Climate-driven changes to ocean circulation and their inferred impacts on marine dispersal patterns
Abstract
Aim
The dispersal and distribution patterns of many marine organisms are driven by oceanographic conditions, which are influenced by global climate. Climate-driven oceanographic changes are thus likely to result in biogeographical changes. We assess how recent and predicted oceanographic changes affect the dispersal capacities and distributions of ecologically important (especially habitat-forming) marine organisms.
Location
We include studies from tropical, temperate and sub-polar regions to draw globally relevant conclusions.
Methods
We review biogeographical, biological and oceanographic studies to critically evaluate emerging trends in biogeographical responses to climate-driven oceanographic changes, and predict how future changes will affect marine ecosystems.
Results
Many oceanic dispersal pathways are being altered by climate change. These changes will affect marine ecosystems by differentially affecting the replenishment potential and connectivity of key habitat-forming species. In particular, the length of propagule pre-competency periods, propagule behaviour and the geographical distance between areas of suitable habitat will be critical in determining how oceanographic changes affect the pattern and success of dispersal events, including the likelihood of species experiencing poleward range shifts in response to a warming climate.
Main conclusions
Future climate-driven oceanographic changes are likely to strengthen or weaken different oceanic dispersal pathways, which will either increase or decrease the potential for dispersal and connectivity in various marine taxa according to the interaction between the local oceanographic, geographical and taxon-specific biological factors. A key focus for future work should be the development of fine-scale near-shore ocean circulation models that can be used to assess the dispersal response of key marine taxa under various marine climate change scenarios.
Introduction
Many marine species are predicted to undergo considerable range shifts in response to climate change as they attempt to track their preferred environmental envelopes. Broadly, species are predicted to move toward the poles and/or to greater depths with a warming climate (Wernberg et al., 2011; Vergés et al., 2014). However, the potential for such range shifts is heavily dependent upon the dispersal capacity of species, which is the key mechanism by which populations can persist through major environmental changes (Berg et al., 2010; Banks et al., 2011; Sorte, 2013; Bates et al., 2014). While evidence is emerging that many marine species show active dispersal capabilities during their early life history (e.g. coral reef fishes; Stobutzki & Bellwood, 1997; Fisher et al., 2000), many other taxa have limited motility, often during the earliest stages of their planktonic development, and so their broad-scale dispersal can be influenced by how ocean circulation patterns shape patterns of drifting (Winston, 2012). Such organisms often play key roles within ecosystems, for example as habitat-forming species (e.g. corals, macroalgae, seagrasses) that underpin the biodiversity and function of entire marine communities (Wernberg et al., 2011; Vergés et al., 2014). When environmental changes affect such foundation species, impacts can cascade throughout associated communities (Vergés et al., 2014). Consequently, understanding and predicting changes that may affect the dispersal of species is critical for the adaptive management and conservation of marine ecosystems (Berg et al., 2010; Lett et al., 2010; Sorte, 2013; Bates et al., 2014; Thomson et al., 2015; Kleypas, 2015).
Marine species often rely on a high propagule output, extended propagule persistence, propagule survival and extrinsic transport mechanisms (e.g. atmospheric and oceanic circulation) to achieve broad-scale dispersal, maintain population connectivity and colonize new territory (O'Connor et al., 2007; Banks et al., 2011; Kendrick et al., 2012; Sorte, 2013). However, these factors can be sensitive to environmental change. For example, elevated temperatures have been found to reduce gamete production in many marine taxa, including corals (Albright & Mason, 2013), macroalgae (Rothäusler et al., 2009) and seagrasses (Reed et al., 2009; Caputi et al., 2014). Likewise, pelagic larval duration is often temperature dependent (O'Connor et al., 2007; Byrne & Przeslawski, 2013; Figueiredo et al., 2014). In general, higher temperatures hasten larval development and shorten the pre-competency period, increasing the likelihood of larvae settling at shorter distances from their natal populations (Byrne & Przeslawski, 2013; Kendall et al., 2013; Figueiredo et al., 2014). Larvae that are exposed to temperatures above their optimal threshold also frequently display developmental abnormalities, which prevent them from respiring, swimming, feeding and/or settling effectively, often increasing mortality rates (Randall & Szmant, 2009; Byrne, 2011; Edmunds et al., 2011; Ericson et al., 2012). High sea surface temperatures (SSTs) and intense UV radiation can also accelerate the degradation of macrophyte rafts (Díaz-Almela et al., 2009; Macreadie et al., 2011; Rothäusler et al., 2011; Graiff et al., 2013; Tala et al., 2013).
Key drivers of dispersal in marine systems, such as atmospheric and oceanic circulation, are similarly sensitive to environmental change. Climate change models predict that many ocean currents will undergo changes in intensity and path under predicted future climatic conditions. While dispersal ability is not always directly evident from oceanographic influences or species life-history traits (e.g. pre-competency period, presence of a larval phase) (Johannesson, 1988), species that do disperse are often influenced by oceanographic forces (Gillespie et al., 2012). Therefore, these predicted changes could considerably alter the dispersal patterns and distributions of marine species, and affect their ability to undergo range shifts (Lett et al., 2010; Banks et al., 2011; Johnson et al., 2011). However, to date, climate-driven changes to atmospheric and oceanic circulation patterns have predominantly been interpreted from a climatic perspective, with the biogeographical impacts of such changes only being more recently realized. The idea that future climate-driven changes to ocean circulation patterns have the potential to alter dispersal pathways, potentially affecting the ability of species to track and adapt to climate change, has been briefly speculated on (e.g., Munday et al., 2009; Keith et al., 2011; Chamberlain et al., 2012) but has not yet been examined in detail. Here we review some of the predicted changes in large-scale ocean circulation patterns under future climate change scenarios, with a particular focus on Southern Hemisphere examples. We then relate these changes to the ability of marine species to achieve range shifts, and the possible impacts of such shifts on diverse marine communities, synthesizing our current understanding of the potential effects of changing ocean circulation on species dispersal and distribution patterns.
Consideration of oceanographic dispersal barriers and pathways, and how species respond to these, is crucial for estimating species dispersal potential, dispersal patterns and impending range shifts under future climate scenarios (Keith et al., 2011). However, it is important to acknowledge that an array of post-recruitment processes, such as species interactions, habitat availability, colonization ability and propagule tolerances, can also influence the potential for range shifts in marine species. These other factors can potentially inhibit settlement and population establishment of marine organisms, even in the presence of oceanic dispersal pathways. Many of these post-recruitment factors have been discussed extensively elsewhere (e.g., Keith et al., 2011; Bates et al., 2014; Keith et al., 2015; Sunday et al., 2015) and are largely outside the scope of this review. Instead, we focus on how shifts in some major ocean circulation patterns, arising from global climate change, may impact on this important component of broad-scale marine dispersal.
Climate Change Will Affect Oceanographic Drivers of Dispersal and Distribution
Prevailing wind patterns and the resulting ocean surface currents, eddies and ocean fronts are key determinants of the dispersal patterns of planktonic larvae/propagules and species distributions for marine organisms (Sorte, 2013; Sunday et al., 2015). Global circulation models predict that changes in the strengths and locations of wind patterns will lead to changes in the paths and intensities of large-scale surface currents and the positions of ocean fronts (Oke & England, 2004; Cai, 2006; Roemmich et al., 2007; Luo & Rothstein, 2011; Sorte, 2013; Zhang et al., 2013). These changes have the potential to drastically alter dispersal patterns, connectivity and recruitment dynamics in marine communities (Cai, 2006; Lett et al., 2010; Banks et al., 2011; Johnson et al., 2011). Of critical concern is whether climate-driven oceanographic changes will modify propagule transport patterns between latitudinal climatic zones (e.g. tropical to temperate or temperate to sub-polar regions). For example, if poleward currents weaken, the ability of marine species to undergo the range shifts necessary to track their climatic envelopes could be severely compromised. In the Southern Hemisphere, a number of ocean currents [e.g. the Indonesian Throughflow (ITF) and Leeuwin Current] are predicted to weaken under future climate change scenarios (Poloczanska et al., 2012; Sun et al., 2012). By contrast, western boundary currents (WBCs) (e.g. the East Australian Current, EAC) are predicted to intensify (Fig. 1) (Chamberlain et al., 2012; Sun et al., 2012). Furthermore, circumpolar fronts in some parts of the Southern Ocean are inferred to have shifted southward in response to recent climate change (Kim & Orsi, 2014), and the Southern Ocean eddy field has intensified in recent decades (Hogg et al., 2015). These trends are likely to continue in response to a warming climate (Gersonde et al., 2005; Sokolov & Rintoul, 2009a).

Major surface ocean currents. Ocean currents discussed in this paper are labelled. Colour/tone denotes currents that are predicted to weaken (red/dashed black in greyscale) and strengthen (green/solid black in greyscale) under future climate change scenarios. ACC, Antarctic Circumpolar Current; EAC, East Australian Current; ITF, Indonesian Throughflow; NEC, North Equatorial Current; NECC, North Equatorial Counter Current; NQC, North Queensland Current; SEC, South Equatorial Current.
Known Shifts in Species Distributions via Changing Ocean Currents
A number of ocean currents in the Pacific Ocean have already displayed rapid warming and intensification as a result of climate change (Wu et al., 2012; Vergés et al., 2014). These changes are driving poleward range shifts, range expansions and/or contractions of many key marine species. Although some species are restricted by the availability of suitable habitat, larval tolerances and other factors, and are unable to achieve such range shifts (Wu et al., 2012; Morris, 2013; Sorte, 2013; Vergés et al., 2014). The Kuroshio and Tsushimo Currents, for example, are experiencing acceleration as a result of climate-driven changes in the wind stress over the North Pacific Ocean, a trend that is predicted to continue into the foreseeable future (Fig. 1) (Tanaka et al., 2012; Yatsu et al., 2013). This acceleration is increasing the transport of warm, tropical water and marine propagules to the north, driving the rapid expansion of tropical species (Yamano et al., 2011; Tanaka et al., 2012). Poleward range expansions along the coast of Japan have been documented for tropical corals since the 1930s (Yamano et al., 2011) and for macroalgae since the 1970s (Tanaka et al., 2012) (Fig. 2a). Such range expansions of tropical marine species map closely to changing thermal gradients arising from the strengthening Kuroshio and Tsushimo Currents, highlighting the profound effects of changing ocean current systems on the dispersal of tropical species into subtropical and temperate waters (Yamano et al., 2011; Tanaka et al., 2012).

(a) Range extension of marine species associated with intensification of the Kuroshio and Tsushima Currents. 1Historical and current distribution of coral Acropora solitaryensis adapted from Yamano et al. (2011), photo reproduced with permission of Dr Charlie Veron. 2Historical and current distribution of macroalgae Sargassum ilicifolium adapted from Tanaka et al. (2012), photo reproduced with permission of David Burdick. (b) Intensification of the East Australian Current (EAC) and expansion of associated species. 1Historical and current distribution of sea urchin Centrostephanus rodgersii adapted from Banks et al. (2011) and Ling et al. (2009), photo reproduced with permission of Peter Southwood. 2Historical and current distribution of barnacle Austromegabalanus nigrescens adapted from Pitt et al. (2010), photo reproduced with permission of Dr Elvira Poloczanska.
Similar changes are being observed along the east coast of Australia, where wind-driven strengthening of the South Pacific Gyre has enhanced the southward extension of the EAC (Fig. 1) (Cai, 2006; Duan et al., 2013; Cetina-Heredia et al., 2014). Long-term trends indicate that the intensification of the EAC extension has advected temperature and salinity properties c. 350 km southward over the past 60 years (Ridgway, 2007; Hill et al., 2011; Sun et al., 2012). Future projections indicate that the EAC system may expand further southward under climate change scenarios (Matear et al., 2013; Oliver & Holbrook, 2014). These increased southward flows of warm water carrying marine propagules have led to the establishment of new populations of both tropical and temperate marine species well beyond their known historical distributions (Pitt et al., 2010; Banks et al., 2011; Johnson et al., 2011). Sedentary marine animals, such as barnacles and echinoderms, have undergone southern range expansions of 20–710 km along the south-east coast of Australia (Fig. 2b) (Pitt et al., 2010; Poloczanska et al., 2011). Similarly, many tropical macroalgae are undergoing southward range expansions in south-east Australia, as well as corresponding southward shifts of temperate macroalgae species (Wernberg et al., 2011). Photographic evidence also suggests that tropical corals (e.g. Plesiastrea versipora; Johnson et al., 2011) and seagrasses (e.g. Halophila minor; Poloczanska et al., 2012) have undergone poleward range expansions of 300–650 km facilitated by the strengthening of the EAC.
Future Oceanographic Changes to Drive Species Distributions and Range Shifts
Current observations of the changing dispersal and distribution patterns of marine species highlight the potential for further effects in other regions that are projected to see major changes in oceanographic conditions. These early biogeographical impacts of climate change are predominately being observed in areas identified as climate change hotspots – regions that are warming at a rate faster than 90% of the world's oceans (Hobday & Pecl, 2014). The Southern Hemisphere encompasses many such areas, including parts of the Southern Ocean that have been identified as some of the most rapidly warming areas in the world (Bromwich et al., 2013; Hobday & Pecl, 2014). Projected alterations to the strength of major ocean currents, eddies and circumpolar fronts in the Southern Hemisphere are likely to profoundly affect species dispersal (Hobday & Lough, 2011; Kirtman et al., 2013). For instance, weakening of ocean currents has the potential to inhibit the dispersal of marine propagules, strengthen present-day dispersal barriers and/or create new barriers to dispersal (Keith et al., 2011). Strengthening ocean currents, however, are likely to transport water and associated propagules further and/or at a faster rate. Increased directional transport may increase the likelihood of propagules reaching areas at the extremes of their present-day dispersal ranges and, where suitable habitat is available, facilitate the colonization of new areas (Johnson et al., 2011). Such changes have the potential to increase species dispersal capacities by enabling propagules to traverse present-day dispersal barriers (Fig. 3). Importantly, these changes are likely to produce cross-latitudinal effects on species distribution patterns and connectivity via the dispersal stage, both within and among tropical, temperate and subantarctic regions.

Effects of predicted climate-driven circulation changes on species dispersal capacity. As evidence arises these effects will need to be interpreted in the light of additional biological and ecological factors (e.g. longevity of dispersive stage, reproductive output, temperature suitability), some of which may also change with climate change.
Weakening ocean currents will reduce southward dispersal in north-western Australasia
Tropical reef ecosystems in the central Indo-Pacific are internationally recognized as having some of the highest levels of marine biodiversity in the world (Carpenter et al., 2011). However, evidence suggests that many tropical species within the marine biodiversity hotspot of the Indo-Australian Archipelago (IAA) are highly vulnerable to climate change as they are already living at temperatures near their lethal threshold (Somero, 2010; Byrne, 2011; Nguyen et al., 2011). In order to maintain environmental equilibrium in a changing climate, many tropical marine species are predicted to undergo current-facilitated poleward range shifts (Cheung et al., 2009). However, oceanographic current systems around the IAA are predicted to decrease in intensity and transport volume under future climate scenarios, potentially inhibiting such shifts.
The oceanography of north-western Australasia is dominated by the ITF and the Leeuwin Current (Fig. 1), both of which are likely to be sensitive to climate change. The ITF is a unidirectional surface current that carries up to 19 sverdrups from the Pacific Ocean, through the IAA, into the Indian Ocean (Sprintall et al., 2009). From here the majority of ITF waters are advected westward with the South Equatorial Current (SEC) (Barber et al., 2006; Knittweis et al., 2009; Carpenter et al., 2011; Chan et al., 2014). The ITF also transmits thermocline and sea level abnormalities that induce pressure gradients responsible for driving the Leeuwin Current, which flows southward down the west coast of Australia (Coleman et al., 2011; Sun et al., 2012). Both the ITF and the Leeuwin Current are expected to undergo a significant decrease in flow due to weakening of the easterly trade winds across the Pacific Ocean (Sun et al., 2012). Sun et al. (2012) predicted that by 2060 the ITF will have weakened by 20% of its 1992 strength, while the Leeuwin Current is estimated to undergo a 15% decrease in its Australian winter transport volume from the 1990s to the 2060s. Such predictions are consistent with observed trends (Feng et al., 2004; Wainwright et al., 2008; Sun et al., 2012).
These current systems have been shown to play an important role in driving dispersal and population connectivity in many marine taxa. For example, the ITF has been found to drive the dispersal of the coral Heliofungia actiniformis (Knittweis et al., 2009) and the mantis shrimp Haptosquilla pulchella (Barber et al., 2006) within the Indonesian Archipelago, as well as widespread dispersal of tropical macroalgae such as Sargassum polycystum (Chan et al., 2013) and Sargassum aquifolium (Chan et al., 2014). Similarly, the Leeuwin Current facilitates southward movement of marine species (e.g. the coral Pocillopora damicornis; Thomas et al., 2014), transporting tropical species into the cooler waters of south-western Australia.
Due to the complex oceanography between Indonesia and north-western Australia, the potential oceanographic drivers of dispersal and associated dispersal patterns in this region are not yet clear (Underwood et al., 2013). Nonetheless, the large network of reefs and shoals between Indonesia and north-western Australia are thought to act as dispersal ‘stepping-stones’ for species with longer propagule pre-competency periods or those with long-term active dispersal capabilities (Underwood et al., 2013). However, the large expanse of ocean, and the complex circulation patterns in this region, may act as dispersal barriers for species with limited active dispersal capacities and short propagule pre-competency periods, such as corals. While there has been little empirical research into the dispersal of marine organisms between Indonesia and north-western Australia, dispersal models suggest that there is potential for coral larvae to be transported throughout southern Indonesia and the north-eastern Indian Ocean (Wood et al., 2014). However, phylogeographical studies have found Acropora tenuis populations on offshore and coastal north-west Australian reefs to be geographically isolated (Underwood, 2009). These findings suggest that connectivity between these coral populations has been absent over ecological time-scales and very rare over evolutionary time-scales, supporting the hypothesis that dispersal between marine populations in this area is limited (Underwood, 2009).
The ability of species to undergo range shifts is strongly dependent on the strength and nature of dispersal limitation (Underwood, 2009). Therefore, species and populations that already have limited dispersal opportunities will be more affected by weakening ocean currents than those which are less restricted by present-day dispersal barriers (Sorte, 2013). Evidence suggests that dispersal facilitated by the ITF and the Leeuwin Current predominantly occurs via a series of ‘stepping-stones’ from reef to reef over multiple generations (Knittweis et al., 2009; Underwood et al., 2013). As weaker currents take longer to transport propagules from an upstream to a downstream location, the weakening of the ITF and Leeuwin Current is likely to decrease the geographical distance that can be successfully traversed within the propagule pre-competency period (Underwood, 2009). Species with relatively long pre-competency periods and/or which actively select to be in fast-flowing water layers may still be able to undergo poleward range expansions under the influence of weaker current flows, albeit at a slower rate (Martínez-Quintana et al., 2015). However, species with shorter pre-competency periods or homing behaviours, upon which geographical distance already acts as a dispersal barrier, face a probable reduction in successful dispersal events between areas of available habitat, inhibiting poleward range shifts (Metaxas & Saunders, 2009; Underwood, 2009; Selkoe & Toonen, 2011; Sorte, 2013). In addition, the potential lack of dispersal pathways between Indonesia and north-western Australia suggests that southern Indonesian and offshore north-west Australian reef populations may have very few opportunities to undergo poleward range expansion at the rates required to track their suitable climatic envelopes under a changing climate (Underwood, 2009). For example, Sen Gupta et al. (2013) indicated that some parts of the north-west Australian coastline will experience rapid (10–20 km year−1) migration of SST isotherms under future climate change scenarios. Limited dispersal pathways in this region, combined with anticipated weakening of the ITF and the Leeuwin Current over the next 50 years (Sun et al., 2012) is likely to eliminate opportunities for species to move poleward at rates similar to, or faster than, this rapid migration of SST patterns. Instead, populations may become increasingly isolated, and vulnerable to the warming climate and disturbances (e.g. thermal bleaching and storms) (Underwood, 2009).
Changing Coral Sea currents will alter dispersal along the Great Barrier Reef
Oceanic circulation in the Coral Sea and the equatorial eastern Pacific Ocean is dominated by currents that respond to changes in the wind stress curl over the South Pacific Gyre (Choukroun et al., 2010; Bell et al., 2011; Luo & Rothstein, 2011). The SEC flows westward across the equatorial Pacific Ocean, through the Coral Sea to the east coast of Australia (Fig. 4a). Here the SEC bifurcates to form the northward flowing North Queensland Current (NQC) and the stronger southward flow that feeds into the EAC and continues down the east coast of Australia into the Tasman Sea (Choukroun et al., 2010; Bell et al., 2011; Luo & Rothstein, 2011; Ridgway & Hill, 2012). The SEC is predicted to significantly weaken in the future (Fig. 4a), and potentially undergo a southward shift (Cai et al., 2005; Bell et al., 2011; Luo & Rothstein, 2011). Yet the predicted response of the northern EAC to climate-driven changes in wind circulation patterns over the South Pacific Ocean remains ambiguous and is often resolution dependent. For example, Ridgway (2007) inferred from large-scale ocean circulation models that the flow of the EAC (north of c. 30° S) will decrease as the climate warms, compensating for the increase in EAC flow off south-west Australia. Conversely, a strengthening trend across the entire EAC system has been predicted using a downscaled, regional circulation model (Sun et al. 2012). Further clarification of these changes is required before their impacts on the dispersal of local marine organisms' can be determined.

(a) Representation of predicted changes in ocean current velocity at 30 m and sea surface temperature (SST) (°C) in the tropical Pacific Ocean by 2100, based on data from Luo & Rothstein (2011). Arrow size reflects the relative differences in ocean current strength and velocity, and colour/tone denotes present day (blue/dark grey in greyscale) and predicted (white/light grey in greyscale) current flow. EAC, East Australian Current; NEC, North Equatorial Current; NECC, North Equatorial Counter Current; NQC, North Queensland Current; SEC, South Equatorial Current. (b) Dispersal of important reef-building corals, such as Acropora spp. (photo reproduced with permission of David Excoffier) and Seriatopora hystrix (photo reproduced with permission of David Burdick), and coral-dependent species, such as coral trout, Plectropomus leopardus (photo reproduced with permission of Leonard Low) are likely to be affected by these changes in ocean current velocity.
Oceanographic and phylogeographical studies suggest that the bifurcation of the SEC, and the associated switch from northward to southward transport along the Great Barrier Reef (GBR), tends to act as a dispersal barrier between northern and southern populations of coral reef species (James et al., 2002; Bode et al., 2006; Ceccarelli, 2012; Mantovanelli & Heron, 2012). South of the bifurcation point, the EAC rapidly transports tropical waters and associated marine propagules southward down the GBR toward the subtropical reefs of northern New South Wales (Noreen et al., 2009; Malcolm et al., 2011; Condie et al., 2012; Mantovanelli & Heron, 2012). This predominantly unidirectional flow is thought to strongly influence larval delivery to southern GBR populations, and provide a key biophysical link between tropical and temperate waters (Noreen et al., 2009). Conversely, to the north of the bifurcation point the dominant northward flow of the NQC results in extremely slow and rare southward transport, inhibiting oceanic exchange between northern and southern regions of the GBR (Choukroun et al., 2010). Consistent with these expectations, northern and southern GBR coral populations, for example Acropora millepora (Choukroun et al., 2010; Van Oppen et al., 2011) and Seriatopora hystrix (Kininmonth et al., 2010; Wood et al., 2014), have been found to be genetically distinct and isolated from each other (Fig. 3b).
Much like on the west coast of Australia, reefs on the east coast also serve as dispersal ‘stepping-stones’ for coastal marine species. These stepping-stones will be critical for any southward range expansions in response to climate change (Noreen et al., 2009; Ceccarelli, 2012). However, as discussed above, the success of such stepping-stones in facilitating southward dispersal with changing ocean currents will be dependent upon the relationship between current strength, geographical distance, the duration of propagule pre-competency periods and propagule behaviour (Selkoe & Toonen, 2011). Strengthening of the northern EAC may enhance dispersal, potentially allowing species to overcome present-day dispersal barriers and undergo poleward range shifts. Conversely, weakening of the northern EAC and/or a change in the bifurcation of the SEC may limit dispersal, in particular of species with short pre-competency periods, homing behaviours and/or those that remain in slow-flowing water layers. Weakening of the EAC is likely to strengthen the already strong dispersal barriers in this region, further isolating populations to the north of the SEC bifurcation point from those to the south, and further inhibiting poleward range shifts (Choukroun et al., 2010; Kininmonth et al., 2010; Van Oppen et al., 2011; Wood et al., 2014). Which marine species and/or populations are most at risk of genetic isolation and inhibited range shifts due to their geographical location, nearby oceanographic features, propagule pre-competency periods and behaviour remains poorly understood. Additional research that sheds light on the relative dispersal abilities of organisms in these ecosystems will be fundamental to directing future conservation and management efforts.
Strengthening western boundary currents and shifts of circumpolar fronts may drive dispersal into the Southern Ocean
Long-distance poleward dispersal events in the temperate Southern Hemisphere are predominately driven by ocean currents that flow along the coastlines of the major continents (Vergés et al., 2014). These include the southward-flowing WBCs: the EAC, the Agulhas Current, and the Brazil Current (Fig. 1) (Vergés et al., 2014). These ocean currents have been undergoing systematic intensification and/or poleward expansion since the 1900s (Wu et al., 2012). The associated increase in the poleward transport of warm low-latitude waters has facilitated southern range shifts of many species in the temperate region. These shifts have driven rapid changes in the composition of temperate marine communities, as discussed by Wernberg et al. (2011) and Vergés et al. (2014). Such oceanographic changes and associated shifts in species ranges are predicted to continue into the foreseeable future (Wernberg et al., 2011; Wu et al., 2012; Vergés et al., 2014).
The potential for coastal species to undergo continued poleward dispersal beyond the end of the continents (Australia, South Africa and South America) is strongly limited (Wernberg et al., 2011). It is largely presumed that climate change will drive coastal temperate species toward, and beyond, the limits of available habitat at the southern margins of the continents, resulting in extinction events (Wernberg et al., 2011; Vergés et al., 2014). However, this may not always be the case. The WBCs inject subtropical and temperate waters into the Southern Ocean via eddies and interact with the Antarctic Circumpolar Current (ACC) (Clarke et al., 2005; Sokolov & Rintoul, 2009a, 2009b; Dencausse et al., 2011). Eddies within the ACC play a dominant role in driving the transport of marine propagules southward in the Southern Ocean (Fach & Klinck, 2006). Therefore, sustained movement of species beyond the southern tips of the southern continents, across large expanses of open ocean to sub-Antarctic islands, may be possible via long-distance dispersal facilitated by the eddy field (in the southward direction) and the ACC itself (eastward) (Clarke et al., 2005). Predicted poleward intensification of the WBCs is expected to increase the southward transport of temperate waters via these pathways; this in turn may enhance transport of temperate marine propagules into the Southern Ocean (Sen Gupta et al., 2009). Here they may be influenced by the ACC, which is thought to facilitate the dispersal of many marine species in the Southern Ocean (e.g. rafting macroalga, Durvillaea antarctica and associated organisms; Fraser et al., 2009; Nikula et al., 2010; Fraser et al., 2011).
The ability of temperate coastal marine taxa to successfully colonize the sub-Antarctic islands and/or Antarctic continent is largely unstudied (Frenot et al., 2005). However, Frenot et al. (2005) predicted that climate change will drive an increase in the introduction and colonization success of terrestrial organisms on both the sub-Antarctic islands and the Antarctic continent (Clarke et al., 2005; Frenot et al., 2005). While this colonization effect is likely to also apply to marine species (Frenot et al., 2005), additional factors, such as the presence and absence of sea ice and changes in salinity, need to be taken into account when assessing the colonization ability of temperate coastal marine species in Antarctic waters (Chown et al., 2012).
The success of such dispersal events is strongly dependent on the availability of climatically suitable habitat to the south (Clarke et al., 2005). In addition to the significant warming events being experienced by the majority of the world's oceans, the temperature of the Southern Ocean is also influenced by circumpolar ocean fronts (Fig. 5). These ocean fronts mark important temperature boundaries. For example, the Sub-Tropical Front (STF) separates northern, warm subtropical waters from the relatively cool southern, sub-Antarctic waters, and the Antarctic Polar Front (APF) separates sub-Antarctic from Antarctic water (Dong et al., 2006; Hamilton, 2006; Fraser et al., 2012). The strong temperature gradient across the fronts is thought to act as a biological barrier to the dispersal and distribution of a range of marine taxa, including zooplankton, crustaceans, echinoderms and macroalgae (Fraser et al., 2012; Janosik et al., 2011; Garden et al., 2014; Shetye et al., 2014).

Approximate positions of the major ocean fronts in the Southern Ocean and major ocean currents which drive dispersal into and within the Southern Ocean, adapted from Gersonde et al. (2005). ACC, Antarctic Circumpolar Current; APF, Antarctic Polar Front; EAC, East Australian Current; SACCF, Southern Antarctic Circumpolar Current Front; SAF, Subantarctic Front; STF, Subtropical Front. Insets show the southward (poleward) shifts of the SAF, APF, and SACCF between 1992 (solid lines) and 2007 (dashed lines) in the Australian–Antarctic Basin and the Drake Passage, adapted from Sokolov & Rintoul (2009a).
Circumpolar fronts in the Southern Ocean have migrated southward over the last two decades (Gersonde et al., 2005; Kim & Orsi, 2014). For example, Sokolov & Rintoul (2009a) showed that the Antarctic circumpolar fronts have shifted south by an average of 60 km between 1992 and 2007, due to changes in sea level and ocean circulation (Fig. 5). The migration is non-uniform, only occurring in the Atlantic and Indian sectors of the Southern Ocean, implying that migration is driven largely by southward extensions of WBCs (Kim & Orsi, 2014) and not by climate-induced latitudinal shifts in the Southern Hemisphere westerly winds.
Studies have also suggested that fronts have varied during periods of past climate change; during historical glaciation events the STF had a relatively northern position while during warmer interglacial periods the front shifted southward (Trend-Staid & Prell, 2002; Sikes et al., 2009; Sokolov & Rintoul, 2009a; Kohfeld et al., 2013; Shetye et al., 2014). These past changes suggest that under predicted future climatic conditions the circumpolar fronts in the Southern Ocean will continue to shift southward in some areas. Furthermore, recent evidence supports the notion that Southern Ocean eddies have increased over the last two decades in response to an intensification of the Southern Hemisphere westerlies (Hogg et al., 2015). A stronger eddy field is likely to strengthen southward heat transport and southward dispersal in the ACC belt.
The combination of predicted intensification and extension of the WBCs, southward shifts in some ACC fronts and an enhanced eddy field may facilitate more successful poleward dispersal events in the Southern Ocean (Vergés et al., 2014). This process is particularly relevant to species with long dispersal phases, such as rafting macroalgae and associated organisms. Furthermore, landmasses that lie close to the boundaries (such as Gough Island, just south of the STF, or Bouvet Island, just south of the APF) could switch climates from subantarctic to temperate, or antarctic to subantarctic, with drastic consequences for community composition (Smith, 2002).
Near-shore circulation patterns will add a further layer of complexity to changing dispersal patterns
Fine- and meso-scale near-shore circulation patterns play important roles in the dispersal of coastal marine species (Siegel et al., 2008; Andutta et al., 2012; Chamberlain et al., 2012). In coastal waters the interaction of water flow, including tidal currents, with the seafloor and nearshore topographic features, such as islands, bays and reefs, induce friction and create turbulence. Combined with temporal variation in flow patterns, this can result in complex nearshore circulation (Limouzy-Paris et al., 1997; Cetina-Heredia & Connolly, 2011; Suthers et al., 2011; Andutta et al., 2012). These flows can dictate propagule trajectories by altering propagule retention and flushing times (Limouzy-Paris et al., 1997; Cetina-Heredia & Connolly, 2011; Andutta et al., 2012; Vaz et al., 2013). For example, eddies can promote propagule retention by recirculating and trapping propagules in their centre, thus restricting dispersal, even in the presence of strong directional offshore currents (Limouzy-Paris et al., 1997; Andutta et al., 2012; Vaz et al., 2013; Cetina-Heredia et al., 2015). Propagules can also be transported along the edge of the eddy, where transport velocities are greater (Vaz et al., 2013; Cetina-Heredia et al., 2015). This increased transport can provide faster advection, facilitating along-shore dispersal and the advection of propagules away from the coast, potentially aiding the dispersal of coastal species by offshore currents (Limouzy-Paris et al., 1997; Andutta et al., 2012; Vaz et al., 2013; Cetina-Heredia et al., 2015. The levels of propagule retention relative to those of flushing and dispersal are dependent on circulation regime, bathymetry and topography, and vary with location and time (Andutta et al., 2012; Cetina-Heredia et al., 2015). In general, it is thought that smaller eddies, in particular those that are short-lived (i.e. lasting a few hours to days at the most), are unlikely to affect propagule dispersal across time-scales relevant to propagule pre-competency periods, while larger and/or longer-lived eddies are more likely to affect propagule dispersal (Andutta et al., 2012). However, some studies suggest that eddies with smaller spatial scales do affect propagule transport and no direct correlations between the presence and/or size of an eddy and retention and dispersal rates have been found (Vaz et al., 2013; Smith et al., 2015).
While very few existing simulation-based climate change projections fully resolve fine- and meso-scale near-shore circulation patterns, downscaled models are beginning to be made (Suthers et al., 2011; Matear et al., 2013). These models predict that eddy activity and patterns will change under future climate models, a factor that is currently absent from large-scale, low-resolution global climate models (Chamberlain et al., 2012; Matear et al., 2013). For example, Matear et al. (2013) showed that with increased EAC transport under future climate change scenarios, as discussed above, eddy activity to the south of the EAC separation point (c. 32° S) and into the Tasman Sea will increase. Given that particle trajectories in this region are often determined by the presence or absence of eddies, such changes may have considerable effects on species dispersal and distribution patterns (Suthers et al., 2011). When this model was applied to other WBC regions, a similar increase in eddy activity could be predicted in all current systems, suggesting that similar impacts on propagule dispersal and marine species distributions may occur around the globe (Matear et al., 2013).
Due to the complex and often opposing impacts of eddies on particle transport patterns, a general prediction regarding the impact of increased eddy activity under climate change scenarios on species dispersal capacity cannot be made (Fig. 3) (Vaz et al., 2013). However, differences in the projected circulation patterns between downscaled models and those of low-resolution global models highlight the importance of considering fine- and meso-scale, near-shore circulation patterns under future climatic conditions (Munday et al., 2009; Chamberlain et al., 2012; Qin et al., 2014). Moreover, we also need the capacity to model how more extreme and/or frequent stochastic coastal events can affect coastal dispersal pathways, such as offshore plumes of freshwater and terrigenous sediments ejected from flooded rivers (Gibson et al., 2003). While running a global climate model with eddy resolution and/or other near-shore hydrodynamic events would be ideal, incorporating such fine-scale resolution across a global scale presents a substantial logistical challenge (Cetina-Heredia & Connolly, 2011). A more realistic prospect is the continued development of downscaled climate change models that focus on a relatively small area of a single current system (Cetina-Heredia & Connolly, 2011). However, because large-scale range changes can require long-distance dispersal, small-scale models that account for eddies will need to be integrated into larger-scale models to be useful in predicting dispersal and species distribution patterns under future climate change (Cetina-Heredia & Connolly, 2011; Chamberlain et al., 2012).
Conclusions and Future Research Priorities
Using currently available climate change models, in combination with relevant biogeographical and biological data, we predict that climate-driven oceanographic changes will either enhance or reduce species dispersal by strengthening, weakening and altering the structure of oceanic dispersal pathways. Changing oceanic and coastal circulation patterns are already altering the dispersal pathways of many marine taxa, driving major changes in marine ecosystems (Pitt et al., 2010; Banks et al., 2011; Poloczanska et al., 2012; Tanaka et al., 2012; Cetina-Heredia et al., 2015). Given projections for near-future climate-driven oceanographic changes (Hobday & Lough, 2011; Kirtman et al., 2013; Hobday & Pecl, 2014), our review indicates that changes in marine population connectivity and range shifts will continue into the future, probably at accelerated rates. Given the capacity for some marine taxa to engage in active dispersal (especially during late stages of planktonic development; Stobutzki & Bellwood, 1997; Fisher et al., 2000), it is likely that a complex mix of oceanographic, geographical and biological factors will influence these changes in a species- and location-specific manner (O'Connor et al., 2007; Kleypas, 2015). We infer that the length of propagule pre-competency periods, propagule behaviour and the geographical distance between areas of suitable habitat will be critical factors in determining the capacity for marine species to make poleward range shifts. By combining oceanographic, biogeographical and biological data we can enhance our predictions and risk assessments of marine species range shifts in a more ecologically relevant manner. This process is critical for identifying species, populations and locations most at risk of experiencing limited poleward dispersal, genetic isolation, extinction events and/or shifts in community composition. An in-depth understanding of these factors and how they interact should be a major research priority.
Challenges
The usefulness of future predictions is contingent on collecting more accurate and appropriately scaled empirical data on two main factors: the dispersal biology of key marine taxa, and ocean currents. Due to a paucity of data on how the dispersal capability of species changes with respect to environmental conditions, in many cases it is not yet possible to make specific predictions about changes to particular species and/or marine communities under future climate scenarios (Metaxas & Saunders, 2009). For the latter, much more work is needed to improve predictions for other major surface ocean currents that influence marine dispersal (e.g. Eastern Boundary Currents; Hsieh et al., 2009; Wang et al., 2010).
Improved near-shore circulation data and models are also needed to address critical questions relating to larval transport, retention and settlement into adult habitats. Many regional and small-scale ocean currents are poorly represented in global climate change models, largely due to rather coarse spatial and temporal resolution, as well as a lack of high-resolution bathymetric data required to resolve tidal and current forcing patterns (Smith & Sandwell, 1997; Becker et al., 2009; Kirtman et al., 2013). Downscaled regional ocean circulation climate projections that capture finer-scale features, the potential for important stochastic events (e.g. ejecta from flooded rivers) and more realistic estimates of transport volume are beginning to address this gap (e.g. Luo & Rothstein, 2011; Echevin et al., 2012; Sun et al., 2012; Matear et al., 2013).
Remote sensing plays a critical role in providing data on ocean currents and validating oceanographic models, but is not without its own challenges. A variety of data sources from land-based, airborne and satellite platforms can be used to track water motion, yet there will always be a necessary trade-off between spatial resolution, spatial extent and temporal resolution. For example, near real-time products, such as those delivered by the National Center for Environmental Prediction (NCEP) Climate Forecast System Reanalysis (CFSR) (Saha et al., 2014), offer the potential to monitor changes in surface transport patterns. These products are ideal for global-scale monitoring, investigating large- to mesoscale currents that exhibit displacements (Klemas, 2012). However, their spatial resolution renders them inappropriate for monitoring fine-scale and near-shore current patterns. Unmanned airborne systems may offer the potential for finer-spatial-scale observations, but are severely limited by their range and consequent imaging extent (Watts et al., 2012). Therefore satellite observations remain the most viable option for remotely tracking water motion and monitoring ocean current patterns, but the challenges of spatial and temporal resolution need to be reconciled.
There remain fundamental gaps in our understanding of depth-specific changes in major ocean currents, and areas of highly complex oceanography (e.g. between Indonesia and north-western Australia), that are often important for shaping hotspots of marine diversity and endemicity (Roberts et al., 2002; Fisher et al., 2011) Likewise, many subsurface areas of marine habitat are poorly mapped and described, providing little information about their function as dispersal stepping-stones or receptor environments for climate change-driven colonization. As such, future interdisciplinary marine ecology and oceanography research should seek to address the dispersal abilities of ecologically key species, specific biogeographical regions and small-scale flows particularly relevant to the dispersal of marine species.
Dispersal and connectivity are critical to the sustainable management of our marine natural resources, for both fisheries (e.g. recruitment) and conservation (e.g. marine protected area networks) (Bell et al., 2011; Underwood et al., 2013; Gerber et al., 2014). Adaptive management protocols will need to take into account shifting dispersal pathways arising from oceanographic changes, both within and beyond their region of interest. For instance, with improved near-shore oceanographic models we may identify marine species and areas at risk of reduced/increased propagule delivery, and develop adaptive spatial (e.g. protection zoning, adjusted catch limits) and/or temporal (e.g. seasonal spawning closures to maximize propagule production) approaches. Given the dearth of empirical information, it is likely we will need to prioritize research on key species (e.g. habitat facilitators, apex predators, important fishery targets) and regions in order to deliver this increased adaptive management capacity to meet the challenges imposed by global climate change.
Acknowledgements
LJW was supported by a scholarship and research funding from the North Australian Marine Research Alliance (NAMRA).
References
Biosketch
Laura Wilson is a PhD scholar at the Fenner School of Environment of Society, Australian National University. Her research focuses on the population genetics, dispersal and connectivity patterns, and phylogeography of seaweed species in north-western Australasia. Laura led the development and writing of the literature review. All authors contributed to the development of ideas and concepts.