In Europe and elsewhere, marine plants and animals have undergone climate-driven shifts in their distributions (Sunday et al. 2012; Poloczanska et al. 2013), and major changes in assemblage structure and ecosystem function are projected to occur as a result (Helmuth et al. 2006; Hawkins et al. 2009). While patterns of ecological change, and the processes driving them, have been well documented in both intertidal (Helmuth et al. 2006; Hawkins et al. 2009) and pelagic (Richardson and Schoeman 2004) systems, there is currently limited information from subtidal benthic systems, especially from hard-bottom habitats that cannot be routinely trawled, dredged, or cored. This was highlighted by the recent “Marine Climate Change Impacts Knowledge Gaps” report, which stated that knowledge of large scale benthic species distributions within UK waters is required, to detect changes over large areas of the seabed and patterns of benthic response to climate change. This understanding is urgently needed to maintain healthy and biologically diverse seas (MCCIP 2012).
Kelps are cool-water species that are stressed by high temperatures (Steneck et al. 2002), so that seawater warming will affect the distribution, structure, productivity, and resilience of kelp forests (Dayton et al. 1992; Wernberg et al. 2010; Harley et al. 2012). Poleward range contractions have been predicted for several more northerly distributed kelp species (e.g., A. esculenta, L. digitata, L. hyperborea) in response to ocean warming in the Atlantic (Hiscock et al. 2004; Muller et al. 2009; Raybaud et al. 2013). It is evident that the relative abundance of several kelp species changes with latitude along NE Atlantic coastlines, which corresponds to a regional-scale temperature gradient, and that several habitat-forming kelps are at their range edge in the UK and Ireland (e.g., L. ochroleuca at its northernmost limit, A. esculenta at its southernmost limit, Fig. 3). Because of these distribution patterns and because the distributions of some intertidal species have shifted, several authors have predicted that relatively southerly distributed species will increase in abundance, while more northerly species will decrease in abundance and/or undergo range contractions in the UK and Ireland (Breeman 1990; Hiscock et al. 2004). There is some evidence to suggest that more southerly distributed kelp species (e.g., L. ochroleuca and S. polyschides) have increased in abundance and have undergone poleward range-edge expansions, while conversely, northern species (e.g., A. esculenta) have decreased in abundance in response to recent warming (Simkanin et al. 2005; Brodie et al. 2009; Birchenough and Bremmer 2010). However, the evidence base is largely based on anecdotal reports and unpublished survey data, and detailed historical examinations of distribution patterns are lacking.
As changes in the identity and abundance of habitat-forming species can have wide-ranging consequences for community structure and ecosystem functioning (Jones et al. 1994), there is a pressing need to examine climate-driven distribution shifts and their wider implications. For example, if a cool-water habitat former is replaced by a warm water species that is functionally and structurally similar, it is plausible that the wider community or ecosystem will be relatively unimpacted (e.g., Terazono et al. 2012). Conversely, if a structurally or functionally dissimilar species becomes dominant or habitat formers are lost and not replaced, then widespread changes in biodiversity patterns and ecological processes are likely to ensue (Ling 2008; Thomsen et al. 2010). In the UK and Ireland, a range contraction of A. esculenta, the dominant species on very exposed shores and an important midsuccessional species in more sheltered locations (Hawkins and Harkin 1985), would impact community structure and functioning as there is no warm water equivalent. Alaria esculenta is particularly susceptible to climate fluctuations, having disappeared from much of the English channel during a warm period in the 1950s and not recovering as conditions became cooler in the 1960s (Southward et al. 1995). Replacement of L. hyperborea with L. ochroleuca, which are more similar both structurally and functionally, may have less knock-on effects, although subtle differences in kelp species traits have been shown to influence local biodiversity patterns (Blight and Thompson 2008). Most dramatically, the predicted increase in the relative abundance of S. polyschides (Birchenough and Bremmer 2010) could have major implications for kelp forest structure and functioning as it is a fast-growing, annual species with distinct morphological and ecological traits (Table 1). Similarly, increased abundance of another annual, U. pinnatifida, relative to perennial species would also represent a major ecological shift from a stable habitat to one dominated by boom–bust cycles, with significant knock-on effects for biodiversity and productivity (see Pedersen et al. 2005 for relevant fucoid example). As kelps make a significant contribution to coastal primary production, facilitate export of carbon from high to low-productivity systems, and fuel entire food webs, changes in the quality or quantity of detrital material resulting from climate-driven changes in kelp species identity, abundance, or productivity could have far-reaching consequences (Krumhansl and Scheibling 2012). In the UK and Ireland, the wider implications of shifts in kelp species identity and abundance for kelp forest productivity, trophic linkages, and ecosystem functioning are almost entirely unknown.
It may be possible to predict the future structure of kelp forests under continued ocean warming in the UK and Ireland by examining the current structure of kelp forests under warmer conditions further south. For example, coastal waters off northern Portugal are some ~3°C warmer than off southern England and some ~5°C warmer than in northwest Scotland, which is within the projected range of NE Atlantic warming within the next 50–80 years (Philippart et al. 2011). The structure of kelp forest habitats off northern Portugal and Spain is strikingly different from those in UK waters (Hawkins and Harkin 1985; Fernandez 2011; Tuya et al. 2012). Most obviously, the geographic range of L. digitata does not extend further south than France and therefore does not form dense stands in the low intertidal and shallow subtidal zones. Laminaria hyperborea is present southward to north Portugal, but is generally much smaller and lower in abundance, forming isolated patches rather than dense canopies under warmer conditions. Conversely, L. ochroleuca is more abundant and often larger, while S. polyschides is generally more abundant across a wider depth range. However, recent observations suggest that S. polyschides (Fernandez 2011; Diez et al. 2012; Voerman et al. 2013), L. ochroleuca (Fernandez 2011; Diez et al. 2012; Voerman et al. 2013), and L. hyperborea (Tuya et al. 2012; Voerman et al. 2013) have undergone range contractions and/or declines in abundance in recent decades in response to seawater warming along the Iberian Peninsula. Loss of canopy-forming macroalgae at large spatial scales will have major implications for biodiversity and ecosystem goods and services (Voerman et al. 2013). It is very likely that kelp forest biomass and productivity will be diminished under warmer, stormier conditions (Krumhansl and Scheibling 2012), although direct measurements of kelp forest structure, biodiversity, productivity, detritus production and export, and resistance and resilience to perturbation along a regional-scale temperature gradient along the NE Atlantic coastline are lacking. Comparative experimental work along regional-scale temperature gradients is a promising approach in climate change ecology and can yield critical information on the mediation of ecological processes by ocean climate (Wernberg et al. 2010, 2012). Comparative kelp research along a regional-scale temperature gradient along Western Europe, spanning from Portugal (average sea temperature ~16°C) to Norway (average sea temperature ~8°C), would significantly enhance our understanding of climate change impacts on kelp forest structure and functioning.
In conjunction with ocean warming, observed and predicted increases in storminess (Lozano et al. 2004; Weisse et al. 2005) and ocean acidification (Connell and Russell 2010; Koch et al. 2013) will also impact kelp forests. As canopy-forming macroalgae may be damaged and dislodged during periods of intense wave action (De Bettignies et al. 2013), increased storminess will affect the structure and functioning of entire kelp habitats, by altering patch dynamics (Dayton and Tegner 1984) and potentially driving ecological phase shifts (Dayton et al. 1999; Wernberg et al. 2011). With regard to ocean acidification, experimental work on noncalcifying macroalgae lags considerably behind research focussed on calcifying algae and invertebrates (Connell and Russell 2010; Wernberg et al. 2012), but some generalized responses are emerging. From a physiological viewpoint, noncalcifying fleshy algae such as kelps can utilize elevated CO2 concentrations to increase growth rates (Harvey et al. 2013; Koch et al. 2013; Kroeker et al. 2013) and, probably, increase thermal optima for key physiological processes to potentially offset the impacts of increased temperature (Koch et al. 2013). Thus, increased CO2 concentrations may benefit kelp species. However, from an ecological viewpoint, the competitive balance between kelps and noncalcifying turf-forming algae may be shifted toward the latter in a high CO2 world (Connell and Russell 2010). When kelp canopies are removed under conditions of thermal stress, poor water quality, or intense wave action, mats of turf-forming ephemeral algae can replace them to form an alternative, degraded habitat type. Under certain conditions, including poor water quality (see “Land–sea interface” section), turfs can persist in space and time to inhibit kelp recruitment and consequently restrict kelp forest recovery. Experimental evidence and predictive theory both suggest that turf-forming algae will prosper under elevated temperature and CO2 (Connell and Russell 2010), increasing the likelihood of large-scale shifts from structurally diverse kelp canopies with associated calcified and noncalcified flora to simple habitats dominated by noncalcified, turf-forming seaweeds. The ramifications of such shifts are far-reaching and include regional biodiversity patterns, trophic linkages, nutrient cycling, and habitat provision for socioeconomically important marine organisms (e.g., fish and crustaceans).
Finally, two key knowledge gaps concerning the climate change ecology of kelp forests. First, there is a paucity of information on the capacity of local kelp populations to acclimatize or even adapt to climate-mediated change. It is clear that kelp populations can maintain physiological processes under a wide range of environmental conditions through local adaptation (e.g., Delebecq et al. 2013), but the rate at which kelp species can respond to rapidly changing temperatures and other localized stressors is unclear. Second, seaweed populations are particularly susceptible to short-term extreme warming events (Dayton and Tegner 1984; Smale and Wernberg 2013; Wernberg et al. 2013), which may increase in magnitude and frequency as a consequence of anthropogenic climate change (Jentsch et al. 2007; Feng et al. 2013). Short-term climate variability may pose greater threat to kelp populations at lower latitudes (i.e., toward range edges) than those within midlatitude temperate regions. For example, southerly distributed kelp forests off Spain and Portugal, which are subjected to environmental variability driven by the strength of coastal upwelling, comprise species at thermal maxima with dynamic range edges (Fernandez 2011; Tuya et al. 2012; Voerman et al. 2013). Anomalous warming events also have the potential to cause stepwise changes in the structure and functioning of kelp forests in midlatitude systems, and greater understanding of the resistance and resilience of kelp populations and their associated communities to such events is of ever-growing importance. Moreover, the effects of short-term temperature variability will likely be compounded by additional simultaneous stressors, such as nutrient loading, pollution, disease, or fishing pressure, which may interact with extreme climatic events to reach ecological tipping points (Crain et al. 2008).
As macrophytes are restricted to the photic zone, kelp forests form nearshore, coastally fringing habitats that are strongly influenced by connectivity between land and sea. Light is well known as the main driver of the distribution, depth, and abundance of kelp (Kain 1979; Dayton 1985), and contemporary declines in water clarity associated with coastal urbanization and land use have impacted macroalgal-dominated habitats across Europe (see Airoldi and Beck 2007 for review). Human activities across much of the world's temperate coastlines have increased sediment and nutrient loading into the coastal environments, which has been consistently linked with the widespread disappearance of kelp forests (e.g., Eriksson et al. 2002; Connell et al. 2008). Burrows (2012) recently showed that the distribution of L. hyperborea in the UK is strongly linked with ocean color (indicative of both oceanic phytoplankton content and terrestrially derived material), as greater light attenuation results in decreased depth penetration and abundance of kelp species and their associated communities. Off the coast of Norway, a recent large-scale disappearance of S. latissima, which has been replaced by ephemeral turfing algae, has been attributed to chronic eutrophication combined with increased temperatures (Moy and Christie 2012), although further work is needed to clarify these mechanisms. Clearly, processes acting across the land–sea interface can detrimentally impact the structure and functioning of kelp forests, and sustainable management of these habitats depends on integrated approaches spanning multiple ecosystems. In the NE Atlantic, these impacts will likely be exacerbated by both climate change, as precipitation rates and extreme climatic events are projected to increase (Philippart et al. 2011), thereby enhancing runoff, and by continued coastal development and land use.
Crucially, multiple concurrent stressors (climate and non-climate-related) do not act in isolation, but often combine synergistically in their effects, so that the total impact is far greater than the sum of individual factor effects (Crain et al. 2008; Harvey et al. 2013). Synergism can cause “ecological surprises”, where unexpected regime shifts occur quickly because a tipping point is exceeded (Crain et al. 2008). In kelp forests, multiple stressors can cause shifts from complex, biologically diverse habitats to simple turf-dominated “barrens” (Dayton and Tegner 1984; Ling et al. 2009; Russell et al. 2009). It is evident that increased nutrient loading and turbidity can interact with climate change factors to increase the competitive ability of ephemeral turf species, which can form an alternative stable state and inhibit the recovery of kelp forests (Russell et al. 2009; Moy and Christie 2012). The effects of multiple stressors on temperate algal communities are, however, poorly understood as only 20% of marine climate change experiments have focussed on primary producers and most have been single-factor laboratory experiments comprising few species (Wernberg et al. 2012). Continued research effort addressing the interactive effects of multiple climate and non-climate-related stressors under both laboratory and field settings should remain a priority.
Overgrazing by invertebrate herbivores, particularly sea urchins, can decimate kelp forests and cause phase shifts from structurally and biologically diverse habitats to depauperate “barrens” (reviewed by Steneck et al. 2002). In the North Atlantic, the green sea urchin Strongylocentrotus droebachiensis has deforested extensive areas of kelp forest in eastern Canada (Mann 1977), Iceland (Hjorleifsson et al. 1995) and northern Norway (Leinaas and Christie 1996), with major consequences for ecosystem structure and functioning (Steneck et al. 2002). At lower latitudes, the importance of grazing by the purple sea urchin Paracentrotus lividus on macroalgal assemblages has been recognized along Mediterranean and Atlantic coastlines (Bulleri et al. 1999; Hereu et al. 2004; Tuya et al. 2012).
In the UK and Ireland, the extent of deforestation by urchin grazing is generally restricted and patchy, although heavily grazed areas are more common in Scotland. Urchin grazing can certainly be important in setting local distributions of macroalgae, including kelps. Some of the earliest grazing work was done in the Isle of Man (Jones and Kain 1967), which showed that the edible sea urchin Echinus esculentus may determine the lower depth limit of L. hyperborea stands through intense grazing of young sporophytes. Similarly, P. lividus, which is relatively common along the west coast of Ireland, influences the distribution of macroalgae within Lough Hyne through grazing activity (Norton 1978; Kitching and Thain 1983). Recent resurveys of Lough Hyne have suggested that since classification as a marine reserve in 1981, the abundance of several urchin predators (i.e., crabs and sea stars) has increased, leading to declines in P. lividus abundance and consequent changes in macroalgal assemblages (O'Sullivan and Emmerson 2011). The green sea urchin Strongylocentrotus droebachiensis, which is only found in the north of Scotland, may also cause restricted patchy deforestation, but extensive barren formation has not been attributed to this species.
Harvesting and cultivation
The demand for kelp for human consumption, alginate production, aquaculture feed, and (potentially) biofuel has increased in recent decades and will almost certainly continue to grow. Direct removal of kelps has major implications for kelp population structure, whole community dynamics, and wider ecosystem functioning (Christie et al. 1998; Vásquez 2008; Krumhansl and Scheibling 2012). There is some evidence to suggest that due to the rapid recruitment and growth of kelps and their associated species, industrial-scale wild harvesting of kelps can be achieved sustainably. For example, in both Norway and Chile, some 130,000–200,000 tonnes is extracted annually and has been for some time (Vásquez 2008; Vea and Ask 2011). However, while a limited natural harvest may be sustainable if properly managed with appropriate fallow periods, the potential for impact on the other services provided by kelp may be considerable. Although kelps recruiting into harvested areas may reach preperturbed densities and sizes within a few years, their associated assemblages may take considerably longer to recover (Christie et al. 1998). Kelp harvesting also negatively impacts the abundance of gadoid fishes and reduces the area of habitat preferred by foraging seabirds (Lorentsen et al. 2010), for example.
Across Europe, the potential for kelp biomass to be used for conversion to biofuels has reignited interest in large-scale kelp harvesting. A realistic contribution to energy markets through bioethanol production may require more kelp than can be wild harvested from natural habitats, prompting efforts to develop methods of farming kelp. Mariculture of kelps is commonplace in Asia, particularly in China, where demand for seaweeds for human consumption is high. It is clear that intense kelp farming can impact local patterns of water movement and may cause organic enrichment of sediments and anoxia (Krumhansl and Scheibling 2012). However, many researchers are championing integrated aquaculture practices that utilize seaweeds as biofilters within multitrophic farming operations (Neori et al. 2004; Troell et al. 2009). In northwest Scotland, for example, cultivation of kelps and other seaweeds adjacent to salmon farms can generate significant yields of algal biomass while simultaneously removing waste nitrogen (Sanderson et al. 2012). However, the impacts of large-scale kelp cultivation in nonenriched systems are poorly known and may be detrimental. The Crown Estate recently commissioned an independent investigation into the wider ecological effects of proposed seaweed mariculture off the west coast of Scotland (Aldridge et al. 2012). Using ecosystem-based modeling approaches, the authors concluded that; the effects of the proposed farming activity on nutrient concentrations are expected to be ‘marginally significant’……and might become ‘certainly significant’……The observable effects of nutrient removal would be a lower nutrient concentration in the water, decreased productivity and energy fluxes through the pelagic system, decreased flux of organic material to the seabed, and subtle alteration to community structure. (Aldridge et al. 2012). It is beyond doubt that large-scale kelp production, through both wild harvesting and mariculture, has the potential to impact kelp populations, their associated benthic communities, and wider ecosystem structure and functioning. While it is recognized that a conservative ecosystem-based management approach is a prerequisite for achieving sustainable production, the baseline knowledge on the structure and functioning of kelp ecosystems at regional scales needed to underpin such an approach is currently lacking.