The Earth’s climate is changing and observed climate changes are causing a wide array of impacts [1
]. In marine systems, poleward shifts of many tropical species within the cnidarians, fishes, molluscs, echinoderms and crustaceans have been attributed to climate change [2
]. Additionally, behaviours such as breeding and migration are expected to respond to climate change [13
]. On a global scale, the upper 75 m of the oceans have warmed by 0.11 (0.09 to 0.13) °C per decade over the period 1971 to 2010, and the latest projections indicate warming will continue in all oceans across all depths [14
]. As a result, further range shifts and changes to biological patterns and processes are predicted to occur across a wide range of taxa [15
Temperate marine habitats can provide climate refuges for tropical species and potentially offset temperature induced local extinctions [7
]; however they may also be negatively impacted at the ecosystem level. There is growing concern that the tropicalisation of cool water marine communities will lead to the loss of some of Australia’s unique temperate marine biodiversity. Although Australian cool waters have lower species diversity than northern tropical waters, they harbor much higher proportions of endemic species. The south-west region of Western Australia (WA) is part of the southern temperate faunal province, a region extending across the entire southern coast of Australia and into southern New South Wales, and includes Tasmania and New Zealand [17
]. This region has high levels of endemism estimated to be between 75% [18
] and 95% [19
] for molluscs. These high endemic percentages are reflected across the entire southern Australian province in many other marine groups including fishes (85%); echinoderms (31%), sponges (56%) and macroalgae (50%) [21
Southern Australia has experienced pronounced increases in ocean temperatures [14
] and coastal waters in south-western Australia have also warmed appreciably, increasing by 0.013 °C·year−1
since 1951 [23
]. Any climate induced changes in species abundance and/or distribution patterns may have deleterious impacts on temperate taxa. Attrition of taxonomic diversity has already been observed in association with the poleward expansion of the barren-forming sea urchin, Centrostephanus rodgersii
from NSW southward to eastern Tasmania [3
], where a minimum of 150 taxa associated with macroalgal beds are estimated to have been lost [24
]. Tropicalisation of temperate reef communities can also lead to deleterious shifts in community structure [15
]. Of paramount concern are the intrusions of tropical herbivorous fishes, which can lead to overgrazing and the loss of foundation species such as canopy-forming algae [25
]. Climate induced changes are also implicated in impacts to the abundance and recruitment dynamics of commercially important species such as the Panulirus cygnus
(Western Rock Lobster) [28
Furthermore, warming waters may facilitate the establishment, spread and impact, of invasive exotic species in temperate marine systems [29
]. For example, the exotic Carcinus maenas
(European green shore crab), a predator in intertidal and shallow subtidal habitats, has extended its range from Victoria into Tasmania in response to increasing ocean temperatures and the strengthening of the East Australia Current [30
]. Dosidicus gigas
(Humboldt or Jumbo squid) was rarely observed beyond San Diego a decade ago, but now has been observed as far north as Alaska [31
]. This range shift has been attributed to climate change and an increase in areas referred to as “dead zones”, which the squid are able to exploit due to low oxygen tolerance. Fisheries scientists are monitoring this voracious, top predator species, and where schooling populations are high, impacts to prey species are expected to be significant [32
Climate is not, however, the only reason for change in marine environments, other factors including population growth, urbanisation, pollution, overfishing and shipping all interact to influence the composition and structure of marine communities [33
]. In order to fully understand the extent of changes to species distribution and abundance patterns, and to effectively manage and conserve temperate resources, quantitative baseline data are urgently required. Western Australia provides an ideal natural laboratory to monitor and examine the effect of climate change on marginal reef communities [35
]. The State occupies roughly one third of the Australian continent, and has 24,000 km of continuous north-south aligned coastline [36
] spanning nearly 22° of latitude [37
]. In contrast to the east coast of Australia, where a similar expanse of coastline is split across three separate States (Queensland, New South Wales and Victoria), potentially complicating effective monitoring due to the need for coordination among jurisdictions, in WA marine resource monitoring and management efforts are coordinated by multiple partner agencies within a single jurisdiction. One third of the coast lies in the tropics (Pilbara and Kimberley coasts, 12°–23.5° S), and there is also substantial subtropical habitat (Shark Bay, Houtman Abrolhos Islands to Jurien Bay 23.5°–31° S), shifting to temperate coastline (Perth to Esperance, 32°–34° S) in the south.
Temperate reefs near Perth (~32° S) occur in the transition zone between subtropical and temperate habitats. In situ temperature measurements at a monitoring station off Rottnest Island indicate that the mean ocean temperature has increased ~0.6 °C over the past five decades [23
]. While reefs in the Perth metropolitan area occur well outside the tropics and do not form coral reefs, they can support a high level of hard coral cover (e.g., 72.5% coral cover was recorded at Hall Bank) [38
]. To date, 25 scleractinian coral species have been recorded in the vicinity of Perth and Rottnest Island [37
], and this region forms the southernmost boundary for half of these species. Rottnest Island also forms the southernmost boundaries for many reef fish species including twenty that were first observed at the island in 1991 [39
]. The marine crustacean fauna of Rottnest also has a strong tropical influence with 39% being Indo-West Pacific species and 48% being southern Australian endemics [40
]. The flora and fauna of Rottnest Island were examined in detail in the 1990s [41
] and these volumes provide a detailed qualitative baseline of the many marine taxa, however, comprehensive biodiversity assessments for some groups (e.g., molluscs, sponges) are still lacking.
Roe Reef, Rottnest Island and Minden Reef, Fremantle are two reef systems that occur in close proximity to the Perth metropolitan area. To our knowledge Minden Reef has never been formally surveyed, and comprehensive surveys of benthic fauna at Roe Reef are lacking, hence no quantitative information about the faunal and floral composition of these reefs is available. In order to fill this knowledge gap, and to provide a baseline of information for the level of benthic cover and diversity within eight taxon groups (hard corals, soft corals, sponges, molluscs, crustaceans, echinoderms, fishes, algae), we undertook the first quantitative diversity assessments of these two locations in 2013 and 2014. Furthermore the Western Australian Museum database, Ozcam, Atlas of Living Australia, WORMS and other databases (where relevant) were queried to examine the known distribution for the flora and fauna we recorded in order to determine the extent of tropicalisation occurring in our study area. Here we report the findings of these baseline surveys, which will assist detection of future changes in community composition and structure at these locations.
Here we report the presence of 427 species of marine flora and fauna from eight higher taxon groups occurring in the Perth metropolitan area. Three of these species appear to be new to science and many more may be revealed with further investigation because over 110 specimens collected in this survey require further taxonomic and molecular investigation to confirm their identity (sponges, soft corals, molluscs and crustaceans). We record for the first time 15 species that have not previously been recorded from this latitude in Western Australia and the presence of 24 species that are endemic to south-west Australia (Table 2
Although the flora and fauna of Roe Reef has not been the subject of quantitative scientific investigation it is a popular recreational dive site and the fish, algae and crustacean communities have been examined on various occasions [42
]. However, Minden Reef is a little known location that is rarely visited by recreational divers and to our knowledge has never been the focus of a formal marine survey. The results of these surveys add to our knowledge of local nearshore marine environments and support a growing body of evidence of a diverse and regionally significant marine fauna in temperate WA.
The overall diversity of fauna was slightly higher at Roe Reef (218 species vs.
187 species), although a similar number of species was recorded at sites within reefs. Slight differences in survey effort and survey depths prevent statistical comparisons within and between sites. Nevertheless, it was clear from our surveys that the two reefs, despite being only 21 km apart, contained very different communities. For example, the biomass of hard corals at Minden Reef is approximately three times higher than recorded at Roe Reef and the biomass of soft corals is two times higher at Minden Reef. The temperate water specialist Coscinarea marshae
, was only observed at Roe Reef, and Pocillopora damicornis
and Montipora mollis
were not encountered at Minden Reef, even though these coral species were abundant at Roe Reef. Other than Chromodoris westraliensis
, the charismatic sea slug or nudibranch fauna was either a “Minden” fauna or a “Roe” fauna (Table S1
). Further, Trachinops brauni
(Bluelined Hulafish), which is endemic to south-west WA was observed in abundance at Roe Reef, but not recorded at Minden Reef. Similarly, Trachinops noarlungae
(Yellowhead Hulafish) was almost fifty times more abundant at Roe Reef than Minden Reef. It also appears that Minden Reef may be a nursery for juvenile Choerodon rubescens
, the heavily fished Baldchin Groper.
The spatial heterogeneity of community composition is further reinforced by the finding that no species of hard coral, soft coral or sponge were recorded at all survey sites. One mollusc (Chromodoris westraliensis
), one echinoderm (Centrostephanus tenuispinus
), one decapod (Ancylocheles gravelei
) and 12 species of fishes were the only organisms recorded at all sites (Table S1
). The spatial differences in community composition relate, in part, to the different topographic and environmental settings of the reefs. For example, Minden Reef lies just 2 km offshore whereas Roe Reef is 20 km offshore and 1.6 km from Rottnest Island. Minden Reef is a small, low profile reef compared to Roe Reef’s topographical complexity incorporating caves and vertical reef faces. Minden Reef’s inshore location, protected by more seaward reefs between Garden and Rottnest Islands, combined with its close proximity to Fremantle Harbour, has resulted in a high sediment regime. As a result marine communities are dominated by abiotic material (sand, silt, rock and maritime wreckage, see Figure 2
J), and on the southern side of the reef, by seagrass. In contrast, Roe Reef generally experiences comparatively clear water conditions with a higher energy regime and is dominated by a variety of flora, including turf algae, brown algae and kelp.
Amongst our results are numerous important findings, particularly the discovery of undescribed species on the doorstep of another one of Australia’s major capital cities [57
]. Firstly, a new species of red algae has been determined from fresh tissue collected at Roe Reef. Secondly, two new species of sponge barnacle were collected, one each from Minden and Roe Reefs. One of these provides the first record of the genus Pyrgospongia
in Australia. Our data also extends the known range of 15 species and in numerous instances, thousands of kilometres southwards. For example, the xanthid crab, Zalasius dromiaeformis
from Minden Reef has not previously been recorded from Western Australia. The barnacle Acasta fenestrata
collected from both Minden and Roe Reefs, and the shrimp Synalpheus bispinosus
, collected from Minden Reef, have not previously been recorded south of the Dampier Archipelago. Similarly, a species of brittle star, Amphiura (Amphiura) brachyactis
, collected from Minden Reef, has not previously been recorded south of Roebuck Bay, Kimberley. The cemented clam Chama pacifica
collected at Roe Reef has not previously been recorded south of Broome. The distribution of Alertigorgia mjöbergi
, a tropical octocoral species previously known as far south as Exmouth, has been confirmed at Minden Reef. This study also provides the first record of the sponge Protosuberites epiphytum
, collected from Roe Reef, occurring south of Shark Bay. Numerous other species had their known range extended south from the Houtman Abrolhos Islands (e.g., Favites rotundata
, Cyclax spinicinctus
, Liomera margaritata
, Armatobalanaus allium
, Pyrgoma cancellata
) as a result of this survey. These latter two species are only the second and third species of coral inhabiting barnacles to be reported in the Perth area, as well as representing some of the most southerly records for this group. The only other coral barnacle species from the region is Pyrgomina djanae
, a southern WA endemic [58
Few studies have examined the sponge fauna of the temperate regions of Western Australia. The sponges of Minden Reef were reported in a newspaper article in 1907 [59
], which mentioned the high incidence of sponges found at Minden Reef (and surrounding temperate locations) in the context of development of a bath sponge industry. At that time the most commonly used species for this industry were those belonging to the genera Spongia
. Species of both of these genera were reported in our survey, but they were not common—one and five individuals respectively. Based on previous studies [60
] it can be determined that of the 67 sponge species reported here, 11 are known to occur in local habitats, two are new records and the remaining 54 species require additional taxonomic comparisons.
It is important to note that only one introduced species was recorded in these surveys, the east Australian scallop Scaeochlamys livida
]. This species is now established in Western Australia and was recorded at both reefs (based on dead valves). While increased incidence and movement of certain pest species with climate change is expected globally this is subject of ongoing study [62
] and requires further attention in the Perth transition zone.
Cypraeidae or cowries are a widespread, attractive and diverse family of gastropod molluscs, many of which are targeted by collectors. In Western Australia, an active group of professional collectors has been monitoring distribution and abundance of several target taxa, including the highly sought after Australian endemic genus Zoila
, for many years. Because of this longstanding interest, cowries represent a viable model system to track range shifts and new arrivals through citizen science partnerships, much like coordinated nudibranch surveys have recently galvanized community efforts to monitor faunal change on the east coast [10
]. The survey at Minden Reef documented both of the two species known to occur there, the direct developing Zoila friendii friendii
and Z. venusta
. These two species have been observed in the area for at least the previous 25 years. To date there are no records of any settlement by free drifting veligers of other Cypraea
species at Minden Reef. No cowries were observed during the survey of Roe Reef. This in contrast to historical data that indicate quite a diverse cowrie fauna of approximately 22 species exists at Rottnest Island (of which Roe Reef is a part). Of these, ten have been confirmed to represent breeding populations (e.g., egg masses observed) and almost all represent temperate species and/or temperate endemics (see Table S2
). In contrast, the other twelve species are considered to be recent arrivals and possibly tropical vagrants; all twelve are well-known tropical (IWP) species. Most of these are known only from dead collected shells indicating that these records are unlikely to represent established populations, however several species have been found live, including the soft coral predator, Ovula ovum
or egg cowrie (see Table S2
). Although these are preliminary findings, there is a clear signature emerging of a slow, but constant influx of tropical species arriving at Rottnest Island over the past 50 years. One species that has potentially disappeared from Rottnest Island is Notocypraea pulicaria
. It is a southern Australian endemic, like all members of the genus. This species was recorded from the island in the 1930s and has not been observed live at Rottnest Island since.
This data set verifies that some level of tropicalisation of reef communities in the Perth metropolitan area is occurring. The past two decades has seen strengthening of the Leeuwin Current [63
], which may explain the occurrence of some of these and other more tropical species, such as Pterois volitans
(Common Lionfish), Anampses geographicus
(Scribbled Wrasse), Siganus fuscescens
(Black Rabbitfish), and Labroides dimidatus
(Common Cleanerfish) occurring adjacent to Perth [64
]. Previous reports indicate the southern distributions of many species of tropical fishes are extending as a consequence of warming sea temperatures in Western Australia [65
]. A growing body of evidence suggests that climate driven range extensions of tropical herbivores, especially fishes, can dramatically alter algal ecosystems [67
]. Excessive grazing by new tropical colonists can reduce temperate reefs previously densely covered with macroalgae to a canopy-free state dominated by turf species, especially when the algae are also under warm water stress [27
]. Although the strengthening of the Leeuwin Current has been well documented in recent decades, it is expected to weaken under climate change predictions into the future [63
]. These counterintuitive shifts in abiotic variables will dictate similarly complex responses in biotic systems.
While some of the new records may be attributed to a lack of previous survey effort, there are many examples of significant range extensions in this dataset (Table 3
). Chama pacifica
can be cryptic, with a shell overgrown by epibionts, but the Perth metropolitan area has been visited extensively over several decades and this species has not previously been recorded. If the reef communities in the Perth metropolitan area are becoming tropicalised as suggested by this study, this will have ecological and perhaps evolutionary ramifications for locally endemic species. Ordinarily, the decadal cycle of Leeuwin Current weakening (with lessening propagule pressure) would result in many tropical vagrants disappearing from southern limits. However, since the predicted Leeuwin weakening will coincide with continued global seawater warming, it is likely that many vagrants will remain and establish as sustainable populations.
The species abundance and distribution records presented here are conservative. We acknowledge our surveys covered only a small area, and in some cases the high level of Ecklonia
cover obscured other benthos making the documentation of some benthic fauna, such as sponges, problematic. Furthermore, it is difficult to survey molluscs and crustaceans during daylight hours as many are nocturnal and cryptic by nature. Furthermore, some fauna such as nudibranchs may be seasonal. As a result, diversity of these taxa has almost certainly been under-represented. Surveying the fish fauna at one point in time is unlikely to capture the true diversity of fishes, other than those that are site attached. Although the hard and soft coral records provide a reliable representation of the diversity of those taxa at Minden and Roe Reefs, additional species may be encountered with further surveys. It is interesting to note that that all but four of the scleractinian coral species recorded from a previous survey at nearby Hall Bank (~3 km north of Minden Reef; [38
]) were recorded in this survey. Two species present at Minden Reef in the present study (Goniopora pendulus
, Turbinaria peltata
), were not recorded at Hall Bank, despite the latter reef having a far greater benthic cover of corals (>67%). Taxonomic biases are not likely to explain this result as both species are distinctive and easily identified. This finding is perhaps a reflection of the heterogeneity of communities in the Perth metropolitan transition zone and substantiates the need for thorough survey effort to detect diversity.
As marine environments and their associated faunal assemblages continually adjust to global warming trends, the accumulation of baseline biodiversity data is critical. This is particularly important along the margins of recognised transition zones on continuous coastlines, such as the temperate reefs surveyed here. Although, range shifts can be subtle and often difficult to document, other effects are clearer, with the impacts of heightened sea surface temperatures already apparent along the WA coast. Endemic corals growing on high latitude reefs have proven susceptible to thermal bleaching from anomalous warming events as was observed with Coscinaraea marshae
at Rottnest Island [69
] after the 2011 “marine heat wave” that affected Western Australian coral reefs across 12° of latitude [70
]. Ongoing monitoring of community composition is the only way to better understand the conservation implications of climate changes affecting local environments. The positions of these sites adjacent to a major metropolitan center should enable monitoring by the marine science community, to not only further document baseline biodiversity but also track the incidence of warm water invasive species. Ongoing monitoring of the composition and structure of marine communities in the Perth metropolitan transition zone is needed to quantify the extent and rate of change and to justify any future management interventions that may be initiated to protect potentially at risk endemic south-west fauna [72
Observation in different locations and seasons is required to more accurately capture the diversity of marine life in the Perth metropolitan area. For example, many nudibranchs are temporally and/or spatially rare and may not be detected in an area for long periods (e.g., Goniobranchus aurigera, known only from the holotype collected in 1986 at Quinns Rocks). Many species known from historical records were not observed during the two surveys and whether this indicates a true absence or statistical absence due to a lack of survey effort remains to be established. Furthermore, the roving nature of pelagic and mobile fishes, the cryptic and nocturnal nature of crustaceans, and the seasonal nature of marine algae may have contributed to the level of diversity being underreported which would increase with further work.