In this section we bring forward key findings that outline how climate change may impact on soil invertebrate communities of the polar regions, information that is useful for understanding potential future states. However, as will be apparent shortly, there is a dearth of in-depth knowledge across taxa and many groups have received little attention. We will focus on broad-scale responses but recognize that fine-scale responses may differ from those observed at larger scales. For instance, local soil moisture availability will depend on topography and vegetation type in addition to the interaction between changed precipitation and temperature (e.g. Peck et al. 2006).
Climate related impacts on soil invertebrates
Soil invertebrates of the polar regions are well adapted to variation in climatic conditions at daily to annual time scales, and it has therefore been suggested that climate changes may not have a substantial direct impact on these organisms (e.g. Peck et al. 2006). There is however mounting evidence that climate change can impact soil invertebrate communities in the polar regions. In this section we discuss how warming and increased precipitation as well as other climate change related factors, such as freeze-thaw cycles (FTCs), depth of active layer and UV-B radiation, might influence soil invertebrate communities. Impacts more closely associated with microbial and vegetation responses to climate change will be considered in the next section.
A review of the current literature revealed a handful of studies that report on long term trends of soil invertebrates associated with observed climate changes and a limited number of studies that report on the impacts of climate change manipulations. After excluding articles that presented results covered elsewhere, we were left with seven studies with 58 observations (Kennedy 1994; Sinclair 2002; Convey et al. 2002; Convey 2003; Bokhorst et al. 2008; Day et al. 2009; Simmons et al. 2009) and five studies with 15 observations (Coulson et al. 1996; Reuss et al. 1999; Sjursen et al. 2005; Dollery et al. 2006; Tsyganov et al. 2011) that report on soil invertebrate responses to experimental warming and/or water addition in Antarctic and the Arctic, respectively. None of these studies report on the response of soil invertebrates to water addition in the Arctic, while three studies report on the impact of water addition in the Antarctic with two of them further reporting on the interactive effect of warming and water addition. The observed responses (positive, neutral, negative) to experimental warming and/or water additions of all soil invertebrate groups investigated are summarized in Fig. 2 (mean effect sizes presented as ln response ratios of dominant groups are presented in Table S1). Care should be taken in interpreting observed responses to climate change simulations as such manipulations are often poor proxies for natural climate change (Bokhorst et al. 2011), and high variability in soil invertebrate communities within ecosystems may obscure significant patterns under low replication. However, a general pattern emerges in the climate change responses reported to date (Fig. 2). Across groups of soil invertebrates and ecosystem types the response to climate warming is highly variable with negative, positive and no responses observed, while water additions generally have no impact or a positive influence on soil invertebrate communities. Closer examination of the reported impacts indicate that most negative warming responses are displayed by invertebrate groups sensitive to low soil moistures (e.g. collembolans; Day et al. 2009) in ecosystems where soil moisture availability may already be a limiting factor (e.g. Coulsen et al. 1996; Simmons et al. 2009). Similarly, in both of the studies that reported on the interaction between warming and water addition, negative impacts of warming on nematodes (Simmons et al. 2009) and collembolans (Day et al. 2009) were alleviated by water addition.
Figure 2. Summary of observed soil fauna group responses to warming and/or water addition in (a) the Antarctic and (b) the Arctic. Only studies that experimentally increased mean temperature (i.e. not extreme events) and/or added water to plots are included. An observation is defined as the response of a distinct soil fauna group (including Protista, Rotifera, Tardigrada, Nematoda, Collembola, Acari, Diptera) to climate change within a single site or habitat type, i.e. one study can have multiple observations if it reports on responses of several soil fauna groups, ecosystems and/or sites (soil fauna responses were considered significant if P < 0.05). When multiple studies were found to report results from the same experiment, only the most recent, longer-term, data were included.
Download figure to PowerPoint
Combining experimental data and long term records of soil invertebrate community responses to observed climate change allows us to draw some conclusions on potential belowground impacts. Experimental warming has been shown to increase nematode densities in the Arctic and maritime Antarctic (Convey 2003), while a local cooling trend observed between 1986 and 2000 (Doran et al. 2002) had a negative impact on densities of the dominant nematode Scottnema lindsayae in the MDVs (Barrett et al. 2008b). These results suggest that warming will have a largely positive impact on local nematode densities but may change community composition (e.g. Nielsen et al. 2011b) although local responses may vary depending on local soil characteristics (e.g. Simmons et al. 2009). Warming has also been shown to increase mite densities in Polar ecosystems (Kennedy 1994; Sjursen et al. 2005; Dollery et al. 2006), although other studies found no response (Coulson et al. 1996; Webb et al. 1998) suggesting that the response is both habitat and group specific. For instance, Sjursen et al. (2005) found an increase in oribatid mites and a decrease in mesostigmatid mites under increased temperatures, respectively, although the pattern was only significant in a glade and a fellfield and not in a heath. By contrast, collembolans often respond negatively to experimental warming in the Arctic (Sjursen et al. 2005; Dollery et al. 2006) while both positive (Kennedy 1994; Day et al. 2009) and negative responses have been observed in maritime Antarctica (Bokhorst et al. 2008). The idiosyncratic responses of mites and collembolans to warming are supported by laboratory studies, which show that the mites are generally more tolerant to high temperatures than collembolans (Block et al. 1994; Hodkinson et al. 1996). Moreover, it should be mentioned that there is often a change in community composition under changed climate even if there were no overall response in density of the group (e.g. Webb et al. 1998; Sjursen et al. 2005; Dollery et al. 2006). There is very little information on the response of other taxa for either the Arctic or the Antarctic. It is however established that water stress is likely to impact groups that live in moist to wet soils or on water films on soil aggregates (Convey et al. 2003), such as nematodes, rotifers, tardigrades, enchytraeids and Diptera larvae (Hodkinson et al. 1999), and soft bodied animals such as collembolans and prostigmatid mites as well as nymphs of other mites (e.g. Day et al. 2009; Bokhorst et al. 2012). In short, increased temperatures are likely to increase the abundance of soil invertebrates in both the Arctic and the Antarctic as long as it is not associated with a significant decrease in soil moisture availability, and it is furthermore likely to lead to a shift in the composition of the belowground communities as soil organisms differ in their ability to cope with climatic changes.
As mentioned earlier a ‘novelty’ of predicted climate change scenarios is the potential increase in the frequency and magnitude of extreme events (ACIA 2005; Tebaldi et al. 2006; Krinner et al. 2007), and there is evidence that such events may have a disproportionally large impact on soil communities in the polar regions. For example, a particularly warm summer increased soil moisture availability throughout Taylor Valley (MDVs, Antarctica), which was evident for several years, and caused a change in nematode community structure and an increase in the abundance of a nematode species associated with more moist soils (Barrett et al. 2008a). However, life in the polar deserts of the MDVs is generally limited by water due to low temperatures, and increased temperatures may have opposite impacts elsewhere (Nielsen et al. 2012). Accordingly, an unusually hot and dry Arctic summer caused high mortality of collembolans in an area with relatively low precipitation (Coulson et al. 1996), indicating that drought during hot summers may have negative impacts on belowground community composition in the Arctic. Similarly, the Ward Hunt Island region in the Canadian high arctic experienced a hot summer in 2008 (air temperatures up to 20.5 °C), and high temperatures had a significant influence on the landscape scale distribution of key soil microhabitats with potential large-scale impacts on soil communities (Vincent et al. 2009). Such extreme events may therefore lead to alternate ecosystem states with long lasting effects. However, very few studies have experimentally assessed the impacts of extreme events in the polar regions and we could learn a lot from such manipulations. For instance, recent work has shown that extreme winter warming has a strong negative impact on microarthropod communities (e.g. Bokhorst et al. 2012). This work highlights potential negative implications of increased winter temperatures or through decreased snow cover which buffers temperature variability in the soil.
Other climate change related factors that may influence on belowground communities include the frequency of FTCs, the development and duration of ice layers, and increased UV radiation. It is likely that increased physiological stress associated with FTCs could have a negative impact on soil invertebrates (Turner et al. 2009a) but the impact will depend on the severity of the FTCs. A study conducted in the sub-arctic found that the frequency of FTCs had a positive influence on the abundance of collembolans and oribatid mites at the site where the increase in FTCs were most pronounced (Konestabo et al. 2007). As the authors hypothesize it is likely that an increase in time where the soil is thawed in the plots with increased FTCs compared with continuously frozen may outweigh the negative effects of the FTCs themselves. In particular, it needs to be considered that most polar region biota are well equipped to survive and even metabolize at temperatures well below the point of freezing, and only severe frost may have any significant impact on their survival. Moreover, the impact of increased variation in air temperature on soil temperature depends strongly on local characteristics, and it has for instance been shown that the temperature of Antarctic dry soils are much more responsive to changes in air temperature than moist or wet soils due to the buffering capacity of water (Lewis Smith 1999). By contrast, the aforementioned study by Bokhorst et al. (2012) indicated that an increase in FTCs negatively impacted mite and collembolan densities. The authors ascribe the contrasting results to a higher frequency of FTCs than that imposed by Konestabo et al. (2007). Hence, the impact of FTCs on belowground communities will depend on frequency and how low the realized soil temperatures are. Moreover, the susceptibility to FTCs is group specific with laboratory studies suggesting that collembolans are less tolerant to FTCs than oribatid mites (Coulson et al. 2000). Changes in the frequency of FTCs may thus cause a shift in soil communities.
Also altered depth and extent of snow cover, permafrost melt and recession of glaciers may impact belowground communities. Snow packs provide insulation against climate extremes, and soils underneath packs are typically warmer and show lesser variation in temperature (Walker et al. 1999). Snow packs also influence soil moisture availability and soil chemistry, and through this have a large impact on the distribution of soil biota (Gooseff et al. 2003) but impacts may be both positive and negative depending on local conditions and more knowledge is needed to make good predictions of future impacts. Warming is expected to increase the depth of the active layer by 30–40% as permafrost melts in the Northern Hemisphere by the end of the twenty-first century (Stendel & Christensen 2002). This may increase the area of high biological activity in the soils and enhance soil biodiversity. However, the melt of permafrost may also have severe impacts on soil communities through the formation of thermokarsts, a problem very evident in the Arctic (e.g. Osterkamp et al. 2000; Vogel et al. 2009). Also the large-scale retreat and melt of glaciers (Cook et al. 2005) will influence belowground communities. For example, the water released from glaciers in arid regions, such as the polar deserts of Antarctica, may enhance water availability and thereby promote soil organisms, and glacial recessions may provide new areas to colonize (Kennedy 1995; Kaufmann 2002).
Finally, it has been suggested that the increased UV radiation associated with the formation of the ozone hole over Antarctica could limit soil communities (Kennedy 1995). We find it questionable that this should have a large direct impact on growth and survival rates of soil biota although there is some evidence that UV exposure can influence belowground invertebrate communities (e.g. Rinnan et al. 2005; Tosi et al. 2005). However, there may be significant indirect effects through impacts on soil microbes and plant communities. When plants are subjected to high UV irradiation they produce UV-B absorbing compounds which increase their metabolic cost and contribute to a potential limitation of aboveground biomass and plant height (Newsham & Robinson 2009). Moreover, incubation studies suggest that UV-B can inhibit the growth of several Antarctic fungi in the surface layers of the soil (Hughes et al. 2003). This could indirectly influence belowground communities by altering resource quality (i.e. plant palatability) and quantity, and through this nutrient cycling (Convey et al. 2002).
Vegetation and microbial driven climate change effects
There is a strong link between the plant community composition aboveground and belowground communities (Wardle 2002), and climate change is likely to have strong effects on soil invertebrate communities through changes in vegetation type. While the strength of the link between above and below ground communities has not been explicitly investigated in the polar regions there is some evidence that support such a link. For instance, as discussed in Nielsen et al. (2011ab) Antarctic soils with vegetation harbor substantially greater abundances of soil fauna and support different communities than does bare ground. It is also apparent that soil fauna communities differ between vegetation types in the Arctic (e.g. Coulson et al. 2003).
Vegetation composition responses to recent climate changes are widespread in the Arctic (Callaghan et al. 2011) and the Antarctic (e.g. Fowbert & Smith 1994; Parnikoza et al. 2009). For example, in the Arctic, soil temperature directs changes in the functional composition of the vegetation (Brooker & van der Wal 2003) and increasing temperatures generally enhance net and gross primary production (Marchand et al. 2004). In maritime Antarctica, the observed local range expansion of vascular plants as well as bryophytes (Fowbert & Smith 1994) appears to be tightly linked to local temperature trends (Parnikoza et al. 2009). Climate warming is also expected to have a positive influence on net carbon assimilation in Antarctica, as the temperature is generally lower than the temperature required for optimum carbon assimilation shown by native plants (Kennedy 1995). Moreover, a warmer and wetter climate is likely to enhance successful colonization events as well as germination of plants residing in the local soil seed bank (Lewis Smith & Ochyra 2006). Hence, considering the current low biomass, productivity and diversity of plants in Antarctica, warming and increased precipitation is very likely to lead to the development of more favorable habitats for soil organisms thus enhancing the complexity of the soil food webs in Antarctica.
Similarly, climate change responses of the microbial communities could have cascading impacts on the soil invertebrate communities, particularly in the polar deserts dominated by vast expanses of bare ground where the soil invertebrates rely almost solely on the microbial communities for resources (e.g. Nielsen et al. 2011b). There is some evidence that microbial communities respond to imposed long-term climate changes (Rinnan et al. 2007; Timling & Taylor 2012), but whether directional responses exist is still unknown. However, several studies have found that warming tends to increase nutrient availability and enhance decomposition rates (e.g. Schmidt et al. 2002; Hill & Henry 2011), which is likely to positively impact invertebrate communities. For example, Rinnan et al. (2007) speculates that limited microbial biomass responses to climate change manipulations could be due to increased grazing by soil invertebrates. This is supported by the large increase in nematode populations observed in the plots (Reuss et al. 1999). Hence, greater microbial activity may contribute to trophic cascades. By contrast, it seems that a warmer and wetter climate will have a more consistent impact on microbial communities in Antarctic ecosystems. A recent study used open top chambers to investigate the impacts of warming on microbial communities across a latitudinal gradient in Antarctica (Yergeau et al. 2012). The authors found that the soil microbial communities show consistent responses to short-term climate warming with an increase in both bacterial (under vegetation only) and fungal abundances, and changes in the bacterial community composition likely reflecting greater nutrient availability (Yergeau et al. 2012). Increased microbial biomass would likely be able to support more soil invertebrates, while changes in the microbial community composition may impact on the species composition of soil invertebrates.
Implications of belowground community responses to climate change for C dynamics
The climate change impacts on soil biota in the polar regions may have implications for global C dynamics through enhanced rates of decomposition and faster turnover of nutrients. While this is mainly driven by enhanced microbial activity, it appears likely that the role of soil invertebrates in decomposition processes, which is currently limited mainly due to climatic constrains of biological activity (e.g. Wall et al. 2008), will increase as the climate becomes more favorable. This is a concern considering the potential positive feedback on future climate changes (Chapin et al. 2005). Already it appears that permafrost melt and greater biological activity associated with climate warming during the late 20th century have caused the Arctic as a whole to switch from being a C sink to a C source (e.g. Schuur et al. 2009; Jahn et al. 2010). However, there is still large uncertainty about the scale of C efflux as warming has been found to both increase (Biasi et al. 2008), decrease (Sjögersten et al. 2008) or have no impact on soil C efflux locally (Lamb et al. 2011). These idiosyncratic responses appear to be related to differences in soil moisture. For example, a study explored soil respiration rates to warming in the Arctic, and found that dry tundra is more responsive to climate warming than moist and wet tundra (Oberbauer et al. 2007). The authors hypothesize that the response of respiration rates in wet and moist tundra to climate warming are dampened due to higher water tables and soil moisture contents (Oberbauer et al. 2007). Similarly, a recent study suggests that Arctic wetland ecosystems generally act as C sinks rather than C sources (Lund et al. 2010). Hence, it is likely that warming will have the most pronounced impact on C stocks in drier habitats initially and may only later, if at all, impact C stocks in moist to wet habitats. Moreover, a study of old C release from Arctic soils showed that the release of old C increased with time since permafrost thaw because initially the increase in C efflux is offset by increased uptake of C by plants (Schuur et al. 2009). This suggests that short-term measurements of soil respiration are unlikely to capture the true extent of C release, and we might therefore have underestimated the potential impacts of Arctic climate warming. By contrast, the minimal organic content of Antarctic soils suggests that these soils are likely to become a C sink under a warmer and wetter climate. Indeed, a recent study found that warming had a positive effect on aboveground biomass as well as C stocks of soils dominated by vascular plants in maritime Antarctica (Day et al. 2008).