Spatial Distribution and Temporal Patterns of Cassin's Auklet Foraging and Their Euphausiid Prey in a Variable Ocean Environment

Abstract

Krill (Euphausiids) play a vital ecosystem role in many of the world's most productive marine regions, providing an important trophic linkage. We introduce a robust modeling approach to link Cassin's auklet (Ptychoramphus aleuticus) abundance and distribution to large-scale and local oceanic and atmospheric conditions and relate these patterns to similarly modeled distributions of an important prey resource, krill. We carried out at-sea strip transect bird surveys and hydroacoustic assessments of euphausiids (2004-2013). Data informed separate, spatially-explicit predictive models of Cassin's auklet abundance (zero-inflated negative binomial regression) and krill biomass (two-part model) based on these surveys. We established the type of prey responsible for acoustic backscatter by conducting net tows of the upper 50 m during surveys. We determined the types of prey fed to Cassin's auklet chicks by collecting diet samples from provisioning adults. Using time-depth-recorders, we found Cassin's auklets utilized consistent areas in the upper water column, less than 30 m, where krill could be found (99.5% of dives were less than 30 m). Birds primarily preyed upon two species of euphausiids, Euphausia pacifica and Thysanoessa spinifera, which were available in the upper water column. Cassin's auklet abundance was best predicted by both large scale and localized oceanic processes (upwelling) while krill biomass was best predicted by local factors (temperature, salinity, and fluorescence) and both large scale and localized oceanic processes (upwelling). Models predicted varying krill and bird distribution by month and year. Our work informs the use of Cassin's auklet as a valuable indicator or krill abundance and distribution and strengthens our understanding of the link between Cassin's auklet and its primary prey. We expect future increases in frequency and magnitude of anomalous ocean conditions will result in decreased availability of krill leading to declines in the Farallon Islands population of Cassin's auklets.

Fig 1. Study area within the Gulf of the Farallones and Cordell Bank National Marine Sanctuaries, California, USA, showing ACCESS program transect lines and oceanographic sampling stations visited for CTD casts and zooplankton tows.

Fig 2. Schematic study modeling methodology showing binned and environmental data, predicted to 1 km² prediction cells, linked with Cassin’s auklet diet data, Cassin’s auklet TDR dive data, and zooplankton tow samples from the upper water column.

Fig 3. Predicted krill biomass (g m^-2), 2004–2013: variations in modeled habitat use for May (a, predictions averaged over n = 8 years), June (b, n = 7), July (c, n = 7), and September (d, n = 8); each gradation represents a decile, monthly mean values with all years pooled.

The 200 m isobath and boundaries of National Marine Sanctuaries included in blue and black, respectively.

Fig 4. Predicted krill biomass (g m^-2), 2004–2013 (a-j, sequentially): variations in annual modeled habitat use (predictions averaged over n = 3 months except 2010, n = 4, and 2013, n = 2); each gradation represents a decile, annual mean values with all months pooled per year.

The 200 m isobath and boundaries of National Marine Sanctuaries included in blue and black, respectively.

Fig 5

Annual variation in predicted (a) Cassin’s auklet abundance (birds km^-2) and (b) krill biomass (g m^-2), yearly averaged values for (c) the Southern Oscillation Index (SOI), (d) the Pacific Decadal Oscillation (PDO), (e) the North Pacific Gyre Oscillation, and (f) Upwelling Index within the Gulf of the Farallones and Cordell Bank National Marine Sanctuaries by year (2004–2013) with all months pooled.

Fig 6. Predicted (a) krill biomass (g m^-2) and (b) Cassin’s auklet abundance (birds km^-2), 2004–2013: top habitat use model applied across all months and years; each gradation represents a decile.

The 200 m isobath and boundaries of National Marine Sanctuaries included in blue and black, respectively.

Fig 7. Predicted Cassin’s auklet abundance (birds km^-2), 2004–2013: variations in modeled habitat use for May (a), June (b), July (c), and September (d); each gradation represents a decile, monthly mean values with all years pooled.

The 200 m isobath and boundaries of National Marine Sanctuaries included in blue and black, respectively.

Fig 8. Predicted Cassin’s auklet abundance (birds km^-2) by year, 2004–2013 (a-j, sequentially): variations in annual modeled habitat use; each gradation represents a decile, annual mean values with all months pooled per year.

The 200 m isobath and boundaries of National Marine Sanctuaries included in blue and black, respectively.

Fig 9. Intersection of high predicted spatial co-occurrence in red for krill and Cassin’s auklets combined over May-July and September 2004–2013 at top 20% habitat use for each predictive model.

Consistent high use areas (Cordell Bank and the Farallones Escarpment) are noted. 200 m isobath and boundaries of National Marine Sanctuaries included in blue and black, respectively.

Fig 10. Intersection of high predicted spatial co-occurrence in red for krill and Cassin’s auklets for May (a), June (b), July (c), and September (d) at top 20% habitat use for each predictive model.

200 m isobath and boundaries of National Marine Sanctuaries included in blue and black, respectively.

Fig 11. Intersection of high predicted spatial co-occurrence in red for krill and Cassin’s auklets 2004–2013 (a-j, sequentially) at top 20% habitat use for each predictive model.

200 m isobath and boundaries of National Marine Sanctuaries included in blue and black, respectively.

Fig 12. Annual proportion of major zooplankton taxa, by number, in the chick provisioning diet of Cassin’s auklets on Southeast Farallon Island, 1985–2013.

Numbers above the bars indicate sample sizes.

Fig 13. Percent frequency of occurrence of E. pacifica and T. spinifera adults and juveniles in surface hoop net samples (2004–2011).