBR-mud oyster association further adds to the uniqueness and ecological significance of the Western Port BBRs.

The Victorian Government is actively seeking to increase reactional fishing activity (Target 1 Million initiative), which may increase the pressure associated with physical damage from anchors and fishing equipment (potentially visible in Figure 6D). GPS coordinates for the site, known to fishers as “The Corals”, are widely available. Recreational fishing activity should be monitored in terms of spatial intensity (has commenced as part of the Western Port Bryozoan Reefs Project and as per recommendation #4 in Ford and Gilmour, 2013) and linked to a program to monitor colony damage. Responsible anchoring practices and the use of new technological advancements in trawling motors, for example, could be investigated to minimise physical impacts. Precedence for this exists for the Port Phillip Bay shellfish reef restoration project (Gillies et al., 2018), where fishers were educated about techniques to use current and wind conditions to maximise fishing on-reef structures while placing anchors in appropriate off-reef positions.

In WP, bryozoan reefs are potentially threatened by sedimentation and scouring. Sediment transport models have shown that the BBR site is exposed to southerly transport of sediment particles via ebb tidal water circulation from the north-eastern sector of the bay (north of French Island), through the tidal divide and into the Rhyll segment of East Arm. Once deposited, these fine sediment particles can become resuspended by wind-driven wave energy. No data are available on sedimentation specific to the BBRs, however, previous studies identified the Rhyll segment of East Arm as an area of very fine sediments (Marsden et al., 1979) and depositional processes (Hancock et al., 2001). Our field studies identified the presence of thick deposits of fine silt between the rows of LBRs that is differentiated from the muddy sands further away from the LBRs, suggesting localised sorting and trapping of fine particles. Initiatives currently underway to decrease sediment inputs include rehabilitation of mangrove and saltmarsh habitat and replanting of intertidal seagrasses and riparian vegetation (Department of Environment, Land, Water and Planning [DELWP], 2017; Melbourne Water, 2018). While these initiatives are predominantly focussed on protecting seagrass meadows, they will also benefit BBRs.

To date, very limited marine pest incursions or outbreaks have been documented in the WP region. The algae wakame, Undaria pinnatifida, now prevalent throughout neighbouring Port Phillip Bay, presents a potential risk to BBRs as they present a large area of hard substrate that may be colonised by the algae. The Northern Pacific seastar, Asterias amurensis, also prevalent throughout Port Phillip Bay, presents a risk to the abundant bivalve community associated with the BBR and a potentially large risk to bryozoans directly. Other Asteroidea, such as Patiriella regularis and Coscinasterias calamaria have are known to consume bryozoan zooids (Gordon, 1972), thus is it reasonable to speculate that Northern Pacific seastar could do likewise.

Increasing industrialisation and growing needs for larger and more frequent commercial and passenger shipping operations within WP, in addition to increasing recreational vessel activity, increases the risk of potential marine pest introductions. Monitoring efforts should expand in step with this increasing risk, both at port locations and at highly sensitive locations such as the BBR complex.

Fortunately, the BBR of WP have, until now, have largely escaped anthropogenic destruction and so conservation efforts can focus on preservation rather than later restoration. In stark contrast, there are several examples in New Zealand whereby, despite many years of protection, various bryozoan populations have failed to sufficiently recover from past destructive fishing practises (Cranfield et al., 2004; Rowden et al., 2004; Wood et al., 2012, 2013; Wood and Probert, 2013; Mello et al., 2021).

The Western Port Bryozoan Reef Project is engaged in concurrent studies of; reef macrofauna biodiversity, bryozoan colony age and growth, bryozoan reef fish biomass and citizen science acoustic monitoring methods. These studies are occurring at multiple spatial scales from the sub-meter scale to investigate biological interactions, to area-based management scales. The Victorian Government is embarking on a process of identifying indicators of Good Environmental Status (GES), adopting the general framework of the European Union’s Marine Strategy Framework Directive (European Commission [EC], 2008). Good Environmental Status is a quantitative or qualitative description of the condition/character of the environment that managers wish to maintain or drive toward improvement (see Table 2). The GES is linked to ecosystem models and indicators for monitoring and establishing monitoring objectives. GES indicators that could be applied to BBR are listed in Table 3. Non-destructive techniques are required and data to inform these indicators can be derived from acoustics, imagery, physicochemical studies. Poor underwater visibility at the BBR location put a focus on acoustic methods of monitoring which has strongly informed related studies of the project.

These unique BBR complexes fall outside of any marine protected areas and are not afforded any legislated protection. The Western Port Bryozoan Reefs Project has nominated the bryozoan reefs for listing under the Victorian Flora and Fauna Guarantee Act 1998 as a community that is prone to a variety of future potential threats that may lead to extinction. Other formal legislative conservation options available include listing the community on the Environment Protection and Biodiversity Conservation Act 1999 (EPBC Act) List of Threatened Ecological Communities.