2   Global warming and marine ecosystems

2.1 Climatic factors

2.1.1    Rising water temperatures
2.1.2    Retreat of Arctic sea ice
2.1.3    Changes in ocean currents

2.2  Impacts of global warming on marine ecosystems

2.2.1    Natural climate variability
2.2.2    Human-induced climate change

2.3  In focus: Climate and fisheries

2.3.1    Changes in fish populations
2.3.2    Regional prognoses of impacts on fisheries
2.3.3    Global prognoses of impacts on fisheries

2.4  In focus: Climate and coral reefs

2.4.1    Warming impact on corals
2.4.2    Acidification impact on corals
2.4.3    Measures for coral conservation

2.5  Guard rail: Conservation of marine ecosystems

2.5.1    Recommended guard rail
2.5.2    Rationale and feasibility

2.6  Recommendations for action: Improving the management of marine ecosystems

2.6.1    Fischeries management   
2.6.2    Marine protected areas  

2.7   Research recommendations 

2.1   Climatic factors

2.1.1   Rising water temperature

The temperatures in the ocean influence sealife as well as the solubility of carbon dioxide in the water. They change the density of seawater, thereby influencing the currents and the sea level: the thermal expansion of water contributes considerably to sea-level rise. The sea surface temperature also affects the atmosphere in a multitude of ways. The mild Atlantic air that is often felt in Europe during the winter obtains its heat from the relatively warm ocean. High water temperatures also lead to increased evaporation, which is an important energy source for the atmosphere (for example, in tropical cyclones) and a source of water for many intensive precipitation events (among others, the Elbe river flood of 2002 in Germany).
     Significantly improved data sets of global ocean temperatures covering the past 50 years have become available to researchers in recent years through international efforts in the exchange of data (NODC, 2001). Based on over 7 million measured temperature profiles, Levitus et al. (2005) have reconstructed a time series of the heat content of the world ocean. They report an increase in the amount of stored heat of 15 . 1022 joules from 1955 to 1998. This corresponds to an average heat absorption of 0.2 watts per m2 for this time period when averaged across the entire surface of the Earth. For the period from 1993 to 2003 heat absorption was even greater, at 0.6 watts per m2 (Willis et al., 2004). This increase of heat in the ocean indicates that the Earth is presently absorbing more energy from the sun than it can radiate back into space. This reveals a state of disequilibrium in the heat budget of the Earth, as is to be expected due to the anthropogenic greenhouse effect (Hansen et al., 2005).
     Averaged globally and throughout the entire water column, the temperature of the ocean has only risen by 0.04 ºC since 1955. So far only the surface mixed layer with a thickness of a few hundred metres has warmed, while the average ocean depth is 3800m. The amount of sea-level rise caused by thermal expansion of the water so far is therefore only a small fraction of what will result when the warming extends into the deep sea over the coming centuries (Section 3.1.1).
     Figure 2.1-1 shows the variation of the sea-surface temperature, which is very important for the climate system. It shows a strong similarity to the development of air temperatures, but the warming is not as pronounced (0.6 °C since the beginning of the twentieth century). These two facts are not surprising. Thermally, the sea surface is closely coupled to the overlying atmosphere. Making up 30 per cent of the Earth’s surface, the land masses, because of their lower heat capacity, warm up more quickly than the oceans, so the global mean air temperature rises generally more quickly than that of the ocean. A data set of air temperatures measured by ships at night above the sea surface (Parker et al., 1995) shows a pattern very similar to the water temperatures. These data support the fact of a warming trend in the ocean surface waters and once more confirm the global warming measured by weather stations.

 

Figure 2.1-1
Globally averaged sea-surface temperature, according to three data centres: The UK Met Office (UKMO, blue), the US National Center for Environmental Prediction (NCEP, black), and the US National Climatic Data Center (NCDC, red).
Source: IPCC, 2001a

     Figure 2.1-2 shows the increase in surface temperatures in the North Atlantic, which in large part range between 0.3 and 1 °C over the indicated time period. A significantly stronger warming of several degrees is seen in Arctic latitudes, primarily because of the positive (strengthening) feedback with the shrinking sea ice (Section 2.1.1). Some small areas, however, show a cooling trend due to dynamic changes in the sea. This is particularly true of the Gulf Stream region off the coast of the USA and in regions near Greenland. The reason is probably natural internal fluctuations in the circulation, which superimpose the general warming trend caused by greenhouse gases.
The increase of sea temperatures in tropical latitudes is of particular interest because it influences tropical storms. This will be discussed in Section 3.1.2.

Figure 2.1-2
Development of sea-surface temperatures in the North Atlantic and European marginal seas. Temperature changes of the yearly average between 1978 and 2002 are shown (as a linear trend). Based on the GISST data set of the British Hadley Centre.
Source: PIK, based on Hadley Centre, 2003

    

2.1.2   Retreat of Arctic sea ice

An especially strong warming of seawater has been observed in the Arctic region in recent decades. This was described in 2004 in detail along with its impacts in an international study (Arctic Climate Impact Assessment; ACIA, 2005).
     The study concludes that a significant reduction of the Arctic sea ice has occurred that can not be explained by natural processes but only by human influences. The ice retreat can be clearly seen in satellite photographs (Fig. 2.1-3). The satellite time series from 1979 to 2005 shows a decline in the ice area of 15 to 20 per cent. The lowest ice extent ever measured was recorded in September 2005. Using a compilation of observations from ships and coastal stations, this development can be extended back to the time before satellite measurements were available. These kinds of observations go back to the year 1900, and cover about 77 per cent of the area of the Arctic region. The long-term data strongly suggest that the present shrinking of the ice cover is a unique event in the past hundred years.

Figure 2.1-3
Satellite photos of the Arctic ice cover, (a) September 1979 and (b) September 2005.
Source: NASA, 2005

Changes in the thickness of Arctic ice are more difficult to observe than its lateral extent. With the end of the Cold War, measurements by military submarines that had patrolled beneath the Arctic ice became available. These data indicate that the ice thickness may have already decreased by 40 per cent (Rothrock et al., 1999). Other investigations suggest smaller decreases in the thickness. Johannessen et al. (2005) report a decrease of 8–15 per cent, so the actual changes still have to be regarded as uncertain.
     Additional knowledge for the Arctic Ocean is obtained from computer models with high spatial resolution, driven by observed weather data. For recent decades they show a decrease in ice extent that is in agreement with the satellite data as already discussed. In these models the ice thickness decreases more strongly, about 43 per cent since 1988 (Lindsay and Zhang, 2005). Maslowski et al. (2005) obtained similar numbers. If the warming continues unchecked, the scenarios produced by global models indicate that the Arctic Ocean will be practically ice-free in the summer by the end of this century (MPI für Meteorologie, 2005). According to the regional models mentioned, this condition could occur even earlier.

2.1.3   Changes in ocean currents

Since the 1980s scientists have begun to address the question of possible abrupt changes in Atlantic currents and their effects on the climate (Broecker, 1987). The basic problem – a possibly strong nonlinear response of the current to freshwater influx – was recognized as early as the 1960s (Stommel, 1961). In recent years there has been an increased focus by researchers on the probability and the possible impacts of such events. However, the research is still at an early stage and many questions have not yet been answered. The danger of changes in the marine currents was brought to the attention of the public through the ‘Pentagon Report’ by Schwarz and Randall (2003), which featured in the media in 2004. This report presented a worst-case scenario in which, during the next 10 to 20 years, the North Atlantic Current stops flowing, which would lead to a severe cooling in the North Atlantic region within just a few years. This is, however, a speculative and extremely improbable scenario. In the present situation there is no evidence to support an imminent change in the currents. But in the longer term, and with continued climate warming, this could develop into a serious danger by the middle of this century.
     Huge masses of water currently sink from the surface to great depths in the Nordic Seas and the Labrador Sea. From there the water flows southwards at depths of 2–3 km to the Southern Ocean (Figure 2.1-4). Balancing this loss of water, warm surface water flows from the south into the northern latitude regions. This results in a large-scale turnover of water in the Atlantic, in which around 15 million m3 of water per second are transported. Like a central-heating unit, the ocean transports 1015 watts of heat to the northern Atlantic region through this process, which is equivalent to 2000 times the total output of Europe’s power stations.

Figure 2.1-4
The system of global ocean currents, primarily showing the ‘thermohaline’ circulation that is driven by temperature and salinity differences.
Source: after Rahmstorf, 2002

     Global climate change affects this water flow by decreasing the density of seawater in two ways: first, the temperature increase causes a thermal expansion of the water and, secondly, increased precipitation and meltwater input dilute the seawater with freshwater. This density decrease can retard the sinking of water in the northern Atlantic, the so-called deep-water formation. Particularly in the Nordic Seas a salinity decrease has already been observed in recent decades (Curry and Mauritzen, 2005), although according to modelling calculations this trend is still too weak to have an impact on Atlantic current patterns.
     British researchers have recently reported measurements suggesting that the circulation in the Atlantic may have already weakened by 30 per cent (Bryden et al., 2005). The interpretation of these data, however, is still contested in professional circles, in part because they do not agree with modelling calculations or with changes in sea-surface temperatures (Figure 2.1-2), where such a weakening of heat transport should be accompanied by a noticeable cooling. But if the trends of warming and salinity decrease should continue to strengthen in the coming decades, this may actually lead to a noticeable weakening of the Atlantic current over the course of this century, and in an extreme case possibly even to a total cessation of deep-water formation.
     In all probability the consequences would be severe. The North Atlantic Current (not the Gulf Stream as is often too simply stated) and the greater part of the Atlantic heat transport would be shut down. This would significantly change the temperature distribution throughout the entire Atlantic region. Depending on the degree of warming that has taken place before, it could even lead to regional cooling to levels below today’s temperatures. Southern Hemispheric warming would then be even stronger.
     As a result of dynamic adaptation of the sea surface to the altered currents, sea level in the North Atlantic would quickly rise by up to 1 m and slightly fall in the Southern Hemisphere. This redistribution of water would not have an immediate impact on the global sea-level average (Levermann et al., 2005). But over the long term the global average would rise by an additional 0.5 m due to the gradual warming of the deep ocean after the loss of input of cold water. In addition, the tropical precipitation belt would very likely shift because the ‘thermal equator’ would drift southward (Claussen et al., 2003). This is indicated both by model simulations and historical climate data.
Initial simulation computations also show a reduction of the plankton biomass in the Atlantic by half (Schmittner, 2005; Section 2.2.2.2). Because of thermohaline circulation the Atlantic is presently one of the most fertile marine regions and most productive fisheries areas of the Earth. In addition, the interruption of deep-water formation would reduce the ocean’s uptake of anthropogenic CO2 (Chapter 4).
     A breakdown of the North Atlantic Current is a risk that is difficult to calculate, but which would have severe adverse effects. One critical factor is the amount of freshwater that enters the northern Atlantic in the future. This will depend in large part on the speed at which Greenland’s ice sheet melts. A reliable prediction is not possible with the present state of knowledge; at best, a risk estimation can be attempted. For this purpose the Potsdam Institute for Climate Impact Research together with the American Carnegie Mellon University questioned a dozen of the world’s leading experts in the autumn of 2004, in detailed interviews lasting around six hours each. Their estimations of the risk of a total stop of deep-water formation and the associated currents varied considerably, but some were surprisingly high (Zickfeld et al., submitted). With an assumed global warming of only 2 ºC by the year 2100, four of the experts estimated the risk at greater than 5 per cent; with 3–5 ºC of warming, four of the experts indicated a risk exceeding 50 per cent.

2.2   Impacts of global warming on marine ecosystems

This section focuses on the impacts of climate warming (see Section 2.1) on marine ecosystems. WBGU considers this to include the entire marine realm, from the high seas to aquatically dominated coastal ecosystems. WBGU has deliberately only selected factors that are important to the subject of this special report. Overfishing, considered to be the most significant adverse anthropogenic impact today (Pauly et al., 2002; MA, 2005b), is not discussed. Also not treated here are direct destruction of marine ecosystems, pollution and alien species invasions (GESAMP, 2001; UNEP, 2002). Acidification of the sea is treated in Chapter 4. Together, these anthropogenic impacts have already strongly reduced the resilience of many marine ecosystems (Jackson et al., 2001).
     The most productive areas in the oceans, the shallow continental shelves (<200 m water depth) are the most intensely affected by these impacts. Although the shelves make up less than 7 per cent of the ocean’s surface, this is where the greatest proportion of the primary and secondary production takes place, and where the most productive fishing grounds are found (Section 2.3). The primary production of the seas by algae (phytoplankton) is limited to the translucent upper water layer, the euphotic zone (down to approx. 200 m water depth). A multitude of secondary producers live from these primary producers, especially zooplankton, fish and marine mammals, both in open water (pelagic) and at or below the sea floor (benthic). All organisms are linked to one another through a complex food web (Figure 2.2-1). For its energy source, the fauna of the dark deep sea is dependent on the organic carbon from the primary production, which sinks to the depths as dead biomass (‘biological pump’). Only in the vicinity of hydrothermal vents in the deep sea do bacteria form an independent basis for higher life forms through chemosynthesis.
     The coastal ecosystems are also of great biological and economical importance. In addition to their economic utility, some species-rich coastal systems such as wetlands, mangrove forests and coral reefs play a special role in protecting the coasts from flooding and erosion (Section 3.2).

2.2.1     Natural climate variability

The natural variability of abiotic factors in marine ecosystems such as water temperature or ocean currents is relatively great, and often follows non-linear or cyclic patterns. Studying the effects of natural climate variability can provide valuable information about the impacts of global warming. Compared to terrestrial systems, marine ecosystems react more sensitively and quickly to changes in climatic conditions, with unpredictable consequences for species compositions, spatial shifts of populations, or restructured food webs (Steele, 1998; Hsieh et al., 2005; overview by Brander, 2005). As Klyashtorin (2001) has shown, many Atlantic and Pacific fish stocks exhibit close correlations with climate patterns over many decades (Figure 2.2-2), for example, with the atmospheric circulation index, which describes atmospheric conditions in the Atlantic-Eurasian region. Even small natural climatic changes can have significant effects on marine ecosystems and fish stocks – through direct temperature effects, as a result of changes in primary production, or through impacts on important development stages (e.g., juvenile fish stages: Attrill and Power, 2002).

Figure 2.2-2
Correlation of the catch of various economically important fish stocks with the atmospheric circulation index.
Source: compiled on the basis of Klyashtorin, 2001

For example, the cod stocks off Greenland reacted to a warming of the North Atlantic in the 1920s and 1930s with a rapid expansion to the north (approx. 50 km per year) and a considerable increase of stock size, which later decreased again as a result of overfishing and deteriorating climatic conditions (Jensen, 1939). Plankton-feeding fish species in particular, such as sardines or anchovies, show strong natural stock fluctuations, in which large-scale climatic variations play an important role (Barber, 2001; PICES, 2004). The short-term disturbances of the ENSO events (El Niño/Southern Oscillation), for example, have far-reaching, 2- to 3-year effects on the marine ecosystems of the Peru-Humboldt current system (decreased nutrient supply causing lower primary production, partial collapse of fish populations: Barber, 2001) and on the most productive fish stock in the world (Peruvian anchovies: FAO, 2004; Bertrand et al., 2004). The impacts of the ENSO events are, however, reversible, with ‘normal’ conditions being re-established as a rule within a few years (Fiedler, 2002).
     Ignoring small interannual variations, however, regional climatic conditions, along with the structure and dynamics of the ecosystems within a marine region, can also remain relatively stable over a period of many years or decades, defining what is generally referred to as a regime. When this kind of relatively stable situation changes rapidly, within the course of one or two years, it is called a ‘regime shift’ (King, 2005). Along with these regime shifts, considerable structural changes in the affected marine ecosystem occur, from the phytoplankton up to the highest trophic levels in the food web, including large predatory fish.
     Regime shifts have been observed often and in various marine regions (King, 2005). In the North Sea in the late 1980s, for example, a regime shift occurred that was related to abrupt changes in surface temperature, wind conditions and a multitude of biological parameters (Reid et al., 2001; Beaugrand, 2004; Alheit et al., 2005). Due to an increase in westerly winds the influx of warm water into the North Sea was strengthened causing, among other things, a degradation of living conditions for North Sea cod. There is probably a connection between this persistent change in the North Atlantic Oscillation and anthropogenic climate warming (Gillett et al., 2003). In the North Pacific off the coast of California, alternating regimes with a period of around 60 years have been documented covering almost two millennia (Baumgartner et al., 1992). They cause a distinct restructuring of the marine ecosystems (Hare and Mantua, 2000; King, 2005).
     How regime shifts are triggered and what effects they have in the food web of an ecosystem are not yet thoroughly understood, even though quite detailed observations of changing ecosystem structures do exist. The energy fluxes originating in the phytoplankton, at the base of the food web, often seem to play an important role (‘bottom up’ control: e.g., Richardson and Schoeman, 2004). However, structural changes can also be controlled ‘top down’, caused by the collapse of the population of predatory fish, either by overfishing (Worm and Myers, 2003; Frank et al., 2005) or by climatic changes (Polovina, 2005), and reaching down to the lower levels of the food web by trophic coupling.

2.2.2     Human-induced climate change

Although the natural variability can be very large regionally, the global warming trend already predominates in most areas (Figure 2.1-2). The anthropogenic impact on various climatic factors has already had observable effects on the distribution of marine organisms and the species assemblages of marine ecosystems (overview by Brander, 2005). Climate impacts have been described for all levels of the ecosystem, from primary production (Section 2.2.2.2) to zooplankton (e.g., Richardson and Schoeman, 2004) and small pelagic fish species (sardines), all the way up to the large predatory fish (tropical tuna: Lehodey et al., 2003).

2.2.2.1     Effects of water temperature on the physiology of marine organisms

According to the latest findings, temperature has a significantly greater influence on the distribution of animal and plant species than was previously assumed, and this is independent of the position of the organisms in the food web (Huntley et al., 2004). To a large degree, the window of thermal tolerance in which a species can survive, grow and reproduce determines its distribution (Pörtner, 2005). An increase in water temperature (Section 2.1.1) influences the life of marine organisms both directly and indirectly. A direct physiological impact is seen when the upper limit of the temperature tolerance range for a species is exceeded. This applies, for example, to tropical corals (Section 2.4.1). An indirect influence of increasing water temperature is observed, for example, when organisms previously available temporally and spatially as food for a species are no longer present due to changes in the species assemblage of an ecosystem caused by temperature differences (Section 2.2.2.5). Both of these effect chains can lead to shifts of populations, the invasion of alien species, and even the disappearance of species.

2.2.2.2     Phytoplankton and global primary production

The climatic factors altered by human activities (Section 2.1) initially affect the phytoplankton and therefore primary production. The total marine ecosystem, all the way up through the various trophic levels to the large predators such as tuna and sharks, feeds in principle from the primary production. Therefore, through this coupling, a change in primary production will have an effect on the higher trophic levels of the food web and will be reflected in changed species assemblages or biomasses in the total ecosystem. The primary production is influenced by many climatic factors (Fasham, 2003):

• Temperature: Growth and species composition of the phytoplankton are strongly dependent on temperature. Initially, primary production is directly stimulated through warming. But the increased temperature can also indirectly slow down production, for example due to a decrease in nutrient supply resulting from prominent temperature stratification.
• Light: Changes in the ice or cloud cover of the surface water have a direct influence on the primary production because the phytoplankton require light as an energy source. The light supply for phytoplankton also diminishes with increased mixing of the surface water.
• Nutrients: Climate change can also indirectly influence the supply of nutrients to the phytoplankton (primarily nitrogen and phosphorous, but also ‘micronutrients’ such as iron: Jickells et al., 2005). Through the sinking of dead organisms from the productive upper layer of the ocean, organic material and thereby nutrients are continuously exported to the deep sea (‘biological pump’: Falkowski et al., 2003). The return transport to the upper layers occurs through upwelling currents and vertical mixing, which are influenced by climate in the form of temperature stratification, as well as wind and current conditions (for example, Sarmiento et al., 2003).

In addition to all these factors, climate warming is largely a result of rising CO2 concentrations. In many phytoplankton species this leads to a direct increase in the rate of photosynthesis, although the various species groups benefit to differing degrees (Section 4.3.1). These various factors are also all coupled with one another. The warming of the surface layers not only increases photosynthesis rates, it also promotes a more stable layering of the water column, which decreases the nutrient supply and weakens the plankton production. The stronger stratification can also destabilize the dynamics of phytoplankton production (Huisman et al., 2006). An increased wind speed, on the other hand, counteracts the temperature effect upon stratification. In the northeast Atlantic the sum of these counteracting effects produces a net increase of phytoplankton in cold-water regions (because here with the good nutrient supply and higher turbulence the improved metabolism rates due to temperature increase are predominant), and a net decrease in warm-water regions (because stronger stratification, under limited nutrient availability, worsens the growing conditions; Richardson and Schoeman, 2004). It is therefore not surprising that these effects are difficult to model and vary greatly from region to region.
     Satellite-based observations of the phytoplankton biomasses derived from the chlorophyll content of seawater reveal that the global annual primary production has decreased in nine of twelve ocean regions since the 1980s, and the global average by more than 6 per cent (Gregg et al., 2003). The high northern latitudes account for 70 per cent of the global decline, presumably caused by the worsening nutrient supply due to the rise in temperature. Only three tropical ocean regions (northern and equatorial Indian and the equatorial Atlantic) exhibited an increase. For the North Atlantic, long-term data series based on physical samples show an increase in phytoplankton north of 55 °N and a decrease south of 50 °N (Richardson and Schoeman, 2004). The projections for a future with global warming show contradictory trends. The modelling of Bopp et al. (2001) suggests a reduction of the global marine export production (which correlates well with primary production) by about 6 per cent in the next 65–75 years with a doubling of the atmospheric CO2 concentrations. Production in the tropics would decline due to stronger stratification and the resulting decrease in nutrient supply, while increasing in the subpolar regions. In contrast, the models of Sarmiento et al. (2004), with large uncertainty, show a slight increase in the global primary production. Again, the effects are regionally highly variable. The authors of the Arctic Climate Impact Assessment consider it probable that moderate warming would promote primary production in the Arctic, mainly due to the reduction of sea ice (ACIA, 2005).
     So the available findings are, at least in part, contradictory, and regional observations are not always in agreement with model prognoses. Obviously, our understanding of the critical processes, such as the temperature sensitivity of primary production, is insufficient. The quality of coupled climate, ocean, and ecosystem models presently does not allow any robust conclusions (Sarmiento et al., 2004), although some regional models are already able to represent the connections between the changes in ocean currents and primary production (examples in Brander, 2005).
     It is improbable that climate change will lead to the breakdown of the North Atlantic Current, but this possibility cannot be excluded (Section 2.1.3; Rahmstorf, 2000; Curry and Mauritzen, 2005). The simulations of Schmittner (2005) show a completely altered ecosystem situation for this scenario: the biomasses of phyto- and zooplankton in the North Atlantic would decrease by half due to sharply reduced nutrient supply in the surface waters, with corresponding large impacts on ecosystem productivity and structure.

2.2.2.3     Zooplankton

Primary production by phytoplankton is the nutritional basis for the zooplankton (secondary production: often small crustaceans), which is in turn significant as food for the growth of fish populations. Fish larvae in particular are dependent on the synchronous and high availability of appropriate zooplankton, so that fish stocks can replenish and production is maintained. The following examples show that for the zooplankton too, noticeable changes can already be identified as a result of anthropogenic climate change.
     In the North Atlantic the distribution of copepods, an important group in the marine food web, has shifted to the north by around 10° of latitude as a result of a combination of changes in the North Atlantic Oscillation (NAO) and human-induced climate change (Beaugrand et al., 2002). For the North Sea cod these changes, along with overfishing, have contributed to poor conditions encountered by the fish larvae and a steady decline of the population (Beaugrand et al., 2003).
     Krill (Euphausia superba) in the Antarctic have declined significantly since 1976, while other zooplankton species (salps) have increased, which can probably be attributed to the climate-driven reduction of sea ice around the Antarctic peninsula (Atkinson et al., 2004). Because krill is an important food source for fish, penguins, seals and whales, this has led to significant changes in the food web in the Southern Ocean. Investigations of planktonic foraminifera in sediments covering the past 1400 years have revealed an anomalous change in the species assemblages in recent decades. This suggests that the anthropogenic warming of the ocean has already exceeded the range of natural variability (Field et al., 2006).

2.2.2.4     Marine mammals

The warming also causes a decrease in the geographic extent of the Arctic sea ice. This especially affects animals such as polar bears and ringed seals, which are directly dependent on this habitat in their feeding habits and for the rearing of their young (ACIA, 2005).
     Polar bears feed almost exclusively on seals, which are bound to the ice habitat. Female polar bears bear their young in caves on the land. In the spring after their winter sleep, in order to reach their hunting areas on the ice, the mother and her young are dependent on ice corridors, because the young animals cannot cross large areas of open water. If the ice continues to recede, they will not be able to reach their hunting grounds. Adult polar bears are good swimmers, and can cover distances in the water of over 100 km. Monnett et al. (2005), however, report a doubling of the number of polar bears sighted swimming in open water within a 20-year observation period, as well as most recent finds of four drowned polar bears near Alaska in a location where the ice was over 200 km to the north of its normal seasonal limit. Around Canada’s Hudson Bay, the area of their southernmost occurrence, the polar bear population has declined by 22 per cent since 1987 (Carlton, 2005). With the loss of the summer sea-ice cover, polar bears are forced into a life on land, where they encounter competition with brown and grizzly bears and increased contact with humans, which reduces the chances for survival of this species .
Scenarios for the Baltic Sea also indicate that the ice cover here will significantly decrease over the next 30 years. The Baltic ringed seal requires a firm ice layer with a snow cover for at least two months for rearing its young.      Of the four former breeding areas in the Baltic Sea with separate populations only one suitable area will remain available in the future: the northern Bay of Bothnia (Meier et al., 2004). Initial observations have been reported in the Antarctic, too, that can be attributed to climatic changes. For the past 20 years birth rates of the cape fur seal have been in decline. This decrease correlates with unusually high temperatures of the surface water subsequent to the abundant El Niño events between 1987 and 1998, and it has presumably been intensified by the lowered nutrient supply – primarily krill – for the female seals (Forcada et al., 2005).
     These examples illustrate how changes in the ice habitat caused by climatic change can drastically impact the highest trophic levels.

2.2.2.5     Ecosystem impacts

Temperature increases and other factors related to climate change affect groups of organisms in different ways, so that population shifts can occur at different rates and intensities to separate species that previously inhabited the same region or were present at the same time. This decoupling of previously synchronous trophic levels (‘trophic mismatch’) can produce considerable changes in the ecosystem structure (for example, in the North Sea: Edwards and Richardson, 2004). Climate-induced spatial changes in the phytoplankton distribution can affect both the herbivorous zooplankton and the carnivorous zooplankton, so that fish, seabirds and mammals also have to adapt to the new conditions (Richardson and Schoeman, 2004).
     These kinds of large-scale shifts have already been observed at different levels in the food web, for example in the North Atlantic (Beaugrand and Reid, 2003). After an anomalous temperature increase in the 1980s, populations of cold-water species such as euphausids and copepods shifted northward and the stocks decreased, while the smaller warm-water species showed a corresponding increase (Beaugrand et al., 2002). This then led to a decline in the salmon population. For the future, Beaugrand and Reid (2003) expect a continuing decline in the number and distribution of the salmon population, especially at the southern edge of its geographic distribution (Spain and France).
     As a result of the displacement of distribution areas of many species toward the poles, the pressure on marine ecosystems increases in the polar regions due to the immigration of new species, while the inhabitants of these regions, adapted to cold temperatures, cannot move to cooler latitudes. They are therefore particularly sensitive to climate change, so that losses of habitats and species are to be expected, especially in the polar marine sea-ice ecosystems (Smetacek and Nicol, 2005; ACIA, 2005). In addition, regional expressions of global warming are especially evident in the Arctic, in part because there is a particularly strong feedback with regional temperatures there as a result of the albedo changes due to retreating sea ice.
It is likely that primary production in the Arctic Ocean will increase due to climate warming, albeit from a low initial level (ACIA, 2005). The increased production can either be exploited by the zooplankton or fall out as sedimentation and provide nutrients for the benthic fauna. In regions with seasonal ice cover, a temporal shift could occur between the phytoplankton bloom and the massive occurrence of zooplankton as well as between the zooplankton and fish larvae due to the climate-dependent changes in the start of ice melting in spring. Such a lack of synchronization would result in a lower share of the primary production being available for higher levels of the food web. Reliable predictions cannot be made, however, concerning the effect of increased primary production on fish, bird and mammal populations (ACIA, 2005).
     An important question is whether anthropogenic climate change can influence naturally occurring regime shifts (Section 2.2.1). With the low resilience of marine ecosystem structures and the intensity of expected anthropogenic climate signals (Section 2.1.1) the possibility that regime shifts in the future will exhibit a different quality, occur more often or rarely, or occur in regions where they have not been previously seen can absolutely not be excluded. The observed acceleration in the periodicity of regime shifts in the North Pacific (King, 2005) could be indicative that there is a link with anthropogenic climate change, although a conclusive judgement cannot be made at this time (Brander, 2005).
     The new findings in marine ecology support the increased climate change mitigation efforts that have been called for in previous WBGU reports (e.g. WBGU, 2003), because if climate warming continues unchecked, severe unpredictable and undesirable changes in marine ecosystems cannot be discounted.

2.3     In focus: Climate and fisheries

For over 2600 million people fish is the basis of at least 20 per cent of their protein supply (FAO, 2004). Industrial fisheries are growing and are coming into increasing competition with the 30 million traditional fishers, who are often faced with losses in income (World Bank, 2004). World fish production in recent years has remained stagnant at around 130 million tonnes per year, whereby the proportion of fish caught in the sea has slightly decreased and the share of aquaculture has risen (FAO, 2004). Fish stocks are very unequally distributed in the sea: less than 7 per cent of the ocean’s area is represented by the continental shelves (water depths <200 m), but these account for more than 90 per cent of the global fish catch (Pauly et al., 2002).
     At the same time, the human impact on the marine ecosystems of the continental shelves is especially great: overfishing (FAO, 2004; MA, 2005b), including illegal or unregulated fishing (Gianni and Simpson, 2005), degradation and destruction of marine and coastal habitats (such as corals, Section 2.4), the invasion of alien species, pollution of the world’s oceans (GESAMP, 2001), and, as a new threat, acidification (Chapter 4) endanger the health of the ecosystems and the sustainability of their use. Poor fisheries management and resulting overfishing are certainly more critical factors for the fish stocks than the anthropogenic climate change observed so far (Worm and Myers, 2004; ACIA, 2005). In the future, however, the latter could also cause a considerable additional burden for marine ecosystems (IPCC, 2001b; Richardson and Schoeman, 2004; Section 3.1.4).

2.3.1     Changes in fish populations

Similar to the terrestrial realm, the species in marine ecosystems often respond to anthropogenic warming with a poleward shift (Parmesan and Yohe, 2003). This is also the case for many of the fish populations in European shelf waters, with increasing evidence of a northward shift due to warming (Fig. 2.3-1). The stocks of cod in the North Sea are decreasing at a rate that cannot be explained by overfishing alone. Today, the upper limit of the thermal tolerance window has already been reached there, with the result that populations are moving northward. The decrease of cod correlates significantly with the changed species assemblage, stock decline and smaller average body size of the zooplankton (Beaugrand et al., 2003), which can probably be attributed to climate change. Fundamental changes in the pelagic ecosystem have been observed in the North Sea from 1925 to 2004, with a clear shift of many populations northward and the immigration of southern species (Beare et al., 2004). These systematic long-term trends correlate with the rising sea temperature. From observations in the North Sea, Perry et al. (2005) conclude that a further increase in temperature will result in additional changes in the species composition and ecosystem structure that cannot be predicted in detail, but will probably put considerable adaptive pressure upon commercial fisheries. In various Arctic marine regions, with a regional warming of 1–3 °C, northward displacements of fish populations can be expected, along with the establishment of discrete populations (e.g., cod near Greenland) as well as the immigration of southern species (ACIA, 2005).

2.3.2     Regional prognoses of impacts on fisheries

For some marine regions, especially for waters in the northern latitudes, our understanding of the ecosystem structures and their response to natural climate variability is good enough to discuss possible impacts of climate change. The Norwegian Sea, for example, is a very well studied area (Skjoldal, 2004). Based on experience with natural climate variability it can be assumed that regional temperature increases of 2–4 °C could increase the primary and secondary production of the sub-Arctic part of the Norwegian Sea and therefore improve conditions for fish production (Skjoldal and Sætre, 2004). At the same time, however, the spectrum of species would experience a shift, i.e. warm-water southern species would be introduced to the area.
     The authors of the Arctic Climate Impact Assessment (ACIA) also assume that a regional warming of 1–3 °C would improve the conditions for some economically important fish populations, such as Atlantic cod or herring, because the retreat of sea ice would increase both the primary and secondary production as well as allowing these species to spread northward (ACIA, 2005).
Regime shifts with distinct changes in species composition (Section 2.2.1) are not ruled out by the ACIA, but its authors deem that the adaptation of the fisheries sector to the new conditions should not present a great expense. The ACIA comes to the general conclusion that the type and effectiveness of fisheries management – in particular the prevention of overfishing through application of the precautionary principle – will have a greater impact on production in the Arctic than the moderate regional climate change of 1–3 °C that is projected for the 21st century. Accordingly, significant economic or social effects at the national level are not expected, even though individual Arctic regions that are heavily dependent on fisheries could be clearly impacted.
     This assessment changes, however, in the case of a more substantial regional climate change (>3 °C). While the authors of the ACIA consider it possible that negative consequences could result for fisheries in some Arctic marine regions, they hesitate to make predictions for most Arctic regions due to the incomplete understanding of ecosystem structure and dynamics. Although the Arctic waters have been comparatively well studied, no reliable ecosystem model coupled with climate scenarios currently exists. The assessment of ecosystem impacts therefore must remain speculative. In order to answer the remaining questions, research efforts will have to take a more ecosystem-based approach. Improved numerical ecological models based on integrated environmental monitoring will make an important contribution (Skjoldal and Sætre, 2004; Section 2.7).

2.3.3     Global prognoses of impacts on fisheries

The Food and Agriculture Organisation of the United Nations (FAO) refers to future anthropogenic climate change as an example of the uncertainty justifying a precautionary approach to fisheries management (FAO, 2000). In its report ‘The State of World Fisheries and Aquaculture 2002’, FAO draws attention to the importance of natural long-term climate variability for the development of fish stocks in a chapter dedicated to this subject. It also points out that global warming could have significant impacts – positive or negative – on some, if not most of the commercial fish stocks (FAO, 2002). It concludes that stocks drastically reduced by overfishing are more vulnerable to climatic changes than sustainably exploited stocks (FAO, 2004). However, FAO’s long-term projections are, even today, still based inter alia on the assumption that environmental conditions, including the climate, are not changing significantly.
     The Intergovernmental Panel on Climate Change (IPCC, 2001b) points to the increasingly acknowledged relationship between natural climate variability and the dynamics of fish stocks and concludes that global warming complicates these relationships and will make fisheries management more difficult. Climate change therefore has the potential, during the coming decades, to become an important factor in the management of marine resources, although the effects will vary widely depending on the region and ecosystem characteristics (IPCC, 2001b).
     The authors of the Millennium Ecosystem Assessment also warn about the consequences of climate change, although they have not carried out a detailed analysis. They describe current knowledge about the effects of climate change on marine ecosystems as inadequate. They point out in particular that the response of fish stocks to environmental influences depends, not least, on population size. Healthy stocks with large production of fish larvae can adapt better to population displacement and changes in ecosystem structure. Stocks that are greatly reduced due to overfishing respond more sensitively to environmental influences such as climate change (MA, 2005b) because there is a greater probability that the minimum stock level for reproduction is not attained.
     Despite the lack of scientific data, several general recommendations for the management of marine ecosystems and fisheries management can be made. These will be discussed in Section 2.6.

2.4     In focus: Climate and coral reefs

Tropical coral reefs are recognized as the most species-rich of marine biotopes, not so much because of the abundance of species of the reef-building corals themselves (over 835 species have been described), but because of the biological diversity of organisms that live on and from coral reefs, representing an estimated 0.5–2 million species (Reaka-Kudla, 1997). Coral reefs provide important products such as fish and building materials (blocks of coral limestone). They also offer protection from the effects of tsunamis and coastal erosion, and at the same time, because of their aesthetic and cultural value, they are an important source of income from tourism. Although coral reefs only cover 1.2 per cent of the global continental shelves, it is estimated that more than 100 million people are economically dependent on them (Hoegh-Guldberg, 2005). A status report on worldwide coral reefs (Wilkinson, 2004) provides information on their development since the 1950s and raises vital concerns with its estimation of the future trends:

– 20 per cent of all coral reefs have been effectively destroyed and show no immediate prospects of recovery,
– 24 per cent of all coral reefs are under imminent risk of collapse through human pressures,
– a further 26 per cent are under a long-term threat of collapse. .

The changes of the past 20–50 years are referred to as the ‘coral reef crisis’ because the adaptive capacity of corals and the animals and plants associated with them to changing environmental conditions has been exceeded worldwide (Hoegh-Guldberg, 1999; Pandolfi et al., 2003). The pressure from human activities is locally generated, first through poor land management practices, whereby sediments, nutrients and pollutants are released and washed into the sea and damage the reefs. In addition, overfishing, primarily the fisheries using destructive methods (dynamite, cyanide, heavy fishing rigs), reduces the populations of key species on the reef, damaging the function of the ecosystem and reducing productivity. After ecosystem damage, macroalgae have an advantage over the coral in their growth because the feeding pressure by selectively caught fish that normally feed on these algae declines.
In addition to the local stress factors, two results of global climate change are becoming increasingly important to the condition of coral reefs and will therefore be investigated in more detail in this section: the increase in seawater temperature and the acidification of seawater. These two factors contribute individually as well as synergistically, together with the local anthropogenic stressors, to the destruction of coral reefs.
It is only in recent decades that coral reefs were also discovered to exist in deep, dark, cold-water zones in practically all of the world’s oceans (Freiwald et al., 2004). Their ecosystems and the serious dangers to them, particularly from bottom-trawl fishing, are currently being researched. Whether they are also threatened by the effects of climatic changes such as temperature change and changes in the availability of calcium carbonate is not clear.

2.4.1     Warming impact on corals

Coral reefs dominate tropical coasts at latitudes between 25 °N and 25 °S, which corresponds to a seawater temperature range of 18–30 °C (Veron, 1986). Along with the atmosphere, the surface layers of the ocean have also warmed in recent decades (Section 2.1.1). In seven tropical regions where corals occur, a warming of 0.7–1.7 °C has been measured in the 20th century (Hoegh-Guldberg, 1999).
     Since 1979 a new phenomenon has been described with increasing frequency and geographic extent, called coral bleaching. This refers to the loss of single-celled algae that live in symbiosis with the corals. If a coral is subjected to a stress situation, which either in nature or in the laboratory can be produced by high or low temperatures, intensive light, changes in salinity or other physical, chemical and microbial stress factors, the algae will be expelled from the coral tissue. The living tissue of the corals is transparent without algae cells, so the white limestone skeleton will show through – hence the term coral bleaching. This phenomenon is to some extent reversible because algal cells can be taken up again by the body tissue. But after extended periods of coral bleaching the corals die.
     Abundant occurrences of coral bleaching were first described in the scientific literature in the early 1980s. Strongly increasing worldwide occurrences correlate with higher surface temperatures of seawater and with disturbances related to El Niño events (El Niño/Southern Oscillation, ENSO). The most intense event by far occurred in 1997–1998, resulting in the death of 16 per cent of all tropical corals worldwide. Regionally the values were even higher, e.g. at 46 per cent in the western Indian Ocean (Wilkinson, 2004).
The strength and duration of the temperature anomalies are important values for predicting coral bleaching. The ‘Degree Heating Weeks’ (DHW), which aggregate the thermal stress over 12 weeks, were developed as an indicator. One DHW is equal to one week with a temperature of 1 °C above the summer maximum during the previous 12 weeks. The USA’s National Oceanic and Atmospheric Administration (NOAA) provides an operational early warning system for this. Analyses of the measurement series show that 8 DHW led to coral bleaching in 99 per cent of all cases. Coral bleaching can be predicted today with over 90 per cent probability several weeks before the event occurs (Strong et al., 2000). The worldwide area of coral reefs affected by DHW >4 is continuously increasing (Wilkinson, 2004). Modelling calculations based on IPCC scenarios indicate that between 2030 and 2050 events similar to the anomalous year of 1998 could occur annually, spelling the end of coral-dominated ecosystems (Hoegh-Guldberg, 2005). By combining the data of the NOAA early warning system with global circulation models, Donner et al. (2005) arrived at similar conclusions. According to their research, in 30–50 years coral bleaching will occur every one to two years in the large majority of all coral reefs if the corals do not adapt their temperature tolerance by 0.2–1 °C per decade.
     An important observation is that the threshold value of seawater temperature for triggering coral bleaching at many locations is only 1–2 °C above the maximum summer temperature. Tropical corals are therefore living very close to the highest temperature at which they can survive (Hoegh-Guldberg, 1999). Assuming that the near-surface seawater temperatures will continue to increase, the question is how corals could respond to this temperature increase. Hughes et al. (2003) describe possible responses: a single threshold value for all coral species is unlikely; instead, the threshold values vary within a certain bandwidth depending on coral species, water depth and location. A model in which the different threshold values for the demise of the corals change with time through acclimatization and evolution seems to be most realistic. Symbiotic algae that occur in various genotypes are adapted to different upper temperature limits, for example. After coral bleaching has occurred, heat-tolerant algal groups could be taken up into the coral tissue, offering improved protection against future temperature peaks, and thereby providing some limited adaptation to climate change (Baker et al., 2004; Rowan, 2004). Hoegh-Guldberg (2005), however, expresses the concern that the evolutionary adaptation of corals and algae cannot keep pace with the rapid environmental changes taking place over just a few decades.
     Coral reefs could also respond to increased seawater temperatures with a shift of their distribution or a change in their species assemblage. A poleward displacement of the distribution region, however, could only amount to a few degrees of latitude at most, because both the light (for photosynthesis of the symbiotic algae) and the aragonite supersaturation (for calcification) are limiting factors (Buddemeier et al., 2004).

2.4.2     Acidification impact on corals

The acidification of the sea through hydrolysis of CO2 in seawater (Section 4.1) influences the carbonate chemistry and thereby also affects the corals, which produce skeletons of calcium carbonate (Orr et al., 2005). The formation of limestone (calcification) is not only the foundation for the growth of the coral reefs but also helps to counteract the process of reef erosion. The CO2-determined impairment of calcification hampers the spread of coral reefs to cooler marine regions. Consequently, both increased temperatures and increased CO2 concentrations must be expected to drastically constrain the distribution areas of the present-day coral reefs (Hoegh-Guldberg, 2005).
     In laboratory experiments simulating a doubling of the CO2 concentration in the atmosphere, the calcification rate for corals dropped by 11–37 per cent (Gattuso et al., 1999). Modelling calculations by Kleypas et al. (1999) confirm these results. According to their findings, calcification today has already fallen by 6–11 per cent compared to pre-industrial rates. With a doubling of CO2, a further drop of 8–17 per cent compared to today’s rates was calculated. Decreased calcification results in slower expansion of the coral skeleton and therefore a decreased competitive capacity for space in the coral reef. In addition, skeletons of lower density are produced, which are more delicate and vulnerable to erosion.
     The calcification rate is influenced not only by CO2 concentrations but also by water temperature. Increased seawater temperatures can lead to higher metabolic activity and increased photosynthesis rates of the symbiotic algae and thus to increased calcification by the corals (Lough and Barnes, 2000). McNeil et al. (2004) conclude from in-situ investigations and model calculations that the calcification rates of corals in the year 2100 could be at as much as 35 per cent above the pre-industrial rates in spite of decreasing aragonite saturation due to marine warming, presuming an adaptation of the corals to higher seawater temperatures. These hypotheses are scientifically contested (Kleypas et al., 2005). For calcification to increase over the long term the temperature rise of seawater has to remain below the thermal tolerance limit of the corals. So the key question here is whether the tropical corals and their symbiotic algae can genetically adapt their temperature tolerance quickly enough to keep pace with rising seawater temperatures. The question of possibly increased calcification would be moot if the corals die from heat stress.

2.4.3     Measures for coral conservation

Due to the specialization of tropical coral reefs within a narrow range of temperatures, aragonite supersaturation, and high light availability conditions, climate change, in addition to local anthropogenic stress factors, poses a great threat to them. Increasing occurrences of coral bleaching highlight the need for rigorous implementation of climate policy measures. Even the healthiest reefs are not immune to these impacts, as the status report on coral reefs points out (Wilkinson, 2004). It has, however, been found that ‘healthy’ reefs located in pristine areas have the greatest chance of surviving coral bleaching episodes. So it makes good sense to strengthen the resilience of coral communities through protective measures.
     For this purpose the establishment of marine protected areas (MPAs) is considered to be especially effective, preferably in their most stringent form as No-Take Areas, which are closed to fishing (Hughes et al., 2003; Bellwood et al., 2004; Section 2.6.2). The focus on protected areas, however, should not lead to neglect of the remaining much larger reef areas that are not designated as protected. The critical functional groups (communities of particular, often regionally different species that maintain the ecosystem) have to be protected at a regional level; otherwise, the area loses resilience.

2.6     Recommendations for action: Improving the management of marine ecosystems

Anthropogenic climate change has the potential to cause considerable additional stresses to marine ecosystems in future (Section 2.2–2.4). It is likewise possible that it will have an impact on commercial fishing, given that the naturally occurring variability in climate already plays a major role in the fluctuation of fish stocks. In some regions, anthropogenic temperature change is on the point of exceeding the highest levels ever reached by natural variability (e.g. in the Arctic: ACIA, 2005). Given the current state of knowledge, however, it is virtually impossible to make globally aggregated forecasts of the impact of climate change on marine ecosystems. As no comparable historical data or empirical figures are available, forecasts would amount to little more than speculation.
     Mitigation of climate change, particularly by substantially reducing greenhouse gas emissions (WBGU, 2003; Schellnhuber et al., 2006), is crucial if additional stresses on marine ecosystems are to be limited. One mitigation option of direct relevance to oceans is sub-seabed storage of CO2 (Chapter 5). Due to the geophysical lag effects of the climate system, however, adaptation measures will be unavoidable even if rigorous efforts are made to reduce emissions. For this reason, adaptation to climate change will be the focus of this section. Priorities set by WBGU in this context are fisheries management and marine protected areas. It makes sense to adopt adaptation measures for other reasons, too, as climate change is only one of many ways in which human influence degrades marine ecosystems (overfishing, destruction and pollution of marine ecosystems, invasion of alien species, etc.; GESAMP, 2001; UNEP, 2002). Even considered separately, each of these factors poses a considerable challenge to the international community.
     Coupling and synergistic effects between the different factors call for particular attention (Brander, 2005). A coral reef that has suffered prior damage due to pirate fishing with poison or dynamite will be particularly sensitive to periods of unusually high temperatures (Section 3.3; Wilkinson, 2004). Where stocks of a species have been heavily depleted by overfishing, they will regenerate much more slowly if coastal ecosystems that serve as nursery grounds are exposed to severe stresses as a result of infrastructure measures or pollution, or if there is an additional stress in the form of warming. It is easy to find other examples to add to this list (for an overview, see Brander, 2005). Consequently, it is vital to consider the various factors in an integrated manner if management of marine ecosystems is to be successful. In the coming years, it will therefore become all the more important to rein in current overfishing practices and other destructive anthropogenic factors at the same time, so that marine ecosystems will have sufficient resilience to cope with climate change (Brander, 2005).
     For these reasons, the ecosystem approach for promoting conservation and sustainable use of ecosystems and their living resources that was developed under the Convention on Biological Diversity and reaffirmed at the World Summit on Sustainable Development (WSSD) is vitally important (see e.g. OSPAR, 2003). In order to implement this approach, research and monitoring of marine ecosystems and ocean regimes must be improved and this knowledge applied to the assessment and management of fish species of commercial interest (FAO, 2003; Section 2.7). Current knowledge regarding the exceedingly complex interactions between climate, physical and chemical conditions in the sea, marine ecosystems and fishery is inadequate for making reliable predictions relating to how marine systems are likely to respond to climate change (ACIA, 2005; Section 2.7). Inadequate knowledge must not, however, serve as a pretext for delaying conservation and management measures. On the contrary: in accordance with the precautionary principle, action must be taken even if uncertainty prevails. This precautionary principle is already enshrined in multilateral fisheries policy, e.g., in the United Nations agreement on migratory fish stocks.

2.6.1     Fisheries management

National and international institutions face the challenge of dealing with the complex set of anthropogenic factors that currently characterizes the fisheries sector, of making decisions regarding sustainable management of the sector on this basis and, not least, implementing these decisions on the ground. Up to now, the situation has been less than satisfactory: calls for sustainable management of fish stocks, which we have been hearing for decades now and are reiterated time and again at international conferences, have hardly brought about any improvement in the overall situation (although there are some important regional exceptions) (Section 2.3). Half of all fish stocks are fully exploited, while a quarter of fish stocks have already collapsed as a result of overfishing (FAO, 2004). Illegal and unregulated fishing on the high seas continues to be an unresolved problem despite international efforts (FAO, 2001). In the future, this already very difficult situation will be exacerbated by climate change. In addition, new technologies have extended the boundaries of the feasible ever further in fishing, for example fish finding using much-improved sounding technology, and reaching great depths or particular stocks using modern catching methods. Nowadays, virtually no ocean habitat is inaccessible to fishing activities.
     For these reasons, management of fishing grounds based on the ecosystem approach and on the precautionary principle is urgently needed in order to maintain the resilience of marine ecosystems (Scheffer et al., 2001; Pikitch et al., 2004). The Agreement on Conservation and Management of Straddling Fish Stocks and Highly Migratory Fish Stocks in the high seas applies the precautionary principle. In FAO programmes, too (e.g. in the Code of Conduct for Responsible Fisheries; FAO, 1995), the precautionary principle and ecosystem conservation have played a major role for some time. The EU’s strategy for protection of the marine environment names the ecosystem approach as a key component (EU-Kommission, 2005), although it excludes fisheries policy from this strategy. Moreover, the binding requirements it sets out are very vague, with the result that the strategy’s effectiveness is likely to depend largely on how it is implemented by the Member States.
     Broad-based enforcement of sustainable fisheries management is long overdue (Fujita et al., 2004). The scientific and conceptual foundations have already been laid and have been reaffirmed repeatedly in the international policy arena. In many cases legislative provision at national and regional level is already adequate. In the European Union, for example, the Common Fisheries Policy has been endowed with a legal framework that is perfectly acceptable in environmental policy terms. It has yet to be rigorously implemented, however, most notably implementation of adherence to the scientifically-based recommendations from the International Council for the Exploration of the Sea regarding catch quotas. Rapid elimination of excess fishing fleet capacity has also yet to take place (SRU, 2004). It is not the purpose of this section to discuss all the issues relating to global fisheries management and its shortcomings. The aim is rather to formulate recommendations or reinforce existing recommendations relating to the additional problem of climate change and its impact on fisheries.

• A paradigm shift away from publicly subsidized overfishing (SRU, 2004) to a sustainable fisheries sector is long overdue. In order to achieve this, efforts must be urgently intensified to resolve the primary problems of the marine fisheries sector, namely excess fishing fleet capacity, destructive fishing practices, excessive bycatch, inflated catch quotas, illegal or unregulated fishing in the high seas, habitat destruction in coastal ecosystems, and pollution. An increased drive to promote labelling of sustainable marine products is also urgently needed. Implementation of the goals adopted by the World Summit on Sustainable Development (WSSD) is a key yardstick in this context.
• Eliminating subsidies in the fisheries sector is an effective means of slowing overfishing and putting an end to it altogether in the long term. Estimates of subsidies to the fisheries sector worldwide range from US$15–30 thousand million annually (Milazzo, 1998; Virdin and Schorr, 2001). These subsidies should be cut in order to reduce incentives to overexploit the marine environment. At the same time, public funds would be set free for investment in activities that include protecting the marine environment.
• Recent efforts to reduce fisheries subsidies in the context of the WTO are welcomed by WBGU. This relates particularly to subsidies in the OECD countries and especially in the EU (SRU, 2004). The possibility of negative social and ecological consequences arising as a result of cuts in subsidies, particularly in developing countries, due to the search for new ways of earning an income or alternative ways of exploiting the natural environment, must be explored and, where appropriate, taken into account. This must not, however, be allowed to hold up implementation of a swift and consistent change in international policy on subsidies.
• Due to the complex interaction of many factors, both anthropogenic and natural, the integrated ecosystem approach for promoting conservation and sustainable use of ecosystems and their living resources developed under the Convention on Biological Diversity and reaffirmed at the WSSD is vitally important. On the one hand, monitoring of ocean regimes and ecosystem parameters (e.g. indicator species) must be improved; on the other, the resulting knowledge concerning the state of the ecosystem must be integrated into the process of assessing and managing commercially important fish stocks (FAO, 2003).
• The precautionary principle must be rigorously applied as the basis for fisheries management. Particularly when forecasting fish stock dynamics and calculating catch quotas on the basis of these, safety margins should be included to ensure that stocks do not fall below the minimum required for reproduction and that the age structure of the fish population remains healthy, even in the event of a regime shift induced by climate change (King, 2005). Fisheries management must be enabled to respond to regime shifts in good time and with appropriate strategies (Polovina, 2005). One example of the need to adapt in this way is the cod fishery in the North Sea (Section 2.3.1).
• For short-term management (1–5 years), although anthropogenic climate change will have relatively little impact, interannual variability and climatic events such as El Niño may trigger major effects (Barber, 2001). Assessing and forecasting these factors is an important area where research is needed.
• The role of the future climate is currently largely ignored when developing management strategies for the medium term (5–25 years), either because it is seen as something that can be disregarded or because it is considered impossible to foresee. Since climate change can have considerable impact on recruitment and distribution of fish stocks in the medium term, it will become necessary for fisheries management to take these effects into consideration. At present, the impact of climate variability and climatic events on fish stocks can only be analysed after the event. In view of the fact that climate change is already apparent, forecasting capacity needs to be developed in future and used to conduct risk analyses. This applies particularly to sensitive fish populations on the edge of their natural distribution area.
• When developing the models that are used as the basis for setting quotas, there needs to be a shift away from analysing and modelling individual fish populations of commercial interest to ecosystem-based models that take into account the dynamic interactions between climate, ocean and marine ecosystems (Pikitch et al., 2004). The usefulness of static concepts based on the assumption of unchanging environmental conditions is becoming increasingly questionable.
• In the case of terrestrial ecosystems, subdividing the area in question into zones with varying intensities of use is a long-established procedure for solving land-use conflicts (WBGU, 2001). For the oceans too, in the context of marine spatial planning systems, zoning is increasingly recognized as a useful instrument for sustainable, ecosystem-based fisheries management (Pauly et al., 2002; SRU, 2004; Pikitch et al., 2004; Boersma et al., 2004). Marine protected areas have a special role to play as a component of marine spatial planning in this context because, in conjunction with other measures, they represent an important tool for implementing the ecosystem approach. Recommendations relating to this are discussed in more detail in the next Section 2.6.2.

2.6.2     Marine protected areas

2.6.2.1     Definition and motivation

Climate change, ocean acidification and sea-level rise will have considerable impact on the marine environment (Sections 2.2–2.4). These ‘new’ anthropogenic factors, moreover, are affecting marine ecosystems which, in many regions, have already been significantly weakened by overfishing, contamination, invasive species and other human-induced influences. Recommendations for improving fisheries management were presented in Section 2.6.1 above. This section will deal with marine protected areas (MPAs), which – like their counterparts on land – are one of the most important instruments available for ecosystem protection (IUCN, 1994; Kelleher, 1999; Murray et al., 1999).
     IUCN defines marine protected areas as ‘any area of the intertidal or subtidal terrain, together with its overlying water and associated flora, fauna, historical and cultural features, which has been reserved by law or other effective means to protect part or all of the enclosed environment’ (IUCN, 1988).
     MPAs play a particular role in protecting the marine environment, as they are a direct means of implementing the ecosystem approach, and one of the easiest to apply (Royal Commission on Environmental Pollution, 2004). Although they cannot halt either climate change or acidification, nor altogether prevent invasion by alien or highly migratory species, they are an important tool for enhancing the resilience and adaptive capacity of ecosystems. They are also important because they allow mitigation of anthropogenic factors such as overfishing or habitat destruction within their boundaries by means of management or prohibition (e.g. Mumby et al., 2006). MPAs are thus the most important means, for example, of dealing with coral bleaching (Section 3.3.4), because although they do not tackle the underlying cause, they can enhance the general resilience of the reef (Grimsditch and Salm, 2005). However, enhanced scientific understanding of the relationship between resilience, anthropogenic influence and biological diversity is needed (Section 2.7). For coastal protection, near-natural ecosystems are also important: for example, the Asian tsunami of 26 December 2004 was able to penetrate much further inland in places where the mangroves or coral reefs had been destroyed than elsewhere (Danielsen et al., 2005; Fernando and McCulley, 2005). Protected coastal ecosystems are therefore also an important component of strategies for adapting to climate change (Section 3.4.1).
     In addition to their role in ecosystem protection, MPAs can also be useful as a fishery management tool for conservation of commercial fish stocks, e.g. where traditional management has failed and overfishing has resulted, or to safeguard against future mistakes of this sort (Bohnsack, 1998; Pauly, et al., 2002; Gell and Roberts, 2003). Even conservation-based fishery can have a range of negative effects on marine ecosystems, and establishing MPAs can mitigate these (Palumbi, 2003). MPAs can also provide a retreat for species that are fished, but are not subject to monitoring or management. Coastal ecosystems and estuaries are also important for protecting the nursery grounds of many fish species against climate variability (Attrill and Power, 2002). MPAs should be viewed in conjunction with the traditional tools of fishery management, for one reason because quota setting could also be affected by the establishment of a large-scale network of MPAs if fishery activities are restricted to the areas outside MPAs (Hilborn, 2003).
A graded system of protected area management categories is applied in the marine environment too. This ranges from total protection (marine reserves where extractive use is prohibited) to areas serving primarily to uphold sustainable and/or traditional use of marine resources (IUCN, 1994). Areas that are closed to fishing activities (no-take areas) represent a special type of MPA. Different categories of protected area often sit side by side, with core areas under strict protection and peripheral zones with fewer restrictions relating to use (Agardy et al., 2003). The effectiveness of MPAs can be improved if they form part of a protected areas system geared towards ensuring ecological representativeness and creating networks.
     Although differences of opinion still persist with regard to the optimum design and management of marine protected areas (NRC, 2001), there is broad consensus that adaptive management, linking of individual MPAs to protected areas systems, participation or co-management and an integrated view of the relationship between MPAs and the intensively exploited areas outside them are important aspects of MPA design and management.

2.6.2.2     International policy objectives

Wecause of the double use of MPAs for ecosystem conservation on the one hand and as a fishery management tool on the other (Lubchenko et al., 2003), as a guard rail, WBGU recommends designating 20–30 per cent of the marine area for inclusion in a linked system of MPAs (Section 2.5.1). Current protected areas coverage amounts to less than 1 per cent of marine habitats. Considerable catching up is therefore required and it is only very recently that this has led to the formulation of a number of policy objectives in this area:

• At the WSSD, the international community set itself the target of establishing an ecologically representative and effectively managed network of marine protected areas by 2012 (WSSD, 2002).
• The World Parks Congress reaffirmed this goal in 2003 and made it more specific with the recommendation that at least 20–30 per cent of every marine habitat should be strictly protected (WPC, 2003a).
• In the context of its programme of work on protected areas, the Convention on Biological Diversity has adopted the WSSD target for protected areas, albeit without specifying a coverage percentage (CBD, 2004a).
• A regional example is the OSPAR/HELCOM Convention, which has also set itself the target of creating a well managed and ecologically coherent system of marine protected areas by the year 2010 (OSPAR, 2003).

• The Convention on Biological Diversity – whose objectives also cover marine ecosystems – envisages protected areas as an in situ conservation measure (Art. 8 a).
• The United Nations Convention on the Law of the Sea (UNCLOS) makes explicit mention of special protected areas only in the following context: Art. 211, para. 6 allows tightening of protective regulations in clearly defined areas in connection with measures to prevent pollution of the marine environment from vessels. Coastal states wishing to apply this provision can designate such an area within their respective exclusive economic zone, and apply to the competent international organization, the International Maritime Organization (IMO), requesting affirmation of the need for special measures to protect this area. Reasons may be related to the area’s oceanographical and ecological conditions, its utilization or protection of its resources. Under this provision, however, special measures are restricted to regulations for the prevention of pollution from vessels.
• In various agreements in the field of international law on protection of the marine environment, there are terms and concepts that have similar objectives to MPAs. The International Convention for the Prevention of Pollution from Ships (MARPOL Convention), for example, provides for the establishment of ‘Particularly Sensitive Sea Areas’ (PSSAs). The purpose of the protected areas in this case is to ensure protection from pollution from ships in particularly vulnerable areas that are to be designated accordingly. Further examples are the Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR Convention) and the Convention for the Protection of the Mediterranean Sea against Pollution (Barcelona Convention), both of which have a specific additional annex (OSPAR) or protocol (Barcelona Convention) on the establishment of ‘Specially Protected Areas’. IMO has adopted a Resolution (A.885 (21)) setting out procedures for the establishment of PSSAs (Hohmann, 2001).
Requirements for designating protected areas vary according to the maritime area in question. The limits set out under international law depend primarily on the maritime area in which the protected area is to be established in a given case (Box 2.6-1; Proelß, 2004).
• In principle, in the case of internal waters and territorial sea, the coastal state has the freedom to decide on designation of MPAs. This may be limited to a certain extent, but only with regard to restrictions on shipping activities (Box 2.6-1).
• The legal situation with regard to exclusive economic zones (EEZ) is similar. UNCLOS accords particular sovereign rights to coastal states in this maritime area concerning exploitation and conservation of the living and non-living natural resources in the maritime area in question, including the seabed and its subsoil. In relation to establishing MPAs, this means that the coastal state has the freedom to adopt measures so long as these measures are aimed at restricting exploitation of natural resources. However, the establishment of an MPA in this area may not, for example, restrict the right of innocent passage of foreign vessels (Box 2.6-1).
• On the high seas, although the establishment of MPAs is not ruled out in principle (Proelß, 2004), it entails certain legal problems (Platzöder 2001; Warner 2001). These are discussed in Section 2.6.2.4.
• In contrast to agreements that are limited to particular regions (e.g. the OSPAR Convention, the Convention for the Protection of the Marine Environment of the Baltic Sea – the Helsinki Convention, or the Barcelona Convention) there is currently no global instrument of international law that specially promotes designation of cross-border MPAs or places any obligation on states to do so.

Box 2.6-1 Maritime zones under international law The United Nations Convention on the Law of the Sea (UNCLOS) lays down the fundamental provisions under international law with respect to delimitation of maritime zones, and this is also relevant for designation of MPAs. In brief, the maritime zones under UNCLOS are organized as follows (Figure 2.6-1): • High seas: According to Articles 86 and 89 UNCLOS, ‘high seas’ are those parts of the sea that are not subject to the sovereignty or jurisdiction of any state and as such constitute ‘an area under international administration’. The principle of freedom of the high seas applies in this area. This comprises primarily freedom of navigation and of overflight, freedom to lay submarine cables and pipelines, freedom to construct artificial islands and other installations, freedom of fishing and freedom of scientific research. These freedoms may be exercised by all states, including land-locked states. In the area constituting the high seas, no state may validly purport to impose restrictions of any sort on other states relating to use of the high seas. Furthermore, international agreements between individual states can always only bind the states that are party to the agreement, and not third states. A distinction is drawn between the high seas and maritime areas that are subject to varying degrees of territorial jurisdiction by the coastal state. According to Article 86 UNCLOS, the maritime areas over which the coastal state has varying degrees of jurisdiction are: • Exclusive economic zone: Here, in the zone that lies on the boundary with the high seas, the coastal state begins to have the right to exercise jurisdiction based on its territorial rights. The corresponding rights of sovereignty, however, are limited insofar as they relate exclusively to exploitation and conservation of the living and non-living natural resources in the zone in question, including those in the seabed and its subsoil. • Continental shelf: This term too is defined in terms of the right to exploit natural resources, although in this case it is specifically those of the seabed and subsoil close to the coast. Thus, according to Article 77 para. 1 and 2 UNCLOS, the coastal state exercises over the continental shelf exclusive sovereign rights for the purpose of exploring it and exploiting its natural resources. The definition of resources that applies in the case of exploitation of the continental shelf, however, is somewhat restricted by comparison with that of the exclusive economic zone (non-living resources of the seabed and its subsoil and ‘immobile’ organisms). • Territorial sea: Here, the special rights of coastal states are no longer limited to exploitation of marine resources, but are identical to actual territorial sovereignty. • Internal waters: This is where the sovereign rights and jurisdiction of a coastal state are most extensive; this maritime zone forms an integral part of a state’s territory. Figure 2.6-1 Maritime zones under the United Nations Convention on the Law of the Sea (UNCLOS). NM = nautical mile (1 NM = 1.852 km). Source: Gorina-Ysern et al., 2004

2.6.2.4     Marine protected areas in the high seas

There are significant deficits in the legislation pertaining to establishment of marine protected areas in the high seas (CBD, 2005a). The regional multi-functional maritime conventions only cover a very limited range of marine areas outside national jurisdiction, with the result that large areas of the world’s oceans are not covered. In addition, existing regional fishery management regimes are limited to particular fished species such as tuna, while species that are not intensively fished are excluded. Application of the ecosystem approach under these regimes is also inadequate.
     The United Nations Convention on the Law of the Sea (UNCLOS) reaffirms the right to freedom of navigation, which forms part of customary international law (i.e. it is a right that, in principle, cannot be restricted) (Art. 87 UNCLOS). It is thus out of the question to establish an MPA in the high seas that entails the prohibition or restriction of shipping activities. In addition, states may not conclude agreements establishing MPAs in the high seas that might be detrimental to third states that are not parties to the agreement. A corresponding commitment agreement among states in a particular region that are the main users of the high seas thus has no binding effect on third states. ‘Freedom of the high seas’ also includes, for example, the fundamental right of every state to use the marine resources of the high seas (e.g. fishing). In contrast to the case of freedom of navigation, however, this right is not unrestricted, and correspondingly there is a range of international conventions regulating the use of living marine resources in the high seas, especially relating to particular species. Examples include the prohibition on fishing of anadromous species (e.g. salmon, which spawns in freshwater but lives in seawater) in the high seas in accordance with Art. 66, para. 3(a) UNCLOS, or the whale sanctuaries provided for under the International Convention for the Regulation of Whaling (Gerber et al., 2005).
     Urgent problems such as the increasing destruction due to fishery activities of sensitive undersea structures that are particularly rich in biological diversity (e.g. seamounts or cold-water coral reefs; UNGA, 2004; CBD, 2004b), and the scale of illegal and unregulated fishing (FAO, 2001), necessitate rapid identification and implementation of solutions for marine protection in the high seas. In view of the clear will of the international community to step up use of marine protected areas as a tool, action is needed to improve provisions pertaining to MPAs in the high seas under international law. In developing a regime for marine protected areas in the high seas, the following specific requirements should be met (CBD, 2005a):

• Moving beyond approaches that focus on specific species or regions, the aim should be to arrive at an integrated approach enabling the creation of large-scale networks for protecting marine ecosystems in the high seas too. Freedom of access to marine protected areas in the high seas should also be guaranteed for scientific research, insofar as this does not run counter to conservation objectives.
• In view of the problem of illegal and unregulated fishing in the high seas – which cannot be tackled by individual states because they do not have territorial jurisdiction to enforce the law in this maritime zone – mechanisms for enforcing the relevant conservation requirements in the high seas must be considered (Platzöder, 2001; Warner, 2001).
• In view of the need to create networks that cover a broad area, efforts should be made to ensure that the establishment of MPAs in the high seas – contrary to what has happened hitherto under the various specific conventions – takes place in a coordinated fashion (CBD, 2005b).

2.6.2.5     Negotiation processes

At global level, negotiations on MPAs are taking place notably in two parallel political processes:

• In the Convention on Biological Diversity, MPAs are on the agenda of a Working Group on Protected Areas, including protected areas beyond the limits of national jurisdiction. However, attempts to agree specific areas of the high seas that might be suitable for designation as MPAs or set a specific target of establishing 5–10 MPAs in the high seas by 2008 have so far failed due to resistance on the part of a few fishing nations (e.g. Iceland, Norway, New Zealand).
• In 2004 an informal Working Group of the General Assembly of the United Nations was established (UNGA, 2004) with a broad mandate relating to marine biodiversity conservation beyond areas of national jurisdiction. This Working Group convened for the first time in February 2006. Although the positions of the different country groups on MPAs still diverge widely, many states acknowledge the need to act to close this gap in international law.

2.6.2.6     Recommendations for action relating to marine protected areas

Despite the importance of marine protected areas, protection of the marine environment must not be reduced to this one instrument alone. Adherence to the ecological guard rail (20–30 per cent of marine ecosystem areas under protection; Section 2.5) is indispensable for conservation of the marine environment, but areas outside MPAs must also be managed sustainably using the ecosystem approach. A particularly important precondition for the success of MPAs is the urgently needed enforcement of sustainable fisheries management (Section 2.6.1). Another precondition is adherence to the guard rails on climate change and ocean acidification (Box 7-1), without which even an excellent protected areas system will lose most of its impact. In addition, it is not sufficient to plan, designate and link MPAs; they must also be well managed and adequately funded. Only in this way is there a chance that the target of significantly reducing the rate of loss of biological diversity by 2010 can be met in marine environments too.

Implementing international targets

• The rate of growth of MPA coverage, currently3–5 per cent per year, is much too low to meet the internationally agreed targets on time, given that current coverage is less than one per cent (Wood et al., 2005). Efforts in this regard need to be intensified significantly.
• MPAs should be sufficiently large and linked with one another in protected area systems; they should encompass zones with different forms and intensities of use and be part of an integrated system of management that includes neighbouring continental shelf and coastal areas. In addition, they should be designed to be flexible and adaptive, since climate change may impose the need for reorganization if ecosystem processes change or shift in location (Soto, 2002). Improving the scientific basis in this context should be carried out in parallel with ongoing management activities. Adaptive management strategies and flexibility are crucial to deal with local impacts of climate change that are difficult to forecast.
• In the territorial sea and in EEZs, states can begin to implement international objectives now without coming up against problems relating to international law. The EU Habitats Directive and Birds Directive are both fully applicable in EEZs. In Germany this has already taken place: in the context of the Natura 2000 network, around 30 per cent of the German maritime area of the EEZ has been registered as protected areas with the European Commission. However, imposing restrictions on fishing is not a straightforward matter at present, as this is an area in which EU competences apply. MPAs should be used more effectively as a tool for sustainable fisheries, for example by setting long-term or temporary limits on fishing activities in protected areas (SRU, 2004).
• In developing countries particularly, there is a great deal of catching up to be done. In these countries, not only are there very few MPAs, but the designated protected areas in many cases amount to little more than ‘paper parks’ in which effective protection is not or cannot be enforced. Development cooperation should therefore make designation and management of MPAs a priority. In doing so, cooperation should take place between protected areas specialists and fishery representatives, and the local population should be involved in planning and management.

Securing funding

Due to the difference between the guard rail for protection of marine ecosystems (20–30 per cent coverage) and current coverage (less than 1 per cent of marine ecosystems are currently protected; Section 2.5.2), the need for additional funding is considerable. According to the findings of Balmford et al. (2004), the annual costs associated with protecting marine ecosystems on this scale would be in the region of US$ 5–19 thousand million. This figure includes both ongoing costs and one-off costs relating to implementation. Indirect costs, however, such as costs to businesses in the fisheries sector arising as a result of fishing exclusion, are not included.

• WBGU considers it to be the responsibility of national governments and the international donor community to ensure adequate funding for management of marine protected areas. Up to now, it has often proved impossible to ensure availability of adequate long-term financing: payments from public donors are often small and there are competing demands on these resources. The same applies to international transfer payments such as those made by the Global Environment Facility (GEF) or by donors in the context of bilateral development cooperation (OECD, 2002; GEF, 2005a). WBGU calls upon public donors to undertake additional efforts to make adequate funding available on a sustained basis. Complementary funding may be raised by means of instruments such as user charges or promoting private donations for conservation measures (Emerton, 1999; Morling, 2004).

Protected areas in the high seas: Closing gaps in international law

The ongoing process of political negotiations to develop an instrument for designating and managing protected areas in the high seas is to be welcomed, and the German Federal Government should give this process vigorous support. The basis for this is UNCLOS. Despite the fact that UNCLOS lays greater emphasis on rules pertaining to usage than on protection and conservation of marine resources, it nevertheless also provides the primary legal framework for protection of the marine environment (Platzöder, 2001). Fundamentally changing UNCLOS is not a political option, whereas adding moderate supplementary provisions to the current law of the sea seems feasible in both political and legislative terms. Options for doing this include the following:

• The primary task is to develop a multilateral agreement on designation of protected areas and corresponding systems in the high seas and append this to UNCLOS, either as an additional protocol or as a supplementary convention. A precedent already exists for this type of procedure in the form of the Agreement relating to the Conservation and Management of Straddling Fish Stocks and Highly Migratory Fish Stocks that is tied to UNCLOS, and which is concerned primarily with the exploitation of relevant fish species in areas outside national jusrisdiction.
• It would make sense to assign the monitoring and coordination tasks to the same – yet to be established – international regime. From a legal perspective, the mechanism should be laid down primarily in the above-mentioned multilateral agreement. With regard to this, the proposal to establish a Global Oceans Commission, put forward at the first International Marine Protected Areas Congress in 2005 in Geelong (Australia), can be considered as being along the right lines.
• The Convention on Biological Diversity (CBD) should also be amended or supplemented accordingly, whereby care will need to be taken to avoid overlap. The CBD has developed considerable expertise in the field of biodiversity conservation, and should therefore be involved in the UNCLOS negotiation process, for example by providing expert input. At the same time, the new regime’s functions and areas of responsibility should be explicitly recognized by the CBD. In order to achieve this, the relevant CBD negotiation processes should be intensified with the aim of securing a key role for the CBD in the design of MPAs in the high seas in terms of content, e.g. deciding on appropriate instruments and criteria for the selection of MPAs. With a view to supporting cross-border conservation efforts, it would be useful to explore whether it would be worthwhile developing a protocol to the CBD on protected areas in the medium term based on the findings of the current Working Group. Such a protocol should cover the whole spectrum of protected areas and not merely MPAs.
• The informal Working Group of the General Assembly of the United Nations on marine biodiversity in areas beyond national jurisdiction has taken the first step towards closing the gap in international law regarding MPAs in the high seas. At the next UN General Assembly, the German Federal Government should urge that this sound basis be used to ensure continuation of the negotiation process.

2.7 Research recommendations

Research into climatic factors

• Behaviour of sea ice: Possibilities for monitoring changes in the thickness of Arctic sea ice in particular remain inadequate, and sea ice simulation models need to be developed further in order to improve estimation of future sea ice dynamics.
• Behaviour of continental ice: The future behaviour of Greenland’s continental ice sheet is likely to be decisive for the future dynamics of ocean currents in the Atlantic. Methods of modelling continental ice sheet dynamics need to be improved significantly.
• Stability of Atlantic circulation and risks of changes in currents: Climate models continue to differ considerably as regards the information they provide on the future stability of ocean currents. Reasons for this include internal oceanic processes (such as mixing), which as yet are poorly understood, and interactions with other climate components (e.g. the freshwater budget of the North Atlantic), which are difficult to quantify. Observation and enhanced modelling efforts could help to reduce the uncertainties in this area.

Research into marine ecosystems, fisheries and marine protected areas

• Monitoring: Observation data covering large areas of the oceans over long periods are vital, particularly as regards the nutrient situation and plankton (especially zooplankton). Because they provide, among other things, important inputs for modelling marine ecosystems, support should be given to appropriate monitoring programmes (e.g. using the Continuous Plankton Recorder).
• Understanding the systems involved: Too little is known about the structure and dynamics of marine ecosystems to be able to make reliable estimates of the impact of climate change. Examples include the significance of temperature effects on primary production, the impact of sea-ice retreat, or decoupling of trophic levels as a result of disparities in species’ responses to climate change (e.g. migration, adaptation). Increased emphasis should be given to ecosystem-based research approaches with a view to enhancing understanding of the relationships between anthropogenic disturbance, biological diversity and resilience of marine ecosystems and incorporating this into new ecosystem models. The international research project GLOBEC and the new IMBER project have drawn up a detailed catalogue of issues (GLOBEC, 1999; IMBER, 2005). Promotion of these interdisciplinary research initiatives by national research promotion agencies should be stepped up.
• Modelling marine ecosystems: In order to gain a better understanding of the impact of changes in climatic factors (temperature, wind and ocean current patterns, etc.) on marine ecosystems, knowledge regarding the various ecosystem components must be integrated into improved ecosystem models and coupled with recent climate/ocean models.
• Improving the basis of fisheries management: In order to implement the ecosystem approach in fisheries management, there needs to be a shift in focus in terms of model development away from examination of individual fish species under the assumption of constant environmental conditions to a more integrated ecosystem modelling approach. Qualitative models should also be used to achieve this, incorporating expert knowledge regarding processes in dynamic systems (Kropp et al., 2005). Special attention should be paid to the impact of natural climatic variability and anthropogenic climate change on fish population dynamics, as well as to the socio-economic consequences and possible adaptation measures.
• Design and management of marine protected areas: Development of the theoretical basis for the design of marine protected areas should move away from the study of individual species towards multi-species, ecosystem approaches. Of particular importance in this context is the networking of MPAs with one another and with concepts for sustainable use of the surrounding coastal and marine areas. There are many unanswered questions regarding the design of MPAs in view of climate change and the potential for adaptation. In line with the principle of adaptive management, research and monitoring issues need to be given greater consideration in the design and management of MPAs. The basis for defining guard rails and coverage targets needs to be improved, particularly the basis for designating the proportion of an area that should be strictly protected (no-take areas). Furthermore, there is a need to increase research evaluating participatory approaches (e.g. community-based management) and the use of traditional knowledge, and to evaluate approaches that draw on management experience within the local population.