Significant northern range extension of the non-native batillariid Zeacumantus subcarinatus in Australia, with observations of new habitat usage

The New Zealand batillariid gastropod Zeacumantus subcarinatus has been recorded as an invasive species in Australia since at least 1924, with populations established in rocky shore habitats at multiple locations in Greater Sydney. We observed a large population (10,000+ individuals) of Z. subcarinatus at an estuarine tidal mudflat on the New South Wales mid-north coast, representing a ~ 250 km northern range extension, and the first known record of this species in a sheltered, muddy habitat in Australia. We discuss the possible establishment means of this new population.


Introduction
Hundreds of non-native and cryptogenic marine species have been detected in Australian waters, with records distributed around almost the entire country's coastline in almost every possible habitat (Thresher 1999;Hayes et al. 2005;Sliwa et al. 2009). Although some of these species are well-established and pose significant threats to native species and ecosystems, such as the Northern Pacific seastar (Asterias amurensis Lutken, 1871), European shore crab (Carcinus maenas (Linnaeus, 1758)) and Japanese kelp (Undaria pinnatifida (Harv.) Suringar) (MPSC 2018), many have few known negative impacts (DAWR 2015), although in some cases this may be due to a poor understanding of their ecology and distribution in Australian waters rather than a true absence of impacts (Glasby and Creese 2007).
One such species is Zeacumantus subcarinatus (G.B.Sowerby II, 1855). Native to New Zealand, Z. subcarinatus is a small species (~ 15 mm maximum height) in the family Batillariidae. It is widely distributed within its native range, with records around almost the entire New Zealand coastline, including the Chatham Islands, and can often be found in incredible numbers, with some studies reporting densities of over 18,000 individuals per square metre (McClatchie 1979). It mostly occurs in relatively protected habitats, including intertidal mudflats, rock pools and bays (McClatchie 1979;Keeney et al. 2013).
Although the earliest known record of Z. subcarinatus in Australia is 1924, with specimens held in the Australian Museum collected from Gordon Bay in the eastern suburbs of Sydney (AMS C.50204), it was likely introduced earlier than this. Finlay (1926) noted that "[Zeacumantus subcarinatus] has gained an Australian foothold; it is extremely common and spreading rapidly at Freshwater, near Manly, Sydney", and Iredale (1936)  In the ~ 100 years since, Z. subcarinatus has remained well-established in the Sydney region, and indeed is now present in all four of Sydney's major estuarine systems (Andrews et al. 2010). Across 32 surveyed locations, Andrew et al. (2010) found live individuals at nine locations, almost all of which were sheltered, rocky sites, especially high tidal pools. At one of these sites (Kurnell), the population was estimated to exceed 10,000. Interestingly, surveys failed to detect Z. subcarinatus at eight locations at which it was previously recorded, indicating possible local extinctions, although TM observed empty shells at one of these sites (Long Reef) in 2020 and 2022. Aside from a 2013 collection from Lake Illawarra (AMS C.525073.001), and two 2022 records of empty shells from Jervis Bay and Ulladulla uploaded to the online citizen science platform iNaturalist (https:// www. inatu ralist. org/ obser vatio ns/ 11951 1138, https:// www. inatu ralist. org/ obser vatio ns/ 12012 2126), Z. subcarinatus has previously never been detected outside the Greater Sydney region, with the northernmost record a 1973 collection from the Bouddi Peninsula (AMS C.316563).

Study site
This study was conducted along the Camden Haven Inlet on the New South Wales mid-north coast, ~ 300 km north of Sydney (Fig. 1a). This inlet is the easternmost section of the Camden Haven River and feeds into the Pacific Ocean. It is highly trained, with the mouth permanently open. There are three intertidal mudflats along the northern bank of the inlet where it flows past the town of North Haven (Fig. 1b); these flats are inundated during high tide, with water from the inlet flowing beneath the northern training wall, although the water level in each flat on any given day is highly variable depending on the tide and environmental conditions such as cloud cover and rainfall (Fig. 1c, 1d). The three flats are characterised by a rich community of estuarine crustacean and mollusc species (Supplementary Table 1).
The first mudflat (herein, mudflat 1) covers ~ 8500 m 2 and includes large patches of estuarine plants such as creeping brookweed (Samolus repens (J.R.Forst. & G.Forst.) Pers.), sand couch (Sporobolus virginicus (L.) Kunth), and seablite (Suaeda australis (R.Br.) Moq.), and small patches of grey mangrove (Avicennia marina (Forssk.) Vierh.). During high tide, water depth can exceed 1 m. There are also extensive patches of green and brown algae across the centre of mudflat 1. Mudflat 2 covers ~ 2800 m 2 and includes significantly less vegetation cover, especially centrally, with large patches of Sporobolus virginicus mostly restricted to the edges of the flat. Water depth typically reaches just 20-30 cm even during large tides; this flat is the only one of the three to lose all water during some low tides. Mudflat 3 covers ~ 3400 m 2 and is almost entirely covered by a dense layer of Sporobolus virginicus, with a large strip of Avicennia marina running perpendicular through the centre of the flat. Water depth generally reaches a maximum of ~ 50 cm during high tide. The substrates in mudflats 2 and 3 are much sandier compared to mudflat 1.

Methods and results
On 22 December 2022, TM opportunistically observed an empty Zeacumantus subcarinatus shell on the eastern edge of mudflat 1 at full low tide (~ 0.1 m). Thirty minutes of searching at mudflat 1 on each of 23 and 24 December (~ 30 min before full low tide, ~ 0.1 m), on bare mud and under and around stones and deadwood, revealed a further twenty empty shells in total, but no live individuals were observed. Given Andrews et al. (2010) only observed Z. subcarinatus on rock platforms and in rockpools in Sydney, thirty minutes of searching was also conducted on rock platforms at the nearby Wash House Beach, ~ 1.3 km southeast of the study site (mid-tide, during a rising tide, ~ 1 m), but no specimens were observed (either live individuals or empty shells).
On 30 December 2022, a search of mudflat 2 at 2:40 PM during a full high tide (~ 1 m) immediately revealed thousands of live Z. subcarinatus (Fig. 2). Based on a rough visual estimate, over 10,000 individuals were present across an inundated area of ~ 250 m 2 , however, it was difficult to accurately assess the population size due to uneven spatial distribution; many individuals were strongly clustered around the bases of large rocks, whilst others were more disparately spread across the bare mud.
A number of individuals appeared to be deposit feeding.
Thirty minutes of searching on 31 December 2022 at mudflat 3 at full low tide (~ 0.4 m) did not reveal any live specimens or empty shells.

Discussion
Our observations represent a ~ 250 km northern range extension of this species in Australia, and the first record of live individuals outside the Greater Sydney region. Perhaps the most pressing questions raised by our findings are (1) is Z. subcarinatus present at other sites between Sydney and North Haven, and (2) if yes, how widely distributed and well-established is it along the coastline between these two locations? Given Z. subcarinatus lacks a pelagic larval stage, with eggs hatching into crawl-away larvae (Fredensborg et al. 2005;Keeney et al. 2013), there seem to be three scenarios regarding the North Haven population: 1. It represents a disjunct population established via a second, independent introduction event from New Zealand. 2. It represents a disjunct population established from the Sydney population. 3. It represents the most northern (known) population of a distribution stretching from Sydney to North Haven, with populations between these two regions yet to be detected.
It is unclear exactly how Z. subcarinatus was first introduced to Sydney (Hutchings et al. 1987), but the proposed mechanism is via shipping (Powell 1979), with individuals transported in ballast or attached to ships' hulls (Fleming et al. 2018). Whilst scenario 1 is therefore possible, with a New Zealand vessel accidentally transporting individuals to the North Haven region independent of Sydney, we believe this is unlikely due to the lack of trans-Tasman vessels porting in Camden Haven Inlet, in contrast to the high ship traffic of Sydney Harbour (Mayer-Pinto et al. 2015). It is possible that a disjunct population was introduced via recreational watercraft from Sydney, however, this is difficult to confirm. Andrew et al. (2010) observed Z. subcarinatus in Sydney on macroalgae such as Neptune's necklace (Hormosira banksii (Turner) Decaisne), and posited that dispersal may occur via rafting (either as adults or eggs) on floating algae or other flotsam. Further, rafting-mediated range expansions of marine species have been documented or posited across a range of gastropod taxa (Ávila et al. 2012;Baptista et al. 2021), including to and from Australia (Donald et al. 2005(Donald et al. , 2015Gordillo and Nielsen 2013), and for species without a pelagic/planktonic larval phase (Winston 2012), as is the case for Z. subcarinatus. We therefore speculate that the North Haven population of Z. subcarinatus established via rafting from the Sydney population.
However, the question remains of whether individuals rafted directly from Sydney, resulting in a single, disjunct northern population (scenario 2), or whether they 'hopped' along the New South Wales coast, establishing a series of currently undetected populations between Sydney and North Haven (scenario 3). Although direct transport from Sydney to North Haven is possible, with modelling showing that particles originating from Sydney can settle at the Camden Haven estuary (Hewitt et al. 2022), we speculate that scenario 3 is more likely given the East Australian Current's predominant southwards flow, and thus that Z. subcarinatus reached North Haven by hopping along the coast and establishing geographically intermediate populations between Sydney and North Haven. Systematic surveys along the New South Wales central coast will be important for detecting these putative populations. Importantly, our observations are also the only current known record of Z. subcarinatus in Australia in a sheltered, muddy habitat, with Andrew et al. (2010) noting that "it is often found in sheltered muddy habitats in New Zealand (W. Ponder, pers. comm.) and it is possible that it could move into this habitat in NSW, although at present there is no evidence for this." Surveys should therefore target both rocky shores and mudflats, as the species is clearly able to establish large populations in Australia in both habitat types.
More surveys around North Haven will also be useful to determine whether mudflat 2 represents the only population in the area; there are other sheltered muddy habitats on the opposite bank of the Camden Haven Inlet, including the large Gogley's Lagoon. Although empty shells were found in mudflat 1, it is unclear whether a population also exists here. Major flooding occurred along the Camden Haven River in March 2021, with the water level reaching 1.70 m at North Haven (Manly Hydraulics Laboratory 2021). During this time, all three mudflats were entirely submerged, and thus the empty shells could potentially have been transported into mudflat 1 from mudflat 2. However, if this did occur, we speculate that at least some empty shells (given the large population size in mudflat 2) would have also washed into mudflat 3. Given TM failed to detect any empty shells in mudflat 3, a live population may be present in mudflat 1. Future surveys for live Z. subcarinatus should target the western third of mudflat 1, as the habitat here is most similar to mudflat 2, i.e., the mud is relatively sandy, there are large expanses of bare substrate punctuated by large rocks, and the water level is highly variable depending on tidal and weather conditions. Conversely, the centre of mudflat 1 is almost permanently inundated (and thus the mud is very soft and boggy) and supports extensive algae patches; Sydney populations of Z. subcarinatus are significantly more abundant on bare rock microhabitats than in rock pools covered with algae (Andrew et al. 2010). Given mudflat 3 is almost entirely covered by dense grass (Sporobolus virginicus) and mangroves (Avicennia marina), with almost zero bare substrate, the apparent absence of individuals from this flat is possibly genuine.
Although Andrew et al. (2010) concluded that Z. subcarinatus currently poses little to no threat to other rocky shore species in Sydney, they did find differences in the abundance of some molluscs in pools with and without Z. subcarinatus. Based on opportunistic observations by TM, the mollusc assemblage in mudflat 2 seems to be less diverse and less abundant compared to mudflat 1 (Supplementary Table 2). However, these differences could possibly be driven by environmental factors; some of the molluscs thus far recorded only from mudflat 1, such as Conassiminea studderti Fukuda and Ponder (2006), are associated with wood and/or mangroves, both of which are absent from mudflat 2. Future surveys in the region should therefore also focus on more rigorously quantifying differences in diversity and abundance of other mollusc species across the three mudflats.