- Top of page
- Materials and Methods
- Conflict of Interest
The Antarctic Circumpolar Current (ACC) is a key feature of the Southern Ocean and a primary factor shaping Southern Ocean ecosystems (Tynan 1998). Recent investigations using coupled ocean-atmosphere climate models suggest that the westerly wind belt that drives the ACC is intensifying and shifting polewards in response to global climate change (Thompson and Solomon 2002; Oke and England 2004; Fyfe et al. 2007). These adjustments are mirrored by a gradual southward migration of the ACC (Fyfe and Saenko 2005; Sokolov and Rintoul 2009; Downes et al. 2011), which has seen a southward shift of up to ~1° Lat since the 1950s (Gille 2002). The sub-Antarctic Prince Edward Islands (PEIs) lie in the direct path of the ACC and provide a critical habitat for up to five million top predators (pinnipeds, flying seabirds, and penguins) that breed on land (Williams et al. 1979; Condy 1981; Ryan and Bester 2008). The abundance and diversity of organisms associated with these islands confers on them elevated ecological and conservation status (Ryan and Bester 2008) and makes the PEIs ideal laboratories to study the sensitivity and adaptation of sub-polar ecosystems to long-term natural and anthropogenic changes (Smith 2002).
The PEIs are situated between the two major oceanic frontal systems that delineate the ACC: the sub-Antarctic Front (SAF) to the north and the Antarctic Polar Front (APF) to the south (Fig. 1), lying in the transition zone known as the Polar Frontal Zone (PFZ). The latitudinal position of the SAF plays a critical role in determining the oceanographic conditions around the islands (Perissinotto and Duncombe Rae 1990; Perissinotto et al. 2000; Ansorge and Lutjeharms 2002). When the SAF is far to the north of the PEIs, the flow rate of the ACC around the islands is slow enough that frictional forces between the current and the islands result in the formation of anticyclonic eddies that become trapped in the shallow region between the islands and in their lee (Perissinotto and Duncombe Rae 1990; Ansorge and Lutjeharms 2002). Freshwater runoff from the islands transports macronutrients, particularly reduced forms of nitrogen produced by top predators, into the surrounding waters (Ismail 1990; Perissinotto and Duncombe Rae 1990). The retention of freshwater runoff in the shelf waters also results in increased water column stability, as the depth of the mixed layer becomes shallower than the euphotic zone, thus promoting the development of dense phytoplankton blooms dominated by large diatoms (>20 μm) between and downstream of the islands, a phenomenon known as the “island mass effect” (Boden 1988; Perissinotto and Duncombe Rae 1990). This freshwater input, however, needs to be retained between the islands for a minimum of 15 days to enable the formation of these blooms (Perissinotto and Duncombe Rae 1990). During these blooms the bulk of the phytoplankton production is not utilized directly, and a large portion sinks to the seafloor, thus providing an important food source for the benthic community in the shallow shelf region of the islands (Perissinotto 1992; Pakhomov and Froneman 1999). In contrast, when the SAF lies farther south and close to the islands, the region around the PEIs experiences much higher flow rates, resulting in a flow-through system that prevents the formation of eddies and hence the development of phytoplankton blooms in close proximity to the islands (Perissinotto et al. 2000).
Figure 1. (A) Map of the Indian sector of the Southern Ocean indicating the average geographical position of the three major fontal systems and the position of the Prince Edward Islands (PEIs). STC, subtropical convergence; SAF, sub-Antarctic Front; APF, Antarctic Polar Front. (B) Location of sampling sites (separated into nearshore and inter-island sites) in the vicinity of the PEIs. Dashed oval indicates the region in which anticyclonic eddies form. Corresponding letters (A, B, and C) denote sample collections from different years but similar locations (see Fig. 3). *Shrimp (Nauticaris marionis) data obtained from Pakhomov et al. (2004); **benthos data obtained from Kaehler et al. (2000); ***shrimp and benthos data obtained from current study.
Download figure to PowerPoint
Despite their location in a generally low productivity region (Laubscher et al. 1993; Froneman et al. 2001), the PEIs support a diverse (~550 species) and biomass-rich inter-island benthic community (Branch et al. 1993; Pakhomov and Froneman 1999) due to the enhanced productivity associated with bloom formation at these islands (Perissinotto 1992; Branch et al. 1993; Pakhomov and Froneman 1999; McQuaid and Froneman 2008). A key component of the benthos of the PEIs is the hyperbenthic shrimp Nauticaris marionis. Adult shrimps feed predominantly on benthic suspension- and deposit-feeders (Perissinotto and McQuaid 1990; Pakhomov et al. 1999), which are largely sustained by the autochthonous phytoplankton blooms (McQuaid and Froneman 2008; Allan 2011). In turn, these shrimp represent a dominant prey item in the diets of land-based predators such as gentoo penguins and Crozet shags that feed primarily within the inshore waters (Adams and Wilson 1987; Espitalier-Noël et al. 1988; Adams and Klages 1989; Crawford et al. 2003a). These shrimp contribute on average 26% and 25% by mass to the annual food consumed by gentoo penguins and Crozet shags, respectively (Espitalier-Noël et al. 1988; Adams and Klages 1989). It is important to note that their dietary contribution appears to vary seasonally, for example, from April to September their contribution ranges between 27–71% and 28–51% by mass (gentoo penguins and Crozet shags, respectively; Espitalier-Noël et al. 1988; Adams and Klages 1989). Furthermore, these shrimp periodically form an important food source for land-based predators that feed within both the inshore and offshore waters such as rockhopper and macaroni penguins (mixed-feeders; Brown and Klages 1987). During chick rearing rockhopper penguins, and to a lesser extent macaroni penguins, predominantly feed inshore (Brown 1987; Brown and Klages 1987). As a result, N. marionis forms an important trophic link between the sessile benthos and certain island-based predators (Perissinotto and McQuaid 1990).
It is becoming increasingly apparent that global climate change is affecting the terrestrial ecosystem of the PEIs through increased air temperatures, decreased precipitation and melting of the ice plateau (Smith 2002; Sumner et al. 2004; le Roux and McGeoch 2008). Evidence for the impacts of climate change on the marine ecosystem in the immediate vicinity of the PEIs is, however, generally lacking due to the difficulty in obtaining long-term data on marine organisms. Nonetheless, data from the vicinity of these islands indicate that the surface water temperatures have increased by >1°C since the 1950s (Mélice et al. 2003), and that this change in temperature has coincided with increased abundances of warm water zooplankton species over the past three decades (Pakhomov et al. 2000). Taken together, these trends suggest a southwards migration of the SAF, which is in agreement with recent studies that the strengthening and shifting of the southern hemisphere westerlies are concomitant to a poleward shift in the ACC (Oke and England 2004; Fyfe and Saenko 2005; Sokolov and Rintoul 2009) and SAF (Downes et al. 2011). As polar marine ecosystems are strongly shaped by large-scale physical processes, long-term shifts in the position of an oceanic front such as the SAF could feasibly have dramatic consequences for sub-Antarctic islands hundreds of kilometers away. We hypothesized that the long-term shift in the latitudinal position of the SAF has affected the primary productivity in the vicinity of the PEIs, due to increased occurrences of flow-through conditions, resulting in a temporal depletion in δ13C values of benthic organisms inhabiting the shallow shelf region of the islands.
- Top of page
- Materials and Methods
- Conflict of Interest
This study represents an analysis of the temporal trends in the isotopic signatures of the inshore benthos of the PEIs in order to identify the impact of regional climate change on this productive ecosystem and its potential implications for top predators that inhabit these islands.
Here, we present evidence of a measureable temporal shift in the carbon signatures of an inshore sub-Antarctic marine ecosystem. With the notable exceptions of the sponges and hydrozoans, the benthic invertebrates (suspension-feeders and scavengers/predators) inhabiting the shallow shelf region of the PEIs showed a significant depletion in their stable carbon isotope signatures (δ13C) from 1999 to 2009 (depletions ranged from 1.96‰ to 4.64‰; Fig. 2). The shrimp N. marionis, for which we have longer term data, too showed a depletion in δ13C values in both nearshore and inter-island individuals over the period 1984 to 2009 (Fig. 3A–C). Over the corresponding period 1999 to 2009, inter-island shrimps showed a depletion in carbon signatures of 2.28‰, which is within the range of that recorded in the inter-island benthos (δ13C depletion of between 1.96‰ and 4.64‰). In contrast, no such temporal depletions were apparent in the carbon isotope signatures of pelagic zooplankton collected at the PEIs (between 1999 to 2009; Fig. 2). As the zooplankton collected at the PEIs were transported there from the upstream region by the easterly flowing ACC, these animals would have been feeding in the oligotrophic open waters of the PFZ prior to their capture (Perissinotto and McQuaid 1992; McQuaid and Froneman 2008). The dissimilar temporal trends in isotopic signatures in pelagic zooplankton and benthic invertebrates (including the shrimp N. marionis) suggests that the gradual depletion of δ13C signatures in the benthos reflects a change in the biology of the primary producers that support the shallow shelf food web at the PEIs.
At the PEIs, phytoplankton form the dominant food source for the benthic community inhabiting the shallow shelf region in the vicinity of the islands (McQuaid and Froneman 2008; Allan 2011). Variations in δ13C signatures of phytoplankton have been linked with changes in isotopic fractionation due to alterations in water temperature, molecular CO2 concentrations, and growth conditions (Bidigare et al. 1997; Laws et al. 1997; Rau et al. 1997; Burkhardt et al. 1999). As oceanic δ13C values tend to become enriched with increasing temperature and depleted with increasing molecular CO2 concentrations (Rau et al. 1997), increased atmospheric CO2 resulting from the burning of fossil fuels could account for some of the observed depletion in δ13C signatures. In sub-polar regions, a depletion of ~0.015‰ year−1 has been calculated in the δ13C signatures of oceanic dissolved inorganic carbon (Suess effect; McNeil et al. 2001). Between 1984 and 2009, the Suess effect would account for only a 0.38‰ depletion in the δ13C signatures of organic matter, a value substantially smaller than the 2.55‰ depletion observed in the shrimps over this period. As a result, anthropogenic CO2 cannot be the sole driver of the variations in δ13C signatures observed in the N. marionis population. Phytoplankton growth rates also have a strong negative relationship with carbon isotope fractionation, with rapidly growing populations having more enriched δ13C values than slow growing forms (Bidigare et al. 1997; Laws et al. 1997; Burkhardt et al. 1999). Dense, rapidly growing phytoplankton communities associated with the “island mass effect” have enriched δ13C values compared to those in the less productive open waters of the PFZ in which the islands lie (~−23.3‰ vs. ~−24.8‰, respectively; Kaehler et al. 2000). Thus, the temporal depletion of δ13C signatures observed in the shrimp and various benthic invertebrates likely reflects a long-term decline of enhanced primary production (frequency of diatom blooms) in the vicinity of the islands.
The sensitivity of the ACC to climate change, in particular to the strengthening and shifting of the westerly wind belt (Oke and England 2004), has been of great interest. Climate models run under global warming conditions suggest that these wind stress changes are concomitant to a poleward shift in the ACC (Oke and England 2004; Fyfe and Saenko 2005; Sokolov and Rintoul 2009) and the three branches of the SAF (Downes et al. 2011). This southward migration of the SAF (indirect effect of climate change) seems to have contributed to a long-term decline in the frequency of phytoplankton blooms (“island mass effect”) at the PEIs, reflected by temporal depletions in the carbon isotope signatures of the various benthic invertebrates that inhabit the shelf region of the islands. The absence of a trend in the isotopic signatures of the sponges and hydrozoans supports this interpretation, as these two taxa showed dissimilar feeding behaviors from the other benthic species. Sponges generally consume small particles in the pico- and nano-size range (<20 μm; Reiswig 1990; van de Vyver et al. 1990; Pile et al. 1996) and are unlikely to feed on phytoplankton blooms consisting of large diatoms, so should be unaffected by shifting frequencies of phytoplankton blooms driven by larger diatoms (>20 μm). The hydrozoans at the PEIs appear to obtain a large proportion of their diet from animal prey such as zooplankton (elevated proportions of the fatty acid 18:1ω9 [>14.0%] and a carnivory index [18:1ω9/18:1ω7] >3.0; Allan 2011). As such, the hydrozoans, like the sponges, are unlikely to utilize large amounts of diatom production or respond directly to changes in the frequency of diatom blooms. As phytoplankton are an important food source for the food webs at the PEIs (Kaehler et al. 2000; Allan 2011), decreased occurrences of diatom blooms between the islands will have strong effects on higher trophic levels.
The PEIs seasonally support up to five million land-based predators that come to the islands to breed and molt (Williams et al. 1979; Condy 1981; Ryan and Bester 2008). Over the last three decades, there has been an overall decrease in the population sizes of predators that predominantly feed inshore such as gentoo penguins, Crozet shags, and rockhopper penguins (Fig. 4A; Crawford et al. 2003b, 2009a,b); while species that feed regularly in both the inshore and offshore regions, such as macaroni penguins, have shown smaller declines in population size over the same period (Fig. 4A; Crawford et al. 2003b, 2009b). Conversely, population sizes of offshore-feeding land-based predators (light-mantled sooty albatrosses, wandering albatrosses, grey-headed albatrosses, northern giant petrels, southern giant petrels, sub-Antarctic fur seals, and Antarctic fur seals) have been stable or increasing over the corresponding period, with the exception of dark-mantled sooty albatrosses (Fig. 4B and C; Hofmeyr et al. 2006; Ryan et al. 2009). Top predator populations are influenced by a variety of direct and indirect factors that include both present population pressures and the relaxation of earlier pressures. Competition both within and among species for resources such as food and space for breeding can be important as populations increase. Fisheries can have direct effects through mortality of animals as by-catch, while enhanced mitigation measures can be followed by population recovery from this effect. Fisheries can also remove prey species, again either directly or indirectly as by-catch. Man can have even more direct effects by targeting the land-based predators themselves and again these populations can show recovery once exploitation ceases. Finally, changing environmental conditions due to the effects of climate change have been suggested as having direct effects on populations, for example, increased wind speeds have been linked to population increases of wandering albatrosses due to improved fitness, foraging performance, and breeding success (Weimerskirch et al. 2012). Changes in predator populations at the PEIs are likely a result of a combination of several of these factors. As the offshore-feeding top predators at the PEIs forage over great distances from the islands, they are not as reliant on local productivity in the vicinity of the islands as the inshore-feeding top predators which we will focus on.
Inshore-feeders (gentoo penguins and Crozet shags) at the PEIs are pursuit-divers that feed mainly on demersal fish (Nototheniidae) and N. marionis (Adams and Wilson 1987; Espitalier-Noël et al. 1988; Adams and Klages 1989; Crawford et al. 2003a), whereas the mixed-feeders (rockhopper and macaroni penguins) feed primarily on pelagic fish (in particular Myctophidae) and crustaceans such as euphausiids and N. marionis (Brown and Klages 1987). Although industrial fishing at the PEIs (from the mid 1990s) has depleted demersal fish stocks dramatically (particularly those of Lepidonotothen squamifrons and Dissostichus eleginoides; Lombard et al. 2007), the continued population decreases of mixed-feeders that do not depend on these Nototheniidae fish species indicates that the decreases in inshore-feeding predator populations are not primarily driven by human exploitation of prey species. Fishing activities at the PEIs have been prohibited since December 2004 (Lombard et al. 2007), which is correlated with an apparent increase in the populations of fish-eating Crozet shags and gentoo penguins but not of mixed-feeders. The depletion of fish stocks, however, most likely resulted in a shift in the diet of the Crozet shags and gentoo penguins as L. squamifrons, their dominant dietary fish in the 1980s (Blankley 1981; Espitalier-Noël et al. 1988), was replaced by Gobionotothen marionensis and Lepidonotothen larseni by the late 1990s (Crozet shags; Crawford et al. 2003a). Conversely, the dietary contribution of N. marionis remained the same over this period (Espitalier-Noël et al. 1988; Crawford et al. 2003a).
Nauticaris marionis, which consume benthic suspension- and deposit-feeders (Perissinotto and McQuaid 1990; Pakhomov et al. 1999), represents an important food source for land-based predators that feed inshore (Brown and Klages 1987; Espitalier-Noël et al. 1988; Adams and Klages 1989). In addition, these shrimps together with benthic polychaetes are the dominant food sources for the fish G. marionensis (Bushula et al. 2005), which in turn is an important prey species to Crozet shags (Crawford et al. 2003a). As a result, the benthic and hyperbenthic communities of the shallow shelf region at the PEIs represent a crucial link between primary producers and certain top predators. Since the initial photographic assessments in the 1980s (Branch et al. 1993), there have been no additional quantitative studies on the biomass and abundance of the benthic community around the PEIs. Consequently, we cannot determine whether shifts in the regional hydrodynamics around the islands have resulted in changes in the biodiversity, abundance, and biomass of the benthos. Nevertheless, declines in the abundances of predators that rely heavily on prey inhabiting the inshore areas of the PEIs suggest changes in prey availability within this island ecosystem. Changes in prey availability are likely driven by a combination of local factors such as fisheries impacts on prey populations and changes in competitive interactions among predators for food. In addition to these local factors, our results indicate that changes in prey availability may also arise indirectly through the effects of regional climate change on the marine ecosystem at the PEIs. In this case, regional climate change has caused a southward shift in the position of a major current system (the SAF), which has affected the primary productivity in the vicinity of the PEIs (decreased frequency of diatom blooms) due to increased occurrences of flow-through conditions. The shift in dominant primary production at the PEIs from diatom-dominated communities to less productive small cell-dominated phytoplankton communities is reflected isotopically in the primary consumers (benthic suspension- and deposit-feeders) inhabiting the inshore region of the islands. In addition, the shift in primary producers consumed at the base of the food web is also apparent in higher trophic levels such as ophuiroids, asteroids, and the shrimp N. marionis, that inhabit the inshore region. These results indicate a fundamental shift in the balance between allochthonous and autochthonous trophic pathways within the PEI region, thus highlighting the vulnerability of marine ecosystems to large-scale changes in physical conditions. Most importantly, these results indicate that the consequences of climate change may be indirect.
Our results demonstrate that the most powerful effects of climate change on natural ecosystems may be obscured because they are indirect, thus making it difficult to detect cause and effect relationships. This is particularly problematic where appropriate long-term data (such as biodiversity, abundance, and biomass of benthic communities) are generally lacking. Indirect effects of climate change, such as those observed at the PEIs, coupled with the lack of long-term data for key marine ecosystem components highlight the concerns of the International Program on the State of the Ocean (IPSO) that cumulative impacts on the ocean are greater than previously thought (Rogers and Laffoley 2011).