marine bird and cetacean associations with bathymetric habitats … et al. 2004.pdf · marine bird...

www.elsevier.com/locate/jmarsys

Journal of Marine Systems 50 (2004) 79–99

Marine bird and cetacean associations with bathymetric habitats

and shallow-water topographies: implications for trophic

transfer and conservation

Peggy P.W. Yena,*, William J. Sydemana, K. David Hyrenbacha,b

aMarine Ecology Division, PRBO Conservation Science, 4990 Shoreline Highway, Stinson Beach, CA 94970, USAbDuke University Marine Laboratory, 135 Duke Marine Laboratory Road, Beaufort, NC 28516, USA

Received 20 June 2003; accepted 12 September 2003

Available online 18 May 2004

Abstract

We investigated the aggregative response of marine birds and cetaceans to bathymetric features in central California over 4

years, 1996–1997 and 2001–2002. A total of 1700 km2 of ocean habitat was surveyed over six cruises. We considered the

distribution of the most abundant marine birds and mammals in relation to bathymetry. We focused our analyses on eight focal

taxa: Cassin’s auklet (Ptychoramphus aleuticus), common murre (Uria aalge), sooty shearwater (Puffinus grieus), phalarope

species (red, and red-necked: Phalaropus fulicaria, Phalaropus lobatus), Dall’s porpoise (Phocoenoides dalli), Pacific white-

sided dolphin (Lagenorhynchus obliquidens), humpback whale (Megaptera novaeangliae), and Risso’s dolphin (Grampus

griseus). We evaluated associations of top predators with seven bathymetric indices and three distance measurements to shallow-

water topographies. The bathymetric descriptors included (1) median depth, (2) depth coefficient of variation, (3) contour index,

and shortest distance to (4) the mainland, (5) the continental shelf-break (200-m isobath), (6) the continental slope (1000-m

isobath), and (7) pelagic waters (3000-m isobath). The measurements of shallow water topographies included the shortest

distance to: (8) the Cordell Bank seamount, (9) the Farallon Island Archipelago (a breeding colony for auklets and murres), and

(10) Monterey Canyon. We documented two instances of spatial autocorrelation (for Cassin’s auklet and common murre) at lags

(distances) of 0–3 and 3–9 km, respectively, and accounted for this spatial pattern in analyses of habitat associations. We found

similar relationships between cetaceans and bathymetric features at both interannual and weekly time scales. Seabirds revealed

both persistent and variable relationships through time. For the resident breeding murres, we detected an interannual trend in

habitat use, with these birds shifting their distribution offshore over time. Our study demonstrates that resident and migrant

marine birds and cetaceans are associated with bathymetric features and shallow-water topographies, though responses varied

across species and time. In spite of this variability, we contend that bathymetric associations of upper trophic-level predators can

help delineate sites of elevated trophic transfer. An understanding of marine productivity and predator aggregation patterns is

essential to design ecosystem-level conservation plans for protecting marine habitats and species.

D 2004 Elsevier B.V. All rights reserved.

Keywords: Seabirds; Cetaceans; Habitat hotspots; Bathymetry; Cordell Bank, Monterey Canyon, Gulf of the Farallones; Shelf-break, California

current system

0924-7963/$ - see front matter D 2004 Elsevier B.V. All rights reserved.

doi:10.1016/j.jmarsys.2003.09.015

* Corresponding author.

E-mail address: [email protected] (P.P.W. Yen).

P.P.W. Yen et al. / Journal of Marine Systems 50 (2004) 79–9980

1. Introduction

Biological oceanographers have studied the pelagic

dispersion and habitats of marine birds and mammals

for decades (Jesperson, 1924; Murphy, 1936; Jaquet

and Whitehead, 1996). While previous studies have

revealed that these upper trophic-level marine preda-

tors associate with specific physical and biological

processes at distinct spatial and temporal scales (Hunt

and Schneider, 1987; Jaquet and Whitehead, 1996;

Hyrenbach and Veit, 2003), the predictability of

wildlife–habitat associations and the underlying bio-

physical coupling mechanisms remain, for the most

part, poorly understood. Marine birds and cetaceans

associate with specific water masses, hydrographic

fronts (convergence and divergence zones), and other

mesoscale features such as eddies (reviewed by Hunt

and Schneider, 1987; Schneider, 1991; Jaquet, 1996;

Croll et al., 1998). Horizontal gradients in water

density and the degree of vertical stratification pro-

mote the aggregation of weakly swimming prey at

discontinuities, which in turn provides enhanced feed-

ing opportunities for many marine predators (Hunt et

al., 1990, 1996, 1998; Franks, 1992; Ribic and Ainley,

1997; Spear et al., 2001).

Foraging seabirds and cetaceans are also associated

with a variety of bathymetric features, including

shallow banks and continental shelf-slope regions

(Hui, 1985; Hunt and Schneider, 1987; Hunt et al.,

1996; Baumgartner et al., 2001). In particular, conti-

nental shelf-breaks and slopes appear to be highly

productive habitats, which frequently support high

densities of marine predators (Briggs et al., 1987;

Schoenherr, 1991; Springer et al., 1996; Croll et al.,

1998). Furthermore, many continental shelves are

characterized by complex bathymetries, including

submarine canyons, deep basins, and shallow banks

(Allen et al., 2001). These structures influence water

flow and give rise to secondary circulation features

(e.g., shelf-break fronts, eddies), which often aggre-

gate zooplankton and weakly swimming organisms

and make prey available close to the surface to diving

predators (Simard et al., 1986; Schoenherr, 1991;

Hunt et al., 1996, 1998; Croll et al., 1998; Allen et

al., 2001). The anchoring of important hydrographic

processes at bathymetric features has been invoked to

explain relatively persistent habitat associations of

upper-trophic marine predators in coastal systems

(Haney and McGillivary, 1985; Larson et al., 1994;

Hunt et al., 1996, 1998). Quantifying the magnitude

and predictability of these associations is essential to

characterize bio-physical coupling in dynamic shelf-

slope marine ecosystems.

In the highly variable California Current System

(CCS), high levels of ocean productivity are gen-

erally observed over the continental shelf, particu-

larly adjacent to coastal upwelling centers (Huyer,

1983; Kudela and Chavez, 2000; Marinovic et al.,

2002). Intermediate levels of ocean productivity are

apparent over the shelf-break and slope (200–2000

m), with lowest levels in offshore waters deeper

than 2000 m (Pelaez and McGowan, 1986; Fargion

et al., 1993; Hayward and Venrick, 1998; Venrick,

1998). In spite of seasonal variability in community

structure and overall numbers, marine bird and

mammal abundance patterns in the CCS are char-

acterized by similar onshore–offshore gradients,

with highest densities along the shelf-slope region

and lower densities in offshore waters (Briggs et

al., 1987; Wahl et al., 1993; Forney and Barlow,

1998).

Herein, we quantify the association of marine

birds and mammals with bathymetric habitats in the

Gulf of the Farallones and Monterey Bay region

(37.8jN–36.8jN) during the spring-time (May–

June) period of strong coastal upwelling in the central

CCS (Fig. 1a). Hundreds of thousands of marine

birds and tens of thousands of marine mammals

breed, overwinter, or migrate through the Gulf of

the Farallones/Monterey Bay (hereafter GOF/MB),

where they forage on rich, yet highly variable prey

resources on the wide continental shelf (e.g., Briggs

et al., 1987; Oedekoven et al., 2001; Benson et al.,

2002). The springtime marine bird and mammal

distribution and abundance patterns in this region

have been previously described in relation to hydro-

graphic conditions (Allen, 1994; Keiper, 2001;

Oedekoven et al., 2001). However, the predictability

of wildlife–habitat associations, particularly over

temporal scales (e.g., weekly) comparable to the

periodicity of upwelling-favorable wind events

(Huyer, 1983; Strub et al., 1991; Wing et al.,

1995), has yet to be fully investigated. Building on

these previous investigations, we address the hypoth-

esis that upper trophic-level marine predators are

predictably associated with specific bathymetric char-

Fig. 1. The study area in the Gulf of the Farallons, within the California Current system (CCS), showing (a) the different ocean depth domains and the distribution of 3-km survey bins

surveyed during sweep 1 of 2001, and (b) illustrating three major bathymetric features: Cordell Bank, the Farallon Archipelago, and the Monterey Canyon.

P.P.W.Yen

etal./JournalofMarin

eSystem

s50(2004)79–99

81


acteristics of the continental shelf and shelf-break

region, including submarine canyons and seamounts.

2. Materials and methods

2.1. Study area

We conducted this study in the GOF/MB region, an

area characterized by a broad continental shelf and

complex bathymetry in the form of an island chain

(the Farallon Archipelago), a shallow ( < 50-m summit

depth) seamount (Cordell Bank), a major submarine

canyon (Monterey Canyon), and numerous smaller

canyons bisecting the continental shelf (Fig. 1b).

The Farallon Archipelago consists of two main

islands and numerous sea stacks and islets stretching

over approximately 15 km along the outer continental

shelf. The Farallones are a National Wildlife Refuge,

managed by US Fish and Wildlife Service (USFWS),

and host the largest marine bird and mammal colonies

in the contiguous US. Currently, approximately

350,000 birds of 12 species breed at the Farallon

Archipelago (PRBO and USFWS unpublished data).

Approximately 10000 pinnipeds belonging to five

species also reproduce and/or haul out on the archi-

pelago (Sydeman and Allen, 1999). Cordell Bank, a

prominent seamount situated 20 km northwest

(38.02jN 123.44jW) of the westernmost Farallon

islet (Fig. 1b), is approximately 111 km2 in area,

and reaches to within 35 m of the ocean surface from

about 1830-m depth only a few miles away. Waters

surrounding the Farallon Archipelago and Cordell

Bank are encompassed by the Gulf of the Farallones

and Cordell Bank National Marine Sanctuaries (Fig.

1a), respectively, which are managed by the National

Oceanic and Atmospheric Administration (NOAA)/

National Ocean Service (NOS). The Monterey Bay

canyon system, consisting of the Soquel, Monterey,

and Carmel canyons, is located 140 km southeast of

the Farallon Archipelago, within the Monterey Bay

National Marine Sanctuary, and covers approximately

1200 km2 (Fig. 1b).

2.2. Physical oceanography of the region

Coastal upwelling is a dominant physical oceano-

graphic process in the central CCS (e.g., Strub et al.,

1991; Kudela and Chavez, 2000). The resulting en-

richment of surface waters from upwelling can lead to

exceptional levels of marine productivity (Chavez,

1996; Hayward and Venrick, 1998; Schwing et al.,

2000). Coastal upwelling is temporally (e.g., season-

ally) and spatially predictable, often anchored on

capes and headlands. In the northern portion of our

study area (GOF), upwelling is centered at the Point

Reyes peninsula and occurs mostly year-round, with a

seasonal peak in the late spring and early summer

(Bakun, 1975; GLOBEC, 1992; Wing et al., 1995). In

the southern portion of our study area (MB), upwell-

ing is centered near Davenport (Pennington and

Chavez, 2000). During the seasonal peak of upwelling

(May–June), a relatively persistent upwelling plume

extends 15–75 km offshore in the GOF, where it

interacts with meanders and eddies along the inner

boundary of the California Current (Schwing et al.,

1991; Steger et al., 2000). While upwelling is less

vigorous in the Monterey Bay region, cold surface

waters from the Davenport upwelling plume are

advected southward across the mouth the bay (Kudela

and Chavez, 2000) and localized upwelling occurs at

the edges of Monterey Canyon (Shea and Broenkow,

1982).

2.3. Predator surveys

Vessel-based surveys of marine predators were

conducted as part of annual NOAA/National Marine

Fisheries Service (NMFS) Rockfish Recruitment

Studies (RRS) in 1997 (1 sweep, 8–16 June), 2001

(2 sweeps, 11–27 May), and 2002 (2 sweeps, 9–26

May) aboard the RV David Starr Jordan. In 1996,

surveys were conducted aboard the RV MacArthur in

the GOF only (1 sweep, 12–18 May). We used

standard techniques for censusing seabirds and marine

mammals at sea (Tasker et al., 1984; Buckland et al.,

1993). Briefly, all seabirds that entered a 90j arc from

the bow to the beam and out to 300 m on the one side

with best visibility (e.g. lowest sun glare) were enu-

merated and their behavior recorded by two to three

observers stationed on the flying bridge, eye height 10

m above the surface of the water. Most surveys were

conducted while the research vessel was underway at

a speed of 10–12 knots (18.6–22.3 km h� 1), and

observers went off effort when wind speed exceeded

25 knots (46.5 km h� 1). A hand-held range finder and

P.P.W. Yen et al. / Journal of Marine Systems 50 (2004) 79–99 83

binoculars equipped with reticules were used to

ground-truth the width of the 300-m survey strip

(Heinemann, 1981). All marine mammals were

recorded using standard line transect techniques;

sightings from the center line to the horizon were

recorded, and the radial distance and the angle from

the track were estimated for most sightings (Heine-

mann, 1981; Buckland et al., 1993). We used the

relative marine mammal numbers rather than the

absolute numbers, unlike the seabird data, which is

expressed in absolute abundance. Though the field

technicians changed across years, experienced observ-

ers supervised the surveys each year to ensure high

quality data collection (1996 and 1997: M.B. Decker;

2001: C. Oedekoven; 2002: K.D. Hyrenbach).

2.4. General analytical approach

NMFS-RRS repeatedly surveyed a series of regu-

larly scheduled and standardized hydrographic and

mid-water trawl stations in the GOF/MB region.

Completion of a ‘‘sweep’’ through the entire network

of sampling stations requires approximately one week

(e.g., Fig. 1a). NMFS attempts to complete three full

sweeps of all hydrographic/mid-water trawl stations

each year. As such, this afforded an opportunity for

replicate surveys on both yearly and ‘‘weekly’’ time

scales. In 1997, we participated on one sweep through

the study area, whereas in 2001 and 2002 two sweeps

were made. In 1996, we participated on a different

study designed to examine predator–prey relation-

ships in relation to upwelling and relaxation events (a

modified continuation of Wing et al., 1995). Because

this shorter survey occurred entirely within the geo-

graphic range of the larger NMFS/RRS surveys, we

included these data in these data. This paper inves-

tigates habitat associations between marine predators

and bathymetric features at an interannual scale (rep-

licate spring cruises during different years). To ad-

dress interannual variability, we merged the data from

the two sweeps available in 2001 and 2002, and

contrasted the patterns observed during four years

(1996, 1997, 2001, 2002).

2.5. Data synthesis

During our study we surveyed a total of 5660 km

(avgF std; 863F 367 km, n: 6 sweeps) during 52

days at sea. Continuous transects were divided into 3-

km segments and the density of seabirds (number of

individuals sighted per square kilometer surveyed,

given the 300-m strip width) and the encounter rate

of marine mammals (number of individuals sighted

per kilometer surveyed) were calculated. Because we

were interested in delineating habitats where trophic

transfer to marine predators was likely to be taking

place, density calculations included only birds ob-

served feeding or sitting on the ocean. Birds flying or

following the ship were excluded. We selected the

most numerous seabird and mammal species for

analysis (Fig. 2a and b). The seabirds included two

locally breeding residents (Cassin’s auklet, Ptychor-

amphus aleuticus; and common murre, Uria aalge)

(Sydeman et al., 2001), and two seasonal migrants

(sooty shearwater, Puffinus grieus; and phalarope

species (red and red-necked, Phalaropus fulicaria

and Phalaropus lobatus) (Briggs et al., 1987). Of

the four marine mammals selected, two are seasonal

migrants (humpback whale, Megaptera novaeangliae;

and Risso’s dolphin, Grampus griseus), and two are

year-round residents (Dall’s porpoise, Phocoenoides

dalli; pacific white-sided dolphin, Lagenorhynchus

obliquidens) off central California (Forney and Bar-

low, 1998).

2.6. Data analysis

Before we examined bathymetric habitat associa-

tions, we determined whether species-specific distri-

butions were spatially autocorrelated. That is, we

asked whether there was a statistically significant

tendency for sightings of animals of the same species

to occur in the vicinity of each other. Significant

autocorrelation indicates that samples, in this case 3-

km survey bins, are not independent (Schneider,

1990). The lack of sample independence enhances

the likelihood of falsely rejecting the null hypothesis

(type I error), thus finding spurious significant habitat

relationships when there are none (Hurlbert, 1984).

We evaluated patterns of spatial autocorrelation using

Moran’s I, a metric developed to quantify spatial

patterns for highly non-normal distributions (Moran,

1950) and previously used to assess seabird spatial

autocorrelation (Fauchald and Erikstad, 2002). Mor-

an’s I provides a measure of the similarity of predator

density/encounter rates as a function of the spatial

Fig. 2. The cumulative proportional abundance for (a) marine birds

and (b) mammals observed during this study (1996, 1997, 2001, and

2002). The top four seabird species account for 89% of the total bird

abundance. The top four marine mammals account for 77% of the

total abundance. A total of 57 seabird and 20 mammal species were

sighted during this study.


distance (lag) between consecutive segments (in this

case 3 km survey bins). Values of the index range

from + 1 (clumped distribution: adjacent bins have

similar densities), to 0 (random distribution: adjacent

bins not related to each other), to � 1 (uniform

distribution: adjacent bins have dissimilar densities).

Positive and negative Moran’s I values are indicative

of aggregation and over-dispersion, respectively,

while indices close to zero suggest a random distri-

bution where sightings are not clumped nor uniformly

distributed (Sokal and Oden, 1978). We computed

Moran’s I for each species separately during each

year, and for both sweeps in 2001 and 2002 (n:

8 analyses for each species). We considered up to

30 lags in 3-km increments (i.e., range of lags: 3–90

km) for all years and sweeps except for 1996, when

the survey boundary lengths did not exceed 90 km.

The significance of autocorrelation statistics was

quantified using a randomization procedure, whereby

the observed predator spatial distributions were com-

pared to expected distributions arising from a random

rearrangement of the observed densities/encounter

rates among the sampled locations (Manly, 1991).

This procedure allows one to determine where the

observed spatial autocorrelation statistics fall within

the range of values which would be generated by

chance, given a null hypothesis of no spatial patterns

in predator distributions (Fotheringham et al., 2000).

We performed 500 permutations at each spatial lag for

each species, and compared the observed Moran’s I

statistics to those calculated from the randomization

tests. The autocorrelation metrics included all possible

pairwise samples within a specified distance (spatial

lag) from each other, regardless of their relative

orientation (e.g., North–South, East–West). Our anal-

ysis assumed the underlying autocorrelation structure

was isotropic, or characterized by the same spatial

scales of aggregation across and along the shelf. All

autocorrelation tests were processed using the pro-

gram RookCase V.1.6 (Sawada, 1999). Given the

large numbers of tests, we assumed significance at

the a = 0.01 level. That is, the observed Moran’s I

index for a given species and a specific lag was

significant if five or less of the 500 randomized

indices were larger in absolute value than the value

derived from the observed field data (Manly, 1991).

We explored habitat associations using the Arc-

View 3.2 (ESRI, 1996) geographic information sys-

tem (GIS), by overlaying the predator spatial

distributions (seabird densities and cetacean encounter

rates) over bathymetry. We considered three bathy-

metric indices and seven distance metrics to specific

habitat features: (1) median depth, (2) depth coeffi-

cient of variation (CV: S.D./mean), (3) contour index

[CI: (max depth�min depth)/max depth] within each

3 km by 1-km bin. Using the midpoint of each 3-km


survey bin, we measured the shortest distance to (4)

the mainland, (5) the continental shelf-break (200-m

isobath), (6) the continental slope (1000-m isobath),

(7) pelagic waters (3000-m isobath), (8) the centroid

of Cordell Bank (38.02jN 123.44jW), (9) the centroid

of the Farallon Archipelago (37.70jN 123.00jW), and

(10) the centroid of Monterey Canyon (36.70jN122.11jW) (Table 1).

The response variables used to quantify predator

relative abundance (seabird density and cetacean

encounter rate) were skewed, and did not follow

normal or Poisson distributions (Sokal and Rohlf,

1981). Because standard transformations did not

appropriately linearize these data, we used a nonpara-

metric approach to ascertain statistical significance of

habitat associations. We categorized response varia-

bles into four relative abundance levels. First, bins

where species were absent were scored as ‘0’ values.

Next, for each species and all surveys combined, the

presence data was scored as ‘1’, ‘2’, or ‘3’ using the

33rd and 66th percentiles. Thus, a score of ‘1’

implied that the relative abundance ranged between

the minimum and the 33rd percentile. Similarly, a

score of ‘2’ and ‘3’ was assigned to relative abun-

dance data that fell between the 33rd and 66th

percentiles and the 66th percentile and the maximum

value, respectively.

Additionally, many of the bathymetric variables

were heteroscedastic (Sokal and Rohlf, 1981). To

lessen this problem and to assess the functional shape

of the relationships between marine predator relative

abundance and bathymetric characteristics, we evalu-

ated several transformations for each predictor varia-

Table 1

Variable summaries demonstrating the bathymetric environment in the Gu

Variable Description

Dist_Land Distance from bin midpoint to closest land

Dist_200 m Distance from bin midpoint to 200-m isobat

Dist_1000 m Distance from bin midpoint to 1000-m isob

Dist_3000 m Distance from bin midpoint to 3000-m isob

Dist_Cordell Distance from bin midpoint to 38.02jN 123

Dist_Monterey Distance from bin midpoint to 36.70jN 122

Dist_Farallones Distance from bin midpoint to 37.70jN 123

Median_depth Median depth value within 3� 1-km bin

CV_depth Depth coefficient of variation within 3� 1-k

CI_depth Contour index within 3� 1-km bin

Source of the bathymetry data is the California Department of Fish and G

ble. Instead of arbitrarily applying one data trans-

formation to the entire data set, we allowed the data

themselves to determine which transformations were

most appropriate for each variable by fitting an ordered

logistic regression of the ranked abundance against

each of the predictor variables as follows: (1) non-

transformed (linear), (2) logarithmic, (3) squared, and

(4) square root. A ‘‘year’’ term was included, and for

each regression model, the log-likelihood ratio statistic

(LRS) was used to evaluate which transformation was

most appropriate for each predictor variable (Table 2).

The transformation with the best fit, expressed as the

largest (and most significant) LRS value, was selected

and used in a multi-variable ordered logistic regression

analysis.

The next stage of the analysis involved two steps.

First, we used a backwards stepwise ordered logistic

regression model to identify the most important

bathymetric predictor variables. Backward stepwise

logistic regression is ideal for an exploratory analysis

such as this one (Hosmer and Lemeshow, 2000).

Variables were removed from the model when

a< 0.05, thus yielding a reduced model, which

contained only significant predictor variables and

accounted for the greatest amount of the variance.

Next, the selected predictor variables were used to

develop a final model including a ‘‘year’’ term

(Table 3).

Once we had constructed a ‘‘habitat’’ model for

each species, we tested for persistence of the wild-

life–habitat associations by testing for interannual

variability. We fitted additional models including

‘‘year’’, as well as the interactions between this term

lf of the Farallones

min–max MeanF S.D.

312–81912 30,502F 18,615

h 5–59396 16,967F 12,870

ath 3–50740 15,727F 12,223

ath 2–81144 31,651F19,184

.44jW 0–196525 82,249F 54,939

.11jW 448–297552 126,797F 72,425

.00jW 1840–153386 71,623F 37,387

8–3684 1081F1128

m bin 0–83 6F 8

0–2790 33F 84

ames (DFG) Spatial Data Resources (V.2).

Table 2

Results of univariate logistic regression analysis for the marine predator abundance data and logarithmic (LN), square (SQ), square root (SQR),

nontransformed (L) bathymetric characteristics and distance to features, controlling for year

Variable Common

murre

Cassin’s

auklet

Sooty

shearwater

Phalarope

spp.

Dall’s

porpoise

Risso’s

dolphin

White-sided

dolphin

Humpback

whale

BT v2 BT v2 BT v2 BT v2 BT v2 BT v2 BT v2 BT v2

Dist_Land L 544.0 SQ 98.7 SQ 1.4 L 40.5 LN 2.1 SQ 4.1 LN 14.5 LN 0.4

Dist_200 m SQ 24.4 L 144.6 SQ 18.0 SQR 26.9 LN 5.2 SQ 4.2 SQR 0.3 SQ 1.8

Dist_1000 m SQR 59.8 SQ 39.9 SQ 26.1 SQ 15.1 SQ 4.6 SQR 10.1 SQ 6.3 SQ 6.7

Dist_3000 m SQR 366.6 LN 29.9 SQ 3.0 SQR 20.0 L 1.0 LN 2.6 SQ 18.0 SQ 1.7

Dist_Cordell LN 9.0 LN 72.6 SQ 9.5 SQ 74.1 SQR 4.3 SQ 6.7 SQ 2.1 SQR 1.1

Dist_Monterey SQ 38.5 L 10.7 LN 6.8 SQR 67.5 L 0.9 LN 7.6 LN 1.8 SQ 3.2

Dist_Farallones LN 90.8 LN 46.5 SQR 12.0 LN 45.6 SQR 0.4 LN 8.9 SQ 1.6 L 0.3

Median_depth LN 459.7 L 115.1 SQ 2.6 SQ 16.6 SQ 0.3 SQ 4.1 LN 18.3 SQ 0.6

CV_depth LN 10.6 LN 24.4 LN 5.5 LN 17.6 LN 4.8 LN 7.5 LN 0.3 SQ 1.5

CI_depth LN 10.7 LN 27.5 LN 6.1 LN 18.3 LN 5.1 LN 6.8 SQ 0.3 SQ 1.2

Likelihood ratio statistics were used to determine the best transformation (BT) procedure for each species and predictor variable. Results for all

transformations are available from the authors.


and each bathymetric variable. We evaluated these

models with and without interactions by comparing

the resulting LRS values. The likelihood ratio test was

performed for the full unrestricted model (including

all year-interaction terms), and the restricted model

(no interaction terms included). We interpreted signif-

icant interaction terms as evidence for yearly variation

in bathymetric habitat associations. For species with

significant yearly differences, we further examined

interannual effects using individual ‘‘year’’ terms in

the model. This allowed us to determine the variabil-

ity in habitat associations from year to year using the

Table 3

Final bathymetric associations for each species after controlling for year,

Common

murre

Cassin’s

auklet

Sooty

shearwater

Phala

spp.

Dist_Land � �Dist_200 m � �Dist_1000 m

Dist_3000 m

Dist_Cordell �Dist_Monterey �Dist_Farallones � � + +

Median_depth �CV_depth �CI_depth +

Model v2 639.52 264.11 98.01 130.9

Model P < 0.0001 < 0.0001 < 0.0001 < 0.0

Model R2 0.26 0.18 0.03 0.1

Positive relationships indicate higher densities farther away from a featur

heterogeneous bathymetry (CV, CI). Negative relationships signify higher

homogeneous bathymetry.

significance and the trends in the regression coeffi-

cients of the interaction terms.

3. Results

3.1. Spatial autocorrelation

We found no evidence of significant autocorrela-

tion for six out of eight species, indicating that the 3-

km bins provided an adequate scale of analysis (Fig.

3a and b). The Cassin’s auklet and common murre,

showing the nature of the habitat relationships

rope Dall’s

porpoise

Risso’s

dolphin

White-sided

dolphin

Humpback

whale

+ +

��

+

+

+

2 11.21 17.69 41.32 9.66

001 0.0474 0.0005 < 0.0001 0.02

2 0.03 0.10 0.08 0.01

e (distance), at greater depths (median depth), or with increasingly

densities closer to a feature, at shallower depths, or with increasingly


however, displayed significant spatial autocorrelation

at spatial scales between 3 and 9 km. In particular,

because the auklet data indicated significant aggrega-

tions at the 3-km spatial scale, this lack of sample

independence needs to be accounted for when inter-

preting bathymetric habitat models (see below).

3.2. Bathymetric associations

All eight species of marine predators analyzed in

this study revealed unique associations. A synopsis of

marine bird (Table 3) habitat associations is as fol-

lows: (1) common murres were observed at greater

densities closer to land, near the Farallons, and at

shallower depths (Table 3, Fig. 4a); (2) Cassin’s

auklets were found in greater densities near Cordell

Bank and the Farallones, over less variable bathym-

etry, and closer to the shelf-break (200 m). While

auklets displayed spatial autocorrelation, their habitat

model is significant (overall likelihood ratio statis-

tic = 264.1, p < 0.0001), as were each of the individual

terms within the model (all p < 0.0001). Thus, in spite

of the weak autocorrelation for auklets, we believe

our final habitat model is robust; (3) sooty shear-

waters were most numerous away from the Farallones

and over steep and variable bathymetry, suggestive of

slope waters off the shelf; (4) phalaropes were found

in higher densities closer to land, near the 200-m

isobath and Monterey Canyon, and away from the

Farallones (Fig. 4a). Both the sooty shearwater and

phalaropes appeared to concentrate on the northeast

side of MB, near the Davenport upwelling center.

Cetaceans revealed fewer significant bathymetric

associations (Table 3, Fig. 4b): (5) Dall’s porpoises

were more numerous farther from land and over re-

gions of steep and heterogeneous bathymetry (shelf-

slope); (6) pacific white-sided dolphins were found in

highest densities away from the coastline and close to

the 200-m isobath; (7) Risso’s dolphins were more

abundant away from the Farallons and in the vicinity of

the 1000-m isobath; and (8) humpback whales were

most numerous near the 1000-m isobath (Fig. 4b).

3.3. Persistence of habitat associations over multiple

time scales

Despite fewer significant bathymetric associations,

cetaceans demonstrated greater persistence in their

habitat associations than did the seabirds (Table 4).

The cetacean models revealed no significant temporal

interactions across years. In contrast, most models of

seabird density showed significant interactions with

year. Notably, these interactions were significant for

the resident breeding species (common murre and

Cassin’s auklet), but we detected no variability in the

habitat associations for the non-breeding visitors

(sooty shearwater and phalaropes) (Table 4). The

common murre interaction with year and median depth

reflected a shift in distribution towards deeper waters

through time. The Cassin’s auklet interaction between

Cordell Bank and year suggested an increasing aggre-

gation at this seamount through time (1996–2002). We

interpret all other significant interactions as interannu-

al variability between years in bathymetric associa-

tions, with no clear longer-term trends (Table 5).

4. Discussion

4.1. Seabird and cetacean habitat associations

This study has demonstrated that marine top pred-

ators are associated with a variety of bathymetric

features and shallow water topographies. However,

habitat associations varied by species and taxonomic

group, with resident seabirds showing greater tem-

poral variability in habitat associations. For instance,

common murres were consistently found at higher

densities over shallower water close to land, although

multi-year analyses also showed a tendency for this

species to favor deeper waters in the latter years,

2001–2002 (Table 5, interaction between year and

median depth). Previously, Oedekoven et al. (2001)

showed a reverse pattern during a period of continued

ocean warming, with this species moving coastward

over the period 1985–1994. Briggs et al. (1987) and

Oedekoven et al. (2001) documented higher murre

densities over the shelf and near their Farallon breed-

ing colony. These results complement our findings.

We found that the relationship between murre density

and distance to the mainland was linear, whereas the

relationships with median depth and distance to the

Farallones were logarithmic. These latter patterns,

which indicate that murre densities increased as dis-

tance from the colony and median depth decreased,

are consistent with the notion that breeding murres

Fig. 3. Morans’ I index values (meanF S.D.), as a function of the spatial separation (lag) between survey bins for the four most abundant marine birds and mammals. Large positive

indices imply that similar values occur closer together in space, while smaller indices indicate a lack of spatial structure in the data. Negative indices indicate that values repel other

like values more than expected randomly. The fractions denote the number of significant (a= 0.01) Moran’s I indices out of the possible number of surveys, determined by

randomization tests. The shaded histograms identify those species and lags, where more than half of the surveys yielded significant spatial autocorrelation.

P.P.W.Yen

etal./JournalofMarin

eSystem

s50(2004)79–99

88

Fig.3(continued).


Fig. 4. Distributions of marine predators in the Gulf of the Farallones in 2001, showing the 200-, 1000-, and 3000-m isobaths. (a) Seabird

densities and (b) mammal encounter rates are expressed as the number of individuals sighted per square kilometer and per kilometer, respectively.


Fig. 4 (continued).


Table 4

Summary of likelihood ratio statistics and p-values from the comparison of the full model against the formulation including year or sweep

interaction

p-value v2,df

Common

murre

Cassin’s

auklet

Sooty

shearwater

Phalarope

spp.

Dall’s

porpoise

Risso’s

dolphin

White-sided

dolphin

Humpback

whale

Between years

(N= 1888)

< 0.001,

61.12, 9

< 0.001,

40.06, 15

0.01,

16.46, 6

< 0.001,

36.13, 8

0.65,

4.22, 6

0.13,

4.12, 2

0.07,

11.81, 6

0.51,

1.36, 2

Between sweeps

(2001; N= 747)

0.02,

9.82, 3

< 0.001,

23.10, 5

0.01,

9.05, 2

0.02,

11.73, 4

0.49,

1.43, 2

0.40,

1.87, 2

0.71,

1.37, 3

0.14,

2.19, 1

Between sweeps

(2002; N= 607)

< 0.001,

22.74, 3

< 0.001,

31.78, 5

0.18,

3.47, 2

0.80,

1.66, 4

0.64,

0.90, 2

0.42,

1.72, 2

0.92,

0.49, 3

0.69,

0.16, 1

For cetaceans, no significant interactions were found. However, seabirds demonstrate temporal variability between years, as well as between

weekly sweeps.


forage in shallow water near their colonies. We dis-

cuss the importance of central place foraging in greater

depth below.

Cassin’s auklets persistently favored the shelf-break

(200-m isobath) region, but they displayed interannual

variations in their association with Cordell Bank, the

Farallon Archipelago, and other bathymetrically com-

Table 5

Results of the analyses stratified by year

For those variables that showed significant interannual interactions, habitat

habitat model are shown. (L= linear, LN= logarithmic, SQ= square, SQR

*Significant at , significant at , significant at , non

plex habitats (i.e., high bathymetric coefficient of

variation within habitat segments). Oedekoven et al.

(2001) demonstrated similar relationships with dis-

tance to the shelf-break, although in that study associ-

ations with the shelf-break were quite variable from

year to year. Allen (1994) considered Cordell Bank to

be an important feature for this species in the GOF/MB

models were analyzed separately by year. Coefficients from the final

= square root).

significance is clear.


region. This habitat association was also underscored

by our study, with auklets strongly associated with

Cordell Bank in 3 of 4 years. In our study, the

relationship between auklet density and the 200-m

isobath was linear, indicating a constant rate of decline

away from this feature. Relationships with the Faral-

lones, Cordell Bank and highly variable bathymetry

were all logarithmic, indicating aggregation in close

proximity to these features, with a rapid decline with

increasing distance from each feature. The relationship

with highly variable topography is particularly inter-

esting, and indicates that auklets aggregate over loca-

tions with high structural complexity. However, this

relationship was only significant in one of four years

(2001, Table 5) making any assertions regarding the

use of topographically complex habitats tenuous.

The migrant seabirds, sooty shearwater and the

phalaropes, were not associated with the Farallon

Archipelago, where 11 seabird species (including

murres and auklets) were breeding during the study

period (May–June). In contrast to the auklets, shear-

waters were most numerous over steep bathymetry

(a high bathymetric Contour Index) in two of the

four study years, indicating a preference for escarp-

ments such as the shelf-break/slope region and

submarine canyons. Oedekoven et al. (2001) dem-

onstrated that shearwaters in the region were dis-

tributed widely, yet significantly associated with the

shelf-break. Phalaropes were also attracted to bathy-

metric features such as the shelf-break (200-m

isobath) and the Monterey Canyon. Briggs et al.

(1984) showed that most phalaropes were found

over slope waters offshore from the shelf-break,

although high densities were also occasionally en-

countered on the shelf. Notably, shearwaters and

phalaropes were found in high densities on the

northwest side of Monterey Canyon, where the

200-m isobath comes relatively close to shore. In

fact, Monterey Canyon appeared to be the most

important bathymetric feature for the phalaropes. In

summary, three of the four seabird species consid-

ered in this study occurred in higher densities in the

vicinity of complex, steep, or shallow-water top-

ographies. The murres were the only species strong-

ly associated with the shelf and shallow water in

general, but their distribution shifted offshore to-

wards deeper waters during the course of this study.

Our findings for the seabirds corroborate previous

results (Briggs et al., 1987; Allen, 1994; Oedekoven

et al., 2001).

In contrast, cetaceans displayed relatively persis-

tent bathymetric associations through time. All spe-

cies were found offshore, over deep water, with the

Dall’s porpoise and white-sided dolphin at greater

abundances further from the mainland, and Risso’s

dolphins and humpback whales being associated with

the 1000-m isobath. Allen (1994) also documented

offshore, shelf-break/slope habitat associations for

these species in the GOF/MB region. Additionally,

Risso’s dolphin distributions have been previously

associated with the continental slope in the Atlantic

Ocean (Baumgartner, 1997; Baumgartner et al.,

2001). The Dall’s porpoise, a year-round resident of

the GOF/MB region (Allen, 1994; Forney and Bar-

low, 1998), was found in association with steep top-

ographies (high Contour Index, Table 5), suggesting

that this species occurs over bathymetric escarpments.

These cetacean species are apex predators, and their

diets largely consist of fish and squid (Kenney et al.,

1997; Hunt et al., 2000).

4.2. Depth and contour relationships

Bathymetric features such as continental shelf-

breaks and slopes, and shallow banks and seamounts

are often sites of enhanced marine productivity (Ichii

et al., 1998; Lavoie et al., 2000) as well as predator–

prey aggregation (Macaulay et al., 1984; Barange,

1994; Munk et al., 1995; Springer et al., 1996;

Oedekoven et al., 2001). In particular, shelf-break

fronts can be characterized by elevated macrozoo-

plankon abundance, especially euphausiids (Simard

and Mackas, 1989; Hunt et al., 1996), a key ecosys-

tem constituent and important prey of many ceta-

ceans and seabirds off central California (Schoenherr,

1991; Sydeman et al., 2001; Benson et al., 2002;

Marinovic et al., 2002). The Cassin’s auklet is a

planktivorous seabird which feeds almost exclusively

on the euphausiids Euphausia pacifica and Thysa-

noessa spinifera (Ainley et al., 1996; Sydeman et al.,

2001; Abraham and Sydeman, in press). The close

association of auklets with the shelf-break is proba-

bly related to foraging on these species. Abraham

and Sydeman (in press) documented an increased use

of the euphausiid T. spinifera from 1996 to 2001,

indicating, an increase in the availability of this prey


along the shelf-break. Interestingly, as noted above,

common murres also shifted their distribution to-

wards the outer continental shelf/shelf-break region

over the 4 years of our study, suggesting that they

may have relying on euphausiids more strongly in

the latter years. This also corresponds to a hypothe-

sized 1998–1999 shift to a cold-water regime of

enhanced coastal upwelling (Bograd et al., 2000;

Durazo et al., 2001; Schwing et al., 2003). These

results indicate that the spatial overlap in habitat use

between these two breeding seabirds may have

increased from 1996 to 2002.

4.3. Shallow-water topographies

We considered predator dispersion in relation to

three major shallow-water topographic features in the

GOF/MB region: a seamount, an island chain, and a

deep-water canyon system. These three features are

likely areas of prey concentration, which may pro-

mote the aggregation of marine predators in the

vicinity of these habitats (Schoenherr, 1991; Croll

et al., 1998; Oedekoven et al., 2001; Benson et al.,

2002). Upwelling may support elevated levels of

productivity and prey aggregation in proximity to

this seamount (Boehlert and Genin, 1987; Roden,

1987; Genin et al., 1988). Upwelling in the vicinity

of seamounts, banks, and canyons can stimulate local

productivity by bringing cool, nutrient-rich water

into the photic zone, although there is little evidence

that eutrophication can act as a prey concentrating

mechanism without retention of the localized pro-

duction at hydrographic features (e.g., fronts, eddies)

(reviewed by Genin, this volume). Enhanced water

column mixing may also push weakly swimming

zooplankton and larval/juvenile fish close to the

surface, making them available to surface-foraging

and shallow-diving predators such as auklets and

shearwaters (Simard et al., 1986; Genin et al.,

1988; Haney et al., 1995; Hunt et al., 1996, 1998;

Lavoie et al., 2000). Another mechanism entails

entrapment of vertically migrating prey advected

over shallow-water topographies by sub-surface

frontal systems. In particular, topographic features

shallower than the nocturnal depth occupied by

vertically migrating species are sites especially suited

for prey entrapment (Genin, 2004). For example,

Brooks and Mullin (1983) estimated that about half

of the zooplankton biomass in the Southern Califor-

nia Bight migrates vertically through the 56-m iso-

bath. In particular, most euphausiid species off

California spend the day at depth (>200 m) and

migrate to the surface at night (Brinton, 1967).

While we did not determine the mechanisms

responsible for prey aggregation over these shal-

low-water topographies, vertical entrapment and/or

forcing of prey near the surface likely plays a role

in predator aggregation over Cordell Bank and at

the northwest rim of the Monterey Canyon. Auklets,

shearwaters, phalaropes and humpback whales

(Briggs et al., 1984; Chu, 1984; Benson et al.,

2002; Abraham and Sydeman, in press) forage on

zooplankton and larval fishes in the GOF/MB re-

gion. Cordell Bank and the rim of Monterey Canyon

are shallower than the diurnal depth range of many

zooplankton species, especially euphausiids (Brinton,

1967), and could vertically trap these prey species in

shallow regions within the diving depth of many

predators (Schoenherr, 1991; Benson et al., 2002;

Marinovic et al., 2002). Furthermore, upwelling from

Point Reyes and Davenport could force prey towards

the surface at Cordell Bank and Monterey Canyon

(Allen et al., 2001). In addition, hydrographic fronts

along the shelf-break and the periphery of coastal

upwelling plumes could play a role in concentrating

prey in the vicinity of bathymetric features (Larson

et al., 1994; Hunt et al., 1996; Oedekoven et al.,

2001).

4.4. The Farallon Archipelago

In addition to vertical entrapment, other physical

processes may aggregate and retain prey in the vicin-

ity of the Farallon Archipelago. The complex bathym-

etry of this island chain, with five peaks rising from

the shelf (50-m depth) to 10–120 m above the sea

surface, coupled with strong spring-time tides and

upwelling filaments from the nearly Point Reyes

center, likely result in a complex pattern of eddies,

vortices, and convergence zones (Schwing et al.,

1991; Strub et al., 1991; Hickey, 1997). Secondary

circulation features may concentrate prey around the

archipelago in predictable locations, which may be in

turn exploited by central-place foraging seabirds from

the islands (Pereyra et al., 1969; Larson et al., 1994;

Hunt et al., 1996, 1998).

Marine Systems 50 (2004) 79–99 95

4.5. Environmental variability and central place

foraging

One of the most notable results of this study was

the shift in murre distributions from nearshore to

offshore areas (Table 5, interaction between median

depth and year). During the study, murres (and auk-

lets) were ‘‘central place foragers’’. That is, they were

constrained to return to their colonies on Southeast

and North Farallon Islands to provision young or to

exchange incubation duties with a mate during the

study period. As such, foraging costs for these species

should increase considerably with increasing distance

from the colony, since they must return there after

each foraging bout (Schoener, 1979). When feeding

offspring, foraging trip duration for common murres is

between 5 and 10 min when they are feeding on

epipelagic juvenile rockfish (Sebastes spp.), and over

4 h when they are foraging on northern anchovy

(Engraulis mordax) and other neritic species (Ainley

and Boekelheide, 1990). This disparity is reflective of

the travel costs associated with the use of different

prey resources. In contrast, Cassin’s auklets store

euphausiid and larval fish prey in a sublingual gular

pouch and return to the island nightly to feed their

offspring (Abraham and Sydeman, in press). Auklets

generally return around 22:00 h and remain there until

about 04:00 h (PRBO unpublished data), apparently

to avoid Western Gulls (Larus occidentalis) (Nelson,

1989) which act as predators and kleptoparasites on

the auklets.

We hypothesize that murres, in an attempt to

minimize foraging costs, remain in the proximity to

the island breeding colonies whenever possible. In

1996 and 1997, murres provisioned offspring primar-

ily on northern anchovies (Sydeman et al., 2001),

whereas in 2001 and 2002, pelagic juvenile rockfish

had become more prevalent in the diet (Mills et al.,

submitted for publication). We believe that the appar-

ent movement of murres offshore between 1996–

1997 and 2001–2002 occurred in response to a

change in their prey selection. A number of oceano-

graphic conditions changed in the central California

Current System (CCS) with the onset of a period of

cold-water conditions after mid-1998. First, over the

entire 7-year period, there was an increase in upwell-

ing and a decrease in sea surface temperature at the

Farallones (Abraham and Sydeman, in press). These

P.P.W. Yen et al. / Journal of

changes were concurrent with a possible regime shift

to the ‘‘cold phase’’ of the Pacific Decadal Oscillation

(PDO) during 1998–1999 and generally enhanced

upwelling in the CCS (Bograd et al., 2000; Schwing

et al., 2000; Mantua and Hare, 2002). Phase shifts of

the PDO have been correlated with changes in marine

productivity across a wide variety of trophic levels

(Mantua et al., 1997; Hare and Mantua, 2000; Chavez

et al., 2003), including seabirds (Sydeman et al., 2001;

Schwing et al., 2003). Seabird demographic data at

the Farallon Islands and changes in the murre diet

indicate that rockfish productivity may also have been

affected by a potential PDO shift in the late 1990s.

4.6. Conservation implications

In this study we have shown how bathymetric

characteristics may be useful in understanding the

distribution and abundance of marine predators at

sea. In theory, bathymetric features, such as shallow-

water topographies, may provide a means of predict-

ing important foraging habitats for upper trophic-level

marine predators. This study and many others in the

GOF/MB (Schoenherr, 1991; Allen, 1994; Oedekoven

et al., 2001; Benson et al., 2002; Marinovic et al.,

2002) have demonstrated that seamounts, submarine

canyons, and complex and steep topographies are

important to seabirds and cetaceans, possibly as

centers of trophic transfer. While we have yet to

investigate the mechanisms of predator aggregation,

these habitats are likely associated with elevated

ocean productivity and prey retention, thereby making

dense prey patches available to predators. Further-

more, there is some evidence that seabirds and ceta-

ceans persistently aggregate at these habitat features,

though species redistributions over time (e.g., murres)

and inconsistent relationships (e.g., sooty shearwaters)

have also been documented. If persistence can be

clearly demonstrated, shallow-water topographies

may deserve special conservation consideration, as

important foraging ‘‘hotspots’’ in pelagic systems

(Hooker et al., 1999; Hyrenbach et al., 2000). In this

study, we found evidence of persistent seabird and

cetacean associations with the northwestern rim of

Monterey Canyon (all seabirds and Risso’s dolphins),

over Cordell Bank (auklets, shearwaters and Dall’s

porpoise), and in the proximity to the Farallon Archi-

pelago (murres and auklets). Notably, the Monterey


Canyon is located downstream from the Davenport

upwelling center, which may explain, in part, why

seabirds and cetaceans regularly used this region.

While these static bathymetric characteristics do

not change temporally, the distribution of marine

predators and their prey does vary seasonally and

interannually, due to the influence of hydrographic

processes linked or unlinked to bathymetry. For in-

stance, the observed changes in murre dispersion

throughout the study were likely related to changes

in prey resources and overall oceanographic condi-

tions. Therefore, assessing species distributions in

relation to both bathymetric and hydrographic habitats

is essential to obtain a more complete understanding

of the dispersion of upper-trophic marine predators

and the nature and location of habitat ‘‘hotspots’’.

Nonetheless, the wildlife –habitat associations

documented in this study have important implications

for conserving and managing the GOF/MB ecosys-

tem. From a conservation perspective, static bathy-

metric habitats defined by the extent and location of

specific isobaths (e.g., Cordell Bank, Monterey Can-

yon) may be particularly conducive to the design of

wildlife reserves similar to those established in terres-

trial systems. While effective Marine Protected Area

designs may ultimately require dynamic boundaries

(Hyrenbach et al., 2000), static bathymetric features

provide an initial basis for identifying potentially

important habitats for protection in coastal (e.g.,

shelf-slope regions, shallow banks) and pelagic (e.g.,

seamounts) systems.

Acknowledgements

We thank shipboard observers, Mary Beth Decker,

Meredith Elliott, Cornelia Oedekoven and Sophie

Webb, whose tireless efforts made this analysis possi-

ble. Nadav Nur provided valuable statistical advice,

and Brandon McGuigan and Vincent Yen provided

helpful technical support. We thank Matt Merrifield

and The Nature Conservancy for sharing physical data.

Funding was provided by the National Fish and

Wildlife Foundation (NFWF), the National Oceanic

and Atmospheric Administration-National Marine

Protected Areas Center (NOAA-NMPAC), the Pack-

ard Foundation, and the Moore Family Foundation.

We are indebted to the staff and technicians of NOAA-

National Marine Fisheries Service (NMFS) and the

crew of the RV David Starr Jordan for allowing us to

participate on Rockfish Recruitment Studies cruises.

Steve Ralston, Alec MacCall, Ken Baltz, Keith

Sakuma and Tom Laidig of NOAA-NMFS deserve

special recognition for their long-term collaborative

efforts with PRBO. We are grateful to George H. Hunt

Jr. and Elizabeth Sinclair for their constructive remarks

on a previous draft of this manuscript. This is PRBO

contribution no. 1124.

References

Abraham, C.L., Sydeman, W.J., in press. Ocean climate, euphau-

siids, and auklet nesting: Effects of interannual trends and var-

iation in phenology, diet, and growth of a planktivorous seabird.

Mar. Ecol. Prog. Ser.

Ainley, D.G., Boekelheide, R.J., 1990. Seabirds of the Farallon

Islands: Ecology, Dynamics, and Structure of an Upwelling-Sys-

tem Community. Stanford Univ. Press, Stanford, CA.

Ainley, D.G., Spear, L.B., Allen, S.G., 1996. Variation in the diet of

Cassin’s auklet reveals spatial, seasonal, and decadal occurrence

patterns of euphausiids off California, USA. Mar. Ecol., Prog.

Ser. 137, 1–10.

Allen, S.G., 1994. The distribution and abundance of marine birds

and mammals in the Gulf of the Farallones and adjacent waters,

1985–1992. PhD Thesis, Univ. of California, Berkeley, CA,

unpublished.

Allen, S.E., Vindeirinho, C., Thomson, R.E., Foreman, M.G.G.,

Mackas, D.L., 2001. Physical and biological processes over a

submarine canyon during an upwelling event. Can. J. Fish.

Aquat. Sci. 58, 671–684.

Bakun, A., 1975. Coastal upwelling indices, West Coast of

North America, 1967–1973. NOAA Technical Report NMFS

SSRF-693.

Barange, M., 1994. Acoustic identification, classification and

structure of biological patchiness on the edge of the Agulhas

Bank and its relation to frontal features. S. Afr. J. Mar. Sci. 14,

333–347.

Baumgartner, M.F., 1997. The distribution of Risso’s dolphin

(Grampus griseus) with respect to the physiography of the

northern Gulf of Mexico. Mar. Mamm. Sci. 13, 614–638.

Baumgartner, M.F., Mullin, K.D., May, L.N., Leming, T.D., 2001.

Cetacean habitats in the northern Gulf of Mexico. U.S. Fish.

Bull. 99, 219–239.

Benson, S.R., Croll, D.A., Marinovic, B.B., Chavez, F.P., Har-

vey, J.T., 2002. Changes in the cetacean assemblage of a

coastal upwelling ecosystem during El Nino 1997–98 and

La Nina 1999. Prog. Oceanogr. 54, 279–291.

Boehlert, G.W., Genin, A., 1987. A review of the effects of sea-

mounts on biological processes. In: Keating, B.H., Fryer, P.,

Batiza, R., Boehlert, G.W. (Eds.), Seamounts, Islands and Atolls.

Geophys. Monogr., vol. 43. American Geophysical Union,

Washington, DC, pp. 319–334.


Bograd, S.J., DiGiacomo, P.M., Durazo, R., Hayward, T.L., Hyren-

bach, K.D., Lynn, R.J., Mantyla, A.W., Schwing, F.B., Syde-

man, W.J., Baumgartner, T., Lavaniegos, B., Moore, C.S.,

2000. The state of the California current, 1999–2000: forward

to a new regime? Calif. Coop. Ocean. Fish. Investig. Rep. 41,

26–52.

Briggs, K.T., Dettman, K.F., Lewis, D.B., Tyler, W.B., 1984. Phal-

arope feeding in relation to autumn upwelling off California. In:

Nettleship, D.N., Sanger, G.A., Springer, P.F. (Eds.), Marine

Birds, Their Feeding Ecology and Commercial Fisheries Rela-

tionships. Canadian Wildlife Service Special Publication. Cana-

dian Wildlife Service, Ottawa, pp. 51–63.

Briggs, K.T., Tyler, W.B., Lewis, D.B., Carlson, D.R., 1987. Bird

communities at sea off California 1975 to 1983. Stud. Avian

Biol. 11, 1–74.

Brinton, E., 1967. Vertical migration and avoidance capability of

euphausiids in the California Current. Limnol. Oceanogr. 12,

451–483.

Brooks, E.R., Mullin, M.M., 1983. Diel changes in the vertical

distribution of biomass and species in the Southern California

bight. Calif. Coop. Ocean. Fish. Investig. Rep. 24, 210–215.

Buckland, S.T., Anderson, D.R., Burnham, K.P., Laake, J.L., 1993.

Distance Sampling: Estimating Abundance of Biological Popu-

lations. Chapman and Hall, London.

Chavez, F.P., 1996. Forcing and biological impact of onset of the

1992 El Nino in central California. Geophys. Res. Lett. 23,

265–268.

Chavez, F.P., Ryan, J., Lluch-Cota, S.E., Niquen, C.M., 2003. From

anchovies to sardines and back: multidecadal change in the

Pacific Ocean. Science 299, 217–221.

Chu, E.W., 1984. Sooty shearwaters off California: diet and energy

gain. In: Nettleship, D.N., Sanger, G.A., Springer, P.F. (Eds.),

Marine Birds, Their Feeding Ecology and Commercial Fisheries

Relationships. Canadian Wildlife Service Special Publication.

Canadian Wildlife Service, Ottawa, pp. 64–71.

Croll, D.A., Tershy, B.R., Hewitt, R.P., Demer, D., Fiedler, P.C.,

Smith, S.E., Armstrong, W., Popp, J.M., Kiekhefer, T., Lopez,

V.R., Urban, J., 1998. An integrated approach to the foraging

ecology of marine birds and mammals. Deep-Sea Res. II 45,

1353–1371.

Durazo, R., Baumgartner, T.R., Bograd, S.J., Collins, C.A., de la

Campa, S., Garcia, J., Gaxiola-Castro, G., Huyer, A., Hyren-

bach, K.D., Loya, D., Lynn, R.J., Schwing, F.B., Smith, R.L.,

Sydeman, W.J., Wheeler, P., 2001. The state of the California

current, 2000–2001: a third straight La Nina year. Calif. Coop.

Ocean. Fish. Investig. Rep. 42, 29–60.

ESRI, 1996. Using ArcView GIS. ESRI Press, Redlands, CA.

Fargion, G.S., McGowan, J.A., Steward, R.H., 1993. Seasonality of

chlorophyll concentrations in the California current: a compar-

ison of two methods. Calif. Coop. Ocean. Fish. Investig. Rep.

34, 35–50.

Fauchald, P., Erikstad, K.E., 2002. Scale-dependent predator–prey

interactions: the aggregative response of seabirds to prey under

variable prey abundance and patchiness. Mar. Ecol., Prog. Ser.

231, 279–291.

Forney, K.A., Barlow, J., 1998. Seasonal patterns in the abundance

and distribution of California cetaceans, 1991–92: application

of a confidence-interval based bootstrap test. Mar. Mamm. Sci.

14, 460–489.

Fotheringham, A.S., Brunsdon, C., Charlton, M., 2000. Quantita-

tive Geography: Perspectives on Spatial Data Analysis. Sage,

London. 272 pp.

Franks, P.J.S., 1992. Sink or swim: accumulation of biomass at

fronts. Mar. Ecol., Prog. Ser. 82, 1–12.

Genin, A., Haury, L., Greenblatt, P., 1988. Interactions of migrating

zooplankton with shallow topography predation by rockfishes

and intensification of patchiness. Deep-Sea Res. I 35, 151–176.

Genin, A., 2004. Trophic focusing: the role of bio-physical cou-

pling in the formation of animal aggregations over abrupt top-

ographies. J. Mar. Syst. 50, 3-20.

GLOBEC, 1992. Towards the Development of the GLOBEC Core

Program. GLOBEC Int. Rep., vol. 1.

Haney, J.C., McGillivary, P.A., 1985. Midshelf fronts in the South

Atlantic Bight and their influence on seabird distribution and

seasonal abundance. Biol. Oceanogr. 3, 401–430.

Haney, J.C., Haury, L.R., Mullineaux, L.S., Fey, C.L., 1995. Sea-

bird aggregation at a deep north Pacific seamount. Mar. Biol.

123, 1–9.

Hare, S.R., Mantua, N.J., 2000. Empirical evidence for North

Pacific regime shifts in 1977 and 1989. Prog. Oceanogr. 47,

103–145.

Hayward, T.L., Venrick, E.L., 1998. Near surface pattern in the

California current: coupling between physical and biological

structure. Deep-Sea Res. II 45, 1617–1638.

Heinemann, D., 1981. A rangefinder for pelagic bird censusing. J.

Wildl. Manage. 45, 489–493.

Hickey, B.M., 1997. The response of a steep-sided, narrow canyon

to time-variable wind forcing. J. Phys. Oceanogr. 27, 697–726.

Hooker, S.K., Whitehead, H., Gowans, S., 1999. Marine protected

area design and the spatial and temporal distribution of ceta-

ceans in a submarine canyon. Conserv. Biol. 13, 592–602.

Hosmer Jr., D.W., Lemeshow, S., 2000. Applied Logistic Regres-

sion, 2nd ed. Wiley, New York, NY.

Hui, C.A., 1985. Undersea topography and the comparative

distributions of two pelagic cetaceans. U.S. Fish. Bull. 83,

472–475.

Hunt Jr., G.L., Schneider, D.C., 1987. Scale-dependent processes

in the physical and biological environment of marine birds.

In: Croxall, J.P. (Ed.), Seabirds: Feeding Ecology and Role

in Marine Ecosystems. Cambridge Univ. Press, Cambridge,

pp. 7–41.

Hunt Jr., G.L., Harrison, N.M., Cooney, R.T., 1990. The influence

of hydrographic structure and prey abundance on foraging of

least auklets. Stud. Avian Biol. 14, 7–22.

Hunt Jr., G.L., Coyle, K.O., Hoffman, S., Decker, M.B., Flint, E.N.,

1996. Foraging ecology of the short-tailed shearwaters near the

Pribilof Islands, Bering Sea. Mar. Ecol., Prog. Ser. 141, 1–11.

Hunt Jr., G.L., Russel, R.W., Coyle, K.O., Weingartner, T., 1998.

Comparative foraging ecology of planktivorous auklets in rela-

tion to ocean physics and prey availability. Mar. Ecol., Prog. Ser.

167, 241–259.

Hunt Jr., G.L., Kato, H., McKinnell, S.M. (Eds.), 2000. Predation

by Marine Birds and Mammals in the Subarctic North Pacific

Ocean. PICES Sci. Rep., vol. 14.


Hurlbert, S.H., 1984. Pseudoreplication and the design of ecologi-

cal field experiments. Ecology 54, 187–211.

Huyer, A., 1983. Coastal upwelling in the California current sys-

tem. Prog. Oceanogr. 12, 259–284.

Hyrenbach, K.D., Veit, R.R., 2003. Ocean warming and seabird

assemblages of the California Current System (1987–1998):

response at multiple temporal scales. Deep-Sea Res. II 50

(14–16), 2537–2565.

Hyrenbach, K.D., Forney, K.A., Dayton, P.K., 2000. Marine pro-

tected areas and ocean basin management. Aquat. Conserv. 10,

437–458.

Ichii, T., Katayama, K., Obitsu, N., Ishii, H., Naganobu, M., 1998.

Occurrence of Antarctic krill (Euphausia superba) concentra-

tions in the vicinity of the South Shetland Islands: relationship

to environmental parameters. Deep-Sea Res. I 45, 1235–1262.

Jaquet, N., 1996. How spatial and temporal scales influence under-

standing of sperm whale distribution—a review. Mamm. Rev.

26, 51–65.

Jaquet, N., Whitehead, H., 1996. Scale-dependent correlation of

sperm whale distribution with environmental features and pro-

ductivity in the south Pacific. Mar. Ecol., Prog. Ser. 135, 1–9.

Jesperson, D., 1924. On the frequency of birds over the high At-

lantic Ocean. Nature 114, 281–283.

Keiper, C.A., 2001. Marine mammals off central California relative

to hydrography; 1986–94, 1997–99. MS thesis, San Jose State

Univ., San Jose, USA, unpublished.

Kenney, R.D., Scott, G.P., Thompson, T.J., Winn, H.E., 1997. Esti-

mates of prey consumption and trophic impacts of cetaceans in

the USA Northeast continental shelf ecosystem. J. Nw. Atl. Fish.

Sci. 22, 155–171.

Kudela, R.M., Chavez, F.P., 2000. The impact of the 1992 El Nino

on new production in Monterey Bay, California. Deep-Sea Res.

II 47, 1055–1076.

Larson, R.J., Lenarz, W.H., Ralston, S., 1994. The distribution of

pelagic juvenile rockfish of the genus Sebastes in the upwelling

region off central California. Calif. Coop. Ocean. Fish. Investig.

Rep. 35, 175–221.

Lavoie, D., Simard, Y., Saucier, F.J., 2000. Aggregation and dis-

persion of krill at channel heads and shelf edges: the dynamics

in the Saguenay-St. Lawrence Marine Park. Can. J. Fish. Aquat.

Sci. 57, 1853–1869.

Macaulay, M.C., English, T.S., Mathisen, O.A., 1984. Acoustic

characterization of swarms of Antarctic krill (Euphausia

superba) from Elephant Island and Bramfield Strait. J. Crustac.

Biol. 4, 16–44.

Manly, B.F.J., 1991. Randomization and Monte Carlo Methods in

Biology. Chapman & Hall, London.

Mantua, N.J., Hare, S.R., 2002. The Pacific decadal oscillation. J.

Oceanogr. 58, 35–44.

Mantua, N.J., Hare, S.R., Zhang, Y., Wallace, J.M., Francis, R.C.,

1997. A Pacific interdecadal climate oscillation with impacts on

salmon production. Bull. Am. Meteorol. Soc. 78, 1069–1079.

Marinovic, B., Croll, D.A., Gong, N., Benson, S., Chavez, F.,

2002. Effects of the 1997–98 El Nino on zooplankton

within the Monterey Bay coastal upwelling system with

emphasis on the euphausiid community. Prog. Oceanogr.

54, 205–232.

Moran, P.A.P., 1950. Notes on continuous stochastic phenomena.

Biometrika 37, 17.

Munk, P., Larsson, P.O., Danielsen, D., Moksness, E., 1995. Larval

and small juvenile cod Gadus morhua concentrated in the high-

ly productive areas of a shelf break front. Mar. Ecol., Prog. Ser.

125, 21–30.

Murphy, R.C., 1936. Oceanic Birds of South America, vols. 1 and 2.

American Museum of Natural History, New York.

Nelson, D.A., 1989. Gull predation on Cassin’s auklet varies with

the lunar cycle. Auk 106, 495–497.

Oedekoven, C.S., Ainley, D.G., Spear, L.B., 2001. Variable

responses of seabirds to change in marine climate: California

current, 1985–1994. Mar. Ecol., Prog. Ser. 212, 265–281.

Pelaez, J., McGowan, J.A., 1986. Phytoplankton pigment patterns

in the California current as determined by satellite. Limnol.

Oceanogr. 31, 927–950.

Pennington, J.T., Chavez, F.P., 2000. Seasonal fluctuations of tem-

perature, salinity, nitrate, chlorophyll and primary production at

station H3/M1 over 1989–1996 in Monterey Bay, California.

Deep-Sea Res. II 47, 947–973.

Pereyra, W.T., Pearcy, W.G., Carvey Jr., F.E., 1969. Sebastodes

flavidus, a shelf rockfish feeding on mesopelagic fauna, with

consideration on the ecological implications. J. Fish. Res. Board

Can. 26, 2211–2215.

Ribic, C.A., Ainley, D.G., 1997. The relationships of seabird

assemblages to physical habitat features in Pacific equatorial

waters during spring 1984–1991. ICES J. Mar. Sci. 54,

593–599.

Roden, G.I., 1987. Effect of seamounts and seamount chains on

ocean circulation and hermohaline structure. In: Keating, B.H.,

Fryer, P., Batiza, R., Boehlert, G.W. (Eds.), Seamounts, Islands

and Atolls. Geophys. Monogr., vol. 43. American Geophysical

Union, Washington, DC, pp. 335–354.

Sawada, M., 1999. Rookcase: an excel 97/2000 visual basic (VB)

add-in for exploring global and local spatial autocorrelation.

Bull. Ecol. Soc. Am. 80, 231–234.

Schneider, D.C., 1990. Spatial autocorrelation in marine birds. Po-

lar Biol. 8, 89–97.

Schneider, D.C., 1991. The role of fluid dynamics in the ecology of

marine birds. Mar. Biol. Annu. Rev. 29, 487–521.

Schoener, T.W., 1979. Generality of the size–distance relation in

models of optimal feeding. Am. Nat. 114, 902–912.

Schoenherr, J.L., 1991. Blue whales feeding on high concentrations

of euphausiids around Monterey Submarine Canyon. Can. J.

Zool. 69, 583–594.

Schwing, F.B., Husby, D.M., Garfield, N., Tracy, D.E., 1991. Me-

soscale oceanic response to wind events off central California in

spring 1989: CTD survey and AVHRR imagery. Calif. Coop.

Ocean. Fish. Investig. Rep. 32, 47–61.

Schwing, F.B., Moore, C.S., Ralston, S., Sakuma, K.M., 2000.

Record coastal upwelling in the California Current during

1999. Calif. Coop. Ocean. Fish. Investig. Rep. 41, 148–160.

Schwing, F.B., Murphree, T., deWitt, L., Green, P.M., 2003. The

evolution of oceanic and atmospheric anomalies in the northeast

Pacific during the El Nino and La Nina events of 1995–2001.

Prog. Oceanogr. 54, 459–491.

Shea, R., Broenkow, W.M., 1982. The role of internal tides in the


nutrient enrichment of Monterey Bay California, USA. Estuar.

Coast. Shelf Res. 15, 57–66.

Simard, Y., Mackas, D.L., 1989. Mesoscale aggregations of euphau-

siid sound scattering layers on the continental shelf of Vancouver

Island, Canada. Can. J. Fish. Aquat. Sci. 46, 1238–1249.

Simard, Y., Ladurantaye, R., de Therriault, J.C., 1986. Aggregation

of euphausiids along a coastal shelf in an upwelling environ-

ment. Mar. Ecol., Prog. Ser. 32, 203–215.

Sokal, R.R., Oden, N.L., 1978. Spatial autocorrelation in biology 1.

Methodology. Biol. J. Linn. Soc. 10, 199–228.

Sokal, R.R., Rohlf, R.J., 1981. Biometry, 2nd ed. W. H. Freeman,

New York.

Spear, L.B., Balance, L.T., Ainley, D.G., 2001. Response of sea-

birds to thermal gradients in the tropical Pacific: the thermo-

cline versus the Equatorial Front. Mar. Ecol., Prog. Ser. 219,

275–289.

Springer, A.M., McRoy, C.P., Flint, M.V., 1996. The Bering sea

green belt: shelf-edge processes and ecosystem production. Fish.

Oceanogr. 5, 205–223.

Steger, J.M., Schwing, F.B., Collins, C.A., Rosenfeld, L.K.,

Garfield, N., Gezgin, E., 2000. The circulation and water

masses in the Gulf of the Farallones. Deep-Sea Res. II 47,

907–946.

Strub, T.P., Kosro, P.M., Huyer, A., 1991. The nature of cold fila-

ments in the California current system. J. Geophys. Res. 92,

14743–14768.

Sydeman, W.J., Allen, S.G., 1999. Pinniped population dynamics in

central California: correlations with sea surface temperature and

upwelling indices. Mar. Mamm. Sci. 15, 446–461.

Sydeman, W.J., Hester, M.M., Thayer, J.A., Gress, F., Martin, P.,

Buffa, J., 2001. Climate change, reproductive performance, and

diet composition of marine birds in the southern California Cur-

rent system, 1969–1997. Prog. Oceanogr. 49, 309–329.

Tasker, M.L., Hope-Jones, P., Dixon, T., Blake, B.F., 1984. Count-

ing seabirds at sea from ships: a review of methods and a sug-

gestion for a standardized approach. Auk 101, 567–577.

Venrick, E.L., 1998. Spring in the California Current: the distribu-

tion of phytoplankton species, April 1993 and April 1995. Mar.

Ecol., Prog. Ser. 167, 73–88.

Wahl, T., Morgan, K.H., Vermeer, K., 1993. Seabird distribution off

British Columbia and Washington. In: Vermeer, K., Briggs,

K.T., Morgan, K.H., Siegel-Causey, D. (Eds.), The Status, Ecol-

ogy, and Conservation of Marine Birds of the North Pacific.

Can. Wildl. Serv. Spec. Pub., Ottawa, pp. 39–47.

Wing, S.R., Botsford, L.W., Largier, J.L., Morgan, L.E., 1995.

Spatial structure of relaxation events and crab settlement in

the northern California upwelling system. Mar. Ecol., Prog.

Ser. 128, 199–211.

marine bird and cetacean associations with bathymetric habitats … et al. 2004.pdf · marine bird...

Documents