• Antarctic krill;
  • Antarctica;
  • Elephant Island;
  • Euphausia superba;
  • foraging;
  • patchiness;
  • seabirds


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and materials
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We investigate the influence of krill (principally Euphausia superba) patchiness on the foraging distributions of seabirds to understand how variation in krill influences patch dynamics between krill and birds. At sea-surveys were conducted near Elephant Island, Antarctica, for 3 yr (2004–2006) during the annual U.S. Antarctic Marine Living Resources (AMLR) program. Standardized strip-transect surveys were used to map seabirds, and a combination of acoustic and net surveys was used to map krill. We measured patch size of krill and seabirds and elucidated how krill patch dynamics influence foraging seabirds. The spatial association between krill and predators was influenced by the size and arrangement of krill patches. We found a negative relationship between abundance and patchiness of krill and predators, indicating that when krill is less abundant, its predators are less abundant and concentrated. We conclude that annual patch dynamics of krill strongly influences the local abundance and distribution of seabirds. Such information should be used to interpret potential interactions between seabirds and krill fisheries operating near Elephant Island.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and materials
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In the Scotia Sea, Antarctic Krill (Euphausia superba, hereafter krill) are the principal prey for a variety of seabirds (Croll and Tershy, 1998; Croxall et al., 2002), and breeding success of many species is associated with variability in krill abundance or availability (Murphy et al., 1998; Croxall et al., 1999; Boyd and Murray, 2001; Trathan et al., 2003). In addition, commercial krill fisheries may impact krill predators through increased competition (Croll and Tershy, 1998; Boyd and Murray, 2001; Trathan et al., 2003; Reid et al., 2004). Near the South Shetland Islands (Fig. 1), krill abundance is cyclical, declining with consecutive years of poor reproductive success and increasing following years of good recruitment (Loeb et al., 1997; Hewitt et al., 2003, 2004a). Part of the response by birds to krill must reflect krill abundance, but there might be other aspects of krill biology that impact the distribution and abundance of birds, including krill patchiness.


Figure 1.  Western Antarctic Peninsula region and the location of the Elephant Island study area and nearby South Shetland Islands. Black dots indicate position of Conductivity-Temperature-Depth (CTD) and net sampling stations. Bathymetric contours are 2000, 400 and 200 m. Geographic quadrants are NE, Northeast; NW, Northwest; SE, Southeast and SW, Southwest.

Download figure to PowerPoint

Krill swarms are patchily distributed (Macaulay et al., 1984; Miller and Hampton, 1989; Hewitt and Demer, 1993; Powell and Okubo, 1994). Organisms that have an inherent aggregation behavior, such as krill, are distributed in a strikingly hierarchical fashion, where patches are nested within patches. Murphy et al. (1988) suggested the following hierarchical patch structure for krill heterogeneity: (1) At fine scales, individuals are aggregated in swarms with a diameter of 1–100 m at time scales ranging from hours to days; (2) swarms are aggregated into patches with a diameter of 1–100 km for a period of days to months; and (3) patches are further aggregated into concentrations at a scale of hundreds of kilometers for months. To our knowledge, there are no studies investigating changes in krill patch distribution on annual scales, and how this may influence the local abundance of predators.

Seabirds may exploit patchily distributed prey by responding to habitat patches rather than individual prey types (Wiens, 1976; Duffy, 1986; Morris, 1987; Veit, 1999; Greene and Stamps, 2001). Given that krill patches are nested in a hierarchical fashion (Murphy et al., 1988), foraging behavior (e.g. search pattern, foraging success) of krill-dependent predators should be influenced by krill patchiness (Boyd, 1996; Grünbaum, 1998; Fauchald, 1999; Alonzo et al., 2003; Grünbaum and Veit, 2003). That is, seabirds may survey patch types in their environment, and select those in which foraging is most profitable. The profitability of a patch type could be thought of as a balance between energy budget and previous encounters with krill patches (i.e. area-restricted search) (Bernstein et al., 1988; Veit, 1999; With et al., 1999; Zollner and Lima, 1999). Once a patch or patch types are selected, a seabird may forage opportunistically within those patches, and repeatedly return to the regions where encounter rate with prey was highest.

Changes in the spatial distribution of krill may have significant impacts on foraging and reproductive success of krill-dependent predators (Agnew and Phegan, 1995; Marin and Delgado, 2001). During the predator’s breeding season, the adult animals depend on locating sufficient amounts of krill to support the energetic needs of themselves and their offspring. Seabirds track changes in prey distribution at several scales to balance energetic requirements and increase the likelihood of survival and successful reproduction (Fauchald, 1999; Fauchald and Erikstad, 2002; Davoren et al., 2003). It has been postulated that an increase in prey patchiness should increase the foraging efficiency of seabirds, whereas a decrease in prey patchiness may force predators to switch to other prey items (Fauchald et al., 2000; Fauchald and Erikstad, 2002). In the western Antarctic Peninsula region, krill is integral to the food web (Clarke et al., 2007), and there are few, if any, alternative prey species of sufficient biomass for krill-dependent birds to switch to. Obviously, fewer krill should decrease the likelihood of reproductive success by krill-dependent birds, but understanding why they do poorly should also be examined by measuring changes in foraging behavior in relation to krill patchiness. If krill patches are scarce, predators likely spend more time searching for patches, and once they are found, predators should remain in those regions for longer time periods. For example, Veit and Prince (1997) tracked foraging movements of krill-dependent Black-browed Albatrosses (Thalassarche melanophrys) during a year of average krill abundance and during a year of exceptional krill scarcity around South Georgia (Croxall et al., 1999). They found that foraging trip duration and spatial extent were greater in magnitude during the scarce krill year. When krill is scarce, the encounter rate with krill patches declines. This may prevent seabirds from acquiring enough energy to maintain their metabolism as well as that of their chicks, consequently reducing the chance of successful reproduction.

We present an analysis of Cape Petrel (Daption capense), Chinstrap Penguin (Pygoscelis antarctica), and krill distribution data near Elephant Island during January for three consecutive austral summers (2004–2006). We examine two hypotheses: (1) seabirds concentrate their foraging effort where krill patches occur, and (2) foraging seabirds respond to changes in krill patchiness. We ask whether variation in annual krill patchiness influences patch dimensions of predators. We use patchiness as a measure of spatial heterogeneity to gauge the degree of clumping in space of seabirds and krill (Wiens, 1976; Rose and Leggett, 1990; Powell and Okubo, 1994). We also ask whether spatial association of krill and predators varies annually in relation to krill patchiness. For comparative purposes we investigate whether foraging behavior of petrels and penguins differ in regard to krill patchiness. Aerial petrels are highly mobile in contrast to penguins, which swim and dive in search of krill patches, whereas petrels are restricted to feeding at the surface. Differences in mobility between petrels and penguins may force them to respond differently to krill patchiness. Furthermore, in our study, we speak of ‘good’ krill years when krill abundance is high and patchiness is low (i.e. krill is plentiful everywhere). We hypothesize that when abundance is low and patchiness of krill is high (i.e. fewer patches), foraging distributions of krill predators should be spatially associated with krill patches.

Methods and materials

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and materials
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study area

In the Southern Ocean, at-sea surveys are regularly used to examine the distribution patterns of krill and predators (Obst, 1985; Hunt et al., 1992; Veit et al., 1993; Hewitt et al., 2004a; Reid et al., 2004). Our survey occurred over 3 yr (2004–2006) in the Scotia Sea near Elephant Island (61°S, 62°W), an area of high krill biomass that attracts substantial numbers of seabirds (Hunt et al., 1990; Hewitt and Demer, 1993; Agnew, 1997; Croll and Tershy, 1998). Before 1990, the Elephant Island region was an active krill fishery (Everson and Goss, 1991; Jones and Ramm, 2004), and recent data suggest that krill biomass in the Southern Ocean has declined by an order of magnitude over the last ∼50 yr (Atkinson et al., 2004). Current biomass estimates for the Scotia Sea range between 40 and 100 million tons (Hewitt et al., 2004b; Demer and Conti, 2005). Fishery removals over the last 10 yr have averaged ∼100 kTons, with most of this occurring around the South Shetland Islands and near South Georgia and within 50 nautical miles (nmi) of breeding colonies (Hewitt et al., 2004c). Near Elephant Island, krill biomass fluctuates with a period of roughly 8 yr (Hewitt et al., 2003) and successful reproduction and recruitment by krill is related to the extent of sea ice (Loeb et al., 1997). Interannual variability in broad climate patterns associated with sea-ice cycles is correlated with recruitment and population variability in Antarctic Krill (Loeb et al., 1997) and is thought to be responsible for the observed variability in krill abundance (Murphy et al., 2004). The result is an interannual variation in krill biomass with alternating ‘good’ and ‘poor’ krill years influenced by good reproduction and recruitment success. Around Elephant Island, acoustic estimates of krill biomass have fluctuated almost 2 orders of magnitude since 1996 (Reiss et al., 2008), mostly associated with recruitment failure during low ice, and El Niño conditions. Although it is clear that biomass differs annually, it is also important to understand how spatial heterogeneity of krill varies annually, because variation of krill patchiness may be useful for predicting where seabirds forage.

The density and dispersion of krill also varies within a summer season. For example, Hewitt et al. (2004c) investigated krill density and demography in the vicinity of the South Shetland Islands (Fig. 1) during the 1999/2000 austral summer. They found that the areas of highest krill density shift toward the shelf break as the season progressed. They also found that changes in density and dispersion patterns reflected shifts of adult (∼50 mm) and juvenile (∼25 mm) krill, where sexually advanced stages of krill immigrated into the region later in the season and displaced juvenile stages (Hewitt et al., 2004c). Based on the dispersion patterns of krill within that summer, they suggested that prey availability was relatively constant for krill predators, even though numerical density and size composition of krill changed as the season progressed. However, it is not known how foraging predators respond to annual variability of krill patchiness.

Field methods

The survey area surveyed encompassed 43 865 km2 (18 866 mi2), and consisted of seven north-south transects, each 222 km (∼120 nautical miles) in length (Hewitt and Demer, 1993) (Fig. 1). Hydrographic and biological sampling stations were occupied along each transect to characterize the oceanographic conditions, and to calibrate the acoustic estimates of krill. Ship speed during transect sampling was approximately 18.5 knots h−1 (10 nmi). We only analyzed daytime data because we could not observe birds at night.

Physical oceanography

At each hydrographic station (N = 48, Fig. 1), a Conductivity, Temperature and Depth (CTD) cast was made with a Seabird model 9/11 CTD (SBE, Inc., Bellevue, WA, USA) to a depth of 750, or 10 from the bottom at shallower stations. CTD casts were spaced approximately 20 nautical miles (37 km) apart. These data were used to determine the surface temperature across the region, and to describe the location of the shelf-slope frontal area. This frontal area separates warmer Antarctic Surface water from cooler coastal and shelf waters that are a mixture of water from the Weddell and the Antarctic Surface waters that mix in this region (Tynan, 1998).

Net sampling of plankton

Krill and other zooplankton were captured with a 1.8-m Isaacs-Kidd Midwater Trawl (IKMT) fitted with a 505-μm-mesh net. Filtered volume was measured using a flowmeter (model-2030; General Oceanics, Inc., Miami, FL, USA) mounted on the frame in front of the net. All tows were fished obliquely from a depth of 170 m or approximately 10 m above bottom to the surface. The net was towed at approximately 2 knots and the volume of water sampled averaged 2300 m3. Abundance of krill is expressed as numbers per 1000 m3 water filtered.

Acoustic sampling of krill

Acoustic estimates of krill abundance were made with a multi-frequency echosounder (Simrad EK60) operating with 38, 70, 120, and 200 kHz transducers mounted on the hull 7 m below the surface. Pulses were transmitted every 2 s at 1 kW for 1 ms. Positions were logged every 2 s. SonarData Echoview was used to aggregate acoustic data into 1-nmi horizontal intervals and to a depth of 200 m. We use the nautical area scattering coefficient (NASC) as an index of krill and zooplankton abundance (log # nmi−1) (Hewitt and Demer, 1993; Madureira et al., 1993; Hewitt et al., 2004a,b,c;Demer and Conti, 2005). For the remainder of this paper NASC will be referred to as an index of acoustically estimated krill abundance, although we acknowledge that NASC is simply a measure of acoustic backscatter.

Seabird distribution, behavior and abundance

Observers collected data on seabird abundance continuously during daylight hours. Counts of predators were made within an arc of 300 m directly ahead to one side of the ship while underway (Tasker et al., 1984). Each record was assigned a time (to the nearest tenth of a second) and a spatial position from the ship’s global positioning system (GPS). The GPS was synchronized with the echosounder system. Individual birds, or flocks of birds, were assigned a behavioral code. The behaviors were: flying, sitting on water, feeding, porpoising (penguins) and ship-following (Veit, 1999). Ship-following birds were recorded when first encountered and ignored thereafter.

We examined the foraging distributions of Cape Petrels and Chinstrap Penguins because they breed during January on and near Elephant Island and are the most abundant seabirds in the region (Santora et al., 2005). Cape Petrels are medium-sized petrels and are restricted to feeding in the upper few meters of the water column (Warham, 1990). They are highly gregarious and generally forage in dense flocks that are easily monitored for behavioral changes (Veit, 1999). For example, the frequency of their turning rates increases in proximity to krill swarms when krill swarms are detected (Veit, 1999). Chinstrap Penguins are pursuit-diving predators, which are capable of diving to more than 50 m in search of krill, and which forage primarily during daylight hours (Bengston et al., 1993; Croll et al., 2006). Only sitting Chinstrap Penguins were analyzed here because they were more likely to be foraging as opposed to traveling.

Analytical methods

We constructed contour maps of sea-surface temperature (SST) at 2 m from the CTD data using Ocean Data View (Schlitzer, 2004) to visualize the hydrographic conditions around Elephant Island each year. To determine if mean SST varied interannually, the 2-m SST values were pooled by year and analyzed using a 1-way ANOVA. Acoustically determined krill and seabird (flying Cape Petrels and Chinstrap Penguins) abundance (log ind. nmi−1) along each transect were binned into 1-nmi bins. Furthermore, we divided the survey area into four quadrants (NE, NW, SE and SW) reflecting the variability of the insular shelf break surrounding Elephant Island. We then examined the interannual differences in oceanographic conditions, krill and seabird abundance within quadrants. Interannual variability in the mean abundance of seabirds and krill (estimated by net and acoustics) was tested using ANOVA, with a Bonferroni multiple comparison test (Zar, 1999).

We calculated two statistics to examine interannual differences in patchiness. For patchiness, we make the distinction between ‘form’ (using Moran’s I) and ‘intensity’ (using coefficient of variance – CV) by referring to the size of patches rather than the number of patches. First, we calculated all-directional spatial correlograms to examine patchiness of acoustically determined krill and seabird abundance (Rose and Leggett, 1990; Legendre and Legendre, 1998; Reid et al., 2004). We compared plots of Moran’s I versus distance for each year to determine patch size of acoustic estimates of krill and seabirds. We defined the characteristic patch scale of each species as the point along the space domain where the slope of the correlogram was not different from zero (Reid et al., 2004). Secondly, we calculated the CV of abundance of krill and seabirds as an index of patchiness, and asked whether it differed among years.

To determine whether seabirds were spatially associated with krill patches, we used a geostatistical approach involving a spatially explicit regression model (Anselin et al., 2006; GeoDaS Software) solved with maximum likelihood estimation (Burnham and Anderson, 2002). A spatial weight variable (Euclidean distance measure) was determined for each nautical mile sample within each year. The spatial regression model is:

  • image

where y is seabird abundance, K is krill abundance, β is a regression coefficient, and ε is an error term. The model has spatial lag components, where ρ is a spatial autoregressive coefficient, and λ is a lag term. The spatially lagged variable is the sum of spatial weights multiplied with values from observations at neighboring locations. The model takes into account whether the error terms across different spatial units are correlated as a function of geographic distance (i.e. spatial autocorrelation of residuals). We concluded that seabirds were spatially associated with krill patches if krill abundance (K) was a significant predictor of seabird distribution.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and materials
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Hydrographic variability

Mean SST varied among years (F2,138 = 63.3, P < 0.001, Fig. 2) and 2006 was warmer than 2004 (t = 10.27, P < 0.0001) and 2005 (t = 9.04, P < 0.001). In each year, a surface temperature front was observed northwest of Elephant Island (along the shelf edge). In 2004 and 2005, the coolest water (SST < 1.0°C) was found in the southeast part of the area, with the warmest waters near 60°S. In 2006, the only cool water was found on the southern and southeastern flank of the Elephant Island insular shelf (Figs 2 and 3).


Figure 2.  Annual variability of SST (°C averaged over 15 m) near Elephant Island: (a) 2004, (b) 2005 and (c) 2006. Dots indicate position of CTD station. (d) Comparison of mean ± 95% CI annual SST.

Download figure to PowerPoint


Figure 3.  (a–d) Annual variability of abundance (mean ± 95% CI): (a) Euphausia superba, (b) Thysannoessa macrura, (c) acoustically estimated krill (NASC) and (d) seabirds (shaded bars = Cape Petrels). (e–h) Annual variability (mean ± 95% CI) among geographic quadrants: (e) SST (°C averaged over 15 m), (f) krill, (g) Cape Petrel and (h) Chinstrap Penguin. Geographic quadrants are NE, Northeast; NW, Northwest, SE, Southeast and SW, Southwest.

Download figure to PowerPoint

Krill abundance and distribution

Euphausia superba and Thysannoessa macrura were the two numerically dominant krill species caught in nets (Fig. 3). Euphausia superba exhibited similar mean abundances of ∼60 per 1000 m−3 in 2004 and 2005 and ∼25 individuals per 1000 m−3 in 2006 (Fig. 3). Krill larvae (E. superba) varied annually (F2,139 = 7.02, P = 0.001), where mean abundance in 2006 was greater than in 2005 (t = 3.22, P = 0.005) or 2004 (t = 3.25, P = 0.004). T. macrura, were most abundant in 2005 with mean abundances that year in excess of 350 individuals per 1000 m−3. Annual abundance of T. macrura varied significantly (F2,139 = 5.26, P = 0.001), whereas E. superba did not (F2,139 = 1.59, P = 0.21). Thysannoessa macrura was significantly more abundant in 2005 than in 2004 (t = 3.53, P = 0.002) or 2006 (t = 2.88, P = 0.014).

Acoustic estimates and patchiness indices of plankton

Acoustic estimates of krill varied significantly (Figs 3 and 4) among years (F2,1622 = 185.3, P < 0.0001), and was approximately 2 orders of magnitude less in 2006 than in 2004 (t = 14.66, < 0.0001) or 2005 (t = 18.46, < 0.0001). Acoustic estimates of krill were higher in cooler (<2°C) areas of the Elephant Island area. This was most evident in 2006, when krill were present exclusively in the eastern and southeastern flanks of the Elephant Island insular shelf (Figs 3 and 4). In 2004 and 2005, krill patches were concentrated around the periphery of the Elephant Island insular shelf region (Figs 3 and 4).


Figure 4.  Distribution of krill and seabirds near Elephant Island during 2004, 2005 and 2006: (a–c) Krill, (d–f) Flying Cape Petrels and (g–i) Chinstrap Penguins. Bathymetric contours are 2000, 400 and 200 m.

Acoustically determined krill abundance was patchy in all years (Figs 4 and 5). Changes in the patchiness of krill, measured using CV, was negatively correlated with abundance (r = −0.70, ±95% CI: −0.99 to −0.47) (Fig. 6), indicating that when krill is less abundant (i.e. 2006), the intensity of the distribution of krill patches is greater. Patch sizes of krill, using Moran’s I, indicated that the scales of patch sizes of krill were similar during 2004 and 2005, with patch sizes ranging from 1 to 9 nmi (Fig. 5). During 2006, the scale of krill patch sizes was autocorrelated out to 31 nmi, indicating that when krill is significantly less abundant, the scale of patchiness increases (Fig. 5).


Figure 5.  Spatial autocorrelation functions (Moran’s I) of (a) krill, (b) Flying Cape Petrels and (c) Chinstrap Penguins.

Download figure to PowerPoint


Figure 6.  Relationship between abundance and patchiness of: (a) krill, (b) Flying Cape Petrels and (c) Chinstrap Penguins.

Download figure to PowerPoint

Seabird abundance and patchiness

Abundance of Cape Petrels varied among years (F2,260 = 162.70, P < 0.0001), (Figs 3 and 4). Abundance of Cape Petrels was greater in 2005 than in 2004 (t = 4.95, P < 0.0001) and 2006 (t = 18.01, P < 0.0001). Chinstrap Penguin abundance also varied over the 3 yr of the study (F2,1621 = 18.9, P < 0.0001, Figs 3 and 4), but declined over the study period (r = −0.15, P < 0.0001). Chinstrap Penguins were clustered in the vicinity of Elephant and Clarence Islands during each year (i.e. SE quadrant), which is probably related to good foraging locations within proximity to breeding colonies (Figs 3 and 4).

The scale of patch sizes of Cape Petrels, measured using Moran’s I, was similar during 2004/2006, with patches ranging from 1 to 5 nmi in size (Fig. 5). During 2005, patch size was smaller and ranged from 1 to 9 nmi (Fig. 5). On the other hand, changes in patchiness of Cape Petrels, measured using CV, was negatively correlated (r = −0.93, ±95% CI: −0.99 and −0.88) with abundance (Fig. 6).

Patch sizes of Chinstrap Penguins, measured using Moran’s I, was similar during 2004/2005, with patches ranging from 1 to 20 nmi in size (Fig. 5). During 2006, patch size was smaller and ranged from 1 to 8 nmi (Fig. 5). However, changes in patchiness of penguins, measured using CV, were also negatively correlated (r = −0.83, ±95% CI: −0.82 and −0.66) with abundance (Fig. 6).

Association of seabirds with krill abundance and patchiness

Spatial association between seabirds and krill patches differed markedly among years (Tables 1 and 2). In 2006, when krill was scarce but patchiness was highest, Cape Petrels and krill were closely associated (Tables 1 and 2). In 2005, when krill was abundant and patchiness was lowest, Cape Petrels were also closely associated with krill (Table 1). Cape petrels were not spatially linked to krill in 2004 when krill abundance was high but patchiness was low. By comparison, Chinstrap Penguins were associated with krill when abundance of krill was high and patchiness was low (2004 and 2005) (Tables 1 and 2).

Table 1.   Results for test of spatial association of seabirds and krill.
YearVariableR2 coefficient ± SE z-, P-values
  1. Spatial association between seabirds and krill occurred when K (krill abundance and distribution) was found as a significant predictor of seabird spatial distribution. Maximum likelihood estimates of significant parameters for spatial regression model: coefficient is the maximum likelihood estimate, SE is standard error of the coefficient, z is the maximum likelihood test value and P is probability of rejecting null hypothesis given the test value. λ is spatial lag term; --- indicates no parameter found. R2 is coefficient of determination.

Cape Petrel R2 =
λλλ0.25 ± 0.04 6.2, <0.0010.44 ± 0.04 10.8, <0.0010.4 ± 0.04 9.6, <0.001
KK0.1 ± 0.03 3.4, <0.0010.04 ± 0.02 2.0, 0.04
Chinstrap Penguin R2 = 0.470.370.12
λ0.33 ± 0.04 8.7, <0.001
KK0.05 ± 0.02 3.1, 0.0020.05 ± 0.02 2.5, 0.01
Table 2.   Summary of results for annual spatial association of seabirds and krill.
TestResult by year
Krill patchinessMediumLowHigh
Krill abundanceHighHighLow
Spatial association of Cape Petrels and KrillNoYesYes
Spatial association of Chinstrap Penguins and KrillYesYesNo

Correlation between changes in patchiness (CV) and abundance among foraging birds and krill varied with krill patchiness (Table 3, Figs 7 and 8). Both predators were positively correlated with krill abundance (Table 3), but displayed different responses to changes in patchiness of krill. The abundance and patchiness of Cape Petrels tracked changes in krill patchiness (Fig. 7), whereas penguins were more strongly influenced by changes in krill abundance (Table 3, Fig. 8). There was no association between either the abundance or patchiness of penguins and the patchiness of krill (Table 3, Fig. 8).

Table 3.   Correlation between changes in patchiness (CV) and mean abundance of seabirds and krill.
 Krillr-ValueBootstrap (95% CI)
  1. Confidence intervals (95%) were calculated from 5000 randomizations.

  2. *Significant at P < 0.05.

Cape Petrel
 AbundanceAbundance0.80*0.67 to 0.99
 AbundancePatchiness−0.73*−0.99 to −0.54
 PatchinessAbundance−0.87*−0.99 to −0.77
 PatchinessPatchiness0.63*−0.36 to 0.99
Chinstrap Penguin
 AbundanceAbundance0.31*−0.1 to 0.86
 AbundancePatchiness0.040.01 to 0.08
 PatchinessAbundance−0.58*−0.99 to −0.28
 PatchinessPatchiness0.01−0.51 to 0.54

Figure 7.  Relationship between abundance of krill and (a) abundance and (b) patchiness of Cape Petrels. (c–d) Relationship between patchiness of krill: (c) abundance and (d) patchiness of Cape Petrels.

Download figure to PowerPoint


Figure 8.  (a–b) Relationship between abundance of krill and (a) abundance and (b) patchiness of Chinstrap Penguins; (c–d) Relationship between patchiness of krill: (c) abundance and (d) patchiness of Chinstrap Penguins.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and materials
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Krill abundance and distribution differed among years, and influenced the foraging distributions of Cape Petrels and Chinstrap Penguins in the vicinity of Elephant Island, Antarctica. The negative relationship between abundance and patchiness (i.e. changes in patch size) of krill suggests that when krill is more abundant, they are more widely distributed and less patchy with potentially more smaller-sized patches available to foraging birds. However, when krill are less abundant, they are more patchy and it is likely that the distance between krill patches increases. When krill abundance decreased by ∼2 orders of magnitude between 2005 and 2006, our index of krill patchiness nearly doubled.

Response by seabirds to krill patchiness

Our principal finding that patchiness of krill was negatively related to krill abundance enabled us to assess the response of foraging seabirds in relation to low, medium, and high krill patchiness. We asked whether spatial association of seabirds and krill depended on krill patchiness. Our results yielded insight into how variations in size and arrangement of krill patches might impact foraging success of predators.

We demonstrated that foraging distributions of Cape Petrels and Chinstrap Penguins, two of the most abundant seabirds in the Antarctic Peninsula region, exhibited responses to krill patchiness. When krill abundance decreased by 2 orders of magnitude between 2005 and 2006, abundance of Cape Petrels decreased and correlation between bird and krill patches was higher. Interestingly, the probability of spatial association between Cape Petrels and krill was greater when patchiness of krill was either high or low, but not at intermediate levels. During 2005, when patchiness of krill was lowest, Cape Petrels were spatially associated with krill. In 2006, when krill was comparatively scarce, and patchiness was highest, Cape Petrels were also associated with krill patches. Penguins were associated with krill patches when krill abundance was high and patchiness was low (2004).

Abundance of foraging Chinstrap Penguins decreased during the study, which coincided with the decrease in krill abundance and an increase in krill patchiness. Previously, Croll et al. (2006) showed that annual variability in krill abundance is correlated with the reproductive performance (e.g. breeding population size and breeding success), but not foraging effort by Chinstrap Penguins. Furthermore, they suggested that penguins reduce reproductive success rather than increase foraging effort in response to decreases in krill abundance. This makes sense because the distribution of krill patches around the colony was probably not sufficient for breeding penguins. Therefore, we emphasize that foraging effort by krill-dependent seabirds should be evaluated in relation to abundance and patchiness of krill. Our study of the foraging distribution of Chinstrap Penguins was conducted over the entire Elephant Island region, whereas the study by Croll et al. (2006) was focused on five to eight penguins per year that bred at a single colony, and noted that breeding population size of penguins was correlated with krill abundance. We found that abundance and patchiness of foraging penguins throughout the Elephant Island area was correlated with krill distribution. We emphasize that krill patchiness adds additional insight into changes in penguin foraging distribution. In the future, we need to combine information of individual foraging effort, and breeding population size, while simultaneously tracking changes in foraging distribution of penguins at the population level (i.e. thousands).

Krill patchiness affects foraging behavior

Foraging behavior is an ecological process that should be evaluated with consideration of an organism’s activity during a particular time period (Wiens, 1976; Piatt, 1990) and there are several assumptions that must be met if we are to properly test hypotheses regarding the effects of krill abundance or patchiness on feeding ecology of seabirds. First, we know that petrels and penguins breeding near Elephant Island during January are provisioning resources for chick rearing, and therefore their foraging effort should be a balance between the metabolism of their chicks and their personal growth. Additionally, we assume that (1) seabirds are locating and consuming krill, and concentrate their foraging effort where krill patches occur, and (2) seabirds can locate and detect krill patches. There are physical mechanisms (i.e. location of fronts, eddies, bathymetry) that may be important for influencing spatial patterns of krill and seabirds (Hunt and Schneider, 1987; Piatt and Methven, 1992), but we only address the impact of krill patchiness on foraging distributions of birds.

There are two foraging strategies that seabirds can use to locate krill patches: area-restricted search and local enhancement (Kareiva and Odell, 1987; Buckley, 1996, 1997). Seabirds practice ‘area-restricted search’ by concentrating foraging effort for some time in the location of their last successful prey capture (Veit, 1999; Pinaud and Weimerskirch, 2007). Moreover, if foraging success (i.e. prey capture) is related to prey patchiness, then the amount of time spent searching within a location should, in part, depend on abundance and patchiness of prey in that location. In relation to our study, high spatial correlation between birds and krill might result in higher foraging success, because it means that birds were successful at locating and detecting krill patches.

There are potential impacts on the foraging costs of petrels and penguins in their need to exploit dense krill patches (Piatt, 1990; Piatt and Methven, 1992; Davoren et al., 2003). Petrels and penguins need to exploit dense prey patches to minimize the amount of time spent searching for prey patches and to maximize metabolic intake per encounter with each patch they detect. Although they both feed on krill, their ability to exploit krill patches vertically and horizontally may influence where they spend time foraging. Since Cape Petrels cannot dive further than a few meters, they probably choose foraging locations where krill are close to the surface. Unfortunately, we were unable to monitor krill patches near the surface. Krill can be closer to the surface during the evening and early morning (Hewitt and Demer, 1993), and petrels may aggregate in locations where they had previous success feeding the night before (Hunt, 1990). Each year, we observed aggregations of Cape Petrels sitting on the water. These birds were not observed to feed. They may congregate because they found krill patches that were closer to the surface during the previous evening. Congregations could serve as ‘information centers’ whereby birds may learn where others had success in locating prey (Pöysä, 1992; Silverman et al., 2004).

By comparison, penguins are ideally suited for exploiting krill patches that are out of reach of aerial petrels. However, penguins may consume more energy swimming and diving during a foraging bout than an aerial petrel which travels using energy-efficient gliding. Therefore, the difference in foraging cost and their need to search, detect, and exploit dense patches ought to be examined regarding their locomotion during foraging. Our data indicate that changes in the scale of krill patch size are probably more critical for petrels than penguins. As seen in our distribution maps of foraging Cape Petrels, they are spread throughout the Elephant Island region, whereas penguins were basically concentrated in the same region every year. This indicates that petrels must cover a much larger ocean space during a foraging trip than a penguin, perhaps because krill patches close to the surface are more ephemeral.

Breeding petrels and penguins require dense patches in close proximity to their colony. Although there are differences in the locomotion and prey capture technique of both predators, Chinstrap Penguins usually undergo shorter (1–2 days) foraging trips than Cape Petrels (several days). Both predators should exhibit specific sensitivity to changes in krill patchiness in the vicinity of their breeding locations. If a penguin or a petrel must spend more time foraging farther from its colony then it may not be able to balance energetic requirements of its chick and itself. Other studies have proposed that krill fishing be restricted beyond than 50–100 km from penguin colonies (Croll and Tershy, 1998; Reid et al., 2004), and our study reinforces this proposition.

There is a growing body of evidence suggesting that local enhancement is an important part of foraging strategies used by seabirds to locate patchy prey (Harrison et al., 1991; Silverman and Veit, 2001; Silverman et al., 2004). For example, Grünbaum and Veit (2003) found that density dependence of albatrosses might affect the outcome of foraging success of albatrosses foraging for krill. In relation to our study, this may also be true. By comparison, in 2006, when krill patches were scarce, we recorded relatively fewer penguins and petrels, indicating that there were fewer predators searching for krill near Elephant Island. Therefore, if krill patches are scarce, seabirds may spend more time searching for predators.


Foraging distributions of seabirds are linked to krill distribution. It is reasonable to suppose that factors influencing krill distribution ought to alter seabird behavior as well (Veit and Prince, 1997; Fraser and Hoffman, 2003; Forcada et al., 2006). Our study has implications for two aspects for the ecology of krill and seabirds near Elephant Island: (1) climate variability and (2) commercial krill harvesting.

Near the Antarctic Peninsula, air temperature has risen dramatically in the last ∼50 yr, related to changes in winter sea-ice extent (Moline et al., 2004; Clarke et al., 2007; Murphy et al., 2007). Near South Georgia, the occurrence of krill has been linked to ‘warm’ and ‘cold’ (based on air temperature) periods associated with changes in sea-ice extent in the Scotia Sea (Murphy et al., 1998, 2007). We found an interesting result regarding annual changes in sea-surface temperature (°C at 2 m). During this study, SSTs exceeding ∼3.5°C were found throughout the Elephant Island region during 2006, whereas in 2004 and 2005, when krill abundance was 2 orders of magnitude greater, SST was cooler (∼1.2°C). This anomalous temperature pattern may be related to the low abundance and restricted distribution of krill in 2006 (Moline et al., 2004; Clarke et al., 2007). A likely explanation for the warm SST in 2006 was the poleward movement of the southern Antarctic Circumpolar Current front (sACCf); (Tynan, 1998; Constable et al., 2003). Although we cannot say what is ‘normal’ in terms of relating long-term variability of hydrography and the interaction of krill and seabirds, it is possible that future periods of warming will influence krill distribution, which will affect seabird foraging behavior. Future investigations of krill distribution and foraging behavior by seabirds near Elephant Island ought to explore the importance of the sACCf in relation to climate variability.

Krill is an important resource for birds, and is also a resource targeted by commercial fisheries (May et al., 1979; Everson and Goss, 1991; Mangel, 1994; Marin and Delgado, 2001). Successful management and conservation of resources requires description of spatial pattern and predicting how organisms respond to it (Cairns, 1992; Mangel, 1994; Furness and Camphuysen, 1997; Reid et al., 2005). Therefore, it is important to understand how krill-dependent predators forage for krill and how they may respond to variable conditions of krill patchiness.

Land-based krill predators breeding on the South Shetland Islands consume ∼0.83 million tons of krill during the reproductive season (Croll and Tershy, 1998). The krill fishery that operated near the South Shetland Islands and Elephant Island during the 1980s and 1990s targeted krill patches within <100 km of penguin and petrel colonies during their breeding periods (Agnew and Phegan, 1995; Marin and Delgado, 2001), so while the total catch was not great (100 kTons) the distribution of the fishing effort is of great concern (Hewitt et al., 2004c). Potential impacts on krill-dependent predators through commercial harvesting of krill should be addressed by incorporating the foraging demand of predators (Mangel, 1994; Mangel and Switzer, 1998; Marin and Delgado, 2001). Based on our study, we urge that future modeling work emphasize the impact of krill patchiness in relation to predator foraging demand. For example, negative effects such as competition, through the depletion by fishing vessels of patches exploited by predators, may cause predator populations to suffer (Cairns, 1992; Alonzo et al., 2003). Therefore, we should consider the effects of krill harvesting on predators, not only in terms of abundance and biomass, but also considering dynamics involving patch depletion.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and materials
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank the United States Antarctic Marine Living Resources program for facilitating fieldwork during this study. We greatly appreciate the assistance of the captain and crew of the R/V Yuzhmorgeologiya. A. J. Bernick, M. P. Force, and D. J. Futuyma assisted in data collection of predators. We would like to thank R. Hewitt for maintaining and processing acoustic data during 2004. This paper was enhanced by the comments of anonymous reviewers. Seabird data collection was made possible due to NSF-OPP grant (OPP-9983751) to R.R.V.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and materials
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Agnew, D.J. (1997) The CCAMLR ecosystem monitoring programme. Antarct. Sci. 9:235242.
  • Agnew, D.J. and Phegan, G. (1995) Development of a fine-scale model of land-based predator foraging demands in the Antarctic. CCAMLR Sci. 2:99110.
  • Alonzo, S.H., Switzer, P.V. and Mangel, M. (2003) An ecosystem-based approach to management: using individual behaviour to predict the indirect effects of Antarctic krill fisheries on penguin foraging. J. Appl. Ecol. 40:692702.
  • Anselin, L., Syabri, I. and Kho, Y. (2006) GeoDaS: an introduction to spatial data analysis. Geog. Anal. 38:522.
  • Atkinson, A., Siegel, V., Pakhomov, E. and Rothery, P. (2004) Long-term decline in krill stock and increase in salps within the Southern Ocean. Nature 432:100103.
  • Bengston, J.L., Croll, D.A. and Goebel, M.E. (1993) Diving behavior of Chinstrap Penguins at Seal Island. Antarct. Sci. 5:915.
  • Bernstein, C.A., Kacelnik, A. and Krebs, J.R. (1988) Individual decisions and the distribution of predators in a patchy environment. J. Anim. Ecol. 57:10071026.
  • Boyd, I.L. (1996) Temporal scales of foraging in a marine predator. Ecology 77:426434.
  • Boyd, I.L. and Murray, A.W.A. (2001) Monitoring a marine ecosystem using responses of upper trophic level predators. J. Anim. Ecol. 70:747760.
  • Buckley, N.J. (1996) Food finding and the importance of information, local enhancement and communal roosting on foraging success of North American vultures. Auk. 113:473488.
  • Buckley, N.J. (1997) Spatial-concentration effects and the importance of local enhancement in the evolution of colonial breeding seabirds. Am. Nat. 149:10911112.
  • Burnham, K.P. and Anderson, D.R. (2002) Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach, 2nd edn. New York: Springer.
  • Cairns, D.K. (1992) Bridging the gap between ornithology and fisheries science: use of seabird data in stock assessment models. Condor 94:811824.
  • Clarke, A., Murphy, E.J., Meredith, M.P. et al. (2007) Climate change and the marine ecosystem of the western Antarctic Peninsula. Philos. Trans. R. Soc. Lond. B 362:149166.
  • Constable, A.J., Nicol, S. and Strutton, P.G. (2003) Southern Ocean productivity in relation to spatial and temporal variation in the physical environment. J. Geophys. Res. 108:60016021.
  • Croll, D.A. and Tershy, B.R. (1998) Penguins, fur seals and fishing: prey requirements and potential competition in the South Shetlands, Antarctica. Pol. Biol. 19:365375.
  • Croll, D.A., Demer, D.A., Hewitt, R.P., Jansen, J.K., Goebel, M.E. and Tershy, B.R. (2006) Effects of variability in prey abundance on reproduction and foraging in Chinstrap penguins (Pygoscelis antarctica). J. Zool. 269:506513.
  • Croxall, J.P., Reid, K. and Prince, P.A. (1999) Diet provisioning and productivity responses of marine predators to differences in availability of Antarctic krill. Mar. Ecol. Prog. Ser. 177:115131.
  • Croxall, J.P., Trathan, P.N. and Murphy, E.J. (2002) Environmental change and Antarctic seabird populations. Science 297:15101514.
  • Davoren, G.K., Montevecchi, W.A. and Anderson, J.T. (2003) Distributional patterns of a marine bird and its prey: habitat selection based on prey and conspecific behaviour. Mar. Ecol. Prog. Ser. 256:229242.
  • Demer, D.A. and Conti, S.G. (2005) New target-strength model indicates more krill in the Southern Ocean. ICES J. Mar. Sci. 62:2532.
  • Duffy, D.D. (1986) Foraging at patches: interactions between common and roseate terns. Ornis Scand. 17:4752.
  • Everson, I. and Goss, C. (1991) Krill fishing activity in the southwest Atlantic. Antarct. Sci. 3:351358.
  • Fauchald, P. (1999) Foraging in a hierarchical patch system. Am. Nat. 153:603613.
  • Fauchald, P. and Erikstad, K.E. (2002) Scale-dependent predator-prey interactions: the aggregative response of seabirds to prey under variable prey patchiness. Mar. Ecol. Prog. Ser. 231:279291.
  • Fauchald, P., Erikstad, K.E. and Skarsfjord, H. (2000) Scale-dependent predator-prey interactions: the hierarchical spatial distribution of seabirds and prey. Ecology 81:773783.
  • Forcada, J., Trathan, P.N., Reid, K., Murphy, E.J. and Croxall, J.P. (2006) Contrasting population changes in sympatric penguin species in association with climate warming. Glob. Chang. Biol. 12:411423.
  • Fraser, W.R. and Hoffman, E.E. (2003) A predator’s perspective on causal links between climate change, physical forcing and ecosystem response. Mar. Ecol. Prog. Ser. 265:115.
  • Furness, R.W. and Camphuysen, C.J. (1997) Seabirds as monitors of the marine environment. ICES J. Mar. Sci. 54:726737.
  • Greene, C.M. and Stamps, J.A. (2001) Habitat selection at low population densities. Ecology 82:20912100.
  • Grünbaum, D. (1998) Using spatially explicit models to characterize foraging performance in heterogeneous landscapes. Am. Nat. 151:97115.
  • Grünbaum, D. and Veit, R.R. (2003) Black-browed albatross foraging on Antarctic krill: density-dependence through local enhancement? Ecology 84:32653275.
  • Harrison, N.M., Whitehouse, M.J., Heinemann, D.W., Hunt, G.L. and Veit, R.R. (1991) Multi-species foraging flocks of birds and fur seals around South Georgia. Auk 108:801810.
  • Hewitt, R.P. and Demer, D.A. (1993) Dispersion and abundance of Antarctic krill in the vicinity of Elephant Island in the 1992 austral summer. Mar. Ecol. Prog. Ser. 58:2939.
  • Hewitt, R.P., Demer, D.A. and Emery, J.H. (2003) An eight-year cycle in krill biomass density inferred from acoustic surveys conducted in the vicinity of the South Shetland Islands during the austral summers of 1991/92 through 2001/02. Aquat. Living Resour. 16:205213.
  • Hewitt, R.P., Watkins, J., Naganobu, M. et al. (2004a) Biomass of Antarctic krill in the Scotia Sea in January/February 2000 and its use in revising an estimate of precautionary yield. Deep Sea Res. 51:12151236.
  • Hewitt, R.E., Watters, G., Trathan, P.H. et al. (2004b) Options for allocating the precautionary catch limit of krill among small-scale management units in the Scotia Sea. CCAMLR Sci. 11:8197.
  • Hewitt, R.P., Kimm, S., Naganobou, M. et al. (2004c) Variation in the biomass and demography of Antarctic krill in the vicinity of the South Shetland Islands during the 1999/2000 austral summer. Deep Sea Res. 51:14111419.
  • Hunt, G.L. (1990) The pelagic distribution of marine birds in a heterogeneous environment. Polar Res. 8:4354.
  • Hunt, G.L. and Schneider, D.C. (1987) Scale-dependent processes in the physical and biological environment of marine birds. In: Seabirds: Feeding Ecology and Role in Marine Ecosystems. J.P.Croxall (ed.) Cambridge: Cambridge University Press, pp. 741.
  • Hunt, G.L., Heinemann, D., Veit, R.R., Heywood, R.B. and Everson, I. (1990) The distribution, abundance and community structure of marine birds in southern Drake Passage and Bransfield Strait, Antarctica. Continental Shelf Res. 10:243257.
  • Hunt, G.L., Heinemann, D. and Everson, I. (1992) Distributions and predator-prey interactions of macaroni penguins, Antarctic fur seals, and Antarctic krill near Bird Island, South Georgia. Mar. Ecol. Prog. Ser. 86:1530.
  • Jones, C.D. and Ramm, D.C. (2004) The commercial harvest of krill in the southwest Atlantic before and during the CCAMLR 2000 survey. Deep-Sea Res. II 51:14211434.
  • Kareiva, P. and Odell, G. (1987) Swarms of Predators exhibit ‘preytaxis’ if individual predators use area-restricted search. Am. Nat. 130:223270.
  • Legendre, P. and Legendre, L. (1998) Numerical Ecology. Amsterdam, the Netherlands: Elsevier Science.
  • Loeb, V., Siegel, V., Holm-Hansen, O. et al. (1997) Effects of sea-ice extent and krill or salp abundance on the Antarctic food web. Nature 387:897900.
  • Macaulay, M.C., English, T.S. and Mathisen, O.A. (1984) Acoustic characterization of swarms of Antarctic krill (Euphausia superba) from Elephant Island and Bransfield Strait. J. Crust. Biol. 4:1644.
  • Madureira, L.S.P., Everson, I. and Murphy, E.J. (1993) Interpretation of acoustic data at two frequencies to discriminate between Antarctic krill (Euphausia superba Dana) and other scatterers. J. Plankton Res. 15:787802.
  • Mangel, M. (1994) Spatial patterning in resource exploitation and conservation. Philos. Trans. R. Soc. Lond. B 343:9399.
  • Mangel, M. and Switzer, P.V. (1998) A model at the level of the foraging trip for the indirect effects of krill (Euphausia superba) fishers on krill predators. Ecol. Model. 105:235256.
  • Marin, V.H. and Delgado, L.E. (2001) A spatially explicit model of the Antarctic krill fishery off the South Shetland Islands. Ecol. Appl. 11:12351248.
  • May, R.M., Beddington, J.R., Clark, C.W., Holt, S.J. and Laws, R.M. (1979) Management of multispecies fisheries. Science 205:267277.
  • Miller, D.G.M. and Hampton, I. (1989) Krill aggregation characteristics: spatial distribution patterns from hydroacoustic observations. Polar Biol. 10:125134.
  • Moline, M.A., Clasure, H., Frazer, T.K., Schofield, O. and Vernet, M. (2004) Alteration of the food web along the Antarctic Peninsula in response to a regional warming trends. Glob. Chang. Biol. 10:19731980.
  • Morris, D.W. (1987) Tests of density-dependent habitat selection in a patchy environment. Ecol. Monogr. 57:269281.
  • Murphy, E.J., Morris, D.J., Watkins, J.L. and Priddle, J. (1988) Scales of interactions between Antarctic krill and the environment. In: Antarctic Ocean and Resources Variability. D.Sarhage (ed.) Berlin: Springer, pp. 120130.
  • Murphy, E.J., Watkins, J.L., Reid, K. et al. (1998) Interannual variability of the South Georgia marine ecosystem: biological and physical sources of variation in the abundance of krill. Fish. Oceanogr. 7:381390.
  • Murphy, E.J., Thorpe, S.E., Watkins, J.L. and Hewitt, R. (2004) Modeling the krill transport pathways in the Scotia Sea: spatial and environmental connections generating the seasonal distribution of krill. Deep Sea Res. 51:14351456.
  • Murphy, E.J., Watkins, J.L., Trathan, P.N. et al. (2007) Spatial and temporal operation of the Scotia Sea ecosystem: a review of large-scale links in a krill centered food web. Philos. Trans. R. Soc. Lond. B 362:113148.
  • Obst, B.S. (1985) Densities of Antarctic seabirds at sea and the presence of krill Euphausia superba. Auk 102:540549.
  • Piatt, J. (1990) The aggregative response of common murres and Atlantic puffins to schools of capelin. Stud. Avian Biol. 14:3651.
  • Piatt, J. and Methven, D.A. (1992) Threshold foraging behavior of baleen whales. Mar. Ecol. Prog. Ser. 84:205210.
  • Pinaud, D. and Weimerskirch, H. (2007) At-sea distribution and scale-dependent foraging behaviour of petrels and albatrosses: a comparative study. J. Anim. Ecol. 76:919.
  • Powell, T.M. and Okubo, A. (1994) Turbulence, diffusion and patchiness in the sea. Philos. Trans. R. Soc. Lond. B 343:1118.
  • Pöysä, H. (1992) Group foraging in patchy environments: the importance of coarse-level local enhancement. Ornis Scand. 23:159166.
  • Reid, K., Sims, M., White, R.W. and Gillon, K. (2004) Spatial distribution of predator/prey interactions in the Scotia Sea: implications for measuring predator/fisheries overlap. Deep Sea Res. 51:13831396.
  • Reid, K., Croxall, J.P., Briggs, D.R. and Murphy, E.J. (2005) Antarctic ecosystem monitoring: quantifying the response of ecosystem indicators to variability in Antarctic krill. ICES J. Mar. Sci. 62:366373.
  • Reiss, C.S., Cossio, A.M., Loeb, V. and Demer, D.A. (2008) Variations in the biomass of Antarctic krill (Euphausia superba) around the South Shetland Islands, 1996–2006. ICES J. Mar. Sci. 65:497508.
  • Rose, G.A. and Leggett, W.C. (1990) The importance of scale to predator-prey spatial correlations: an example of Atlantic fishes. Ecology 71:3343.
  • Schlitzer, R. (2004) Ocean Data View.
  • Siegel, V. and Loeb, V. (1995) Recruitment of Antarctic krill Euphausia superba and possible causes for its variability. Mar. Ecol. Prog. Ser. 123:4556.
  • Silverman, E.D. and Veit, R.R. (2001) Associations among Antarctic seabirds in mixed-species feeding flocks. Ibis 143:6162.
  • Silverman, E.D., Veit, R.R. and Nevitt, G.A. (2004) Nearest neighbors as foraging cues: information transfer in a patchy environment. Mar. Ecol. Prog. Ser. 277:2535.
  • Tasker, M.L., Jones, P.H., Dixon, T. and Blake, B.F. (1984) Counting seabirds at sea from ships: a review of methods employed and a suggestion for a standardized. Auk 101:567577.
  • Trathan, P.N., Brierley, A.S., Brandon, M.A. et al. (2003) Oceanographic variability and changes in Antarctic krill (Euphausia superba) abundance at South Georgia. Fish. Oceanogr. 12:569583.
  • Tynan, C.T. (1998) Ecological importance of the southern boundary of the Antarctic circumpolar current. Nature 392:708710.
  • Veit, R.R. (1999) Behavioural responses by foraging petrels to swarms of Antarctic krill Euphausia superba. Ardea 87:4150.
  • Veit, R.R. and Prince, P.A. (1997) Individual and population dispersal of black-browed and grey-headed albatrosses in response to Antarctic krill. Ardea 85:129134.
  • Veit, R.R., Silverman, E.D. and Everson, I. (1993) Aggregation patterns of pelagic predators and their principal prey, Antarctic krill, near South Georgia. J. Anim. Ecol. 62:551564.
  • Warham, J. (1990) The Petrels: Their Ecology and Breeding Systems. London: Academic Press Limited.
  • Wiens, J.A. (1976) Population responses to patchy environments. Ann. Rev. Ecol. Syst. 7:81120.
  • With, K.A., Cadaret, S.J. and Davis, C. (1999) Movement responses to patch structure in experimental fractal landscapes. Ecology 80:13401353.
  • Zar, J.H. (1999) Biostatistical Analysis, 4th edn. Upper Saddle River, NJ: Prentice Hall.
  • Zollner, P.A. and Lima, S.L. (1999) Search strategies for landscape-level interpatch movements. Ecology 80:10191030.