Keywords

1 Introduction

Species distributions are dynamic, continuously shifting in responses to changes in biological and environmental drivers. In the earlier history of the Earth, large-scale geological events and long-term climate fluctuations, such as continental drift or warming and cooling associated with planetary cycles, were the predominant drivers of changes to species’ distributions (Wiens and Donoghue 2005). However, over the past millennium humans have increasingly modified the biological and physical properties of the planet (Worm et al. 2006), and we have now entered the Anthropocene , an era where the human influence on the global Earth system rivals or exceeds natural processes (Karl and Trenberth 2003), speeding up important drivers of species distributions influencing the biogeography of organisms across ecosystems . As a consequence, recent changes in the distribution of many marine taxa have been documented on all continents (Perry et al. 2005; Williams and Smith 2007; Sorte et al. 2010; Wernberg et al. 2011a; Poloczanska et al. 2013).

Seaweeds are dominant organisms on many intertidal and shallow subtidal reefs, where their species-specific distributions often shape local reef communities (Wernberg et al. 2003; Buschbaum et al. 2006; Ingólfsson 2008; Tuya et al. 2009). Although the local effects of biotic interactions can generate continental-scale patterns of species associations (Wootton 2001; Irving and Connell 2006), global biodiversity patterns are not explained by biotic interactions alone but are a consequence of both the biotic and abiotic environments (Lüning 1985; Harley et al. 2006; Tittensor et al. 2010). Two mechanisms have been particularly prevalent in driving recent changes in seaweed distributions: species introductions through the direct relocation of species (transported, deliberate or not, by various vectors) and responses to global climate change (Williams and Smith 2007; Wernberg et al. 2011a; Sorte et al. 2013).

Changes in seaweed distributions include both range extensions , where species colonize new, usually adjacent habitats, and range contractions , where species lose previously occupied areas, going locally extinct at the margins of their distribution range (Fig. 3.1) (Wernberg et al. 2011a; Bartsch et al. 2012; Bates et al. 2014). For species introductions and climate impacts, range shifts are underpinned by different mechanisms involving dispersal and recruitment (introductions, climate expansion) and attrition and mortality (climate contraction) (Bates et al. 2014). Moreover, whereas range expansion only requires the successful establishment of one or a few individuals in a new location, local extinction and range contraction requires the demise of all individuals and is often preceded by periods of declining abundance and failed recruitment while adult individuals persist in the unfavourable area (Hampe and Petit 2005; Bates et al. 2014). Conversely, environmental conditions are generally not limiting the expansion of introduced species following primary introduction , whereas climate-driven responses track shifts in the climate envelope (Fig. 3.1) (Pinsky et al. 2013; Sunday et al. 2015). Consequently, even if priority effects and other biological (competition, predation) processes can work against the expansion process (Waters et al. 2013), seaweed range shifts are expected to be faster for expansions than contractions and faster for introductions than climate responses (Sorte et al. 2010). The effects on the respective habitats and communities should, however, be of the same magnitude and direction (Sorte et al. 2010).

Fig. 3.1
figure 1

Reconfiguration of seaweed range boundaries takes place as one of three general processes. Either species expand their boundaries in a new area following initial primary introduction to a site where climate conditions are not immediately limiting. Alternatively, native species can expand into previously unoccupied areas, tracking their climate envelope as changing conditions make these suitable. Similarly, species can retract from occupied areas as changing climate makes these unsuitable. The processes, and underlying physical and biological mechanisms, differ between these processes, with expansions driven by dispersal and recruitment dynamics and contractions by performance and mortality (Bates et al. 2014)

Full size image

Here, we first provide a brief overview of natural and anthropogenic factors that shape the biogeography of seaweeds. We then provide a quantitative synthesis of how fast humans are affecting seaweed distributions through an analysis of the speed of reported human-mediated changes in seaweed range boundaries. We also review selected case studies of seaweed range shifts for both native species that have changed their ranges in response to changing environmental conditions, and introduced species , spreading in their new environment. Finally, we discuss the challenges of identifying range shifts and the necessity for monitoring distributions to detect seaweed range shifts.

2 Drivers of Seaweed Biogeography

Seaweed biogeographers traditionally group the world’s oceans into seven broad regions: the Arctic and Antarctic Polar regions, the cold- and warm-temperate regions of both hemispheres, and the tropical regions of the Atlantic and Indo-Pacific (Lüning 1985; Bartsch et al. 2012). The boundaries between these biogeographic regions are associated with large changes in species composition, maintained by species temperature tolerances (Van den Hoek 1982), natural barriers (Cowman and Bellwood 2013) and species dispersal limits (Wiens 2011). Humans are now assisting seaweeds and other organisms to overcome these geographic boundaries, which previously limited distributions.

2.1 Temperature

Seaweeds are confined to the photic zone, where temperature patterns are reasonably well understood, allowing species distributions to be compared to oceanographic patterns (Adey and Steneck 2001). Distribution limits of individual seaweed species typically follow major marine isotherms (Van den Hoek 1982; Lüning 1985), giving rise to strong relationships with the temperature signatures of major ocean currents (Wernberg et al. 2013b).

For seaweeds, these patterns are a product of two key types of temperature boundaries : lethal boundaries , determined by a species’ capacity to survive during their unfavourable season; and growth and reproduction boundaries , determined by a species’ ability to grow and reproduce during its favourable season (Van den Hoek 1982; Lüning 1985). Seaweeds can be abundant in areas within both boundaries that are within dispersal ranges of the species. However, as thermal windows have changed over geological time (e.g. following ice age cycles) , they have biogeographic boundaries and seaweed distributions (Adey and Steneck 2001).

2.2 Barriers

If the oceans were a continuous open system, then most species should exhibit cosmopolitan distributions within their respective thermal windows (Myers 1997; Gaylord and Gaines 2000). However, barriers limit dispersal, which leads to discontinuities in species distribution (Myers 1997). These barriers can be ‘hard’ or ‘soft’, depending on their underlying mechanism (Luiz et al. 2012; Cowman and Bellwood 2013).

Hard barriers are physical obstacles such as land masses separating marine systems. For example, the final closure of the Tethys seaway around 12 Mya at the northern tip of the Red Sea created a physical barrier which cut off the low-latitude connection between the Indian and Atlantic Oceans (Cowman and Bellwood 2013).

In contrast, soft barriers refer to hydrographical features that disrupt connectivity . Large stretches without suitable substratum, such as deep oceanic basins (Lessios et al. 1998) or extensive beaches (Hidas et al. 2007), can limit the distribution of species with limited dispersal capacity. The greatest example could be the Eastern Pacific Barrier, a 5400-km stretch of deep open ocean between the central and eastern Pacific, likely in existence since the Cenozoic (Grigg and Hey 1992), where only a few marine species are represented on both sides (Lessios et al. 1998). Nearshore gradients in ocean properties, such as the direction and strength of ocean currents , differences in salinity and/or temperature as a result of currents or local upwelling (Luiz et al. 2012; Cowman and Bellwood 2013), can also function as barriers to dispersal. Therefore, many seaweeds show distribution limits concentrated at particular shorelines, often in locations where major currents collide (Gaylord and Gaines 2000; Schils and Wilson 2006; Waters 2008).

Barriers are, however, not permanent especially over geological time scales. Changes in ice cover and sea levels (glaciation, deglaciation, retreating ice caps, historical sea-level alterations) have led to significant alterations in seaweed biogeography . The Baltic Sea , for example, was entirely covered by glaciers during the last ice age, and all present-day seaweeds in the Baltic Sea have colonized following the opening of the Danish Straits about 8000 years ago (Björck 1995). Similarly, recent glacial retreat in the South Shetland Islands has enabled seaweed expansion into newly available habitat in Antarctica (Quartino et al. 2013). Over several glacial cycles , reduced sea levels exposed the Bassian land bridge, a historical barrier between Tasmania and mainland Australia, interrupting connectivity and colonization for several taxa for prolonged periods of time (Burridge et al. 2004; Waters 2008; York et al. 2008). Also, islands emerging due to volcanic activity (e.g. new island formation in Japans Ogasawara Island chain in 2013) create new space for seaweed colonization and can function as stepping stones for long-range dispersers to overcome deep oceanic stretches to reach distant areas (Nogales et al. 2012). Also, dispersal across large sandy stretches can be facilitated by small rocky platforms functioning as intermediate habitats to facilitate dispersal over the barrier (Dethier et al. 2003; Hidas et al. 2007; Mattio et al. 2015).

2.3 Dispersal

Dispersal is a critical process which allows seaweeds to extend their geographical distribution. Seaweeds employ a broad range of dispersal strategies with some species adapted to short-distance dispersal , typically settling close to their parental populations, and others adapted to long-distance dispersal , typically favouring rapid colonization of new habitats (Santelices 1990).

Most seaweeds disperse by small, largely immotile propagules (zoospores or zygotes) that are transported by waves and currents (Norton 1992; Gaylord et al. 2002). The buoyancy of the propagules, storage components, metabolic rates, and the strength and direction of current flow determine how far these microscopic propagules can disperse (Gaylord et al. 2002), before they have to settle onto hard substrata in the photic zone. In addition to microscopic propagules, seaweeds can also disperse as floating fronds, where a parental thallus is dislodged (breakage of stipes, thallus fragmentation, storms, etc.) and transported by winds and currents (Rothäusler et al. 2012). Many positively buoyant seaweeds can survive, float and disperse for prolonged periods of time (Van den Hoek 1987; Norton 1992; Hobday 2000a; Rothäusler et al. 2012). This dispersal mechanism is particularly efficient for dioecious species that do not rely on concurrent dispersal of male and female thalli (e.g. Sargassum muticum) as these species can establish entire new populations from single floating reproductive fronds. Large drifting seaweeds can also function as a raft for smaller negatively buoyant animals and seaweeds (Van den Hoek 1987; Hobday 2000b; Hinojosa et al. 2010; Fraser et al. 2011; Gillespie et al. 2012; Rothäusler et al. 2012; Fraser and Waters 2013). Floating seaweeds can therefore facilitate the colonization of new habitats on remote shores, sometimes by crossing large ocean basins (Fraser et al. 2011; Rothäusler et al. 2012). Dispersal thus depends on both intrinsic seaweed traits such as buoyancy and propagule characteristics, as well as on external factors such as current speed and direction, and environmental conditions that enable survival, settlement and growth (Norton 1992; Hinojosa et al. 2010).

2.4 Species Introductions (Human-Assisted Dispersal)

A characteristic feature of the past millennium has been an explosion in travel for trade and colonization, over increasing distances and at decreasing travel times. Through the process of human-assisted dispersal, non-native seaweeds have spread (intentional or not) to habitats far away from their origins (also see Chapter by Neiva et al. (2016) in this volume).

Introduced seaweeds are species that have been relocated beyond their native range by human activities and have become successfully established at a new location. The introduction of seaweeds is a stepwise process, starting with transport and initial arrival through a vector (primary introduction, Fig. 3.1), which is followed by initial survival, establishment and finally successful reproduction and spread (expansion, introduction , Fig. 3.1) to nearby locales (Sakai et al. 2001; Bates et al. 2014). The main vectors responsible for seaweed introductions include hull fouling and aquaculture, but ballast water , breakdown of natural barriers (the Suez canal in particular) and the aquarium trade have also transported seaweed around the world (Williams and Smith 2007) (see Chap. by Neiva et al. (2016) in this volume). Of all introductions , only a small subset establishes permanent populations in their new habitats. It has recently been estimated that at least 346 seaweed taxa have been introduced to, and successfully established populations in, new regions worldwide (many of these taxa having invaded multiple biogeographical regions), breaking down barriers evolved over millennia (see Chap. by Neiva et al. (2016) in this volume). Many of these taxa have also become invasive with significant effects on native species, biodiversity and ecosystem dynamics (Williams and Smith 2007; Thomsen et al. 2009, 2014) (see Chap. by Neiva et al. (2016) in this volume).

For successfully introduced seaweeds, it is implicitly assumed that climate is not the primary limiting constraint on their distribution (or they would not have successfully become established) and that secondary expansion can proceed largely as fast as dispersal allows. Expansion of introduced seaweeds should therefore be rapid relative to climate-induced range changes.

2.5 Environmental Change (Human-Induced Climate Change)

Climate and temperature, in particular, play pivotal roles in controlling the global biogeography of seaweeds (Sect. 3.2.1) (Lüning 1985). Consequently, changes in temperature, as for example those associated with anthropogenic greenhouse gas emissions, also alter the distribution of seaweeds (Zachos et al. 2008; Wernberg et al. 2011b; Harley et al. 2012).

On average, anthropogenic emissions of greenhouse gases have caused a decrease in ocean surface seawater pH of ~0.1 since the beginning of the industrial era (IPCC 2014) and ocean warming by ca. 1 °C over the past 4–5 decades, although with substantial local variation (Burrows et al. 2011). While a few regions have cooled due to increased upwelling (e.g., causing kelps to expand their ranges Bolton et al. 2012), most regions have warmed (Lima and Wethey 2012; Hobday and Pecl 2013). Importantly, climate change not only causes gradual and slow increases in temperatures and pH, but also in the frequency and intensity of extreme events (Coumou and Rahmstorf 2012; IPCC 2012). Seaweeds respond to these environmental changes through physiological and morphological acclimations (reversible, phenotypic changes on short timescales), adaptation (irreversible, genotypic changes on medium to long times scales), or migration (changes in distribution on medium timescales) (Bartsch et al. 2012).

Overall, climate change has altered local marine environments leading to changes in distribution and diversity of seaweed communities from local to global scales (Wernberg et al. 2011a; Tanaka et al. 2012; Duarte et al. 2013). As seaweed expansions and contractions follow the external driver of changes in the physical environment, changes in species distributions are expected to be slow relative to introductions (Sorte et al. 2010). Moreover, climate-induced contractions will, in contrast to expansions, typically manifest as repeated recruitment failures and subsequent demise of long-lived populations (Hampe and Petit 2005; Bates et al. 2014). Contractions are therefore expected to be slower than expansions. One obvious exception is the rapid response to extreme events , which can alter local ecosystem structure and functioning abruptly (Wernberg et al. 2013b) and lead to rapid changes in seaweed distributions (Smale and Wernberg 2013).

3 Speed of Range Shifts in Seaweeds

In order to determine the rate at which humans have been modifying biogeographic boundaries of seaweeds, we undertook a meta-analysis of the rate of change in distribution limits for recently recorded range shifts for native and introduced seaweeds (range-shift speed). Data bases were searched using key words like ‘climate change’, ‘warming’, ‘extreme events’, ‘temperature anomaly’ ‘heatwaves’ , ‘introduced seaweeds’, ‘successful invaders’, ‘shift in distribution’, ‘shift rates’, ‘spread rates’, ‘range shift ’, ‘range expansion’ and ‘range contraction ’. We also backtracked references from relevant reviews and meta-analytical papers (Sorte et al. 2010; Poloczanska et al. 2013; Bates et al. 2014). We included studies that showed data for the directions, distances and time windows of seaweed range shifts, allowing us to calculate annual spread rates. Literature reporting changes in abundance without changes in location were excluded from the dataset, as were studies that did not report a range shift per se. Where rates were not reported directly, but identifiable locations given, rates were calculated (using the Google Earth distance calculator). For introduced species , we did not consider the initial primary introduction distance, only expansion from site of primary introduction into its new environment. Where time was reported as an interval, the midpoint was used. These strict data inclusion criteria limited the number of range-shifting seaweeds included in our analysis, which therefore represents a constrained view of seaweed range shifts.

Range-shift speeds were compared between three drivers of change (cf. Sect. 3.2.5, Fig. 3.1). (I) range expansions following introductions, (II) expansions caused by climate change (typically warm-water species) and (III) range contractions caused by climate change (typically cool-water species). More specifically, we tested (a) whether expansions generally are faster than contractions, and (b) whether introductions are faster than climate-driven changes. Tests were made with permutation-based analysis of variance (Logx + 1 transformed range-shift speeds, 9999 permutations of residuals), followed by two a priori defined planned contrasts (expansion vs. contraction and introduction vs. climate). These analyses did not include range shifts caused by primary introductions or in response to discrete extreme events . These range shifts were excluded due to their artificial and stochastic nature, respectively.

Our literature search returned 71 individual estimates of seaweed range-shift speed (Appendix). In general, of the studies where range-shift speeds could be assessed, expansions following seaweed introductions were detected over larger distances and shorter time periods than expansions and contractions due to climate change (Table 3.1). Studies returning range-shift speeds were reported from all continents except Antarctica (Fig. 3.2), although we found strong geographical biases in what types of range shifts had been recorded. For example, no studies with sufficient information to calculate range-shift speed were reported for climate-driven range expansions or contractions in North and South America, nor expansions and introductions in Australia. Europe had the greatest concentration of range shifts reported with sufficient information for all three categories (Fig. 3.2).

Table 3.1 Summary statistics for human-induced seaweed range shifts
Full size table
Fig. 3.2
figure 2

Geographical location of studies reporting range shifts in seaweeds with sufficient information to calculate range-shift speed

Full size image

The five areas where climate-induced range-shift speeds are available (SE and SW Australia, Japan, South Africa and SW Europe) are well-known ‘temperature hotspots’ where the rate of ocean warming since 1950 has been in the top 10 % of observations globally (Hobday and Pecl 2013). Interestingly, the range shift reported in the South African warming hotspot was a range expansion of a cool-water kelp (E. maxima). However, this expansion was attributed to increased nearshore upwelling (Bolton et al. 2012), a consistent but small-scale phenomenon not captured by global satellite data (Smit et al. 2013). This example highlights that predictions about global range shifts from large-scale satellite images may not capture local distribution patterns, particular where upwelling occurs.

Range-shift speeds were determined for 40 taxa (n = 13, 22, 9 for each of the categories, respectively), with some genera represented by more than one species (Table 3.1). As might be expected, there was little overlap in taxa between categories but one species (F. serratus) was represented in all three range-shift categories (Appendix) and another two species (F. vesiculosus and H. elongata) in both climate change responses (Appendix). These responses highlight the context dependency of range shifts , with the direction of shift presumably determined by a combination of ecological interactions opening/closing opportunities for change as well as the relative position within the species’ thermal envelope.

We found support for our range-shift hypotheses (Fig. 3.3). The speed of distributional changes in seaweed range limits differed significantly between the different types of shifts (Fig. 3.3, medians = 35.0, 4.5, and 4.0 km year−1, respectively, P = 0.0001, MS 2,66 = 26.6, pseudo-F = 31.6). Expansions were significantly faster than contractions (P = 0.011, MS 1,2 = 10.1, pseudo-F = 6.9) and climate-induced shifts were significantly slower than those caused by species introductions (P = 0.0001, MS 1,2 = 53.1, pseudo-F = 64.1). However, this test did not include range contractions following a discrete extreme event —a large-scale marine heat wave—where two species of seaweeds were found to contract their ranges by ~100 km in one year (Fig. 3.3, Appendix). These shifts remain some of the fastest observed range changes for any seaweed. When the two heatwave -driven contractions were included in the analysis, the difference between speed of climate-driven contractions and expansions disappeared (P = 0.112, MS 1,2 = 4.0, pseudo-F = 2.5) but the difference in speed between introduction and climate-driven range shifts remained (P = 0.0001, MS 1,2 = 45.9, pseudo-F = 44.8).

Fig. 3.3
figure 3

Speed of range shifts in seaweeds. Arrows highlight direction (upwards = expansion; downwards = contraction) and underlying cause (green = after successful introduction ; red = climate change) of range shifts. Red stars indicate shifts caused by an extreme marine heat wave. The very discrete nature of these shifts differs fundamentally from other reported shifts and consequently these have not been included in the analyses of rates (or the box in the plot)

Full size image

4 Case Studies of Seaweed Range Shifts and Ecological Implications

Many range-shifting seaweeds (cf. Table 3.1) are prominent members of their respective communities, where their addition or deletion is likely to have dramatic impacts on ecosystem structure and functioning (Williams and Smith 2007; Thomsen et al. 2010; Wernberg et al. 2013a; Bennett et al. 2015b) (see Chap. by Neiva et al. (2016) in this volume). The scale and nature of these ecological implications depends on the attributes of the shifting species and the impacted habitat (Thomsen et al. 2011). Here, we provide a range of examples of seaweed range shifts and their ecological implications. We also provide an example of a seaweed declining in abundance, a precursor to range contraction (Bates et al. 2014).

4.1 Range Contractions (Native Species)

In 2011 an unprecedented marine heat wave off the coast of Western Australia caused dramatic canopy loss of dominant seaweeds, including a 100 km southward range contraction of one of the main canopy-forming species, the fucoid Scytothalia dorycarpa (Fig. 3.4). During the heat wave, temperatures exceeded the physiological tolerance of S. dorycarpa for many weeks (Smale and Wernberg 2013). The contraction of S. dorycarpa co-occurred with a significant decrease in the densities of the kelp Ecklonia radiata (Wernberg et al. 2013a) and indirectly resulted in changes of the understorey community structure. The net effect was a shift from a dense three-dimensional canopy habitat to reefs with large open patches dominated by much smaller turf forming seaweeds among patches of E. radiata (Smale and Wernberg 2013) (Fig. 3.4). Concurrently, with the loss of seaweeds, the biomass and diversity of tropical herbivores increased, facilitating the new canopy-free state by suppressing seaweed reestablishment (Bennett et al. 2015b). The combined effects of the range contraction of S. dorycarpa and overall loss of seaweed canopies ultimately resulted in habitat and food loss (Wernberg et al. 2013a; Smale and Wernberg 2013) which are likely to have cascading impacts through altered benthic productivity and food web structure to a variety of higher trophic marine organisms including commercially important crustaceans, fishes and mammals (Lozano-Montes et al. 2011).

Fig. 3.4
figure 4

Scytothalia dorycarpa (a) is a large (~1 m) fucoid endemic to southern Australia. During an extreme heat wave in 2011, S. dorycarpa contracted its range by ~100 km in less than 1 year (Smale and Wernberg 2013). Prior to the heat wave (November 2010, b) reefs were covered by a dense mixed canopy of kelp ( Ecklonia radiata) and S. dorycarpa. However, immediately after the heat wave (November 2011, c) the canopy had large gaps where S. dorycarpa had been extirpated. One year later (November 2012, d), the canopy had not recovered and gaps were filling in with foliose and turf algae (All photos © T. Wernberg)

Full size image

In northern Spain, range contractions have been reported for several canopy-forming seaweeds (Appendix), including Fucus serratus and Himanthalia elongata which have moved westwards in the Bay of Biscay since the late nineteenth century as a response to global warming (Duarte et al. 2013). H. elongata changed its range stepwise by 330 km over 120 years, whereas F. serratus retracted 197 km over 114 years but also reduced its abundance dramatically in its remaining range, i.e. in the westernmost part of northern Spain (Appendix). For both species the rate of contraction appears to have accelerated in recent years (Duarte et al. 2013). The ecological implications of these two range contractions are largely unknown (Duarte et al. 2013), although both species (and several other large, retreating canopy-forming seaweeds ) are important habitat formers for smaller epiphytes and mobile animals (Hawkins and Hartnoll 1985; Lüning 1985; Wernberg et al. 2004; Ingólfsson 2008).

4.2 Range Expansions (Native Species)

The warm-water kelp Laminaria ochroleuca was first recorded in England in 1948, and subsequently expanded its range eastwards to the Isle of Wright at a rate of 5.4 km per year, as well as expanded northwards to Lundy Island at a rate of 2.5 km per year (Table 3.2). Recent resurveys of the inhabited area suggest that L. ochroleuca also expanded from the initially colonized sheltered coastline to moderately wave-exposed open coasts, accompanied by a significant increase in abundance, most likely in response to recent warming (Smale et al. 2014). In the area where L. ochroleuca most recently colonized, it competes with the native dominant congener L. hyperborea . As both species appear morphologically and functionally similar, it was initially assumed that they would have similar ecosystem function with little impact on the colonized ecosystem (Terazono et al. 2012). However, even small morphological differences may incur large cascading ecosystem effects. For example, Smale et al. (2014) showed that epiphyte cover on the smoother stipe of L. ochroleuca was dramatically lower than on the rough stipes of L. hyperborea. Thus, a reduction of the epiphytic habitat can be expected if L. ochroleuca replaces L. hyperborea, potentially with dramatic effects on associated fauna (Christie et al. 2009), trophic interactions (Smale et al. 2014) and biodiversity (Thomsen et al. 2010).

Table 3.2 Overview of the discussed five native and three non-native range shifts as well as the example of abundance change, with their according driver, direction and rate of shift and dispersal means
Full size table

Another example of a recent and unusual range expansion of a native seaweed involves the dominant canopy-forming kelp Ecklonia maxima in South Africa (Fig. 3.5). The distribution of E. maxima along the southern coastline of South Africa appeared unchanged for ca. 70 years, but suddenly expanded eastwards (between 2006 and 2008) at a rate of 36.5 km per year (Bolton et al. 2012). It is suggested that gradual cooling caused the distribution expansion of E. maxima, crossing around Cape Agulhas which is considered a major barrier dividing the western and south coast regions (Anderson et al. 2009). As E. maxima is the major kelp along its distributional range, expansion of this species could have substantial ecological consequences (Bolton et al. 2012).

Fig. 3.5
figure 5

The kelp Ecklonia maxima dominates nearshore reefs in the cold waters around southern Africa west of Cape Agulhas. It is a substantial seaweed which can grow to lengths in excess of 15 m (a: Buffels Bay, Cape of Good Hope). Between 2006 and 2008 this species expanded past Cape Agulhas, presumably due to cooling caused by upwelling (b: recently colonized intertidal populations at De Hoop Nature Reserve) (Photos © T. Wernberg)

Full size image

4.3 Range Expansion (Introduced Species )

Range expansions of non-native seaweeds can also alter ecosystem functioning after successful establishment. For example, Caulerpa cylindracea (Fig. 3.6a) is a highly invasive green seaweed which has spread along the Mediterranean Sea and Canary Islands since the early 1990s at an average rate of 11.9 km per year (Ruitton et al. 2005) (Table 3.2). C. cylindracea has invaded both soft and hard substrata and can form dense monospecific stands. The introduction vector is unknown, but where the species has established dense monocultures it has been associated with a decrease in abundance, biodiversity and biotic homogenization of native species (Klein and Verlaque 2008; Verbruggen et al. 2013). By forming multilayered mats that trap sediment, C. cylindracea can lead to burial of communities by sediment (Piazzi et al. 2005). Specifically, Baldacconi and Corriero (2009) determined its impacts on sponge assemblages in the Ionian Sea suggesting decreases in sponge cover following the invasion .

Fig. 3.6
figure 6

Some of the most notorious invasive marine species are seaweeds, which have spread rapidly throughout many regions of the world where they have been introduced. Caulerpa cylindracea (a) growing among turf and foliose seaweeds in its native environment in Western Australia. Sargassum muticum (b) growing in tide pools in northern Spain, where it is now a dominant element of the seaweed flora. Undaria pinnatifida (c) growing on tidal platforms in southeastern New Zealand (Photos a, c: © M.S. Thomsen, b: © C. Olabarria)

Full size image

The brown, canopy-forming seaweed Sargassum muticum (Fig. 3.6b) is also a well-studied invasive seaweed (Engelen et al. 2015). Originating from Asia, S. muticum has spread over the last few decades along coastlines in western Europe and western North America (Pedersen et al. 2005; Engelen et al. 2015). Within invaded locations, S. muticum can spread rapidly and become a dominant seaweed, sometimes leading to the suppression of local species and alteration of community structure (Stæhr et al. 2000). In 1941, S. muticum was first observed outside its native range in the Strait of Georgia (British Columbia, Canada) from where it subsequently spread along the adjacent coastline (Engelen et al. 2015). In 1984 S. muticum was sighted in Denmark for the first time in Limfjorden from where it subsequently spread at a rate of 4.4 km per year. S. muticum became the most abundant seaweed in Limfjorden, leading to a decrease in cover and abundance of several native canopy-forming seaweeds, including Halidrys siliquosa , Saccharina latissima , Fucus vesiculosus and Fucus serratus (Stæhr et al. 2000).

Undaria pinnatifida (Fig. 3.6c) is another high-profile invasive brown seaweed that is native to Japan, Russia and China. In the last 40–50 years, it has invaded Europe (Atlantic and Mediterranean Sea), North America (Pacific coast), south-western Australia, New Zealand and Argentina (Wallentinus 2007). In Argentina, U. pinnatifida was first recorded in 1992 and has since extended its range >1000 km southwards from its original site of introduction at a rate averaging between 35 and 50 km per year. While U. pinnatifida can have negative impacts on some native seaweeds (Casas et al. 2004), positive effects have also been reported on benthic macrofauna and carbon flow (Dellatorre et al. 2014; Tait et al. 2015).

4.4 Abundance Change

Range shifts with a clear change of species distribution at the distribution limits represent extreme transitions from the presence to absence or vice versa. Prior to range contractions , seaweeds will first decrease in abundances within their ranges, where continued reductions in abundance near range limits represent the first steps towards a range shift (Bates et al. 2014). For example, Johnson et al. (2011) documented that previously widespread Macrocystis pyrifera (Fig. 3.7) kelp forests decreased drastically in cover at several sites in eastern Tasmania (Table 3.2), likely caused by a combination of ocean warming and a strengthening of the East Australian Current (characterized by nutrient poor water) over the past six decades (Johnson et al. 2011). Concurrently, the strengthening of the East Australian Current also led to the range expansion of the sea urchin Centrostephanus rodgersii into Tasmanian waters, facilitated by over-fishing of urchin predators (lobsters, Ling et al. 2009). The range-expanding urchins have likely contributed to the decline in M. pyrifera through destructive grazing, which has also negatively impacted other native seaweeds (Johnson et al. 2011). With the decrease in abundance of M. pyrifera, a fast-growing habitat and food provider, dozens of associated species are losing a unique three-dimensional habitat, resulting in loss of taxonomic diversity and food web complexity (Graham 2004; Ling 2008; Byrnes et al. 2011).

Fig. 3.7
figure 7

The giant kelp (Macrocystis pyrifera) is a majestic seaweed often attaining a size of more than 10 m, has declined dramatically in abundance in Tasmania (pictured) over the past couple of decades due to increased warming, nutrient poor water and urchin grazing (Photo © T. de Bettignies)

Full size image

5 Perspective and Conclusion: Human Impacts on Seaweed Biogeography

Range shifts caused by species introductions and climate change need close monitoring as they are potentially irreversible and likely to have great ecosystem impacts (Madin et al. 2012). A critical problem, however, is that information on species’ range boundaries is scarce and largely qualitative due to lack of baseline information and regular surveys (Wernberg et al. 2011b; Bates et al. 2015; Marcelino and Verbruggen 2015). Ecological niche models can assist to identify areas with suitable habitat, anticipate arrival points and predict the potential extent of range change after a successful introduction (Marcelino and Verbruggen 2015) or environmental change (Molinos et al. 2015; Takao et al. 2015). For example, Takao et al. (2015) found that the present distribution of Ecklonia cava around Japan is well represented by SST-based indices. Chronologically observed changes were well in agreement with the projections, and the results further indicated that temperature will be a key factor for distribution of E. cava in the future (Takao et al. 2015).

Monitoring laboratory experiments and models projecting future shifts combined will help to identify likely range-shift pathways of seaweeds. In response to range-shifting species, management is necessary and several management tools already in place can be applied through, for example education, raising awareness and protected areas. But existing management tools are not always sufficient, and especially the limited knowledge on range-shift limits adaptive management responses (Madin et al. 2012). Additionally, for successful monitoring, a more widespread use of molecular methods is necessary to determine origin of species to prevent misidentification based on plastic morphology (Bolton 2010) and to identify loss of genetic variability at range edges (Provan and Maggs 2012; Assis et al. 2014; Neiva et al. 2015). Also, more regular surveys are required to be undertaken to determine range edges of populations and identify early range shifts and species introductions.

Future temperature increases are likely to result in more range shifts of seaweeds, especially along north–south orientated coastlines (Wernberg et al. 2011b; Molinos et al. 2015). These range shifts include poleward range extensions of warm-water tropical species, poleward range contractions of cold-water temperate species (Sorte et al. 2010; Wernberg et al. 2011a) and (potentially bidirectional) range expansion of introduced seaweeds (Sorte et al. 2013). Current models for marine species predict that expansions will be more prominent than contractions, leading to an overall increase in the biodiversity of many extratropical regions (Molinos et al. 2015). Globally, however, the narrative is likely to be different, because there will be a net loss of species (Cheung et al. 2009) as extinctions will be far more rapid than the evolution of new species. For seaweeds, this will be exacerbated by the juxtaposition of global patterns of species richness and endemicity (Bolton 1994; Kerswell 2006), hotspots of warming (Hobday and Pecl 2013) and barriers to range shifts (Wernberg et al. 2011a). In particular, southern Australia has the highest species richness and endemicity of seaweeds in the world, as well as some of the fastest warming regions in the world. However, the southern coastline is oriented east–west with very limited landmasses farther south. As seaweeds are pushed poleward towards the edge of the continent, there is great risk that they will ‘drop off’ to extinction—indeed, it has been estimated that range shifts could result in as much as a 25 % loss of the seaweed flora (Wernberg et al. 2011a).

To determine and model future biogeographic patterns of seaweed distribution, it is also necessary to take into account increasing threats to the coastal environments. Superimposed on temperature increases, increased ocean acidification will also change competitive hierarchies between fleshy, turf and calcifying marine algae, further altering local seaweed communities—and ultimately also range shifts (Hofmann et al. 2012). Also, interactive future effects, especially combined effects of warming and acidification with non-climate stressors, such as reduced water quality, will lower the resilience of communities and single species to perturbations like species invasions and storms (Wernberg et al. 2011a). Concurrently, more frequent and intense discrete events can drive stepwise changes in local environmental structure and cause larger more dramatic range- shifts (Smale and Wernberg 2013). Finally, ecological interactions are influencing the success of introductions and the speed of range shifts , possibly suppressing recovery, enhancing contraction or slowing down expansions (see M. pyrifera and S. dorycarpa case studies above). The extent to which the ecological context can suppress or enhance range shifts is a question in need of much research effort as we progress from simply detecting change to understanding its underlying drivers and mediators. However, the magnitude of range shifts and biological responses from anthropogenic impacts differ widely among species (Poloczanska et al. 2013).

6 Conclusion

There is now substantial evidence that humans have influenced the global biogeography of seaweeds over the last few decades and will continue to do so in the near future. This evidence generally spans timescales of decades, and is unlikely to simply reflect short-term fluctuations such as ENSO events. Humans influence seaweed biogeography through three distinct processes (introductions, climate expansions and climate contractions), which manifest through different processes (dispersal, recruitment and mortality) (Bates et al. 2014) and therefore proceed at different speeds: introduction > expansion > contraction. These changes in seaweed distributions have also been associated with impacts on seaweed-based ecosystems . While we are still to see the long-term ecological and economic consequences, these are likely to be substantial given the ecosystem services derived from seaweed ecosystems (Bennett et al. 2015a).