Annual and decadal surveys of molluscs on intertidal platform reefs in a warming marine environment in the southeastern Indian Ocean provide a baseline for monitoring for future environmental changes

ABSTRACT

Few data sets are available to evaluate short- and long-term perturbations and impacts of increasing anthropogenic stresses and a warming marine environment in coastal regions near cities. The Perth, Western Australia shoreline has extensive, depauperate sandy beaches interrupted by small, biodiverse intertidal reef platforms. We present transect data on mollusc diversity and density on three Perth platforms recorded from 1983 to 1987, and compare it to 2007 and 2021 to analyse assemblage modifications in response to increased anthropogenic pressures. NMDS multivariate analysis showed a consistent drift in assemblage composition in 2007 and 2021, but this is was not statistically significant from the 1980s baseline (PERMANOVA, P = 0.336). There was no significant long-term decline (or increase) in molluscan species richness. Most species were uncommon but eight were recorded on at least 20 of the 21 surveys. The endemic Western Australian mussel Brachidontes ustulatus (Lamarck, 1819) accounted for 71.8% of all individuals. Total mollusc density varied primarily with changes in B. ustulatus. Tropical species were a minor component of the molluscs that did not change in diversity with warming temperatures. These data provide a long-term baseline for monitoring future anthropogenic changes.

KEYWORDS:

• climate change
• marine invertebrates
• global warming
• marine heatwaves
• mussels
• Western Australia
• Leeuwin Current
• mortality

Introduction

The oceanography off the west coast of Western Australia (WA), between North West Cape and Cape Leeuwin, is unique. The Leeuwin Current that flows southward on the outer continental shelf is the only poleward flowing current on the west coast of a continent (Cresswell and Golding 1980). It has profound effects on coastal climate, marine community structures and fisheries (Pearce and Walker 1991; Pearce 2007). One effect of the Leeuwin Current is that tropical species occur much further south than they otherwise would (Wilson and Allen 1987; Morgan and Wells 1991). The coastline of WA consists of three marine biogeographic regions: a tropical north coast that is part of the Indo-West Pacific; a warm temperate south coast continuous with southern Australia; and a west coast overlap zone between North West Cape and Cape Leeuwin where tropical species dominate in the north and temperate species in the south (Wilson and Allen 1987; Morgan and Wells 1991). About 10% of the shallow water molluscs of WA are endemic to the state. These are concentrated in the west coast overlap zone and are neither tropical nor temperate (Wells 1980).

The west coast is a global hotspot of increasing sea surface temperatures (SST) (Pearce and Feng 2007; Hobday and Pecl 2013). In 2011, a marine heatwave caused extensive losses of kelp beds and tropicalisation of marine environments along 12° of latitude from North West Cape to Cape Leeuwin (Pearce and Feng 2013; Wernberg et al. 2013; 2016). Perth on the lower west coast (32°S) is the primary population and industrial centre of WA (2.09 million inhabitants; population 2022) and Fremantle is the largest general cargo port in WA (Fremantle Ports 2022). Previous studies have described detrimental effects such as imposex on molluscan communities in the Perth metropolitan area caused by the presence of tributylin, linked to shipping antifoulants (Wells and Gagnon 2020). The west coast of WA is thus a natural laboratory to develop an understanding of anthropogenic effects on local coastal marine environments.

The Perth region is characterised by a Mediterranean climate with hot, dry summers and cool, wet winters. The local coastline is essentially one long, sandy beach interrupted by small intertidal limestone platforms that range from a few metres to over 100 m in width. The medium intensity wave action and constantly shifting sand are inhospitable to benthic biota and the beaches are depauperate (McLachlan 1985; Ince et al. 2007; Morrison 2015) but the platforms are small scale biodiversity hotspots, with >130 species of molluscs recorded on Perth platforms (Wells and Keesing 1986; Kendrick and Rule 2014).

As small features on the coastline of a major metropolitan area, coastal platforms are under anthropogenic threats such as pollution, global warming, fishing pressures, etc. Molluscs are ideal subjects for environmental monitoring programs as they are speciose, taxonomically well-known and many species are large and abundant. They have a diverse biology, including a variety of feeding mechanisms, reproductive strategies, growth rates, lifespans etc. (Oehlmann and Schulte-Oehlmann 2002). The usefulness of molluscs on intertidal platforms for monitoring environmental changes and for establishing marine reserves in Australia has been demonstrated by Benkendorff and Davis (2002), Gladstone (2002), Smith (2005), Costa et al. (2010) and Alexander and Gladstone (2013). However, none of these studies included data on long-term temporal changes. We examine data on mollusc populations on three intertidal platform reefs in the Perth metropolitan area collected from 1983 to 1987 to compare with similar surveys made in 2007 and 2021. The goal was to provide a baseline for monitoring for changes in future surveys.

Materials and methods

Annual summer samples

Molluscs on the intertidal limestone platforms at Cottesloe, Trigg Point and Waterman were sampled annually during the austral summer from 1983 to 1987, and then again in the summers of 2007 and 2021. The platforms are discrete and surrounded by sand, so there is probably little or no migration of the adults of most species to or from the platforms. Each platform was divided into habitats based on macroalgal assemblages (Figure 1). Three habitats were sampled at Trigg Point and Waterman: an inshore platform with diverse algae; Sargassum zone on the middle of the platform; and a seaward bare zone with little macroalgal coverage. Four zones were defined at Cottesloe: inshore platform; inner Sargassum on the middle of the platform; outer Sargassum at the seaward margin of the platform; and an Ecklonia zone on the southern end of the platform in slightly deeper water (approximately 1 m).

Figure 1. Map of the location of the three study sites along the coastline of Perth, Western Australia (Drawn from Google Maps 2022). Letters identify habitat zones: B bare; S outer Sargassum; IP inner platform with diverse algae; IS inner Sargassum; E Ecklonia.

During the 1980s surveys commenced at the northern end of each zone (Figure 1). Every 10 m all molluscs >5 mm in 4 × 0.25 m² quadrats placed in a square to total 1 m² were identified and counted. Eight stations were made in each algal assemblage on each platform (24 m² total at Waterman and Trigg Point and 32 m² at Cottesloe). Transects were repeated as closely as possible in subsequent surveys, but were not permanently marked. Most of the species identified were small, short-lived gastropods with a one year life cycle such as Prothalotia pulcherrimus (W. Wood, 1827) (Wells and Keesing 1987). Larger species, such as Lunella torquata (Gmelin, 1791) and Haliotis roei (Gray, 1826) are highly mobile. The combination of gastropod mobility and short lifespans suggest that the 1980 samples were largely independent.

Representative specimens of all species were deposited in the Western Australian Museum. To minimise study impacts all other molluscs were identified in the field and returned to the platform. A similar approach was adopted in 2007 and 2021 but to reduce the workload 10 × 0.25 m² quadrats were surveyed in each habitat (7.5 m² total at Waterman and Trigg Point and 10 m² at Cottesloe). The reduced sampling intensity should not have affected density calculations, but could have reduced the numbers of species recorded. Data for each habitat were converted to average density for the overall platforms using estimates of the area of each habitat on each of the platforms (Table 1).

Table 1. Numbers of temperate, tropical and WA endemic species identified on three Perth coastal platforms in surveys taken between 1983 and 2021.

Year
Distribution	1983	1984	1985	1986	1987	2007	2021
Temperate	30	38	43	43	31	29	26
Tropical	5	7	5	5	4	7	5
Endemic	7	7	8	6	5	6	8
Unknown	6	4	3	5	7	4	6
Total	48	56	59	59	47	46	45

Molluscs collected on the platforms were classified into one of the three categories (tropical, temperate or WA endemic) based on the published literature and collections of the WA Museum.

Statistical analyses

Mollusc densities are presented as mean number of individuals ± standard error per square metre (mean ± SE/m²).

Multivariate analyses were conducted using R statistical software, version 4.0 (R-Core Team, 2023). Analyses of community diversity for richness, Shannon’s diversity and Simpson diversity were conducted using the Vegan R package (Dixon 2003). Non-metric multidimensional scaling (NMDS) analysis was performed using the Vegan R package using Bray-Curtis distance. The nMDS analysis incorporated all species records made during the surveys. Similarity between survey sites, and between decades, was established by the permutational multivariate analysis of variance (PERMANOVA) test, using a P-value of 0.05 to determine significance.

Mussel taxonomy

In earlier platform work (Wells and Keesing 1986; Wells et al. 1986), the common mussel was identified as Brachidontes ustulatus (Lamarck, 1819) a species that, with the exception of a single population in South Australia, is endemic to WA. This was the accepted identification of the mussel, e.g., Wells (1984), Wells and Bryce (1986) and Wilson (2002). It is part of a species complex discussed by Huber (2010), who considered the Australian species to be B. sculptus (Iredale, 1939). The change to B. sculptus was accepted by ALA (2023) and was used in Wells et al. (2023). However, subsequent research by Dr Koh-Siang Tan et al. (pers. comm.) has found that B. ustulatus is the appropriate name, and we have used B. ustulatus in this paper. This change makes only a minor difference in the numbers of endemic and tropical mollusc species on the platforms, but a major change in their relative densities. As the species concerned is well known, the taxonomic issue has no effect on the usefulness of the data presented here for future monitoring programs.

Results

A total of 114 species of molluscs was recorded on the three platforms: 12 chitons, 81 gastropods, 20 bivalves and 1 cephalopod (Table S1). Twenty-one of these taxa were not identified with sufficient taxonomic certainty for their distribution to be classified. Temperate species dominated the total diversity of species fully identified (64 species); 16 were tropical and 13 endemic to WA.

Although 69% of the fully identified species were temperate, they constituted only 26.1% of total density over the years. Tropical species were only 0.1% of the total density. WA endemic species were 73.8% of all species, dominated by the mussel B. ustulatus (71.8%). The biogeographic proportions of the molluscs identified to species fluctuated over the surveys (Table 1). Temperate species were always most diverse, ranging from 26 in 2021 to 43 in 1985 and 1986. There were 4–7 tropical species and 5–8 temperate species identified in the various years.

Most of the identified species were found on only one or two occasions, but four [Patelloida nigrosulcata (Reeve, 1855), Notogibbula preissiana (Philippi, 1849), Lunella torquata (Gmelin, 1791) and Euplica bidentata (Menke, 1843)] were recorded in all 21 surveys and another four were recorded 20 times: Haliotis roei Prothalotia pulcherrima (W. Wood, 1828), Dicathais orbita (Gmelin, 1791) and Mitrella lincolnensis (Reeve, 1859) (Tables S2-4).

During the 1980s Cottesloe had the greatest number of species in all years, ranging from 34 to 53. Trigg Point was second, with 28–40 species, followed by Waterman at 26–34 species. The total number of species recorded annually in the 1980s ranged from 47 to 59. Cottesloe had 31 species in 2007 and 32 in 2021. The figures for Trigg Point were 36 and 27, with 27 species recorded at Waterman in 2007 and 32 in 2021 (Table S1).

There is a significant difference (ANOVA followed by Tukey) in the Shannon and Simpson diversity between Waterman and the other two sites (P < 0.048 and P < 0.031 respectively), but the diversity at Trigg Point and Cottesloe were not significantly different from each other (P = 0.26 and P = 0.43 respectively) (Figure 2). The richness (number of species recorded) was significantly higher at Cottesloe than at the other two sites (ANOVA followed by Tukey, P < 0.042) (Figure 2).

Figure 2. Species richness and diversity indices of molluscs on three intertidal limestone platforms in the Perth metropolitan area from 1983 to 1987, and in 2007 and 2021.

Density variation in the 1980s

Total mollusc density fluctuated greatly between years at Cottesloe, where the range was from 459.1 ± 141.8/m² in 1986 to 2752.6 ± 1098.7/m² in 1984, primarily due to fluctuations in year classes of the mussel B. ustulatus. A similar phenomenon occurred at Trigg Point, where the range was from 1248.9 ± 871.3/m² in 1987 to 5094.1 ± 1569.8/m² in 1985. Again, most of the variation was due to mussels, in this case B. ustulatus and Xenostrobus pulex (Lamarck, 1819). Mussels were virtually absent at Waterman, where the density varied from 199.4 ± 42.2/m² in 1984 to 267.5 ± 109.8/m² in 1987 (Figure 3, Tables S1 to S4).

Figure 3. Total mean density of molluscs (3A) and the mussel Brachidontes ustulatus (3B) on three intertidal limestone platforms in the Perth metropolitan area from 1983 to 1987, and in 2007 and 2021.

One noticeable cause of variation on the inshore platforms was the presence of sand. This was most obvious at Waterman. In January 1983 there were clearly three habitats on the platform: inshore platform, Sargassum zone and bare zone. The inshore platform had 19 species of molluscs and a density of 148 ind/m². In January 1984 the southern half of the inshore platform transect was covered with sand, eliminating the molluscs that had lived in that area. Crevices on the northern half of the transect were filled with sand and only the ridges had molluscs on them; mollusc density was only 56 ind/m². A similar phenomenon occurred in 1985, when total mollusc density was only 24 ind/m². The same pattern occurred to a lesser extent at Trigg Point. A large hole just north of Trigg Point was approximately 10 m long, 3 m wide and 1–1.5 m deep in 1983. In 1984 the hole was filled with sand almost to the platform level. The platform itself was not covered with sand as it was at Waterman, but crevices were filled in and only ridges were exposed. Consequently, the number of species and total mollusc density both declined at Trigg Point in 1984. Sand coverage at this site was not as severe in 1985. The number of species collected on the inshore platform increased to 19, but was still lower than the 24 recorded in 1983. Sand was not a problem at Cottesloe and did not cause the decline in B. ustulatus.

Results in 2007 and 2021

The 19 species most commonly recorded on the platforms during the 1980s were also found in both 2007 and 2021 (Tables S1 to S4). Despite the reduced sampling effort, the total number of species recorded on the three platforms in 2007 (46) and 2021 (48) was consistent with the results in the 1980s (Table S1).

Multivariate analysis

The NMDS analysis of mollusc assemblages (stress value = 0.156) shows separation between the three sites within the ordination space (Figure 4A), with a clear association increasing from south to north. There were significant differences between the mollusc assemblages at Cottesloe, Trigg Point and Waterman (PERMANOVA, P = 0.001). There appears to be a consistent drift from the baseline molluscan community assemblage in successive later decades from the 1980s through to 2021 (Figure 4B), but this was not statistically significant (PERMANOVA, P = 0.336).

Figure 4. NMDS plot of three sites for seven surveys conducted between 1983 and 2021. Ellipses represent 95% confidence intervals.

Discussion

A total of 114 species was recorded during the study. Most species were recorded just once or twice on each platform. At the same time, there were consistencies in the species present, with the same species dominating from 1983 through to the last sampling period in 2021. Eight species, Patelloida nigrosulcata, Notogibbula preissiana, Lunella torquata, Euplica bidentata, Haliotis roei, Prothalotia pulcherrima, Dicathais orbita and Mitrella lincolnensis were recorded on at least 20 of the 21 surveys made during the study and just a few species, most notably the mussel Brachidontes ustulatus, dominated in terms of total abundance. The data presented provide a valuable description of mollusc densities on the three platforms: Cottesloe, Trigg Point and Waterman. If the data are to be used to monitor for future changes in the mollusc assemblages, the key question is whether there have been temporal changes from natural conditions on the platforms.

Marsh (1955) surveyed biota on the WA limestone platforms from Albany to Kalbarri from 1948 to 1952. Cottesloe was a key site and was visited numerous times, including seasonal visits. The description of the Cottesloe platform matches conditions experienced during our surveys. After adjusting Marsh’s (1955) records for taxonomic changes, the same species dominated in the late 1940s and early 1950s as at present.

Population declines at Waterman and Trigg Point in 1984 and 1985 were caused by extensive sand coverage on the inshore platform. This is a natural phenomenon. Perth sand beaches are highly mobile, with a general northward ocean current and sand movement during most of the year and a southerly reversal for about two months in winter (Kempin 1953). Clarke and Eliot (1983) examined 16 years of records of beach-width change at six profile stations on Scarborough Beach, just south of Trigg Point. There was a long-term progradation trend in beach width superimposed on cyclic changes with periods of 0.5, 1.0, 3.5 and 7.0 yrs. Beach width changed by up to 25 m during their study, reflecting movement of sand along the shoreline.

Benthic organisms on the three platforms studied have a high degree of protection from human exploitation. The Waterman platform has been protected as a Reef Observation Area since the 1960s. Trigg Point and Waterman were included in the Marmion Marine Park in 1987. The Cottesloe platform was declared as a Fish Habitat Protection Area in 2001. No harvesting of benthic organisms is permitted at Waterman and Cottesloe; abalone (H. roei) fishing is permitted at Trigg Point for one hour each on 4–6 occasions every summer as part of the Perth metropolitan recreational abalone fishery. The reefs are closely monitored for illegal activities by officers from both the Fisheries section of the Department of Primary Industry and Regional Development and the Parks and Wildlife Authority. Fisheries officers are stationed at each of the three platforms during the abalone season. Council rangers and the general public are also aware of the restrictions and actively report suspect activities. While there may be some illegal fishing of benthic organisms on the platforms it is likely to be minor, and concentrated on abalone.

Three types of pollution affecting water quality have been recorded on the Perth metropolitan coastline: outflow from stormwater drains, groundwater flows and tributyltin. Water (2007) investigated outflows from 65 stormwater drains in the Perth region. Most are small and flow intermittently during winter rains. Several drains occur near the platforms studied in this paper but all flow intermittently. A larger drain with more continuous flows occurs at Floreat Beach, ∼ 6 km north of the Cottesloe platform and ∼5 km south of Trigg Point. Heavy metals and bacterial contamination were the primary concerns identified by the Water (2007) study. Unfortunately, there has been no published update to the Water (2007) study. The urban drains discharging small volumes in the vicinity of the platforms surveyed do not appear to affect the health of the mollusc communities, as the same species, and in similar numbers, were found on the platforms in seven surveys between 1983 and 2021.

Imposex is a reproductive abnormality caused by the antifouling agent tributyltin incorporated in vessel paints. It was first found in WA at the west end of Rottnest Island, approximately 20 km offshore, in 1991 (Kohn and Almasi 1993). Field (1993) surveyed a number of sites in the Perth metropolitan area, reporting 100% of female D. orbita at Cottesloe and 87% at Trigg Point had moderate levels of imposex. Reitsema et al. (2003) and Wells and Gagnon (2020) traced the recovery of the D. orbita populations. By 2019 there had been an almost complete recovery throughout the metropolitan region, including 100% recoveries at Cottesloe and Trigg Point. Overall, the three known pollution types occurring on the three platforms studied are not known to have affected mollusc populations.

Huisman et al. (2008) reported that 60 introduced marine and estuarine species have become established in WA, including 46 in the Fremantle marine area that includes Cottesloe. None of these were found in our surveys. Beechey and Willan (2007) recorded a single dead shell of the east Asian dove snail Mitrella bicincta (A. Gould, 1860) at Cottesloe Beach, near the platform.

Three types of temperature variation could affect mollusc populations on the platforms: warming SST, marine heatwaves and short-term summer high air temperatures. Pearce and Feng (2007) reported SST increased by ∼0.6°C between 1951 and 2007 but there has been no apparent effect on mollusc populations on the platforms. The dominant species on the platforms in 2021 were the same as in the 1980s. Temperate species were most diverse, and the WA endemic mussel B. ustulatus dominated density. Tropical species were a minor component of mollusc diversity that did not increase in either diversity or density with warming temperatures (Figure S1). WA endemic species remained in low diversity and dominated density entirely because of B. ustulatus.

The 2011 marine heatwave had profound effects on the marine biota of the WA west coast, changing benthic community structure, affecting fisheries and extending the ranges of many tropical species southward (e.g., Wernberg et al. 2016; Moore et al. 2012; Ruthrof et al. 2018; Kendrick et al. 2019). Most of the known effects were in subtidal environments. A small abalone fishery at Kalbarri was eradicated (Caputi et al. 2019) and abalone populations on the three platforms we surveyed were reduced but subsequently recovered (Hart et al. 2016; Caputi et al. 2019). Mollusc and echinoderm populations on intertidal platforms at Radar Reef and Cape Vlamingh at the western end of Rottnest Island, ̴28 km off the Perth coastline, declined in the order of 90% between 2007 and 2021. While no direct cause could be confirmed, marine heatwaves are thought to be the most likely cause (Wells et al. 2023). In contrast, there were no reductions in mollusc diversity or density on the three coastal platforms we studied that could have been caused by marine heatwaves (Wells et al. 2023; present study). This is supported by multivariate analysis demonstrating no significant difference between mollusc assemblages in seven surveys from 1983 to 2021 (Figure 4).

We conclude that despite being along the shoreline of the major metropolitan region of Perth, mollusc assemblages on the intertidal platforms at Cottesloe, Trigg Point and Waterman are in relatively pristine condition. Our study found no evidence of any significant long-term patterns of either decline or increase in molluscan diversity and abundance that could be attributed to climate change or other anthropogenic causes. Molluscs on these platforms have already been used for environmental monitoring programs, beginning with the study of molluscs, particularly the abalone H. roei, on the platforms in the early 1980s (Wells and Keesing 1986), followed by their use in assessing the Trigg Point and Waterman platforms in the development of the Marmion Marine Park (Wells et al. 1986). Abalone populations on the platforms have been examined in detail (e.g., Keesing and Wells 1989; Wells and Keesing 1989; Hancock 2006; Hart et al. 2013, 2016). The Trigg Point and Cottesloe platforms were included as sites for studies of tributyltin induced imposex in D. orbita (Field 1993; Reitsema et al. 2003; Wells and Gagnon 2020). The population of L. torquata at Waterman is currently being monitored as part of the Marmion Marine Park management program (Dr Innes Leal, DBCA, pers. comm.).

Climate change, invasive species and chemical and plastic pollution all have the potential to alter mollusc abundance and diversity on rocky platforms located close to urban centres, hence it is important to have long term data on these habitats. There was no significant long-term decline (or increase) in molluscan species richness. Total mollusc density varied primarily with changes in B. ustulatus. Tropical species were a minor component of the molluscs that did not change in diversity with warming temperatures. Short-term perturbations included sand burial events from which populations recovered. The understanding of population fluctuations provided by the present study provides a solid basis for future monitoring for any changes due to human activities or extreme climatic conditions in the Perth metropolitan region and can be used as a model for other Mediterranean environments.

Acknowledgements

The research was conducted with appropriate permits from the Western Australian Department of Primary Industry and Regional Development (DPIRD) and the WA Department of Biodiversity Conservation and Attractions (DBCA). We warmly thank the many people who assisted the project in the 1980s and anonymous reviewers of the manuscript. All authors have read and agreed to the published version of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support this study are included in the supplementary information.

Additional information

Funding

Surveys from 1983 to 1985 were funded by the WA Department of Fisheries (now part of DPIRD). The 1986 survey was funded by the WA Department of Conservation and Environment (now part of DBCA). The 2007 study was funded by the Swan Catchment Council (now the Perth region NRM Inc.). The 1987 and 2021 fieldwork did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.