Environmental factors and fisheries influence the foraging...

DRO Deakin Research Online, Deakin University’s Research Repository Deakin University CRICOS Provider Code: 00113B

Environmental factors and fisheries influence the foraging patterns of a subtropical seabird, the Westland Petrel (Procellaria westlandica), in the Tasman Sea

Citation: Waugh, Susan M., Griffiths, James W., Poupart, Timothee, Filippi, Dominique P., Rogers, Karyne and Arnould, John Y. P. 2018, Environmental factors and fisheries influence the foraging patterns of a subtropical seabird, the Westland Petrel (Procellaria westlandica), in the Tasman Sea, The condor, vol. 120, no. 2, pp. 371-387.

© 2018, American Ornithological Society

Reproduced with permission.

Downloaded from DRO: http://hdl.handle.net/10536/DRO/DU:30108544

http://hdl.handle.net/10536/DRO/DU:30108544

Volume 120, 2018, pp. 371–387DOI: 10.1650/CONDOR-17-179.1

RESEARCH ARTICLE

Environmental factors and fisheries influence the foraging patterns of asubtropical seabird, the Westland Petrel (Procellaria westlandica), in theTasman Sea

Susan M. Waugh,1* James W. Griffiths,2 Timothee A. Poupart,1,3 Dominique P. Filippi,4 Karyne Rogers,5

and John Y. P. Arnould3

1 Museum of New Zealand Te Papa Tongarewa, Wellington, New Zealand2 Department of Conservation, Wellington, New Zealand3 School of Life and Environmental Sciences, Deakin University, Burwood, Victoria, Australia4 Sextant Technology, Wellington, New Zealand5 GNS Science Te Pu Ao, Lower Hutt, New Zealand* Corresponding author: [email protected]

Submitted September 10, 2017; Accepted February 3, 2018; Published April 25, 2018

ABSTRACTEnvironmental and anthropogenic influences in the marine environment are primary drivers of behavior anddemographic outcomes for marine birds. We examined factors influencing the foraging patterns of the Westland Petrel(Procellaria westlandica), a highly threatened, endemic petrel that inhabits subtropical water masses primarily in theTasman Sea, with a poorly known at-sea distribution. Risk assessments place the species at moderate risk of populationimpacts from fisheries-related mortality. Studies in the 1990s indicated that trawl fisheries would have an importantinfluence on the Westland Petrel’s foraging behavior. We investigated the influence of climatic conditions, marineproductivity, bathymetry, the core fishery zone, concurrent fishing activity, light conditions, sex, and breeding stage onWestland Petrel foraging patterns. We analyzed the stable isotopes of carbon and nitrogen from blood sampled duringthe incubation period and examined changes in isotopic niche width over a 6-yr period. We found that the WestlandPetrel’s foraging zone varied only slightly between years and that the location of intensively used areas was stronglyinfluenced by bathymetric slope and latitude, and negatively influenced by chlorophyll-a. The core fishery zone had asecondary influence, suggesting that these petrels co-occur with fisheries, but are not dependent on waste for food.Trophic niche width was significantly wider during strong El Nino conditions, indicating that food type, rather thanlocation, was most affected by climatic variation. Consistent use of one marine area across varying times andconditions increases the risk of adverse effects of climate or human-induced impacts on the species. However, marinespatial management tools become viable in these conditions. Further, with rapid increases in sea surface temperaturesand extreme values recorded in the region in recent periods, changes to fisheries zones and distributions of naturalprey of the species are likely to occur and may change the population’s sustainability.

Keywords: Westland Petrel, foraging, environment, fisheries, Tasman Sea, Procellaria westlandica

Des facteurs environnementaux et les peches influencent les patrons de quete alimentaire d’un oiseausubtropical, Procellaria westlandica, dans la mer de Tasman

RESUMELes influences environnementales et anthropiques en milieu marin sont les principaux moteurs des reponsesdemographiques et comportementales chez les oiseaux marins. Nous avons examine les facteurs influencant lespatrons de quete alimentaire chez Procellaria westlandica, un petrel endemique fortement menace qui habite lesmasses d’eau subtropicales principalement dans la mer de Tasman et dont la distribution en mer mal connue. Lesevaluations de risques le placent comme etant a risque modere d’impact des mortalites dues aux peches sur lespopulations. Des etudes datant des annees 1990 indiquent que la peche au chalut aurait une importante influence surle comportement de quete alimentaire de P. westlandica. Nous avons examine l’influence des conditions climatiques,de la productivite marine, de la bathymetrie, de la zone de peche principale, de l’activite de peche simultanee, desconditions lumineuses, du sexe et du stade reproducteur sur les patrons de quete alimentaire de P. westlandica. Nousavons analyse les isotopes stables du carbone et de l’azote a partir d’echantillons sanguins preleves durant la perioded’incubation et nous avons examine les changements dans la largeur des niches isotopiques sur une periode de sixans. Nous avons constate que la zone d’alimentation de P. westlandica variait peu entre les annees et que lalocalisation des zones intensement utilisees etait fortement influencee par la pente bathymetrique et la latitude, etnegativement influencee par la chlorophylle-a. La zone de peche principale avait un influence secondaire, suggerant

Q 2018 American Ornithological Society. ISSN 0010-5422, electronic ISSN 1938-5129Direct all requests to reproduce journal content to the AOS Publications Office at [email protected]

mailto:[email protected]

que les petrels cohabitent avec les activites de peche mais ne sont pas dependants des rejets de peche pours’alimenter. Les niches trophiques etaient significativement plus larges lorsque fortes dans des conditions El Nino, cequi indique que le type de nourriture, plutot que la localisation, etait le plus affecte par la variation du climat. Uneutilisation continue d’une zone marine a differents moments et dans diverses conditions augmente le risque d’impactsclimatiques ou anthropiques negatifs sur l’espece. Cependant, les outils de gestion spatiale du milieu marindeviennent viables dans ces conditions. De plus, avec de rapides augmentations de la SST et des valeurs extremesenregistrees dans la region recemment, des changements aux zones de peches et dans la repartition naturelle desproies de l’espece sont susceptibles de se produire et peuvent affecter la perennite de la population.

Mots-cles : quete alimentaire, environnement, peches, mer de Tasman, Procellaria westlandica

INTRODUCTION

Top predators in marine ecosystems are exposed to varied

anthropogenic and environmental pressures, which influ-

ence their behavior and can lead to changes in population

size and distribution (Gales et al. 1998, Croxall et al. 2012,

Phillips et al. 2016). Human activities in the marine

environment can cause the mortality of breeding-aged

individuals, the survival of which is particularly important

to the population viability of long-lived species such as

marine birds, mammals, and reptiles (Tuck et al. 2001,

Lewison et al. 2004, Anderson et al. 2011).

Analysis of factors influencing the demography and

behavior of apex marine predators often indicates that

there is a mix of environmental and anthropogenic drivers

at play (Barbraud et al. 2008, 2009, 2011, Delord et al.

2010, Rolland et al. 2010). While oceanographic frontal

systems are of primary importance to determining prey

distributions and, hence, foraging activity of top predators

(Bost et al. 2009, Bailleul et al. 2010), some environmental

influences only become apparent in certain climatic

conditions and are difficult to detect over short timeframes

(Rolland et al. 2010). Moreover, some factors can influence

individual survivorship both positively and negatively.

Such is the case with commercial fishing, where fisheries

waste can provide large quantities of food for scavenging

birds (Furness 1982, Thompson 1992), but may also kill

adult birds through entanglement and drowning (Wei-

merskirch et al. 1997a, Walker and Elliott 2006, Debski et

al. 2016). With changing environmental conditions (Rhein

et al. 2013), marine food webs (Pauly et al. 1998, Jackson et

al. 2001, Bond and Lavers 2014), and prey distributions

(Weimerskirch et al. 2012, Thorne et al. 2016), an

uncertain future lies ahead for marine vertebrate species,

many of which are threatened with extinction (Croxall et

al. 2012, IUCN 2017).

In the New Zealand region, particularly in the Tasman

Sea, seabird diversity and density is among the highest in

the world (BirdLife International 2004, Alderman et al.

2011, Waugh et al. 2012). Approximately one-third of the

346 seabird species globally nest in or visit the 4 million

km2 New Zealand Exclusive Economic Zone (NZEEZ), the

4th largest in the world (Mansfield 2006). Of the 92

indigenous seabird taxa resident in New Zealand (Ministry

for the Environment and Statistics New Zealand 2016),

more than half are listed as threatened or at risk of

extinction (IUCN 2017, Robertson et al. 2017). The New

Zealand region holds more threatened single-site endemic

species than any other region globally (Croxall et al. 2012).

Further, rapid changes in sea surface temperature (SST) in

the Tasman Sea have been signaled (Oliver et al. 2014),

with extreme anomalies of temperatures 68C above

average measured in recent periods (Palmer 2017). Such

changes may contribute to the occurrence of extreme

weather events, putting additional pressure on nesting

habitat (Waugh et al. 2015b). The conservation stakes in

managing human and natural impacts on vulnerable

wildlife populations in this region are therefore high.

To provide protection for marine species at risk from

multiple threats, an understanding of core feeding habitats

is essential (Montevecchi et al. 2012,Wakefield et al. 2013).

To this end, remote tracking methods that can accurately

locate tagged individuals in space and time are commonly

used to define core habitats and to delimit spatial

management zones (Awkerman et al. 2005, Peron et al.

2010, Lascelles et al. 2012, Thaxter et al. 2012). Using GPS

tracking, we explored the foraging patterns of a threatened

seabird, theWestland Petrel (Procellaria westlandica), over

a 6-yr period in the Tasman Sea, during different stages of

the species’ breeding cycle and across climatic conditions.

The Westland Petrel is present at its breeding colonies

between March and December each year. Postbreeding

birds migrate to South American waters (Brinkley et al.

2000, Landers et al. 2011). While considerably smaller

(~1,200 g) than the closely related, but better-researched,

albatross species in the region, the Westland Petrel shares

many demographic traits with them, including low

fecundity but high adult survival (Warham 1990, Waugh

et al. 2015a). The Westland Petrel is exposed to numerous

threats on land and at sea (Richard and Abraham 2015,

Waugh and Wilson 2017), including potential predation by

pigs (Sus scrofa) and dogs (Canis familiaris), breeding

habitat degradation from extreme weather events, attrac-

tion to lights of fledgling young, and fisheries-related

mortality in New Zealand and South American waters. In

view of these threats and considerable storm-induced

damage to the breeding areas of 75% of the population, the

species had its threat ranking increased in 2017 to

The Condor: Ornithological Applications 120:371–387, Q 2018 American Ornithological Society

372 Bathymetry and fisheries influence Westland Petrel foraging S. M. Waugh, J. W. Griffiths, T. A. Poupart, et al.

Endangered (BirdLife International 2017), and is consid-

ered Naturally Uncommon under the New Zealand

national conservation prioritization system (Robertson et

al. 2017).

Feeding association with and mortality due to fisheries

are traits that the Westland Petrel (Freeman 1997, 1998,

Freeman et al. 2001) shares with many congeners (Neves

and Olmos 1998, Neves et al. 2005, Delord et al. 2005,

2010, Freeman et al. 2010). Populations of these large

Procellaria species contribute the highest numbers of birds

killed incidentally by fisheries around the Southern Ocean

(Anderson et al. 2011), leading to a high risk of adverse

population impacts (Debski et al. 2016). However, the

adverse effects of fisheries on the Westland Petrel have yet

to be quantified (Phillips et al. 2016, Waugh and Wilson

2017). Diet studies conducted in the 1990s indicated

interactions of the species with a fishery for hoki

(Macruronus novaezelandiae) and hake (Merluccius aus-

tralis), which operates close to its breeding colonies.

Freeman (1997, 1998) found that up to 63% by weight of

chick diet was composed of fish species discarded by this

trawl fishery that were unlikely to be natural prey of the

petrels. The main component of their natural prey of fish

comprises several species of lanternfishes (Myctophidae)that migrate vertically in a diel rhythm and are found near

the surface during nighttime (Robertson and Clements

2015). It has been speculated that the provision of fisheries

waste from the hoki and hake fishery from the 1970s

onward led to an increase in the population of the

Westland Petrel (Marchant and Higgins 1990). This is a

spawning fishery (Ministry for Primary Industries 2016),

indicating a rich and productive region utilized by a range

of marine vertebrates. The fishery has undergone 10-fold

catch fluctuations since the mid-1990s (Ministry for

Primary Industries 2016), which may have resulted in a

reduction in benefit to the petrels from feeding on fisheries

waste in the present day compared with earlier decades.

Population modeling has shown an ongoing increase in the

species’ population over the 42-yr period. Since 1970,

population growth has been most strongly related to the

Southern Oscillation Index (SOI) and not fishery factors

(Waugh et al. 2015a). The core foraging habitat of

Westland Petrels is over the continental shelf of western

New Zealand in subtropical waters (Freeman et al. 1997,

2001, Landers et al. 2011). This zone is shared with several

top marine predators that are threatened with extinction

and endemic to New Zealand (Brager and Schnieder 1998,

Stahl and Sagar 2000a, 2000b, Taylor 2000, Forest and Bird

2014). This region is therefore of high importance for

protected species conservation, and is listed by BirdLife

International as an Important Bird Area (Forest and Bird

2014).

Our study aimed to identify the environmental and

anthropogenic drivers of the foraging distribution of the

Westland Petrel during its breeding period, to help

conservation managers to monitor and mitigate adverse

effects on the species’ population. Therefore, we modeled

the species’ foraging distribution during 4 yr over its

prebreeding (2011), incubation (2011, 2012, 2015, 2016),

and chick rearing (2012, 2015, 2016) periods. We explored

the relative importance of physical and anthropogenic

variables, including bathymetric slope, location (degrees

north and west of the colony), light conditions (daytime and

nighttime), trawl fishing activity that occurred during

foraging trips, the core fishing zone, chlorophyll-a levels

(Chl-a), SST, and an index of the El Nino–Southern

Oscillation (SOI), as well as the biotic factors of sex,

distance traveled, and breeding stage. We were particularly

interested in the influence of fishing activity that occurred

during Westland Petrel foraging trips on foraging intensity.

We expected that concurrent fishing activity would be a

dominant influence on Westland Petrel foraging, based on

previous study of the petrel’s association with trawl fisheries,

and that environmental factors, the core fishing zone,

climate, and breeding stage would have lesser influence.We

also conducted stable isotope analyses to assess the

importance of foraging zone on trophic niche width.

METHODS

Study Site and Study SpeciesThe Westland Petrel is a large procellariiform seabird that

nests in coastal ranges near Punakaiki, Westland, New

Zealand (42.1468S, 171.3418E). Individuals during the

incubation period in July during our 2011–2016 research

sessions weighed on average 1,235 g 6 128 g (n ¼ 147).

The Westland Petrel is 1 of 6 taxa in the genus Procellaria

(Techow et al. 2009, 2016), of which 3 are endemic to New

Zealand, while the others have South Atlantic or circum-

Antarctic distributions (ACAP 2017a). There is an

estimated annual breeding population of 2,800 breeding

pairs (ACAP 2017b). Breeding starts in March with

prebreeding attendance at colonies. Westland Petrels lay

a single egg in late May–early June which is incubated for

~69 days (Waugh and Bartle 2013). Chicks are fed by both

parents until at least November. For this study, birds were

accessed in burrows fitted with inspection hatches at the

Scotsman’s Creek study colony in 2011, 2012, 2015, and

2016. These years were chosen partly in response to

disruption of the study by storm activity, which reduced

the study colony by ~25–50% in burrow numbers in 2014

(Waugh et al. 2015b), and partly to cover the range of

climatic conditions desirable to examine the influence of

the SOI. In 2011, 16 of 21 tracks of prebreeding birds were

from an adjacent colony, ,1 km distant from the main

study area. These 2 colonies have some interchange of

birds and are therefore considered part of the same colony

complex (Waugh et al. 2015a).


S. M. Waugh, J. W. Griffiths, T. A. Poupart, et al. Bathymetry and fisheries influence Westland Petrel foraging 373

Sampling Methods

Adults were sexed by genetic and morphometric analyses,

with at least one bird of each pair sexed genetically.

Breeding success was measured as the number of breeding

attempts (eggs laid) that resulted in surviving chicks,

numbers of which were estimated in September. Standard

error for these rates was calculated by taking the square

root of the variance, accounting for demographic stochas-

ticity, following Akcakaya (2002). Chicks were weighed

between November 14 and 17 each year using a 2,500 g

Pesola balance (with precision 6 20 g; Pesola, Schindellegi,

Switzerland). Breeding success was compared between the

colony where most accessible breeding birds were

equipped with data loggers and one where fewer than

10% of breeding birds were equipped with loggers in a

given year, reported by Waugh and Wilson (2017). The

authors reported no significant differences in breeding

output between these groups.

One hundred and one I-Got-U GT120 loggers (Mobile

Action Technology 2016) repackaged in waterproof tubing,

measuring 44 3 28 3 11 mm and weighing 25 g, were

deployed on petrels. Loggers were attached to the back

feathers of study birds using adhesive tape. The logger mass

was ,2% of the average bird’s body mass and less than the

recommended 3% mass guideline to limit impacts on

species’ behavior (Phillips et al. 2003). Loggers were

deployed in the prebreeding period (mid-March) in 2011,

in the incubation period shortly after laying (late May) to

mid-incubation (early July) in 2011, 2012, 2015, and 2016,

and in the mid chick rearing period (late August to mid-

September) in 2012, 2015, and 2016 (see Table 1 for sample

sizes for each tracking session). Data loggers recorded

information at 5–15 min intervals, as a tradeoff between

accuracy of information and the need to maintain battery

life and to record complete tracks. Loggers were removed

after each bird had completed �1 trips at sea (average¼ 1.8

trips per bird, range¼1–6 trips), after an average of 15 days.

Westland Petrel Tracking Data

Tracking data were imported into the statistical package R

3.1.3 (R Core Team 2016) and filtered to exclude foraging

trip locations where estimated speed was ,4 km hr�1 or

.100 km hr�1, or where distance from the colony was

,500 m. We took this approach because our primary

interest was in the foraging distribution of Westland

Petrels and speeds ,4 km hr�1 are associated with drifting

TABLE 1. Foraging trip statistics for Westland Petrels by year and breeding period for complete trips with both sexes pooled,showing the number of tracks and number of individuals (females [F], males [M]) tracked. Values for duration, speed, and distanceare means 6 standard deviation. Area of the 50% utilization distribution (UD) is in km2, and the between-year comparisons of areaused are shown as Bhattacharyya’s Affinity (BA), indicating the percent spatial overlap between years.

Breeding period

Year

2011 2012 2015 2016

Pre-eggn 21 6 3 [F], 6 11 [M]Duration (days) 2.7 6 1.9Speed (km hr�1) 13.8 6 5.8Maximum distance (km) 166 6 11550% UD area (km2) 22,156

Incubationn 18 6 1 [F], 6 9 [M] 9 6 2 [F], 6 3 [M] 14 6 6 [F], 6 5 [M] 20 6 7 [F], 6 8 [M]Duration (days) 2.6 6 2.1 4.7 6 3.1 7.5 6 4.5 5.1 6 3.5Speed (km hr�1) 14.8 6 4.2 19.1 6 5.4 12.6 6 3.4 19.0 6 2.5Maximum distance (km) 147 6 91 284 6 183 253 6 227 255 6 19350% UD area (km2) 17,822 17,921 13,665 14,047BA (cf. 2011) 69% 69% 86%BA (cf. 2012) 75% 71%BA (cf. 2015) 86%

Chickn 29 6 6 [F], 6 6 [M] 6 6 3 [F], 6 3 [M] 18 6 4 [F], 6 12 [M]Duration (days) 3.0 6 1.8 4.9 6 1.3 2.0 6 1.3Speed (km hr�1) 15.7 6 5.7 11.7 6 3.8 15.2 6 4.2Maximum distance (km) 171 6 106 182 6 50 127 6 6050% UD area (km2) 13,572 7,897 11,101BA (cf. 2012) 85% 83%BA (cf. 2015) 94%

Breeding success 6 SE (n) 0.71 6 0.12 (14) 0.68 6 0.09 (25) 0.70 6 0.80 (33) 0.55 6 0.80 (40)



on the water (Torres et al. 2011), whereas speeds .100 km

hr�1 are considered unrealistic. Relocations ,500 m from

the colony were removed because seabirds often raft

(Awkerman et al. 2005, Guilford et al. 2008) near the

colony at the beginning and end of foraging trips.

After filtering the data, we calculated the total distance

traveled during each foraging trip, the distance of each

location from the colony, and the light conditions associated

with each location. Daytime was defined as the period

during which the sun’s center was ,128 below the horizon

(corresponding to nautical twilight and daytime), and

nighttime was defined as the period when the sun’s center

was .128 below the horizon.

We also derived the utilization distribution (kernel UD;

Worton 1989) of eachWestland Petrel foraging trip within a

standard grid clipped to the 95% minimum convex polygon

of all filtered locations using the adehabitatHR function

kernelUD (Calenge 2006) in program R. To avoid over- or

under-smoothing, the smoothing parameter was set to h¼0.2, the smallest value of h that prevented fragmentation of

the 95% UD contour (Kie 2013) of all filtered locations.

Resultant foraging trip UDs were amalgamated into a raster

stack using R package raster (Hijmans 2016).

Anthropogenic and Environmental DataFisheries data were extracted from Ministry for Primary

Industries (Wellington, New Zealand) databases, and

included all commercial fisheries activities of trawl, longline,

and gillnet vessels in a zone that overlapped with the areaused by petrels, defined by the starting position of each

fishing event. After inspection of the overlap of fishing

activity with 50% UD kernels of tracking datasets for each

petrel tracking session, fishery overlap by method and target

fish species was tabulated. The fisheries for which fishing

events overlapped .5% of the total tracking session (year

and breeding stage) were retained in the analyses. As no

longline or gillnet fisheries met this criterion, they were not

analyzed further. The UD of concurrent fishing activity was

calculated from the locations of trawl fishing starting

locations that occurred during each Westland Petrel

foraging trip with the smoothing factor set to h ¼ 0.2 and

grid size set to grid¼ 1,000. Concurrent fishing activity UDs

were then converted into rasters, resampled to the standard

grid used for Westland Petrel foraging trips, and amalgam-

ated into a raster stack. This fishing index, termed

‘concurrent fishing activity,’ represented fishing events that

occurred during a foraging trip. In addition, a UD of the

starting locations of all trawling events that occurred during

the petrel tracking sessions over the 6-yr study period was

calculated and resampled to the standard grid. This second

index, termed the ‘core fishing zone,’ represented a zone of

strong trawl fishing activity and marine productivity

compared with surrounding areas, rather than specific

fishing events.

Gridded (cell ¼ 0.889 degrees) Chl-a and SST values

averaged across 8 days were extracted from NOAA

(National Oceanic and Atmospheric Administration)

databases for the periods of bird tracking (M. Pinkerton

personal communication), and gridded 250-m bathymetric

data for the New Zealand region (NIWA 2016) were used

to create a bathymetric slope surface using the terrain

function in R package raster (Hijmans 2016). Chl-a, SST,

and bathymetric slope were then resampled to the

standard grid used for resampling the UDs of petrel

foraging trips. However, SSTwas correlated with latitude (r

¼ 0.46, P , 0.001) and caused variance inflation when

included in mixed effects models, so was removed. SOI

data were sourced from Bureau of Meteorology (2016),

and monthly values were averaged for each year.

Relating Westland Petrel Foraging Patterns toEnvironmental and Anthropogenic FactorsUsing similar methods to those described by Marzluff et al.

(2004) and Bjørneraas et al. (2011), we extracted the kernel

density estimate (KDE) for each foraging trip location fromWestland Petrel foraging UDs using the extract function in

R package raster (Hijmans 2016). We treated the KDE as a

probabilistic measure of foraging intensity, where a high

KDE was assumed to be associated with intensive foraging

activity and a low KDE was assumed to be related to less

intensive foraging activity or transit. We linked this

response to the gridded anthropogenic and environmental

data: Chl-a, SST, SOI, location relative to the colony, light

conditions, sex, breeding stage, and distance traveled. We

included these variables as they have been found to

influence pelagic seabird foraging distributions (e.g.,

Hyrenbach et al. 2002, Delord et al. 2010, Weimerskirch

et al. 2010, Torres et al. 2011).

To model Westland Petrel foraging intensity, generalized

linear mixed effects models were fit using R package lme4

(Bates et al. 2015). Models were run in 2 stages. Stage 1

modeled Westland Petrel foraging intensity as a function

of latitude, longitude, both measures of fishing activity, and

biophysical context (Appendix Table 3). Stage 2 extended

the best-fitting model from stage 1 with the intrinsic biotic

parameters of sex, breeding stage, and distance traveled

(Appendix Table 3). In total, 19 candidate models were fit

with a Gaussian error structure. The response variable

KDE was log-transformed to achieve normality and

continuous variables were standardized by subtracting

the mean and dividing by the standard deviation. All

models included log(KDE) as the response variable,

latitude and longitude as fixed effects, and bird identity

(hereafter, band ID) as a random effect.

Models were ranked by Akaike’s information criterion

corrected for small sample size (AICc) using R package

AICmodavg (Mazerolle 2013), and marginal and conditional

R2 were calculated (Nakagawa and Schielzeth 2013) for the



best model to determine goodness of fit. Normalized

randomized quantile residuals and variance inflation were

assessed for the best models to confirm that the residuals

were normally distributed with homogeneous variance and

that there were no multicollinearity issues of concern.

To understand the patterns of behavior of petrels in

relation to habitat and light conditions, data from 2011

were mapped and plotted in relation to land mass location,

bathymetric features, and light conditions. The map

geographic reference frame was WGS1984, and data

defining the coastline and bathymetry were sourced from

LINZ (2015). The resulting animation showed that birds

were active at night and during twilight as well as during

the day (Waugh 2012). Only 1 yr of data was investigated

because of the heavily time-intensive nature of these

analyses, but it showed patterns of behavior (e.g.,

nocturnal activity, movement associated with fishing

vessels) that were generalized across the dataset when we

examined these factors using statistical models.

The 95% UD represents the overall activity area,

including transit and foraging areas combined. However,

the 50% UD represents the places where the largest

concentrations of locations occurred (Wood et al. 2000).

The spatial extent in km2 of the UDs was calculated in

QuantumGIS using info tools (QGIS Development Team

2012; Table 1). To investigate similarities in foraging areas

between years for the same breeding stage, the spatial

overlap between 50% UDs was calculated using thekerneloverlap function. The overlap index was calculated

with Bhattacharyya’s Affinity as it is appropriate to

quantify the overall similarity between UDs within the

same breeding stage (Fieberg and Kochanny 2005).

Stable Isotope AnalysisBlood was sampled from 128 birds during incubation

(2011: n ¼ 7 females and 11 males; 2012: n ¼ 12 and 2;

2015: n¼10 and 11; and 2016: n¼ 9 and 9, respectively) to

measure d15N and d13C values, as a means of tracking

changes in trophic relationships for the species. These

isotope ratios change in a predictable and quantifiable way

along the food chain (Hobson and Clark 1992a, 1992b).

Because d15N increases with trophic level and d13C varies

according to the oceanographic zone that marine prey

come from, isotopes allow the determination of relative

trophic level and detection of prey consumption from

inshore vs. offshore waters (Hobson et al. 1994). Less

intrusive than stomach flushing techniques, the isotopic

niche is a proxy of the trophic niche (Jaeger et al. 2010)

and provides a much longer average picture of prey

consumption than just the last meal. Using whole blood,

stable isotopes reflect a period of dietary integration of ~4weeks (Bearhop et al. 2002). We sampled ~0.1 ml of blood

from each bird using 0.33 ml insulin needles. Samples were

maintained at 5–158C before being frozen within 8 hr of

collection, and were subsequently freeze-dried for pro-

cessing. Subsamples of ground, dried blood were packed

into 4 3 6 mm tin capsules before processing for stable

isotope analyses. Carbon and nitrogen contents (%C, %N)

and isotopic composition (d13C, d15N) were analyzed in

duplicate at the Stable Isotope Laboratory (GNS Science,

Gracefield, Wellington, New Zealand) using an Isoprime

isotope ratio mass spectrometer (Elementar Analysensys-

teme, Langenselbold, Germany), interfaced to a EuroEA

elemental analyzer (Eurovector, Pavia, Italy) in continuous-

flow mode (EA-IRMS). Working reference standards

(leucine, EDTA, caffeine, hair, and sucrose) were calibrated

against international reference materials (IAEA-N1, IAEA-

N2, IAEA-CH6, and IAEA-CH7), and blanks were

included during each run for calibration.

Isotopic ratios (15N/14N and 13C/12C) are expressed as

isotopic deviations (d) defined as:

dð%Þ ¼ Rs � RRef

RRef3 1000;

where Rs is the isotopic ratio measured for the sample and

RRef is that of the international standards. The d13C value is

relative to the international Vienna Pee Dee Belemnite

(VPDB) standard, and the d15N value is relative to

atmospheric air. Raw values were normalized using

working internal standards calibrated to international

standards and results are expressed in d (%) vs. the

specific reference. Analytical precision of the measure-

ments is 6 0.2%, and reproducibility of the results is

within 6 0.2% for carbon and 6 0.3% for nitrogen (1

rn). As for foraging parameters in the KDE models

(above), sex had little influence on d15N values (F1¼2.88, P

¼ 0.10) or d13C values (F1 ¼ 1.72, P ¼ 0.20); therefore,

isotopic niche analyses were conducted using data for both

sexes combined. Isotopic niches were estimated using

Bayesian standard ellipses computed in R package SIBER,

as described by Jackson et al. (2011) and applied by Jackson

et al. (2012) to compare isotopic resource use between

years. We used the Standard Ellipse Area (SEA) and

estimated the width of the isotopic niche of petrels using

the Bayesian Standard Ellipse area (SEAB). We calculated

the degree of isotopic overlap among the groups (years) of

petrel samples (Jackson et al. 2011) using 10,000 iterations

of the ellipses, and the resulting 50%, 75%, and 95%

credible intervals for the ellipses described the probability

that the trophic niches of the groups of birds overlapped.

Pairwise comparison in niche overlap is expressed as the

Bayesian credible interval in percentages.

RESULTS

Twenty-one loggers failed due to leaks in the waterproof

casing or battery failure before birds left for sea. In 12

cases, loggers were not recovered. This high failure rate



was attributed to the challenges of working with a

burrowing species. Measures to reduce disturbance of

the study birds included handling birds a maximum of 5

times during the breeding period. No nest failures over the

course of the study were attributed to handling associated

with logger deployment. During the 6-yr study period,

tracks were sampled from 73 individuals, 10 of which were

tracked in 2 sessions and 1 that was tracked 3 times.

Main Foraging AreasThere was strong consistency between years and breeding

stages in the areas most used by Westland Petrels (Figure

1). Between-year Bhattacharyya’s Affinity values showed a

high degree of overlap within breeding stages (69–86%

during incubation and 83–94% during the chick rearing

period; Table 1). Most foraging activity occurred along the

Hokitika Canyon to the southwest of the colony and the

sloping continental shelf edge which runs northeast from

that canyon. Other areas within the 50% UDs were in Cook

Strait around the Nicholson Canyon and off southwestern

New Zealand around the Arawata and Haast canyons. In

their core foraging areas, petrels fed mainly over the

continental shelf, within 1,000 m water depths. Outside

this region, foraging activity was concentrated in water

FIGURE 1. Main foraging areas of Westland Petrels, showing core fishing zones, bathymetry, and the 1,000 m depth contour. Areasof intensive usage are associated with steep bathymetric slopes south of the colony at Arawata and Haast Canyons (A) (within the2012 and 2015 50% kernel utilization distributions [UD]); at Hokitika Canyon (H) in the central zone in all years; and in Cook Straitover Nicholson Canyon (C) between the North Island and South Island (2011 only).



depths up to 2,000 m (Figure 1). These areas were within

the 95% UD for the species, but outside the 50% UDs for

each tracking session, indicating that they were more likely

to be used in transit or for occasional foraging by

individual birds rather than being a focus of foraging

activity at a population level.

Birds spent on average 2.6–7.5 days at sea (range¼ 0.1–

15.8 days; Table 1), during which time they were within an

average of 147–284 km (range¼ 5–927 km) of the colony.

The duration of trips during chick rearing was normally

distributed with no indication of bimodality in the

distribution (square-root-transformed data, Shapiro-Wilk

Test Statistic ¼ 0.98, P ¼ 0.44, n ¼ 55), which would have

indicated short and long trip feeding strategies (Weimer-

skirch 1998).

Foraging IntensityThe highest-ranked model of Westland Petrel foraging

intensity included all modeled parameters except sex,

which was not informative (Table 2, Appendix Table 3).

The preferred model fit the data well (marginal R2¼ 37%,

condition R2 ¼ 62%). The fixed effects explained ~37% of

the variance in foraging activity; however, 25% of the

variance in foraging activity was attributed to the random

effect of band ID, implying that foraging patterns varied

considerably among individuals.

Model coefficients indicated that Westland Petrel

foraging was concentrated in an area to the southwest of

the colony, in areas where the sea floor sloped steeply, and

in fisheries zones (Figure 2). Models that included both the

core fishing zone and concurrent fishing activity ranked

more highly than those that included just one fishery

variable (Table 2, Appendix Table 3), indicating that both

parameters were influential, but core fishing activity had a

greater influence on foraging intensity. Westland Petrel

foraging intensity was higher at nighttime than during the

daytime and was negatively related to Chl-a and latitude;

the latter was substitutable for sea surface temperature.

Total distance traveled predicted decreased Westland

Petrel foraging intensity, implying that birds that traveled

farther spent proportionally less time foraging. Foraging

intensity was higher during the prebreeding and incuba-

tion stages than during chick rearing.

Isotope AnalysesThe SIBER analyses of petrel blood samples taken during

the incubation period showed that there was a similar

dietary spread in all years (Figure 3) except 2015, when

niche width was significantly wider than in 2011 and 2012

but not 2016 (Figure 4). SEAB probabilities were less than

0.95 (therefore not significantly different) for all pairwise

comparisons except in 2 cases: 2011 vs. 2015 (SEAB¼0.98)

and 2012 vs. 2015 (SEAB¼ 0.95). Values of both d13C and

d15N were broader in 2015 than in 2011 and 2012, but

values were similar between all other pairs of years (Figure4).

Breeding SuccessCompared with the long-term average of 65% for the

species (Waugh et al. 2015a), breeding success was high in

most years of the tracking study (68–71%) except 2016

(55%). Chick mass at fledging was slightly heavier in 2015

(1,530 6 203 g, n¼ 23) than in 2012 (1,453 6 281 g, n ¼16) and 2016 (1,503 6 207 g, n¼ 15), but not significantly

so (F2,51¼ 3.13, P ¼ 0.69).

Fishing Fleets Associated with Petrel ForagingThe trawl fisheries that were dominant in the core foraging

areas of Westland Petrels, in descending order of

importance, targeted hoki, hake, flatfishes (various species

including Rhombosolea plebeian and Colistium nudipin-

nis), red cod (Pseudophycis bachus), and tarakihi (Nem-

adactylus macropterus). There were 825 events that

overlapped with the 50% UD of petrels during prebreeding

TABLE 2. Generalized linear mixed effects models of WestlandPetrel foraging intensity (kernel utilization distribution [KUD])ranked by Akaike’s information criterion corrected for smallsample size (AICc). Models were run in 2 stages. Stage 1 rankedmodels that included anthropogenic, environmental, andgeographic parameters. Stage 2 ranked models that extendedthe highest-ranked model from stage 1 by adding WestlandPetrel breeding stage, sex, and foraging distance traveled. K ¼the number of parameters, wi¼ Akaike weight, and DAICc is thedifference in AICc value from the top model. The candidatemodel set is specified in Appendix Table 3.

Model number K DAICc wi Log-likelihood

Stage 112 11 0.00 a 0.78 �25,816.210 10 2.70 0.20 �25,818.69 10 8.58 0.01 �25,821.57 9 11.52 0.00 �25,824.011 10 48,425.43 0.00 �50,030.06 9 48,446.93 0.00 �50,041.78 9 48,550.62 0.00 �50,093.65 8 48,571.68 0.00 �50,105.12 6 52,405.09 0.00 �52,023.83 6 52,699.83 0.00 �52,171.24 6 54,157.64 0.00 �52,900.11 5 56,259.86 0.00 �53,952.2

Stage 218 14 0.00 b 0.68 �25,050.919 15 1.51 0.32 �25,050.613 12 103.67 0.00 �25,104.716 13 105.56 0.00 �25,104.715 13 1,396.01 0.00 �25,749.917 14 1,397.53 0.00 �25,749.714 12 1,525.49 0.00 �25,815.6

a AICc ¼ 51,654.47.b AICc ¼ 50,129.78.



tracking in 2011, and an annual average of 353 during

incubation and 341 in the chick rearing stage across all

years of our study. The target fisheries that overlapped

during prebreeding comprised flatfish (40% of events

within the 50% UD), red cod (11%), tarakihi (17%), ling

(Genypterus blacodes; 7%), and dark ghost shark (Hydro-

lagus novaezealandiae; 7%). During the incubation and

chick rearing periods, the main trawl fisheries that

occurred within the petrel’s 50% UD targeted hake (19%

during incubation, 32% during the chick rearing stage),

hoki (58% and 28%, respectively), tarakihi (9% and 12%,

respectively), barracouta (Thyrsites atun; 0% and 9%,

respectively), warehou (Seriolella spp.; 2% and 13%,

respectively), and flatfish (3% and 6%, respectively). Thus,

we note that a wide range of species targeted by trawl

fisheries occur within the main foraging zone of Westland

Petrels across different breeding stages and years.

DISCUSSION

Work on seabird foraging globally has highlighted the

importance of considering both environmental and

anthropogenic factors when determining their distribu-

tions (Weichler et al. 2004,Wakefield et al. 2009, Thorne et

al. 2016). We undertook a multiyear study of the foraging

distribution of the Westland Petrel across different stages

of the breeding cycle and in a variety of climatic

conditions. We found that a combination of physical,

environmental, climatic, and fisheries factors influenced

the foraging distribution of the Westland Petrel.

FIGURE 2. Model coefficients for the highest-ranked model (model 18; Appendix Table 3) of Westland Petrel foraging intensity(kernel density estimate [KDE]). Model estimates indicate the response of foraging intensity to each parameter included in themodel. Error bars indicate 95% confidence intervals. Where the error bar excludes zero, the effect is significant. The base level forcomparison is the prebreeding stage and daylight conditions. The magnitude of the effect is indicated by the distance from zero.



During the breeding season study period, tracked birds

spent very little time near the Subtropical Front, which

passes south of the New Zealand mainland in this region

(Heath 1985). The main zone exploited by the petrels

overlapped considerably with fisheries activity. Several

fishing methods are active in this region, but trawl fisheries

made up .95% of the fishing events in the core areas

where Westland Petrels foraged. We examined the

importance of trawl fisheries to the species’ at-sea

distribution, expanding on knowledge gained in the

1990s and 2010s from tracking studies based on small

sample sizes, and confirmed this species’ core feeding area

to lie within ~250 km of its breeding colonies (Freeman

1998, Freeman et al. 2001, Landers et al. 2011).

For the Westland Petrel, the factor that most influenced

foraging intensity was bathymetric slope. Birds foraged

mostly to the southwest of the breeding colony in areas

where the sea floor sloped steeply, often along underwater

canyon features in 3 regions (Hokitika Canyon within 70

km of the colony, Arawata and Haast Canyons 270 km to

the south, and Nicholson Canyon in Cook Strait, 450 km

to the northeast; Figure 1). The Hokitika and Haast regions

each generate 7.7–9.6 billion m3 of freshwater runoff to the

sea each year, which contributes nutrients to the marine

system of the Tasman Sea (LAWA 2017). These areas were

used in multiple years and/or breeding stages by the

petrels, with a high between-year consistency in areas

used, indicated by the spatial overlap in foraging areas of

69–94% (Table 1). The steeply sloping bathymetric

structures are associated with strong outflow of terrestrial

sediments from large rivers (e.g., Haast River and Hokitika

River) and/or strong upwelling caused by the interplay

between ocean circulation and bathymetric features (in

Cook Strait over the Nicholson Canyon; Heath 1985,

Murphy et al. 2001).

In the New Zealand region, peak zones for oceanic

productivity occur over the Chatham Rise and Cook Strait

and along narrow coastal continental shelf areas (Pinker-

ton 2016), in both spring and autumn (Murphy et al. 2001).

The broader regions frequented by foraging Westland

Petrels had high Chl-a levels across seasons (Pinkerton

2016), suggesting stable and predictable foraging areas, but

the areas used for intensive foraging had low Chl-a levels

compared with neighboring areas. Decoupling of seabird

foraging intensity from the biomass of lower trophic level

prey and primary productivity is a common occurrence

(Gremillet et al. 2008). The increased foraging intensity in

areas of locally depleted Chl-a suggests a reliance on prey

that are farther up the food chain than crustacea or

copepods which feed directly on phytoplankton. The

reliance on natural foods such as squids and myctophidae

noted in Westland Petrel diet studies in the 1990s

(Freeman 1997, 1998) would explain a disassociation with

Chl-a levels. For example, in the Indian Ocean, lanternfish

(Electrona antarctica) were scarce in shallow, shelf waters,

and most abundant in areas where the water was deep

enough to allow a .250-m diel vertical migration and in

areas where their main prey (primary and secondary

consumers such as euphausid and copepods) were found.

FIGURE 3. Ellipses of isotope values for Westland Petrels by year,for breeding individuals sampled during the incubation period(n ¼ 128).

FIGURE 4. Stable isotope Bayesian ellipses run in program R(SIBER) to determine dietary niches of Westland Petrels in theTasman Sea by year, based on blood samples from breedingindividuals sampled during the incubation period (n ¼ 128).Smaller ellipse areas indicate more restricted diets.



These associations explained the negative relationship of

lanternfish with Chl-a (Loots et al. 2007). Our study lacks

the prey-density information to assess the trophic factors

which most influence the Westland Petrel’s foraging

behavior, but it seems probable that their fish and squid

prey distributions explain this relationship, and that

decoupling from areas of strong Chl-a concentration

should be expected.

Fisheries activity also had a strong influence on West-

land Petrel foraging patterns, with concurrent fishing

activity and the core fishing zone predicting significant

increases in foraging intensity. However, the core fishing

zone had a larger effect size than concurrent fishing

activity. Similar findings were reported by Freeman et al.

(1998), who studied the foraging activity of Westland

Petrels in the 1990s and found an association in space

with, rather than dependence on, the predominant hoki

trawl fishery in the region (Freeman and Wilson 2002).

The presence of fisheries waste in the diet fed to chicks

(Freeman 1998) suggests that reliance on this food source,

at least to feed chicks, may have been strong in earlier

decades. However, the co-occurrence of seabirds and

fisheries does not imply that interactions between vessels

and birds are occurring, either in terms of incidentalmortality or feeding on waste (Torres et al. 2011). The

nature of such interactions requires detailed behavioral

and fisheries-observer data, such as through concurrent

tracking and monitoring of fisheries (Weimerskirch et al.

2018). The principal commercial fisheries operating in the

core foraging areas of theWestland Petrel during our study

were trawl fisheries. These fisheries should be included in

the set for which enhanced observer activity is conducted,

to examine fishery–petrel interactions. Fisheries targeting

hoki, hake, flatfishes, red cod, and tarahiki should be

monitored during all stages of the petrel breeding cycle. In

the extensively observed hake and hoki fisheries, observer

records indicate very little mortality of Westland Petrels,

but there is little monitoring of the inshore, multispecies

trawl fisheries that occur in the zones heavily exploited by

the petrels during the prebreeding period (Richard and

Abraham 2015). Relatively few fishing events occurred

within the core foraging area of the petrels during chick

rearing (on average, 341 events per 2-week-long tracking

session) and incubation (353 events) compared with the

prebreeding period (825 events).

The new information presented in this study about

prebreeding and breeding season foraging distribution has

the potential to change the Ecological Risk Assessment

ranking for the Westland Petrel. This is because the spatial

distributions used in such assessments have strong

leverage for determining the overall threat levels accorded

to species (Small et al. 2013, Richard and Abraham 2015).

TheWestland Petrel currently ranks 10th most at risk from

population impacts of New Zealand commercial fishing,

and it is likely that this ranking may increase in importance

or severity as the species’ range overlaps with a greater

range of fisheries than was previously known.

The distance that a bird traveled had a strongly negative

influence on its foraging intensity. This is consistent with

tracking research conducted on albatross and petrel

species such as the Wandering Albatross (Diomedea

exulans) and Sooty Shearwater (Ardenna grisea), in which

individuals may adopt a strategy of covering long distances

to increase the encounter rate with prey (Weimerskirch et

al. 1997b, Weimerskirch 1998). In Wandering Albatross,

birds that undertook long foraging trips encountered prey

at twice the rate of those that fed locally but provisioned

chicks more often (Weimerskirch et al. 1997b). However,

Westland Petrels don’t appear to use a ‘short and long trip’

feeding strategy when feeding chicks, as found in some

Procellariiformes (Magalhaes et al. 2008, Shoji et al. 2015),

and they did not exhibit bimodality in foraging duration

during the chick rearing stage. However, it appears that

longer, looping oceanic trips are more common in some

years than in others for the Westland Petrel, possibly in

relation to prevailing winds and prey availability. In the

strong La Nina year of 2011, birds went the shortest

maximum distances and traveled the slowest. In smallerseabirds, such as the Slender-billed Prion (Pachyptila

belcheri) and Common Murre (Uria aalge), favorable

conditions led to reduced foraging trip lengths (Burke and

Montevecchi 2009, Quillfeldt et al. 2010), results which

were not mirrored in our study. This may be because

terrestrial inputs (e.g., runoff and sediments) have an

influence on marine productivity in the coastal areas of our

study zone.

Light conditions had a relatively minor, but nonetheless

significant, influence on the birds’ activities. Birds foraged

more intensively at night than during the daytime, with

movement observable throughout the day–night cycle

(Waugh 2012). Lanternfishes have been identified as the

principal naturally sourced (nonfisheries) prey of Westland

Petrels (Freeman 1998). Lanternfish of several species are

very common in the midwater fish biomass of the

Southern Ocean (Robertson and Clements 2015). These

species make vertical migrations daily, reaching shallow

water in the hours of darkness. Their bioluminescence and

presence in the epipelagic zone at night make them an

accessible prey for the petrels (Freeman 1998). Several

species are abundant in the Tasman Sea, with species in the

family Myctophidae the most common and diverse in the

region (Robertson and Clements 2015). Westland Petrels

also feed extensively on squid (families Cranchiidae and

Histioteuthidae), which make up 18% of their diet by

weight (Freeman 1998). These species of squid are

common in the diets of albatrosses around the Southern

Ocean (Cherel and Klages 1998), which dive to relatively

shallow depths (generally ,5 m; Prince et al. 1994); thus, it



has been speculated that seabirds tend to feed on

moribund squid floating at the surface (Croxall and Prince

1994, Cherel and Klages 1998), although direct observation

of these events remains rare. These squids are not in the

same species or family targeted by New Zealand commer-

cial fisheries (Ommastrephidae; Ministry for Primary

Industries 2015); hence, the squid in the petrel’s diet is

unlikely to come from fisheries sources.

Westland Petrel feeding intensity showed slight but

negative variation in relation to the SOI, with stronger

feeding intensity when the SOI was in an El Nino phase.

There were 2 yr during our study with strong SOI indices:

2011 had a strongly positive (La Nina) index, while 2015

had a strongly negative index (El Nino). The other 2 yr of

our study (2012 and 2016) had neutral SOI indices. Thus,

these results may be considered a tendency only. They are,

however, supported by the results of our stable isotope

analysis, which showed that the niche width of the petrels

increased significantly in the year of strong El Nino

conditions (2015) and also had a wider spread of d13Cvalues, indicating a wider geographical zone used by

foraging petrels (Cherel and Hobson 2007). In this year,

progeny survival and weight tended to be higher than in

other years, indicating a good year for foraging andbreeding performance. Similarly, Quillfeldt et al. (2010),

using stable isotope analysis on New Zealand populations

of Slender-billed Prion, found that this species fed on

higher trophic level prey, favoring squids over crustaceans

in cooler, more favorable conditions.

The differences in foraging intensity between breeding

stages were slight and characterized by high variance

within each stage. Petrels showed lower foraging intensity

in the chick rearing stage than in the prebreeding stage,

but there was no significant difference in foraging intensity

between the incubating and pre-egg or chick rearing

stages. Birds tended to carry out shorter trips in the chick

rearing stage than in the incubation and pre-egg periods.

When seabirds are not constrained by the need to attend

young in the nest, they tend to forage at greater distances

(Thiebot et al. 2011). The sample of only one season of

activity in the prebreeding stage means that our results

may be unrepresentative, and prebreeding foraging activity

warrants further examination.

This study is the first to conduct a multiyear, several-

breeding-stage analysis of the factors driving the foraging

distribution of the Westland Petrel, a threatened endemic

seabird. We found that the species’ foraging activity was

concentrated in the same areas from year to year and was

most influenced by individual bird’s foraging preferences,

secondarily affected by environmental variables (bathy-

metric slope, Chl-a, and latitude, a proxy for SST), and

thirdly influenced by fisheries activity. Predicted changes in

SST and currents in the Tasman Sea suggest an uncertain

future for the Westland Petrel as food resources may

change in location or abundance. Conservation manage-

ment of this highly threatened species should focus on the

elements that have the greatest possibility of being

influenced, specifically the effects of predators on land

and fisheries impacts at sea. Additional research is needed

to resolve land-based and fisheries threats to the species.

The stability of the Westland Petrel’s core foraging areas

makes it possible to institute marine spatial management

to reduce adverse impacts of fishing or seabed mining on

multiple threatened endemic top marine predators in the

region (e.g., Brager and Schneider 1998, Walker and Elliott

2006, Poupart et al. 2017), such as through the imple-

mentation of marine reserves or restrictions on damaging

activities (BirdLife International 2009, Department of

Conservation and Ministry of Fisheries 2011, Lascelles et

al. 2012). Such areas will need to be at a scale that will

cover key habitat requirements to be effective. Current

marine protected areas in the region cover water within 3

km of the coast only (Department of Conservation 2017)

and therefore are unlikely to afford protection to

vulnerable marine wildlife such as the Westland Petrel.

As a wide-ranging species present in the region for 10 mo

of the year, the Westland Petrel presents a good model

species with which to determine the zones of most

importance for the suite of species in the region.

ACKNOWLEDGMENTS

We thank R. Lane, C. Miskelly, J.-C. Stahl, and K.-J. Wilson forassistance with fieldwork. Thanks also to D. Pinaud, and K.Delord for assistance with tracking analyses, and T. Landersand L. Shepherd for genetic sexing data. Kati WaewaeRunanga granted consent for the study to proceed. We aregrateful to Department of Conservation staff, including G.Coleman, S. Freeman, T Shaw, H. Otley, and C. Wood, foradvice and assistance. We extend our gratitute to M.Pinkerton of NIWA for assistance with accessing ocean colordatasets. We appreciate comments from J. A. Bartle, A.Freeman, K.-J. Wilson, P. Doherty, and an anonymousreviewer, which helped to improve the manuscript.Funding statement: The study was funded by Department ofConservation, Te Papa, and Deakin University, with collabo-rative support from GNS Sciences. None of our funders hadany input into the content of the manuscript, nor did theyrequire their approval of the manuscript before submission orpublication.Ethics statement: All work was conducted under permitWC-26677-FAU issued by the New Zealand Department ofConservation, with approved animal handling procedures.Author contributions: S.M.W., J.Y.P.A., and D.P.F. conceivedthe idea, design, experiment (supervised research, formulatedquestion or hypothesis); S.M.W., T.A.P., and J.Y.P.A. performedthe experiments (collected data, conducted the research);S.M.W., J.G., T.A.P., and J.Y.P.A. wrote the paper (orsubstantially edited the paper); all authors developed ordesigned methods; S.M.W., J.G., T.A.P., K.R., and D.P.F.



analyzed the data; and S.M.W., J.Y.P.A., and K.R. contributedsubstantial materials, resources, or funding.

LITERATURE CITED

ACAP (Agreement on the Conservation of Albatrosses andPetrels) (2017a). Species Assessments. ACAP Secretariat,Hobart, TAS, Australia. https://acap.aq/en/resources/acap-species2

ACAP (Agreement on the Conservation of Albatrosses andPetrels) (2017b). Westland Petrel (Procellaria westlandica).ACAP Secretariat, Hobart, TAS, Australia. https://www.acap.aq/en/acap-species/316-westland-petrel/file

Akcakaya, H. R. (2002). Estimating the variance of survival ratesand fecundities. Animal Conservation 5:333–336.

Alderman, R., D. Anderson, J. Arata, P. Catry, R. Cuthbert, L.Deppe, G. Elliot, R. Gales, J. Gonzales Solis, J. P. Granadeiro, M.Hester, et al.(2011). Albatross and giant-petrel distributionacross the world’s tuna and swordfish fisheries. BirdLifeInternational Presentation at the Third Joint Tuna RFMOsmeeting; La Jolla, July 11–15, 2011. http://www.seabirdtracking.org/sites/default/files/KOBE%20brochure.pdf

Anderson, O. R. J., C. J. Small, J. P. Croxall, E. K. Dunn, B. J.Sullivan, O. Yates, and A. Black (2011). Global seabird bycatchin longline fisheries. Endangered Species Research 14:91–106.

Awkerman, J. A., A. Fukuda, H. Higuchi, and D. J. Anderson(2005). Foraging activity and submesoscale habitat use ofWaved Albatrosses Phoebastria irrorata during chick-brood-ing period. Marine Ecology Progress Series 291:289–300.

Bailleul, F., M. Authier, S. Ducatez, F. Roquet, J.-B. Charrassin, Y.Cherel, and C. Guinet (2010). Looking at the unseen:Combining animal bio-logging and stable isotopes to reveala shift in the ecological niche of a deep diving predator.Ecography 33:709–719.

Barbraud, C., K. Delord, C. Marteau, and H. Weimerskirch (2009).Estimates of population size of White-chinned Petrels andGrey Petrels at Kerguelen Islands and sensitivity to fisheries.Animal Conservation 12:258–265.

Barbraud, C., C. Marteau, V. Ridoux, K. Delord, and H.Weimerskirch (2008). Demographic response of a populationof White-chinned Petrels Procellaria aequinoctialis to climateand longline fishery bycatch. Journal of Applied Ecology 45:1460–1467.

Barbraud, C., P. Rivalan, P. Inchausti, M. Nevoux, V. Rolland, andH. Weimerskirch (2011). Contrasted demographic responsesfacing future climate change in Southern Ocean seabirds.Journal of Animal Ecology 80:89–100.

Bates, D., M. Machler, B. M. Bolker, and S. C. Walker (2015). Fittinglinear mixed-effects models using lme4. Journal of StatisticalSoftware 67:1–48.

Bearhop, S., S. Waldron, S. C. Votier and R. W. Furness (2002).Factors that influence assimilation rates and fractionation ofnitrogen and carbon stable isotopes in avian blood andfeathers. Physiological and Biochemical Zoology 75:451–458.

BirdLife International (2004). Tracking Ocean Wanderers: TheGlobal Distribution of Albatrosses and Petrels. Results fromthe Global Procellariiform Tracking Workshop, 1–5 Septem-ber, 2003, Gordon’s Bay, South Africa. BirdLife International,Cambridge, UK. http://www.birdlife.org/sites/default/files/attachments/tracking_ocean_wanderers.pdf

BirdLife International (2009). Designing Networks of MarineProtected Areas: Exploring the Linkages between ImportantBird Areas and Ecologically or Biologically Significant MarineAreas. BirdLife International, Cambridge, UK.

BirdLife International (2017). Procellaria westlandica. The IUCN RedList of Threatened Species 2017. e.T22698155A119318610.h t t p : / / d x . d o i . o r g / 1 0 . 2 3 0 5 / I U C N . U K . 2 0 1 7 - 3 . R L T S .T22698155A119318610.en

Bjørneraas, K., E. J. Solberg, I. Herfindal, B. Van Moorter, C. M.Rolandsen, J.-P. Tremblay, C. Skarpe, B.-E. Sæther, R. Eriksen,and R. Astrup (2011). Moose Alces alces habitat use atmultiple temporal scales in a human-altered landscape.Wildlife Biology 17:44–54.

Bond, A. L., and J. L. Lavers (2014). Climate change alters thetrophic niche of a declining apex marine predator. GlobalChange Biology 20:2100–2107.

Bost, C. A., C. Cotte, F. Bailleul, Y. Cherel, J. B. Charrassin, C.Guinet, D. G. Ainley, and H. Weimerskirch (2009). Theimportance of oceanographic fronts to marine birds andmammals of the southern oceans. Journal of Marine Systems78:363–376.

Brinkley, E. S., S. N. G. Howell, M. P. Force, L. B. Spear, and D. C.Ainley (2000). Status of the Westland Petrel (Procellariawestlandica) off South America. Notornis 47:179–183.

Brager, S., and K. Schneider (1998). Near-shore distribution andabundance of dolphins along the West Coast of the SouthIsland, New Zealand. New Zealand Journal of Marine andFreshwater Research 32:105–112.

Bureau of Meteorology (2016). Monthly Southern OscillationIndex. National Climate Centre, Bureau of Meteorology,Melbourne, VIC, Australia. ftp://ftp.bom.gov.au/anon/home/ncc/www/sco/soi/soiplaintext.html

Burke, C. M., and W. A. Montevecchi (2009). The foragingdecisions of a central place foraging seabird in response tofluctuations in local prey conditions. Journal of Zoology 278:354–361.

Calenge, C. (2006). The package ‘‘adehabitat’’ for the R software:A tool for the analysis of space and habitat use by animals.Ecological Modelling 197:516–519.

Cherel, Y., and K. A. Hobson (2007). Geographical variation incarbon stable isotope signatures of marine predators: A toolto investigate their foraging areas in the Southern Ocean.Marine Ecology Progress Series 329:281–287.

Cherel, Y., and N. Klages (1998). A review of the food ofalbatrosses. In Albatross Biology and Conservation (G.Robertson and R. Gales, Editors). Surrey Beatty & Sons,Chipping Norton, Australia. pp. 113–136.

Croxall, J. P., and P. A. Prince (1994). Dead or alive, night or day:How do albatrosses catch squid? Antarctic Science 6:155–162.

Croxall, J. P., S. H. M. Butchart, B. Lascelles, A. J. Stattersfield, B.Sullivan, A. Symes, and P. Taylor (2012). Seabird conservationstatus, threats and priority actions: A global assessment. BirdConservation International 22:1–34.

Debski, I., F. Hjorvarsdottir, and K. Knowles (2016). Distribution ofhighly at-risk New Zealand seabirds in the Western CentralPacific Fisheries Commission area. Western and CentralPacific Commission Scientific Committee, Twelfth RegularSession, Bali, Indonesia, 3–11 August 2016. Report WCPFC-SC12-2016/ EB-WP-09. https://www.wcpfc.int/system/files/



https://acap.aq/en/resources/acap-species2

https://acap.aq/en/resources/acap-species2

https://www.acap.aq/en/acap-species/316-westland-petrel/file

https://www.acap.aq/en/acap-species/316-westland-petrel/file

http://www.seabirdtracking.org/sites/default/files/KOBE%20brochure.pdf

http://www.seabirdtracking.org/sites/default/files/KOBE%20brochure.pdf

http://www.birdlife.org/sites/default/files/attachments/tracking_ocean_wanderers.pdf

http://www.birdlife.org/sites/default/files/attachments/tracking_ocean_wanderers.pdf

http://dx.doi.org/10.2305/IUCN.UK.2017-3.RLTS.T22698155A119318610.en

http://dx.doi.org/10.2305/IUCN.UK.2017-3.RLTS.T22698155A119318610.en

ftp://ftp.bom.gov.au/anon/home/ncc/www/sco/soi/soiplaintext.html

ftp://ftp.bom.gov.au/anon/home/ncc/www/sco/soi/soiplaintext.html

https://www.wcpfc.int/system/files/EB-WP-09%20NZ%20seabird%20distribution%20paper%20Rev%201%20%284%20August%202016%29.pdf

EB-WP-09%20NZ%20seabird%20distribution%20paper%20Rev%201%20%284%20August%202016%29.pdf

Delord, K., C. Cotte, C. Peron, C. Marteau, P. Pruvost, N. Gasco, G.Duhamel, Y. Cherel, and H. Weimerskirch (2010). At-seadistribution and diet of an endangered top predator:Relationship between White-chinned Petrels and commerciallongline fisheries. Endangered Species Research 13:1–16.

Delord, K., N. Gasco, H. Weimerskirch, C. Barbraud and T. Micol(2005). Seabird mortality in the Patagonian toothfish longlinefishery around Crozet and Kerguelen Islands, 2001–2003.CCAMLR Science 12:53–80.

Department of Conservation (2017). Punakaiki Marine Reserve.Department of Conservation, Wellington, New Zealand.http://www.doc.govt.nz/parks-and-recreation/places-to-go/west-coast/places/punakaiki-marine-reserve/

Department of Conservation and Ministry of Fisheries (2011).Coastal Marine Habitats and Marine Protected Areas in theNew Zealand Territorial Sea: A Broad Scale Gap Analysis.Department of Conservation and Ministry of Fisheries,Wellington, New Zealand.

Fieberg, J., and C. O. Kochanny (2005). Quantifying home-rangeoverlap: The importance of the utilization distribution. TheJournal of Wildlife Management 69:1346–1359.

Forest and Bird (2014). New Zealand Seabirds: Important BirdAreas and Conservation. The Royal Forest & Bird ProtectionSociety of New Zealand, Wellington, New Zealand.

Freeman, A. N. D. (1997). The influence of hoki fishing vessels onWestland Petrel (Procellaria westlandica) distribution at sea.Notornis 44:159–164.

Freeman, A. N. D. (1998). Diet of Westland Petrels Procellariawestlandica: The importance of fisheries waste during chick-rearing. Emu 98: 36–43.

Freeman, A. N. D., and K.-J. Wilson (2002). Westland Petrels andhoki fishery waste: Opportunistic use of a readily availableresource? Notornis 49:139–144.

Freeman, A. N. D., D. G. Nicholls, K.-J. Wilson, and J. A. Bartle(1997). Radio- and satellite-tracking Westland Petrels Procel-laria westlandica. Marine Ornithology 25:31–36.

Freeman, A. N. D., K.-J. Wilson, and D. G. Nicholls (2001).Westland Petrels and the hoki fishery: Determining co-occurrence using satellite telemetry. Emu 101:47–56.

Freeman, R., T. Dennis, T. Landers, D. Thompson, E. Bell, M.Walker, and T. Guilford (2010). Black Petrels (Procellariaparkinsoni) patrol the ocean shelf-break: GPS tracking of avulnerable procellariiform seabird. PLOS One 5:e9236.

Furness, R. W. (1982). Competition between fisheries and seabirdcommunities. Advances in Marine Biology 20:225–307.

Gales, R., N. Brothers, and T. Reid (1998). Seabird mortality in theJapanese tuna longline fishery around Australia, 1988–1995.Biological Conservation 86:37–56.

Gremillet, D., S. Lewis, L. Drapeau, C. D. Van Der Lingen, J. A.Huggett, J. C. Coetzee, H. M Verheye, F. Daunt, S. Wanless,and P. G. Ryan (2008). Spatial match–mismatch in theBenguela upwelling zone: Should we expect chlorophylland sea-surface temperature to predict marine predatordistributions? Journal of Applied Ecology 45:610–621.

Guilford, T. C., J. Meade, R. Freeman, D. Biro, T. Evans, F.Bonadonna, D. Boyle, S. Roberts, and C. M. Perrins (2008). GPStracking of the foraging movements of Manx ShearwatersPuffinus puffinus breeding on Skomer Island, Wales. Ibis 150:462–473.

Heath, R. A. (1985). A review of the physical oceanography of theseas around New Zealand — 1982. New Zealand Journal ofMarine and Freshwater Research 19:79–124.

Hijmans, R. J. (2016). raster: Geographic data analysis andmodeling. R package version 2.6-7. https://CRAN.R-project.org/package¼raster

Hobson, K. A., and R. G. Clark (1992a). Assessing avian diets usingstable isotopes I: Turnover of 13C in tissues. The Condor 94:181–188.

Hobson, K. A., and R. G. Clark (1992b). Assessing avian dietsusing stable isotopes II: Factors influencing diet-tissuefractionation. The Condor 94:189–197.

Hobson, K. A., J. F. Piatt, and J. Pitocchelli (1994). Using stableisotopes to determine seabird trophic relationships. Journalof Animal Ecology 63:786–798.

Hyrenbach, K. D., P. Fernandez, and D. J. Anderson (2002).Oceanographic habitats of two sympatric North Pacificalbatrosses during the breeding season. Marine EcologyProgress Series 233:283–301.

IUCN (2017). The IUCN Red List of Threatened Species, version2017-3. http://www.iucnredlist.org/

Jackson, A. L., R. Inger, A. C. Parnell, and S. Bearhop (2011).Comparing isotopic niche widths among and within com-munities: SIBER – Stable Isotope Bayesian Ellipses in R.Journal of Animal Ecology 80:595–602.

Jackson, J. B. C., M. X. Kirby, W. H. Berger, K. A. Bjorndal, L. W.Botsford, B. J. Bourque, R. H. Bradbury, R. Cooke, J. Erlandson,J. A. Estes, T.P. Hughes, et al. (2001). Historical overfishing andthe recent collapse of coastal ecosystems. Science 293:629–637.

Jackson, M. C., I. Donohue, A. L. Jackson, J. R. Britton, D. M.Harper, and J. Grey (2012). Population-level metrics of trophicstructure based on stable isotopes and their application toinvasion ecology. PLOS One 7:e31757.

Jaeger, A., M. Connan, P. Richard, and Y. Cherel (2010). Use ofstable isotopes to quantify seasonal changes of trophic nicheand levels of population and individual specialisation inseabirds. Marine Ecology Progress Series 401:269–277.

Kie, J. G. (2013). A rule-based ad hoc method for selecting abandwidth in kernel home-range analyses. Animal Biotelem-etry 1:art.13.

Landers, T. J., M. J. Rayner, R. A. Phillips, and M. E. Hauber (2011).Dynamics of seasonal movements by a trans-Pacific migrant,the Westland Petrel. The Condor 113:71–79.

Lascelles, B. G., G. M. Langham, R. A. Ronconi, and J. B. Reid(2012). From hotspots to site protection: Identifying marineprotected areas for seabirds around the globe. BiologicalConservation 156:5–14.

LAWA (Land Air Water Aotearoa) (2017). West Coast region:Water Quantity. www.lawa.org.nz/explore-data/west-coast-region/water-quantity/surface-water-zones/

Lewison, R. L., L. B. Crowder, A. J. Read, and S. A. Freeman (2004).Understanding impacts of fisheries bycatch on marinemegafauna. Trends in Ecology & Evolution 19:598–604.

LINZ (Land Information New Zealand) (2015). NZ Coastlines(Topo, 1:50k). LINZ Data Service, Wellington, New Zealand.https://data.linz.govt.nz/layer/258-nz-coastlines-topo-150k/

Loots, C., P. Koubbi, and G. Duhamel (2007) Habitat modelling ofElectrona antarctica (Myctophidae, Pisces) in Kerguelen bygeneralized additive models and geographic informationsystems. Polar Biology 30:951–959.





http://www.doc.govt.nz/parks-and-recreation/places-to-go/west-coast/places/punakaiki-marine-reserve/

http://www.doc.govt.nz/parks-and-recreation/places-to-go/west-coast/places/punakaiki-marine-reserve/

https://CRAN.R-project.org/package=raster



http://www.iucnredlist.org/

http://www.lawa.org.nz/explore-data/west-coast-region/water-quantity/surface-water-zones/

http://www.lawa.org.nz/explore-data/west-coast-region/water-quantity/surface-water-zones/

https://data.linz.govt.nz/layer/258-nz-coastlines-topo-150k/

Magalhaes, M. C., R. S. Santos, and K. C. Hamer (2008). Dual-foraging of Cory’s Shearwaters in the Azores: Feedinglocations, behaviour at sea and implications for foodprovisioning of chicks. Marine Ecology Progress Series 359:283–293.

Mansfield, B. (2006). Law of the Sea: The Benefits for NewZealand. Te Ara The Encyclopedia of New Zealand. http://www.TeAra.govt.nz/en/interactive/6967/new-zealands-exclusive-economic-zone

Marchant, S., and P. J. Higgins (1990). Procellaria westlandicaWestland Petrel. In Handbook of Australian, New Zealand &Antarctic Birds, Volume 1: Ratites to Ducks; Part A: Ratites toPetrels (S. Marchant and P. J. Higgins, Co-ordinators).Melbourne, Oxford University Press, Melbourne, Australia.pp. 566–572.

Marzluff, J. M., J. J. Millspaugh, P. Hurvitz, and M. S. Handcock(2004). Relating resources to a probabilistic measure of spaceuse: Forest fragments and Steller’s Jays. Ecology 85:1411–1427.

Mazerolle, M. J. (2013). AICcmodavg: Model selection andmultimodel inference based on (Q)AIC(c). R package version1.3-5. https://CRAN.R-project.org/package¼AICcmodavg

Ministry for the Environment and Statistics New Zealand (2016).New Zealand’s Environmental Reporting Series: Our MarineEnvironment 2016, Data to 2015. Publication number ME1272. http://www.mfe.govt.nz/sites/default/files/media/Environmental%20reporting/our-marine-environment.pdf

Ministry for Primary Industries (2015). Fisheries AssessmentPlenary May 2015: Stock Assessments and Stock Status.Ministry for Primary Industries, Wellington, New Zealand.http://fs.fish.govt.nz/Page.aspx?pk¼113&dk¼23803

Ministry for Primary Industries (2016). Fisheries AssessmentPlenary May 2016: Stock Assessment and Stock Status: Hoki.Compiled by the Fisheries Science Group, Ministry for PrimaryIndustries, Wellington, New Zealand. https://fs.fish.govt.nz/Doc/24096/32_HOK%20document%20final.pdf.ashx

Mobile Action Technology (2016). i-gotU GT-120 Manual. http://www.manualsworld.net/manual/261538/i-gotu/gt-120

Montevecchi, W. A., A. Hedd, L. McFarlane Tranquilla, D. A. Fifield,C. M. Burke, P. M. Regular, G. K. Davoren, S. Garthe, G. J.Robertson, and R. A. Phillips (2012). Tracking seabirds toidentify ecologically important and high risk marine areas inthe western North Atlantic. Biological Conservation 156:62–71.

Murphy, R. J., M. H. Pinkerton, K. M. Richardson, J. M. Bradford-Grieve, and P. W. Boyd (2001). Phytoplankton distributionsaround New Zealand derived from SeaWiFS remotely-sensedocean colour data. New Zealand Journal of Marine andFreshwater Research 35:343–362.

Nakagawa, S., and H. Schielzeth (2013). A general and simplemethod for obtaining R2 from generalized linear mixed-effects models. Methods in Ecology and Evolution 4:133–142.

Neves, T., and F. Olmos (1998). Albatross mortality in fisheries offthe coast of Brazil. In Albatross Biology and Conservation (G.Robertson and R. Gales, Editors). Surrey Beatty & Sons,Chipping Norton, Australia. pp. 214–219.

Neves, T. S., L. Bugoni, D. S. Monteiro, L. Nascimento, and F.Peppes (2005). Seabird abundance and by-catch on Brazilianlongline fishing fleet. CCAMLR Report WG-FSA-05/67. Com-mission for the Conservation of Antarctic Marine LivingResources, Hobart, TAS, Australia.

NIWA (2016). New Zealand Regional bathymetry. NIWA TaihoroNukurangi, Auckland, New Zealand. www.niwa.co.nz/our-science/oceans/bathymetry

Oliver, E. C. J., S. J. Wotherspoon, M. A. Chamberlain, and N. J.Holbrook (2014). Projected Tasman sea extremes in seasurface temperature through the twenty-first century.Journal of Climate 27:1980–1998.

Palmer, S. (2017). New Zealand’s sea temperature swing largestin world. Newshub, Auckland, New Zealand. www.newshub.co.nz/home/new-zealand/2017/12/new-zealand-s-sea-temperature-swing-largest-in-world.html

Pauly, D., V. Christensen, J. Dalsgaard, R. Froese, and F. Torres, Jr.(1998). Fishing down marine food webs. Science 279:860–863.

Peron, C., K. Delord, R. A. Phillips, Y. Charbonnier, C. Marteau, M.Louzao, and H. Weimerskirch (2010). Seasonal variation inoceanographic habitat and behaviour of White-chinnedPetrels Procellaria aequinoctialis from Kerguelen Island.Marine Ecology Progress Series 416:267–284.

Phillips, R. A., R. Gales, G. B. Baker, M. C. Double, M. Favero, F.Quintana, M. L. Tasker, H. Weimerskirch, M. Uhart, and A.Wolfaardt (2016). The conservation status and priorities foralbatrosses and large petrels. Biological Conservation 201:169–183.

Phillips, R. A., J. C. Xavier, and J. P. Croxall (2003). Effects ofsatellite transmitters on albatrosses and petrels. The Auk 120:1082–1090.

Pinkerton, M. (2016) Ocean colour satellite observations ofphytoplankton in the New Zealand EEZ, 1997–2016. NIWAreport no. WLG2016-19. Prepared for Ministry for theEnvironment by NIWA, Auckland, New Zealand.

Poupart, T. A., S. M. Waugh, C. Bost, C.-A. Bost, T. Dennis, R. Lane,K. Rogers, J. Sugishita, G. A. Taylor, K.-J. Wilson, J. Zhang, andJ. P. Y. Arnould (2017). Variability in the foraging range ofEudyptula minor across breeding sites in central NewZealand. New Zealand Journal of Zoology 44:225–244.

Prince, P. A., N. Huin, and H. Weimerskirch (1994). Diving depthsof albatrosses. Antarctic Science 6:353–354.

QGIS Development Team (2012). QGIS: A free and open sourceGeographic Information System. Open Source GeospatialFoundation (OSGeo). https://qgis.org/

Quillfeldt, P., A. Michalik, G. Veit-Kohler, I. J. Strange, and J. F.Masello (2010). Inter-annual changes in diet and foraging triplengths in a small pelagic seabird, the Thin-billed PrionPachyptila belcheri. Marine Biology 157:2043–2050.

R Core Team (2016). R: A Language and Environment forStatistical Computing. R Foundation for Statistical Comput-ing, Vienna, Austria. http://www.R-project.org/

Rhein, M., S. R. Rintoul, S. Aoki, E. Campos, D. Chambers, R. A.Feely, S. Gulev, G. C. Johnson, S. A. Josey, A. Kostianoy, and C.Mauritzen (2013). Observations: Ocean. In Climate Change2013: The Physical Science Basis. Contribution of WorkingGroup I to the Fifth Assessment Report of the Intergovern-mental Panel on Climate Change (T. F. Stocker, D. Qin, G.-K.Plattner, M. M. B. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y.Xia, V. Bex, and P. M. Midgley, Editors). Cambridge UniversityPress, Cambridge, UK. pp. 255–315.

Richard, Y., and E. R. Abraham (2015). Assessment of the risk ofcommercial fisheries to New Zealand seabirds, 2006–07 to2012–13. New Zealand Aquatic Environment and Biodiversity



http://www.TeAra.govt.nz/en/interactive/6967/new-zealands-exclusive-economic-zone



https://CRAN.R-project.org/package=AICcmodavg

https://CRAN.R-project.org/package=AICcmodavg

http://www.mfe.govt.nz/sites/default/files/media/Environmental%20reporting/our-marine-environment.pdf

http://www.mfe.govt.nz/sites/default/files/media/Environmental%20reporting/our-marine-environment.pdf

http://fs.fish.govt.nz/Page.aspx?pk=113&dk=23803



https://fs.fish.govt.nz/Doc/24096/32_HOK%20document%20final.pdf.ashx

https://fs.fish.govt.nz/Doc/24096/32_HOK%20document%20final.pdf.ashx

http://www.manualsworld.net/manual/261538/i-gotu/gt-120

http://www.manualsworld.net/manual/261538/i-gotu/gt-120

http://www.niwa.co.nz/our-science/oceans/bathymetry

http://www.niwa.co.nz/our-science/oceans/bathymetry

http://www.newshub.co.nz/home/new-zealand/2017/12/new-zealand-s-sea-temperature-swing-largest-in-world.html



https://qgis.org/

http://www.R-project.org/

Report 162. Ministry for Primary Industries, Wellington, NewZealand.

Robertson, D. A., and K. D. Clements (2015). Family Myctophidae.In The Fishes of New Zealand (C. D. Roberts, A. L. Stewart, andC. D. Struthers, Editors). Te Papa Press, Wellington, NewZealand. pp. 613–680.

Robertson, H. A., K. Baird, J. E. Dowding, G. P. Elliott, R. A.Hitchmough, C. M. Miskelly, N. McArthur, C. F. J. O’Donnell, P.M. Sagar, R. P. Scofield, and G. A. Taylor (2017). Conservationstatus of New Zealand birds, 2016. New Zealand ThreatClassification Series 19. Department of Conservation, Wel-lington, New Zealand. http://www.doc.govt.nz/Documents/science-and-technical/nztcs19entire.pdf

Rolland, V., H. Weimerskirch, and C. Barbraud (2010). Relativeinfluence of fisheries and climate on the demography of fouralbatross species. Global Change Biology 16:1910–1922.

Shoji, A., S. Aris-Brosou, A. Fayet, O. Padget, C. Perrins, and T.Guilford (2015). Dual foraging and pair coordination duringchick provisioning by Manx Shearwaters: Empirical evidencesupported by a simple model. Journal of ExperimentalBiology 218:2116–2123.

Small, C., S. M. Waugh, and R. A. Phillips (2013). The justification,design and implementation of Ecological Risk Assessments ofthe effects of fishing on seabirds. Marine Policy 37:192–199.

Stahl, J. C., and P. M. Sagar (2000a). Foraging strategies andmigration of Southern Buller’s Albatrosses Diomedea b. bulleribreeding on the Solander Is, New Zealand. Journal of theRoyal Society of New Zealand 30:319–334.

Stahl, J. C., and P. M. Sagar (2000b). Foraging strategies ofSouthern Buller’s Albatrosses Diomedea b. bulleri breeding onThe Snares, New Zealand. Journal of the Royal Society of NewZealand 30:299–318.

Taylor, G. A. (2000). Action Plan for Seabird Conservation in NewZealand. Part A: Threatened Seabirds. Threatened SpeciesOccasional Publication No. 16. Department of Conservation,Wellington, New Zealand. http://www.doc.govt.nz/documents/science-and-technical/TSOP16.pdf

Techow, N. M. S. M., C. O’Ryan, C. J. R. Robertson, and P. G. Ryan(2016). The origins of White-chinned Petrels killed by long-line fisheries off South Africa and New Zealand. PolarResearch 35:art.21150.

Techow, N. M. S. M., P. G. Ryan, and C. O’Ryan (2009).Phylogeography and taxonomy of White-chinned andSpectacled petrels. Molecular Phylogenetics and Evolution52:25–33.

Thaxter, C. B., B. Lascelles, K. Sugar, A. S. C. P. Cook, S. Roos, M.Bolton, R. H. W. Langston, and N. H. K. Burton (2012). Seabirdforaging ranges as a preliminary tool for identifyingcandidate Marine Protected Areas. Biological Conservation156:53–61.

Thiebot, J.-B., A. Lescroel, D. Pinaud, P. N. Trathan, and C.-A. Bost(2011). Larger foraging range but similar habitat selection innon-breeding versus breeding sub-Antarctic penguins. Ant-arctic Science 23:117–126.

Thompson, K. R. (1992). Quantitative analysis of the use ofdiscards from squid trawlers by Black-browed AlbatrossesDiomedea melanophris in the vicinity of the Falkland Islands.Ibis 134:11–21.

Thorne, L. H., M. G. Conners, E. L. Hazen, S. J. Bograd, M. Antolos,D. P. Costa, and S. A. Shaffer (2016). Effects of El Nino-driven

changes in wind patterns on North Pacific albatrosses.Journal of The Royal Society Interface 13:20160196.

Torres, L. G., D. R. Thompson, S. Bearhop, S. Votier, G. A. Taylor, P.M. Sagar, and B. C. Robertson (2011). White-cappedAlbatrosses alter fine-scale foraging behavior patterns whenassociated with fishing vessels. Marine Ecology ProgressSeries 428:289–301.

Tuck, G. N., T. Polacheck, J. P. Croxall, and H. Weimerskirch(2001). Modelling the impact of fishery by-catches onalbatross populations. Journal of Applied Ecology 38:1182–1196.

Wakefield, E. D., T. W. Bodey, S. Bearhop, J. Blackburn, K.Colhoun, R. Davies, R. G. Dwyer, J. A. Green, D. Gremillet, A. L.Jackson, M. J. Jessopp, et al. (2013). Space partitioningwithout territoriality in gannets. Science 341:68–70.

Wakefield, E. D., R. A. Phillips, and J. Matthiopoulos (2009).Quantifying habitat use and preferences of pelagic seabirdsusing individual movement data: A review. Marine EcologyProgress Series 391:165–182.

Walker, K., and G. Elliott (2006). At-sea distribution of Gibson’sand Antipodean Wandering Albatrosses, and relationshipswith longline fisheries. Notornis 53:265–290.

Warham, J. (1990). The Petrels: Their Ecology and BreedingSystems. Academic Press, London, UK.

Waugh, S. (2012). Westland Petrels circumnavigate South Island.Museum of New Zealand Te Papa Tongarewa, Wellington,New Zealand. http://blog.tepapa.govt.nz/2012/09/21/westland-petrels-circumnavigate-south-island/

Waugh, S. M., and J. A. Bartle (2013). Westland Petrel. In NewZealand Birds Online (C. M. Miskelly, Editor). http://www.nzbirdsonline.org.nz/species/westland-petrel

Waugh, S. M., and K.-J. Wilson (2017). Threats and threat statusof the Westland Petrel Procellaria westlandica. MarineOrnithology 45:195–203.

Waugh, S. M., C. Barbraud, L. Adams, A. N. D. Freeman, K.-J.Wilson, G. Wood, T. J. Landers, and G. B. Baker (2015a).Modeling the demography and population dynamics of asubtropical seabird, and the influence of environmentalfactors. The Condor: Ornithological Applications 117:147–164.

Waugh, S. M., D. P. Filippi, D. S. Kirby, E. Abraham, and N. Walker(2012). Ecological Risk Assessment for seabird interactions inWestern and Central Pacific longline fisheries. Marine Policy36:933–946.

Waugh, S. M., T. Poupart, and K.-J. Wilson (2015b). Stormdamage to Westland Petrel colonies in 2014 from cyclone Ita.Notornis 62:165–168.

Weichler, T., S. Garthe, G. Luna-Jorquera, and J. Moraga (2004).Seabird distribution on the Humboldt Current in northernChile in relation to hydrography, productivity, and fisheries.ICES Journal of Marine Science 61:148–154.

Weimerskirch, H. (1998). How can a pelagic seabird provision itschick when relying on a distant food resource? Cyclicattendance at the colony, foraging decision and bodycondition in Sooty Shearwaters. Journal of Animal Ecology67:99–109.

Weimerskirch, H., N. Brothers, and P. Jouventin (1997a).Population dynamics of Wandering Albatross Diomedeaexulans and Amsterdam Albatross D. amsterdamensis in theIndian Ocean and their relationships with long-line fisheries:



http://www.doc.govt.nz/Documents/science-and-technical/nztcs19entire.pdf

http://www.doc.govt.nz/Documents/science-and-technical/nztcs19entire.pdf

http://www.doc.govt.nz/documents/science-and-technical/TSOP16.pdf

http://www.doc.govt.nz/documents/science-and-technical/TSOP16.pdf

http://blog.tepapa.govt.nz/2012/09/21/westland-petrels-circumnavigate-south-island/

http://blog.tepapa.govt.nz/2012/09/21/westland-petrels-circumnavigate-south-island/

http://www.nzbirdsonline.org.nz/species/westland-petrel

http://www.nzbirdsonline.org.nz/species/westland-petrel

Conservation implications. Biological Conservation 79:257–270.

Weimerskirch, H., Y. Cherel, F. Cuenot-Chaillet, and V. Ridoux(1997b). Alternative foraging strategies and resource alloca-tion by male and female Wandering Albatrosses. Ecology 78:2051–2063.

Weimerskirch, H., D. P. Filippi, J. Collet, S. M. Waugh, and S. C.Patrick (2018). Use of radar detectors to track attendance ofalbatrosses at fishing vessels. Conservation Biology 32:240–245.

Weimerskirch, H., M. Le Corre, E. Tew Kai, and F. Marsac (2010).Foraging movements of Great Frigatebirds from Aldabra

Island: Relationship with environmental variables and inter-action with fisheries. Progress in Oceanography 86:204–213.

Weimerskirch, H., M. Louzao, S. de Grissac, and K. Delord (2012).Changes in wind pattern alter albatross distribution and life-history traits. Science 335:211–214.

Wood, A., B. Naef-Daenzer, P. A. Prince, and J. P. Croxall (2000).Quantifying habitat use in satellite-tracked pelagic seabirds:Application of kernel estimation to albatross locations.Journal of Avian Biology 31:278–286.

Worton, B. J. (1989). Kernel methods for estimating theutilization distribution in home-range studies. Ecology 70:164–168.

APPENDIX TABLE 3. Candidate sets of models used to examine Westland Petrel foraging intensity (kernel utilization distribution[KUD]). Models were run in 2 stages. Stage 1 models included geographic, anthropogenic, and environmental parameters. Stage 2extended the highest-ranked model from stage 1 (model 12) to test the parameters of breeding stage (default level¼ prebreeding),sex, and foraging distance traveled. Band ID was included in all models as a random effect. Fishing area¼ an index representing azone of strong trawling activity and marine productivity, Fishing activity¼ an index representing fishing events that occurred duringa petrel foraging trip, SOI ¼ Southern Oscillation Index, Chl-a ¼ chlorophyll-a concentration, and Light regime ¼ daytime vs.nighttime (default level ¼ day).

Model number Explanatory variables

Stage 11 Latitude þ Longitude þ (1jBand ID)2 Latitude þ Longitude þ Bathymetric slope þ (1jBand ID)3 Latitude þ Longitude þ Fishing area þ (1jBand ID)4 Latitude þ Longitude þ Fishing activity þ (1jBand ID)5 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ (1jBand ID)6 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI þ (1jBand ID)7 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ Chl-a þ (1jBand ID)8 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ Light regime þ (1jBand ID)9 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI þ Chl-a þ (1jBand ID)10 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ Chl-a þ Light regime þ (1jBand

ID)11 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ Light regime þ SOI þ (1jBand

ID)12 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI*Chl-a þ Light regime þ

(1jBand ID)Stage 2

13 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI*Chl-a þ Light regime þDistance traveled þ (1jBand ID)

14 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI*Chl-a þ Light regime þ Sexþ (1jBand ID)

15 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI*Chl-a þ Light regime þBreeding stage þ (1jBand ID)

16 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI*Chl-a þ Light regime þDistance traveled þ Sex þ (1jBand ID)

17 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI*Chl-a þ Light regime þ Sexþ Breeding stage þ (1jBand ID)

18 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI*Chl-a þ Light regime þBreeding stage þ Distance traveled þ (1jBand ID)

19 Latitude þ Longitude þ Fishing activity þ Fishing area þ Bathymetric slope þ SOI*Chl-a þ Light regime þDistance traveled þ Sex þ Breeding stage þ (1jBand ID)



Environmental factors and fisheries influence the foraging...

Documents

Transcript of Environmental factors and fisheries influence the foraging...