EGG NEGLECT AND ï ï S IMPLICATIONS FOR EGG PREDATION

IN THE RHINOCEROS AUKLET (Cerorhinca monocerata)

Louise K. Blight

B.A., University of Victoria, 1995

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE

in the Department

of

Biological Sciences

O Louise K. BLight 2000

SIMON FRASER UNIVERSITY

July 2000

All rights reserved. This work may not be reproduced in whole or in part, by photocopy

or other means, without permission of the author.

Acquisitions and Acquisitions et Bibliographic Services services bi biiiraphiques 395 WeUington Street 395. nie Wellinglm Onawa ON KlA ON4 -ON K l A M Canada CaMda

The author has granted a non- exclusive licence allowing the National Library of Cariada to reproduce, loan, distniiiie or seîi copies of this thesis in microform, paper or electronic formats.

The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fiom it may be pruited or otherwise reproduced without the author's permission.

L'auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du C d de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/nlm, de reproduction sur papier ou sur format électronique.

L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de ceUe-ci ne doivent être imprimés ou autrement reproduits sans son autorisation.

Depredation of island-nesting seabirds by introduced vertebrates is a

conservation problem world wide, particularly as marine birds have generally

evolved in the absence of terresmal predators. In conhast, at island colonies where

seabirds have CO-evolved with native predators, the evolution of defensive

behaviours is expected. For breeding seabirds, the primary predation risk is often

incurred by their eggs and chi&. For my thesis research, 1 studied egg predation

and parental incubation behaviour in Rhinoceros Auklets (Cerorhinca

rnonocera ta) at the seabird colony on Triangle Island, British Columbia (51°52'N,

129O05'W). Native deer mice (Peromyscus keeni) opened and ate the eggs in up to

34% of monitored auklet burrows. 1 used stable isotope analyses of mice and their

potential prey items to determine the role of seabird eggs in rodent diets. 613C and

615N values of mouse b e r and muscle tissues suggested that for Triangle Island

Peromyscus, seabird eggs are a primary source of protein during spring and

summer. Although mice c m only take eggs when incubating birds are absent,

parents often took lengthy recesses (~18 h) from their nests. Incidents of parental

egg neglect were greater in 1998, a year of poor at-sea food availability, than they

were in 1999, indicating that lower food supply may lead to more frequent neglect.

Incubation behaviour by breeding Rhinoceros Auklets did not appear to take the

risk of egg predation into account, supporting a life history interpretation. In both

years of my shidy, incidents of egg neglect were more frequent early in incubation.

Previous research suggests that this pattern tracks the more stringent

iii

developmental requirements of older embryos. 1 experimentally chilied auklet eggs

for 48 h either early or late in incubation to test this hypothesis. Hatchability of

experimental eggs did not appear to be affected by embryo age at date of chilling.

Observed nest attentiveness patterns in Rhinoceros Auklets are thus more likely

explained by increasing parental investment in the embryo over üme, or as a

response to the loss of adult body reserves during courtship and egg formation.

Dedication

For Peter

This research was funded by an NSERC post-graduate scholarship, a Simon

Fraser University Graduate Fellowship, and an award from the John K. Cooper

Foundation. NSERC operating grants to Tony Williams and Fred Cooke and the

Nestucca Oil Spill Trust Fund all assisted in supporting my work in the field.

Permission to carry out research at the Anne Vallée Ecological Resewe (Triangle

Island) was granted by BC Parks, and pilots at West Coast Helicopter flew crew and

supplies safely to and from the island, often under the most marginal of weather

conditions.

1 would like to th& my cornmittee, Tony Williams and Doug Bertram, for

many hours of conversation, advice, and feedback. Their diverse and

complementary areas of expertise made them idea CO-supervisors. Alton Harestad

assisted considerably in the initial design of my study, and Keith Hobson's

correspondence clarified some of the finer points of stable isotope analysis. Former

Triangle Island researchers Hugh Knechtel and Laura Jones provided me with ideas

and reality checks prior to my first field season. Chris Guglielrno, Tom Chapman

and other lab mates in the Centre for Wildlife Ecology and the Pink Lab provided

ideas, stories, and perspective. 1 would particularly like to thank John Ryder,

Carina Gjerdrum, and the rest of the field crews at Triangle Island for their wami

companionship and esoteric conversations. Finally, special thanks to Boyd Pyper

for his meticulous approach to data collection, continued moral support, and many

hours of document formatting.

Table of Contents

. . Approval ................... ... ......................................................................................................... 11

... Abstract .......................... ,. ....................................................................................................... iii

Dedication ................................................................................................................................... v

Acknowledgments ........................ .... .......................................................... .. ........... v i

. . Table of Contents ...................................... ...... ........................... vil

List of Tables ............................................................................................................................. x

List of Figures ............................................................................................................................ xi

1 . General introduction ............................................................................................................ 1

Biology of the Rhinoceros Auklet ............................................................................. 3

Distribution and population status .............................................................. 3

Breeding biology ............................................................................................. 3

Biology of Perornyscz~s keeni and Microtus townsendii cowani ........................ 4

Populations' origins and taxonomy .............................................................. 4

Breeding biology ................................................................................................ 5

Diet .................................. .. ................................................................................... 6

............. Study area ............................................................................................... .. 8

Objectives ................................................... .... .......................................................... 10

Note on the chapters as publishable papers ......................................................... 11

II . Predation on NUnoceros Auklet eggs by a native population of Peromyscus ..... 12

Abstract .......................................................................................................................... 12

Introduction ................................................................................................................. 12

vii

...................................................... .....................*......*....................* Methods ......, 1 3

Results ............................................................................................................................ 16

Discussion ..................................................................................................................... 20

III . Predation on seabird eggs by Keen's mice (Peromyscus keeni): using stable

isotopes to deupher the diet of a terrestrial omnivore on a remote

......................................................... offshore island 2 6

Abstract .......................................................................................................................... 26

Introduction ........... .. ........................................................................................... 2 7

............................................................................................. Study Site and Methods 30

Study site ....................... ... ................................................................... 3 0

Sampluig rodents and Likely prey types ...................................................... 31

............................................................................................ Isotopic analyses 3 2

S tatistical analyses of isotopic signals .............. .... ................................... 3 3

................................................................. Mu1 tiple-Source Mixing Mode1 3 5

Results .................................. ... ............................ 36

Sample collection .........................................................................................-.. 36

Isotope Analyses of Prey Types .............. .... ............................................ 38

Isotope Analyses of Rodents ........................... ... .................................. 4 1

Discussion ..................................................................................................................... 48

Isotope Analyses of Prey Types ..................................................................... 48

Seabird eggs and rodents on Triangle Island .......................................... 50

Implications for rodent populations ........................... ... ............................. 52

viii

TV . Egg neglect in a member of the Alcidae: an experimental approach .................... 55

Abstract ....................... .... ..... ,... ................................................................. 5 5

Introduction ................................................................................................................ 5 6

Study site and rnethods .......................................................................................... 58

Results .......................... .. .............................................................................................. 62

Discussion ................................. .... ......................................................................... 63

V . Interannual variation in egg neglect and incubation routine of the

...................................................................................................... Rhinoceros Auklet 68

Introduction ............................................................................................................... 68

Study site and methods .............................. ... ............................................................. 70

S ta tis tical analyses ......................................... .... .............................................. 73

Results .......................................................................................................................... 74

..................................................................................................................... Discussion 82

......................................................................................................................... VI . Conclusion 89

Literature Cited ...................................................................................................................... 93

List of Tables

Table 1-1. Age class distribution of Peromyscus keeni trapped at Triangle

Island, British Columbia, 1998-1999 ..........-.. ......~................................................ 7

Table 2-1. Causes of Rhinoceros Auklet (Cerorhinca monocerata) egg loss at

Triangle Island, British Columbia, 1998 .............................~................................ 17

Table 3-1. Morphometrics and population characteristics of Keen's mice

(Peromysczis keeni) and Townsend's voles (Micro tus towsendii

.............................. cowani) from Triangle Island, British Columbia, 1997-1998 37

Table 3-2. Relative contributions (%) of prey types to diet of Keen's mice

(Peromysczis keeni) and Townsend's voles (Microtus towsendii

............................. cowani) on Triangle Island, British Columbia, 1997-1998 46-7

......... Table 5-1. Results for incubated artificial Rhinoceros Auklet eggs, 1998-1999.. .75

List of Figures

Figure 2-1. Mouse-depredated Rhinoceros Auklet (Cerorhinca monocerata)

.............................................................................................................................. egg 1 9

Figure 2-2. Keen's mouse (Peromyscus keeni) opening intact Rhinoceros

Auklet (Cerorhinca monocerata) egg, Triangle Island, British Columbia ...... 23

Figure 3-1. Stable carbon and nitrogen isotope values ( % O ) for prey types

available to Keen's mice (Peromyscus keeni) and Townsend's voles

(Microtus fownsendii cowani), Triangle Island, British Columbia .................. 39

Figure 3-2. Mean stable carbon and nitrogen isotope values ( O h I SD) for

......................................................................................................... sample prey types 40

Figure 3-3. Stable carbon and nitrogen concentrations in Liver tissues of

Keen's mice (Peromyscus keeni) and Townsend's voles (Microtus

townsendii cowani) coilected at Triangle Island, British Columbia,

...................... 1997-1998 ............................................................................................. .. 42

Figure 3-4. Stable carbon and nitrogen concentrations in muscle tissues of

Keen's mice (Perornyscus keeni) and Townsend's voles (Microtus

townsendii cowani) collected on Triangle Island, British Columbia,

1997-1998. ....................................................................................................................... 44

................ Figure 5-1. Temperature profile of an incubated artificial auklet egg, 1998 76

................ Figure 5-2. Temperature profile of an incubated artificial auklet egg, 1999 7

Figure 5-3. Distribution of neglect events in relation to stage of incubation,

Figure 5.4 . Temperature profile of an incubated artificial auklet egg. showing

increase in incubation temperatures over time ................................................... 81

Figure 5.5 . Growth rate anomalies for Rhinoceros Auklet (Cerorhinca

monocerata) chicks, Triangle Island, British Columbia, 1994-1999 .................. 83

xii

Chapter 1

General introduction

All birds are subject to some degree of predation, but whether or not their

populations are limited by predators or by some other factor depends upon a

number of variabies. Even the loss of a substantial proportion of a population may

not cause a change in breeding numbers if the mortality is compensatory (Le.,

replacing other sources of mortality, such as starvation) rather than additive

(adding to other sources of mort&% Newton 1998). Experimental studies that

have removed or excluded predators from the habitat of their avian prey have

appeared to confirm that in some situations, predation alone may limit

populations (e.g., Byrd et al. 1994), while in others it does not (Parr 1993, Garrettson

et al. 1996).

Introduced predators on islands have provided some of the most dramatic

examples of how predators may depress the population of an avian prey species,

sometimes to the point of extinction (Moors and Atkinson 1984, Johnstone 1985).

In general, this occurs because island-nesting birds have evolved in the absence of

terreshial predators, and thus often nest in accessible locations and/or tend to la&

defensive behaviour (Newton 1998). However, in the few island systems where

endemic terrestrial predators and ground-nesting birds have CO-evolved, predators

may not play the same regdatory role, likely because breeding birds have retained a

suite of behavioural traits that act to deter predation or othenvise mitigate agahst

its effects.

I conducted m y thesis research at the seabird colony at Triangle Island,

British Columbia. As suggested above, seabird colonies are generally free of

endemic terrestial predators, and indeed, island nesting in marine birds is thought

to have evolved due to the absence of predators in these environments (Furness

and Monaghan 1987). At Triangle Island, however, two native species of potential

rodent predators also occur: Keen's deer mouse (Peromyscus keeni) and an

endemic subspecies of Townsend's vole (Microtirs fownsendii cowani; Car1 et al.

1951, Hogan et al. 1993). For my thesis research, 1 wished to examine whether

predation by either or both of these rodent species was a sipificant source of egg

mortality for Rhinoceros Auklets (Cerorhinca monocerufa) breeding at Triangle

Island. 1 also wanted to determine whether parental behaviour during incubation

rnight effectively reduce egg depredation, as might be expected due to CO-evolution

of predator and potential prey. However, life history theory predicts a trade-off

between current and future reproductive potential in order for a parent to

maximize its lifetime reproductive output (Williams 1966, Stearns 1992). If

repelling a predator of egg or chi& were too costly, a parent would be predicted to

act to ensure its own sumival rather than that of its offspring. I will present the

objectives of my research project in detail at the end of this section, but first 1 will

describe the biology of the study species and the relevant features of the study area.

Biology of the Rhuioceros Adciet

Distribtition and vovulation status

The Rhinoceros Auklet is a medium-sized member (ca. 500 g) of the Family

Alcidae. It breeds on selected coastal islands throughout the temperate North

Pacific, with colonies located as far south as central California, and north and

westwards from that latitude through the Aleutians to the northem coast of Asia.

In British Columbia, as in Oregon and California, the breeding population appears

to have undergone a recent range expansion. In the 1940s, Rhinoceros Auklets

were known to breed at only three localities along the BC coast (Munro and

McTaggart-Cowan 1947); more recent estimates increase the number of known

breeding sites in the province to at least 30 (Campbell et al. 1990). Although at least

some of this apparent expansion may be attributed to increased human presence

and observer effort along the British Columbia coast, the growth in breeding

numbers has been well-documented at Triangle Island, the site of this study. No

Rhinoceros Auklets were found breeding at the island's seabird colony in 1949 or

1950, despite extensive biological surveys there (Car1 et al. 1951); however, the most

recent estimates suggest a breeding population of approximately 42,000 pairs

(Rodway 1991).

Rhinoceros Auklets nest in long (1-5 m) burrows that they excavate amongst

the ground vegetation in the colony. A single egg is laid in the nest chamber

between mid-April and early June (Gaston and Dechesne 1996), and is incubated by

3

both parents in shifts lasting approximately 24 h. As with many other species of

offshore-feeding seabirds, incubation may sometimes be intermittent, with both

parents away from the nest for a day or more. These lengthy periods of absence,

termed egg neglect (Boersrna and Wheelwright 1979), serve to increase parental

foraging time when food supplies are scarce or widely-scattered (La& 1967). For

NUnoceros Auklets and many other speaes of seabird, the embryo appears resistant

ro chilhg, and egg neglect does not seem to affect hatdiability (e-g., Wilson and

Manuwal 1986, Boersma and Whedwright 1979, Murray 1980). As in other alcids

(Murray et al. 1983, Astheimer 1991), egg neglect in Rhinoceros Auklets is much

more frequent, and of longer duration, early in incubation (pers. obs.; Wilson 1977).

For Cassin's Auklets (Ptychoramphus alenticus), as for other species of birds, it has

been suggested that high incidences of egg neglect early in incubation correspond

with the period of slowest development of the embryo (Webb 1987, Astheimer

1991).

Biology of Peromyscus keeni and Microtus townsendii cowani

Pciaulations' oriqins and taxonomu

The Keen's mice (also hown as Keen's deer mice) and Tomsend's voles a

occurring on Triangle Island are likely relict populations that originated 11-10 ka,

when large areas of British Columbia's continental shelf were above sea level

(Lutemauer et al. 1989). Triangle Island's deer mice were previously accorded sub-

recent DNA work has suggested that the mice of the Scott Island group (see below)

are conspecifics with those on neighbouring areas of northwestem Vancouver

Island (Hogan et al. 1993). Ongoing DNA analyses of mice collected by me on

Triangle Island in 1998-9 (initially for stable isotope analysis; see Chapter 3) will

likely soon confirm this taxonomic status (I.F. Greenbaum, Texas A&M University,

pers. comrn.). The local sub-species of Townsend's vole is not well-studied (BC

Conservation Data Centre, pers. comm.) and nor are the island's Peromyscas, but

like island races of rodents elsewhere (Redfield 1975, Adler and Levins 1994), both

mice and voles at Triangle Island exhibit larger body size than do mainland

populations, and indeed, are some of the largest specimens known for these speues'

ranges (Carl et al. 1951, Cowan and Guiguet 1975; see also Chapter 3, Table 3-1).

As the rodents on Triangle Island exhibit traits cornmon to island rodents in

general, such as gigantism and apparent longevity (Carl et al. 1951), it is likely that

their breeding biology also follows trends described for island rodents elsewhere

(Adler and Levins 1994). For Triangle Island Peromyscus, this means that litters are

likely smaller than they would be for mainland populations and that sexual

maturation is delayed, in response to reduced mortality (Adler and Levins 1994).

While this thesis was not a study of Keen's mice per se, 1 live-trapped mice

(following Sullivan 1997) in order to determine whether changing rates of egg

predation could be correlated with changes in the mouse population at Triangle

Island. Based on live-trapping results, 1 obtained population densities (as per

Dreman et al. 1998) for the late spMg and early summer, and data on intra- and

inter-annual changes in the age structure of the population (Table 1-1). Three age

classes were determined based upon data for late May (the first trapping period) of

both study years, when there was a clear distinction between the weight of adult

mice (36-60 g) and that of juveniles ( ~ 2 0 g). Sub-adult mice (20-35 g) did not appear

in our samples until the second trapping period (late June) of either year of thk

Live-trapping data (Table 1-1) indicate that, in some years at least, juvenile

mice do not comprise an appreuable proportion of the population until after the

beginnùig of June. It seems reasonable to suggest that Peromyscus may commence

breeding on Triangle Island at a later date than do conspecifics on the mainland,

and the timing of availability of seabird eggs may play a role in determining onset

of mouse breeding at a locality where other foods remain scarce until relatively late

in the season (Chapter 1).

Omnivorous deer mice are well-known predators of songbird eggs (Verbeek

1970, Guillory 1987, Rogers et al. 1997), but several studies have previously been

interpreted to indicate that mice are gape-limited and cannot open eggs larger than

those of a small passerine (Roper 1992, Haskell1995, DeGraaf and Maier 1996, Maier

and DeGraaf 2000). Deer mice have nonetheless been obsewed taking eggs at the

colonies of at least two seabird species (Murray 1980, Gaston 1992). One of these

earlier studies found that the rate of egg predation in a California Xantus' Murrelet

Table 1-1. Age dass distribution of Peromyscus keeni trapped at Triangle Island, 1998-1999. Values are n (%).

Age dass of ïive- 28 May - 30 June - 29 May - 02 July - trapped mice 01June1998 01July1998 O1 lune 1999 04 July 1999

juveniles 1 (2) 8 (17) 17 (32) 2 (7)

sub-adult O 16 (34) O 19 (63)

adult 57 (89) 22 (47) 36 (68) 8 (27)

unassessed (escapedi 6 (9) 1 (2) O 1 (3)

(Synthliboramphus hypoleucas) colony varied with vegetation cover, Le.,

availability of altemate food sources (Murray 1980). Çeeds appeared to be preferred

over eggs where large amounts of seed were available, but murrelet eggs were

readily chewed open and eaten where plant foods were not as abundant-

(Laboratory trials also indicate that deer mice may prefer energy-rich, high-

carbohydrate foods to high-protein foods, if given a choice (Vickery et al. 1994)).

Although voles are known herbivores, their predation on eggs and young

chicks has occasionally been recorded for both marine and terrestrial birds (Sealy

1982, Bures 1997). This addition of vertebrate foods to their diet appears to be

virtually unknown (cf. Batzli 1985), but at the onset of this study 1 did consider

voles to be potential egg predators at Triangle Island due to the paucity there of

herbaceous vegetation - and the possibility of resulting nutritional stresses - for

much of the year.

Study area

Triangle Island (51°52'N, lî9OO5 W; 144 ha) and its associated islets are

situated approximately 46 km (25 NM) offshore, at the western extreme of the Scott

Island group. This diain of rocky islands stretches in a north-westerly direction

from Cape Scott, at the northern end of Vancouver Island, British Columbia.

Triangle Island is located on the edge of the cold, oceanic upwellings associated

with the continental shelf break (Kaiser and Forbes 1992) and is Western Canada's

largest seabird colony, with a breeding population of approximately 1.2 million

birds, primarily Rhinoceros and Cassin's Auklets (Rodway et al. 1990).

Fieldwork took place from mid-April to early July, 1998-1999. The research

descnbed here was carrïed out in the vicinity of the Simon Fraser

University/Centre for Wildlife Ecology research cabin at South Bay. Burrow

density for Rhinoceros Auklets in South Bay is high, although Cassin's Auklets

also breed there in low numbers. In 1998, the first year of this project, 1 established

two study plots; these sarne two plots were also used in 1999. Plot 1 was situated

approximately 30 m from the research cabin, and Plot 2 lay about 300 m to the east.

Both plots were located at sirnilar heights above sea level to control for suspected

changes in mouse population with elevation (Car1 et al. 1951).

My a priori expectation was that predation rate would Vary with vegetation

cover, as has been found elsewhere (see above). 1 chose two shidy plots that varied

in this respect in order to test my assumption. The two most abundant plant

species in both plots were tufted hair grass (Deschampsia cespitosa) and

salmonberry (Rzibzis spectabilis), but Plot 1 differed sigruficantly from Plot 2 in the

mean percent cover of Deschampsia (based on four quadrats per study area; 78.0%

vs. 28.25% for Plots 1 and 2, respectively; t5 = 4.13, P~0.01). In Plot 1, this

Deschampsin-Ru bus association dominated the plant community, with other

(primarily herbaceous) plants showing 1.25% cover by species. In Plot 2, however,

sala1 (Gaziltheria shallon; absent from Plot 1) was the third most commonly

occurring plant species (11.25% cover) in a Deschampsia-Rubus-Gaziltheria-fern

(primarily lady fern Athyrium felix-femina) association. The mean percent cover

of Rzibus was variable for Plots 1 and 2 (ll.Ooh vs. 43.7S0h, respectively), but did not

differ significantly between them (t4 = -2.27, P = 0.09). I chose vegetation plot size

9

based upon the type of plant community present, following Mueller-Dombois and

Eilenberg (1974).

Objectives

As outlined above, the overall objectives of my thesis were firstly to

investigate whether or not predation by native rodents is a sigruficant source of egg

mortality for Rhhoceros Auklets breeding on Triangle Island. Secondly, my

research aimed to determine whether parental behaviour by incubating Rhinoceros

Auklets (e-g., egg neglect) served to reduce - or to facilitate - egg depredation.

Timing of egg neglect by incubating birds is hypothesized to be constrained by

developmental requirements of the embryo (Webb 1987); 1 also wished to assess

whether observed patterns of egg neglect were more readily explaîned by this

hypothesis, or by some other factor or combination of factors that might or might

not include parental assessment of egg predation risk.

The specific objectives of my thesis research were as follows:

1) To quantify the effects of egg depredation by native rodents on Rhinoceros

Aukle t breeding productivity at Triangle Island;

2) To use stable isotope analysis of mouse and vole tissue in order to

determine which (if any) of these two rodent species utilized auklet eggs as food,

and to assess whether there were inter-individual or regional differences in rodent

use of seabird eggs as prey;

3) To determine whether relationships existed between egg predation and

other temporal or spatial factors: vegetation cover; density of rodent populations;

stage of incubation; and whether rates of predation showed inter-annual variation

related to these or to other factors;

4) To determine whether Rhinoceros Auklets facultatively adjusted their

periods of egg neglect in accordance with perceived predation risk; and,

5) To apply a life history approach in order to test the assumption that

observed egg neglect patterns in the Rhinoceros Auklet are adaptive, Le., that these

patterns allow parents to increase foraging time at sea during poor food conditions,

or when parental body condition warrants it, while not comprornising embryo

survival.

Note on the chapters as publishable papers

Each main chapter (2-5) of this thesis has been written as a discrete

manuscript intended for journal publication. As senior author on Chapters 2,4

and 5,1 did al1 research, analyses and writing as work contributing to my thesis

requirements, while other authors reviewed and comrnented on these papers.

Chapter 3 was a joint effort between me and the senior author, Mark Drever. Mark

began data collection for this work in 1997 and 1 took it over as part of my thesis

research in 1998. We both agree that we have contributed equally to the

manuscript. As senior author, Mark was responsible for the first draft and the

analysis of stable isotope data, whiie I wrote ail subsequent drafts. We collaborated

on data collection and sample preparation, discussions of how best to present the

data and resuits, and rebuttals and revisions deriving from the peer review process

at Canadian Journal of Zoology.

Chapter 111

Predation on Rhinoceros Auklet eggs by a native population of

Peromyscus

Abs trac t

Predation by Keen's mice (Peromyscus keeni) was the single greatest cause of

egg loss for Rhinoceros Auklets (Cerorhincn monocerata) at the seabird colony on

Triangle Island, British Coiumbia in 1998. Despite studies suggesting that gape-

limited rodents are unable to open large eggs, mouse depredation was likely

responsible for the loss of more eggs than al l other causes combined, with mice

commonly opening and eating eggs of nearly twice their mass. In one study plot,

mice depredated up to 34% of eggs. This high predation rate may be related to

temporary egg neglect by foraging parents. We suggest that egg depredation may

increase in years of low marine productivity, when adults increase foraging t h e .

Key words: Cerorhinca monocerata; Perornyscus; egg predation; egg neglect,

Triangle Island; Rhinoceros Auklet; Keen's mouse.

Introduction

Depredation by deer mice (Perornyscus spp.) is a widely recognized source of

egg mortality for passerines (Verbeek 1970, Guillory 1987, Rogers et al. 1997), and

also has been documented for ground-nesting shorebirds (Maxson and Oring 1978).

This chapter has been published as Blight, LX., J.L. Ryder and D.F. Bertram. 1999. Predation on Rhinoceros Auklet eggs by a native population of Peromysctis. Condor 101:87-876.

Marine birds nestïng on remote islands largely avoid terrestrial predators. Indeed,

predator avoidance may have been importa. in the evolution of island nesüng

(Fumess and Monaghan 1987). To date, egg predation by Peromyscus has been

documented for only two burrow-nesting seabird speües (Murray 1980, Gaston

1992). In this paper we report depredation by native Keen's mice (Peromyscus

keeni) on Rhinoceros Auklet (Cerorhinca monocerata) eggs.

Triangle Island is the site of western Canada's largest seabird colony. Over

one million seabirds breed there, including an estimated 42,000 pairs of Rhinoceros

Auklets (Rodway 1991). Two native species of rodents also occur there: Keen's

mouse (formerly deer mouse P. maniculatus triangnlatis; Cowan and Guiguet 1975,

Hogan et al. 1993) and an endemic subspeùes of Townsend's vole (Microtus

townsendii cowani; Carl et al. 1951). Island populations of rodents have different

demographics and behaviour, and larger body size, than mainland populations

(Adler and Levins 1994), and those populations on Triangle Island contain some of

the Iargest specirnens hown for the species' range (Carl et al. 1951, Cowan and

Guiget 1975). No other mammalian predators occur on the island.

Me thods

In 1998 we monitored Rhinoceros Auklet burrows for depredated eggs

throughout the incubation period at Triangle Island, British Columbia (51°52'N,

129O05'W). In the pre-laying period, a total of 217 burrows in two study plots - 109

in Plot 1 and 108 in Plot 2 - were prepared by excavating the entire burrow at ann's-

length intervals. These access holes were then fitted with cedar shingles and

13

covered with earth. Like most alcids, Rhinoceros Auklets lay a single egg annually.

We checked the stahis of each burrow and egg once per week from 30 April to 25

May. By the latter date, the rate of occupation of new burrows had dropped to

almost zero and any empty burrows were dropped from the study. We monitored

the remaining burrows once per week until their eggs failed or unül chicks hatched

in June. Throughout the monitoring period, presence or absence of an incubating

adult in the nest chamber was noted. Any unincubated, cold eggs were temporarily

removed from the burrow and chedced for signs of attempted depredation by

rodents, such as chew marks or scratches.

Only eggs that had been chewed open and partially or completely eaten by

mice were scored as depredated. Partially eaten eggs were replaced in the nest

chamber, to be checked on the following week for signs of scavenging. We searched

for missing eggs throughout the burrow. Missing eggs were those that had been

present on the previous check. If not found we then scoured the burrow floor for

shell fragments. Missing eggs were recorded as depredated when chewed shell

pieces were found inside the burrow, and as absent and likely rodent depredated

when no shell was found. Eggs that had been found cold on two or more

consecutive checks were considered to be abandoned. These abandoned eggs were

readily distinguished from those that were temporarily neglected as the former

quickly built up a film of moisture while intermittently incubated eggs remained

dry. Corvid-depredated eggs also were readily disthguished from those eaten by

mice. Although Rhinoceros Auklet burrows are generally too deep for avian egg

predators to enter, in 1998, Northwestern Crows (Corvus caurinus) and Common

Ravens (C. corax) learned to unearth Our hidden access shingles and remove both

hcubated and unincubated eggs. Auklet eggs were recorded as being taken by

c o ~ i d s when shuigles near the nest diamber were found tom up and a previously

present egg was either missing or found eaten in the vicinity of the access hole-

In order to test whether observer presence increased rodent predation, 88

burrows in a third, control plot also were fitted with observation shingles. Each of

these control burrows was monitored at 5-day intervals until an egg was found, and

then left unchecked for 30 days. If an egg was still present on this later check, it was

scored in the sarne manner as the eggs in the study plots, and the burrow was then

monitored every 5 days until a chi& was found. As these later Sday checks

necessarily involved observer presence, only those control eggs found depredated

on the initial burrow check or on the check ending the 30 day control period were

used for purposes of comparison with the study plots. We combined data for the

two study plots and used z-tests (Zar 1984) to compare their rate of depredation with

that in the control plot. Missing control eggs were recorded, but as we had very

limited knowledge of incubation history for the control plot, missing eggs were

excluded from all plots for the comparison of control vs. study plots.

In mid-May, to verify that mice could open an intact and viable egg, two

burrows were monitored overnight for two nights each with a remote video

camera attached to an infrared burrow probe (Peeper burrow probe, Christensen

Designs, Manteca, California). The monitored burrows were selected from those in

which a cold but intact and uncracked egg had just been found, and the probe was

positioned so as not to impede any access of a returning parent. We ended video

monitoring after footage of a predation event was obtained from the second

burrow.

Mean rodent body mass was obtained in late May from mice trapped with

Sherman live traps baited with peanut butter. Mouse mass was recorded to the

nearest g with a 100-g Avinet or Pesola balance. Auklet eggs also were weighed, and

measured for width and length to the nearest 0.1 mm using vernier calipers.

In order to determine whether predation rate at Triangle Island fluctuated

with locality, we used z-tests to check for inter-plot differences in predation rates.

We also tested the hypothesis that the risk of egg predation was higher early in

incubation, when the rate of egg neglect is reportedly higher (Wilson 1977). Data

for the two shidy plots were combined and &-square analysis was used to test for

differences in number of egg predation events in the first half of incubation vs.

incidents in the latter half of the incubation period. Mean incubation tirne for

Niinoceros Auklets is 45 days (Wilson and Manuwal 1986) and the mid-point was

rounded down to 22 days. Laying date was calculated conservatively. Burrow

checks were done every 7 days, so egg age of day 1 was backdated to 6 days before an

egg was first found. Eggs that were classified as abandoned and were later eaten

were excluded from these analyses. Values below are reported as mean f SD.

Results

A total of 124 eggs were laid in the 217 study burrows. Causes of egg loss

other than mouse depredation were varied, but totalled to less than all possible loss

to rodents in both plots (Table 2-1). Depredated eggs were always chewed open at

16

TABLI: 1, Causes of Rhinoceros Auklet (Cerorhirica monocerata) egg loss at Triangle Island. Values are n (%).

Rodent Missing Total likely El43 Total

Study Eggs predation eggs rodent pred. Corvid Egg cracked Gone at cg8

plot in plot (A) (B) (A+B) predation abandoneda in burrow the? hatching loss

'Egg found cold on two or more consecutive checks. b ~ n e nest chamber collapse, one adult dead in burrow frorn Peregrine Falcon (Wco peregriniis) att ack .

one end, usually on the srnail end, but on one occasion the large end. The egg was

then opened along the top surface (Fig. 2-1). When found initially, egg contents

were sometimes completely removed but most often were only partidy eaten.

Most (78%) of the depredation on Rhinoceros Auklet eggs took place during the

first half of the incubation period (x2, = 5.7, P c 0.02).

Although predation rate initially appeared to Vary by plot, with 25% of eggs

consumed by rodents in Plot 1 and 13% in Plot 2 (Table l), the differences were not

significant (z = 1.3, P > 0.18). Depredated eggs replaced by us in their burrows were

generally chewed into fragments and widely scattered, often without a trace, by the

following weekly burrow check. It was for this reason that we recorded missing

eggs as having been likely depredated. When missing eggs were included, there

was still no inter-plot variation (z = 0.5, P > 0.6), and rodent predation may have

occurred in as many as 34% and 30% of nests in Plots 1 and 2, respectively (Table 2-

1). Egg loss from all causes totalled to over 50% in each study plot (Table 2-1).

A total of 28 eggs were found cold on at least one occasion throughout May

and June, although not d l depredated eggs were cold on the previous check.

Although many of these cold eggs were later found depredated, 7 of the 28 (25%)

later hatched, indicating that egg-eating rodents have the opportunity to take viable

embryos as well as those that have been abandoned. Of 49 eggs laid in the control

plot, 6 (12%) were found depredated by rodents imrnediately before or after the 30-

day non-monitoring period. Despite the increased likelihood of predation events

being missed due to the length of the control period, the observed rate of egg

depredation in the control plot did not differ signihcantly from the depredation rate

18

Figure 2-1. Mouse-depredated Rhinoceros Auklet (Cerorhinca monocerata) egg. Triangle Island, British Columbia, showing entry point at small end of egg.

for the two study plots (z = 0.8, P > 0.4). We conduded that observer presence in the

study plots did not increase egg predation artifiually, either through disturbance of

uicubating birds or by providing olfactory dues to mice about an egg's location.

On average, auklet eggs were nearly twice the mass of mice. Mouse body

mass was 43.8 + 5.1 g (n = 56), and egg mass was 79.0 f 5.4 g (n = 18). Mean egg size

was 69 x 47 mm (n = 18). Despite thiç size difference, rnice were still able to open

intact eggs. The mouse W e d in our study took approximately 17 min to chew

through the eggshell and begin consuming its contents. While chewing on the egg,

the mouse altemated between lodging it against the camera lens and holding it

immobile against the burrow floor. The mouse also rolied the egg vigorously about

the burrow as well as chewing on it.

Discussion

At Santa Barbara Island, California, deer mice (P. m. elzisus) preyed upon

Xantus' Murrelet (Synthliboramphus hypoleucus) eggs only when parents were

absent from the nest during the early post-laying period (Murray 1980, Murray et al.

1983). Gaston (1992) also reported that Peromysczis predation on the eggs of the

Ancient Murrelet (Synthliboramphus antiquus) occurred during periods of

parental absence. Our observations suggest that the situation is sirnilar for

Nunoceros Auklets. Rhinoceros Auklets may neglect their egg during the first 8

days of incubation, although it is not known whether these neglected eggs are

incubated at night (Wilson 1977). They may also neglect the egg for intervals of 1 to

3 days later in incubation (Wilson 1977, Wilson and Manuwal 1986, L. K. Blight,

20

unpubl. data). In Cassin's Auklets (Ptychoramphus aleuticus), the incidence of

natural egg neglect also is highest early in the incubation period, and lowest later

on during the period of incubation corresponding to the most rapid ernbryonic

development (Astheimer 1991). Consistent with this, rnost depredation on

Rhinoceros Auklet eggs on Triangle Island occurred in the first half of the

incubation period and lücely coincided with the period of greatest egg neglect. An

alternative explanation for these results is that incidents of egg neglect may be

equally spaced throughout the incubation period and mice may simply switch to

new, emerging food sources near the end of May. Given that unattended eggs are a

large, high-energy food source, however, we predict that mice should take eggs

whenever they find them available.

Egg predation by herbivorous voles has been reported elsewhere (Sealy 1982,

Bures 1997), but appears to be rare. We found no evidence that herbivorous

Townsend's voles were present as predators in aulclet burrows. Voles were

observed eating only vegetation, and chew mark patterns on depredated eggs did

not Vary from those known by us to be caused by mice (pers. obsew.). Although

Larivière (1999) points out the problems involved in identifying nest predator

species through examination of eggshell remains, particularly in diverse, terrestrial

ecosystems, we are confident that identification of nest predators on Triangle Island

presented few such difficulties. Introduced rabbits (On~ctolngzrs cuniculus) are the

only other terrestrial, non-avian vertebrates found on Triangle Island (Car1 et al.

1951).

Various experiments have used Japanese Quai1 (Cotumix cotumix) eggs to

study predation by small mammals (Roper 1992, Haskell1995, DeGraaf and Maier

1996). The results of these studies have generdly been interpreted to mean that

small rodents are gape-ümited and cannot open any egg larger than that of a

moderately-sized passerine. Rhinoceros Auklet eggs are about double the size of 33

x 23 mm reported for quail eggs by DeGraaf and Maier (1996). In addition, the shells

of auklet eggs are thicker (0.29-0.40 mm; Gaston and Dechesne 1996) than quail eggs

(0.22-0.23 mm; DeGraaf and Maier 1996). Mice preying on Rhinoceros Auklet eggs

on Triangle Island are handling an item that is considerably larger, and thicker

shelled, than a quail egg and in fact on average approaches two times rodent mass.

The filming of a mouse depredating an auklet egg clearly illustrates that some

murid rodents are able to open an intact egg despite a pronounced size difference

between predator and prey item (Fig. 2-2). Gaston (1992) suggests that most

successful rnouse predation on Ancient Murrelet eggs was on those that were

previously cracked. The mixed rock and earth substrate of auklet burrows may

sometimes assist mice in opening eggs at Triangle Island, although the filmed egg

did not appear to be cracked by rolling. Craig (1998) reports that least chipmunks

(Tnrniaç minimus) were only able to open quaii eggs when the egg was lodged

between the predator's body and the inside of the nest. Once this was achieved, the

small end of the egg was readily bitten through. The consistency with which auklet

eggs were found opened from the small end indicates that whether a depredated

egg is initially cracked or intact may be unimportant. Scarcity of food supply

through the winter months is likely the limiting factor controlling the mouse

population at Triangle Island (Car1 et al. 1951). Egg laying for Rhinoceros

22

Figure 2-2. Keen's mouse (Peromyscus keeni) opening intact Rhinoceros Auklet (Cerorhinca monocerata) egg, Triangle Island, British Columbia. Mrared video- camera image.

Auklets there begins in mid-April, and seabird eggs may provide mice with an

altemate high-protein food source at a time when seeds are unavailable and other

foods st iU scarce.

In conclusion, depredation by Peromyscus can contribute substantidy to egg

mortality for Rhinoceros Auklets nesting on Triangle Island. Total egg loss appears

to be higher there than on islands where mice are absent: at a Washington State

Rhinoceros Auklet colony with no rodent predators, an estimated 81.5% and 91.1%

of eggs hatched over a two-year study (Wilson 1977). Our preliminary data show a

probable link between egg neglect and predation. Given that egg neglect is an

adaptation in marine birds to deal with patchy and distant food sources (La& 1967),

there is likely to be inter-annual variation of neglect with more frequent or longer

periods occurring in years of poor food availability. As El Niiio conditions

appeared to affect prey composition and availability for Triangle Island seabirds in

1998 (Triangle Island Research Station, unpubl. data), the rate of egg predation by

mice may have been higher during our study than in other years. Whether egg

predation shows inter-annual fluctuations related to neglect patterns rernains to be

tested.

Acknowledgements: Thanks to Laura Jones, Hugh Knechtel, and Tony

Williams for helpful comments and discussion. Comments by Mark Drever, Chris

Guglielmo, and two anonymous reviewers improved the manuscript, and Boyd

Pyper provided assistance in the field. This research was funded by gants from the

Nestucca Oil Spill Trust Fund to DFB and G.W. Kaiser (CWS), Natural Sciences and

Engineering Research Council (NSERC) operating grants to F. Cooke and T.D.

Williams, and by a John K. Cooper Foundation research award and an NSERC post-

graduate scholarship to LKB. The Canadian Coast Guard provided ship and

helicopter support.

Chapter III2

Predation on seabird eggs by Keen's rnice (Peromyscus keeni):

using stable isotopes to decipher the diet of a terrestrial omnivore on a

remote offshore island

Abstract

We used stable isotope techniques to analyze tissues of Keen's mice

(Perornysczls keeni) and Townsend's voles (Microtus fownsendii cowuni) and a

subset of prey items at Triangle Island, western Canada's largest seabird colony.

Isotope analysis allowed us to investigate the importance of seabird prey in rodent

diets in a system where seabirds and non-introduced rodents occur sympatricaily.

The 6I5N values for terrestrial plants and terrestrial invertebrates on Triangle

Island exceeded levels found in many terrestrial biomes and are typical of localities

with high inputs of marine-derived nitrogen. We used multiple-source mixing

models to estimate the relative inputs of potential prey items to vole and mouse

diets. 613C and 615N values of liver and muscle tissues of voles indicate that voles

derived their protein primarily from terrestrial plants, with some contribution by

terrestrial invertebrates. In contrast, isotopic values of liver and musde tissues of

mice indicated mice on Triangle Island prey primarily on seabird eggs and

terrestrial invertebrates. Our results show egg predation at Triangle Island is a

This chapter is currently in press as Drever, M.C., L.K. BLight, K.A. Hobson and D.F. Bertram. Predation on seabird eggs by K m ' s mice (Peromysczls keenz): using stable isotopes to decipher the diet of a terrestrial omnivore on a remote offshore island. Can. J. 2001.

26

general phenornenon in the mouse population, rather than occurring in only a few

specialist feeders. Mice appear to feed on eggs once they become available, and

continue to utilize seabird prey throughout the breeding season.

Introduction

Understanding factors that influence reproductive success ui seabirds is

crucial to developing predictive population models for management and

conservation purposes. Variation in seabird reproductive success is controlled by

several factors, induding weather, predation risk, age and breeding experience of

parents, and timing and quality of available food (Furness and Monaghan 1987).

Depredation of eggs, adults and young by introduced rodents, primarily rats Rattus

spp., is well-recognized as playing a significant role in seabird breeding success

(Moors and Atkinson 1984, Hobson et al. 1999). However, depredation by native

rodents has largely been ignored despite the importance of understanding and

quantifyicg the magnitude of such natural sources of mortality. The absence of

native predators is thought to be important in the evolution of island nesting in

seabirds (Fumess and Monaghan 1987), and while localities where native rodents

occur sympatrically with breeding seabirds appear to be rare, predation rates in these

situations can be high. At Triangle Island, British Columbia, egg depredation by

native Keen's mice (Peromyscus keeni) occurs in up to 30% of burrows of Cassin's

Auklets (Ptychornrnphus aleuticus; Morbey 1995; D.F. Bertram, unpubl. data) and in

up to 34% of burrows of Rhinoceros Auklets (Cerorhinca monoceruta; Blight et al.

1999). Native rodents also prey on seabird eggs elsewhere in the world. For

example, Murray et al. (1983) reported that the eggs of Xantus' Murrelets

(Synthliboramphtis hypoleucus) are depredated by deer mice (Perornyscus

rnanicula t ris eltrstis) on Santa Barbara Island, California. Peromyscus also prey on

eggs of the Ancient Murrelet (Synthliboramphus antiquus) at Haida Gwaii (Queen

Charlotte Islands), British Columbia (Gaston 1992).

At Triangle Island, only two species of rodents occur: native mice (P. keeni),

and an endemic race of Townsend's vole (Microttis townsendii cowani). Our study

used stable isotope analysis to determine the relative importance of seabird eggs in

the diet of rodents at this remote island colony. Traditional methods of diet

analysis such as analyses of stornach contents have limited utility in determinhg a

predator's consumption of eggs. Eggshells do not preserve well in stomach acid,

and egg yolk or albumen can be almost impossible to iden* a few hours after their

consumption (Duffy and Jackson 1986). Furthemore, single samples of stomach

contents do not take seasonal dietary variation into account. Analysis of stable

isotope ratios in tissues of predators circumvents this limitation, and when used

appropriately it has several advantages (Hobson and Sealy 1991, Hilderbrand et al.

1996). The stable isotope approach provides information on diet integrated over

various time scales depending on the tissue diosen: isotopic analyses of liver tissue

integrate food assimilation over a week, while analyses of muscle tissue provide

information on diet over about a month (Tieszen et al. 1983; Hobson and Clark

1992). Thus, in contrast to conventional techniques, it is possible to derive

information over ecologically relevant time periods, and specialists feeding

exclusively on one or two prey types may be readily identïfïed.

Because 615N analyses provide information on trophic Ievel whereas 6I3C

analyses provide information primarily on the source of nutrients to the diet

(Peterson and Fry 1987), the use of these two isotopes (or more) typically provides

better resolution in tracing the diets of consumers and elhinates the ambiguities

that accompany the use of a single isotopic tracer (Peterson et al. 1985). The ocean is

generally enriched in 13C relative to terrestrial ecosystems (Peterson and Fry 1987),

and the relative abundance of 13C remains comparatively unchanged with each

trophic transfer (DeNiro and Epstein 1978; Schoeninger and DeNiro 1984). Thus,

the measurement of I3C can serve as a usefd tracer of marine protein in the diet of

animals (Hobson 1986; Hobson and Sealy 1991). Conversely, 615N values in marine

and terrestrial organisms show a step-wise eruichment due to the preferential loss

of the lighter isotope, "N, during excretion (Peterson and Fry 1987), and serve as

good indicators of trophic level of the sample organism (Minagawa and Wada 1984;

Hobson et al. 1994). Stable-nitrogen isotope ratios in marine systems may also be

typically enriched compared to terrestrial food webs (Schoeninger and DeNiro 1984;

Wada and Hattori 1991; Michener and ScheIl 1994).

Stable isotope analysis has recently been used at another seabird colony to

examine diets of Norway rats (Rattus norvegïcus) preying on Ancient Murrelets at

Langara Island, British Columbia (Hobson et al. 1999). Murrelet eggs, chicks, or

adults had stable isotope ratios geatly enriched in both I5N and I3C cornpared to

other food types available, and this allowed a direct estimate of the proportion of

the rat population feeding on seabirds. We predicted that rodents eating seabird

eggs at Triangle Island would have tissues similarly enriched in both isotopes in

29

contrast to tissues of those rodents eating exclusively plant-based diets. Our goals in

using stable isotope analysis were to ask (i) whether egg depredation at Triangle

Island is the work of a few individuals specializing on seabird prey or is a general

phenomenon among all mice, and (ii) whether seabird eggs are also consurned by

voles. Although egg and chi& depredation by herbivorous voles does occur (Sealy

1982)' it appears to be rare, and we hypothesized that it would be unükely on

Triangle Island. We also used stable isotope analysis to ask (5) whether rodent

sub-populations showed dietary differences based upon season or locality.

Study Site and Methods

Stztdv site

Our study took place from Mardi - May 1997 and April - August 1998, at

Triangle Island, British Columbia (50'52' N, 12g005' W; 144 ha). Triangle Island is

an ecological reserve, and lies 46 km to the northwest of Cape Scott at the northern

end of Vancouver Island, British Columbia, Canada. The island is western

Canada's largest seabird colony, and provides nesting habitat for 1.2 million birds of

12 species, primarily Rhinoceros and Cassin's Auklets (Rodway et al. 1990).

Cassin's Auklets begin arriving at the colony in March, and laying typically begins

in that month. Rhinoceros Auklets begin laying in mid- to late April (Triangle

Island Research Station, unpubl. data). In 1997, Our study focused on three

localities: (1) West Bay, where Cassin's Auklets nest at high density; (2) Calamity

Cove, where both Cassin's Auklets and Rhinoceros Auklets nest in moderate

densities; and (3) South Bay - Cabin site. Burrow density for Rhinoceros Auklets in

South Bay is high, while Cassin's Auklet burrows are found there in lower

numbers (Rodway et al. 1990). In 1998, w e established two study plots in South Bay.

Plot 1 was situated approximately 30 m from 1997's cabin site, and Plot 2 lay about

300 m to the east.

Snmpl in~ rodents and likelv vreu m e s

In early May 1997, we set snap traps within 50 m of the shore for two nights

at two localities (West Bay, n=17 traps; Cdamity Cove, n=9 traps). Traps were set 15

m apart under cover of vegetation or Logs and were baited with peanut butter. Mice

from the Cabin site were collected in March - April 1997, before the majority of egg-

laying activity began. These mice were trapped for routine hygiene measures or

were found dead in the area. In 1998, permit restrictions preduded the use of snap

traps. Instead, mice were trapped using Sherman Iive traps baited with peanut

butter, and euthanized using CO,. Five mice were collected from each of our two

study plots in South Bay between 31 May - 03 June (hereafter "Early"); another five

mice were collected from each plot in the same manner on 22 August ("Late"). One

mouse found dead in May in South Bay was added to the Plot 1 Early samples. As

we did not receive permission to coUect voles (red-listed in British Columbia), the

samples that we analysed were from three dead animals, salvaged from the three

study localities oves the months of Aprïl - June. AU rodents were sexed, weighed,

measured and their reproductive state recorded.

In 1998, we ais0 carried out live trapping in both South Bay study plots in

order to estimate mouse population density. A total of 45 live traps (25 in Plot 1

and 20 in Plot 2) were placed on a grid pattern 10 m apart. Trapping was carried out

over 3 nights between 28 May - 01 June, and for three nights from 30 June - 02 July.

Trapping effort was suspended on rainy nights, and we exduded 30 June's trapping

results for Plot 2 from our analyses, due to the high proportion (47%) of traps found

tripped but empty on that night. Mouse collection was separated in time and/or

space h m density estimate live trapping in order not to confound our population

estimates. We express al1 capture rates of mice in captures pei 100 trap-nights

(C/lOOTN).

Ln 1997 and 1998, we also collected samples of prey representative of those

available to mice and voles, including prevalent terrestrial plants, terrestrial

invertebrates (1998 only), intertidal organisms [marine algae Fucus sp.; isopod and

amphipod spp.], and abandoned (i.e., found cold for > 5 d ) Cassin's Auklet eggs

(hereafter CAAU eggs; 1997 only) and Rhinoceros Auklet eggs (RHAU eggs). These

items were not meant to exhaustively represent all food available to mice and

voles, but rather represented the broad categories of prey types to be found on the

island.

Rodents were presemed in the field at -IO0 C in a propane freezer. Liver and

muscle tissues were later sampled and freeze-dried, and ground into powder using

a Wiggle-Bug dental amalgam mil1 (Crescent Dental Manufacturing, L yons, Illinois,

USA). Lipids were extracted via a dilorofonn methmol r ime as modified fkom

Bligh and Dyer (1959), and samples were then loaded into tin cups, and combusted

at - 1800 OC in a Europa Robo-Prep Elemental Analyzer ushg a helium carrier gas

interfaced with a Europa 2020 isotope ratio mass spectrometer (IRMS). This

continuous flow IRMS technique provided 615N and ouC values with errors of r

0.2%0 and I 0.3 %O for carbon and nitrogen isotopes, respectively. Invertebrates of

the same species were pooled and processed as one sample per speües in each year

of collection, so values of invertebrate samples represent a species mean- The

standards were on PeeDee Belemnite for carbon, and atmospheric air for nitrogen.

Al1 samples were analyzed at the Prairie and Northem Wiidlife Research Centre

and Department of Soi1 Science, University of Saskatchewan.

The isotopic composition of any tissue is expressed as a ratio of the heavier to

the lighter isotope relative to a standard. Cornmonly, 6 notation of parts per

thousand (%O) is used where:

'' = [ ( k m P I e / & h d & ) - ltoo0

and X is 13C and 15N, and R is the corresponding ratio 13C/'2C and "N/'%J. Thus,

tissues with larger 6X values are enriched with the heavier isotope relative to

tissues with smaller 6X values.

S fatistical analyses o f isotopic simals

We used a k-nearest neighbour randomization test to examine differences in

6 ' ' ~ and S13C values among prey types (Rosing et al. 1998). This test treats 6I5N and

613C values as spatial data and calculates the probability that two observed samples

are derived from the same population compared with the probability that samples

with the same values were generated at random. We used 10,000 randomizations

of the data to calculate probabilities. Using this approach ensured that prey types

are bivariately and sipifmintly different from each other (Ben-David et al. 1997af

b), and thus enabled us to use dual-isotope, multiple-source mixing models to

determine the relative contribution of each prey type to diets of rodents (see below).

We used multivariate analysis of variance (MANOVA) to test the null

hypothesis of no differences among groups of rodents in S1'N and 613C values of

tissues, and Tukey's multiple comparison tests to contrast levels where factors were

significant at aS0.05 (Zar 1996). To avoid problems with the assumptions of

multivariate normality and homogeneity of variances necessary for MANOVA, we

used a permutation analysis to calculate the sigxuficance of Wilks' h statistics

(Hobson et al. 1999). We generated 10,000 permutations of the data, randomly

assigning each bivariate point to a group of rodents. We then analyzed each

permutation separately and compiled a randomization distribution of Willcs' h

statistics, and calculated the P-value as the proportion of randomized values as

extreme or more extreme than the observed value (Manly 1991).

We reasoned that the variance in isotopic signals for each group of rodents

was analogous to inter-individual variance in diet. Thus, i f rodents from one area

had low variance, then we conduded that group specialized in the same prey

type(s). A group with a larger variance indicated a range of dietary strategies among

individuals in that group.

Multiple-Source M i x i n ~ Model

In order to determine the relative contribution of each prey type to diets of

individual mice, we used a multiple-source mixing model (Kline et al. 1990, 1993;

Ben-David et al. 1997a, 1997b). This model uses the mean 6 1 5 ~ and 613C values for

each prey type (type A, B, C, etc.) corrected for the enridunent in the tissues of the

consumer (i.e., the fractionation factor; A', B', C'; DeNiro and Epstein 1978, 1981;

Tieszen et al. 1983). These corrected values represent the 615N and 613C values that

would occur if the predator's diet consisted enürely of that prey type, and as the

model uses means, it does not place undue emphasis on any extreme values for a

given prey type. We used a fractionation factor of +2 %O for carbon when eggs and

plant matter were consumed, and +1 %O when marine and terrestrial invertebrates

were consumed, based on results from feeding experiments in captivity on mink

and bears (Hilderbrand et al. 1996; Ben-David et al. 1997a, 1997b). We used a

fractionation factor of +3 %O for 6 " ~ values of al1 prey types (DeNiro and Epstein

1981; Schoeninger and DeNiro 1984). The isotopic contribution of each prey type to

the diet of each individual predator (P) is calculated as:

X, = (l/PX')/(l/PA' + l/PBO + l/PC'),

where X' is A', B' or C' and PX' is the Eudidean distance (z=(x2 + f)'3 between the

615N and 613C values of each predator (P) and the corrected prey types. The model

assumes that individual predators consume all possible types of prey, and thus

35

tends to overestimate the proportion of food types rarely consumed and

underestirnate the proportion of commonly used food items (Rosing et al. 1998).

Therefore, the mode1 provides indices of food consumption rather than actual

proportions in the diet. We induded all prey types (terrestrial plants, terrestrial

invertebrates, intertidal organisms, RHAU eggs, and CAAU eggs) in examining the

diet of mice from South Bay and Calamity Cove sites, but omitted RHAU eggs for

mice from West Bay, since this prey type was not available at that locality.

Resul ts

Samvle collection

h 1997, nine rnice were captured in West Bay (52.9 C/100 TN), seven from

Calamity Cove (77.8 C/100 TN), and 17 mice were obtained from the South Bay -

Cabin area over a 5-day perïod. In 1998, live trapping for mouse density in South

Bay yielded 64.4 C/100 TN for the May/June trapping, and 58.9 C/100TN for the

June/July trapping. No voles were captured by our live trapping effort. Rodent

samples fell into eight groups based on speues, and location and time of capture:

the three groups of mice captured in 1997 (South Bay - Cabin, Calamity Cove, and

West Bay), the four groups from 1998 (South Bay: Plot 1 Early and Plot 2 Early; Plot 1

Late and Plot 2 Late) and the group of voles salvaged in 1998. Morphometrics and

population characteristics are detailed in Table 3-1.

Table 3-1. Morphoinctrics and population characteristics of Keeii's mice (Peroi~ri~sciis k w i ) and Towiiseiid's voles

(Microtirs iowilseiidii corunrii) from Triangle Island, British Columbia, 1997-1998.

Wcight (g) Total Leiigth 'l'ail to vent Sex ratio Proportion in

(ciii) Lcngt h (cm) (M:F) brccding sratc

Dates o f capture I I Mean SD Meari SD Mean SD

Cabiii u V

Calaiiiity Covc

West Bay

Soiitli Bay - Pot I

Soiitli Bay - I'lot 2

South Bay - Plot 1

Soiitli i3ny - I'lot 2

Towiiscntl's Valcs

29 March - 2 April 1997

7 May 1997

3 May 1997

9 May, 3 1 May - 3 Jiinc 1998

3 Juiic 1998

23 Aiigust 1998

23-24 Aiigust 1998

13 April - 29 Jiine 1998

Isotope Analyses o f Prev Tvaes

Prey types differed primarily in their 613C values (Figures 3-1 and 3-2)3.

Intertidal organisms were most enriched in 13C, foUowed by seabird eggs, terrestrial

invertebrates, and terrestial plants (Figure 3-1). CAAU and RHAU eggs had

similar 613C values (-19.9 & and -20.2%, respectively), but RHAU eggs were more

enriched in I 5 N . 615N values had similar means among prey types, and generally

showed larger variance than 613C values (Figure 3-2). Terrestrial invertebrates had

the highest mean 615N value, although highest individual 615N values were found

in terrestrial plants, particularly salmonberry (Figure 3-1). Differences in mean 6I3C

values provided good segregation among prey types. Stable isotope ratios differed

significantly arnong prey types primarily due to differences in 6I3C values (Figure 3-

2; K-nearest neighbour randomization test, P<0.001 for al1 painvise comparisons

among prey types), with the exception of terrestrial invertebrates (K nearest

neighbour randomization test, P = 1.000). This lack of difference however likely

resulted from the small sample size (n = 2) for this prey type. Using the MANOVA

and permutation analysis to caiculate the P-values (a procedure less consewative

than the K-nearest neighbour randomization test), and Tukey's multiple

cornparison tests, resulted in 613C values of terrestrial invertebrates differing

Stable carbon and nitrogen isotope values for rodent prey types on Triangle Island, BC, are archived at the NRC's data depository. These data may be accessed by contacting the Depository of Unpublished Data, Document Deiivery, CISTI, National Research Council of Canada, Ottawa, ON K1A 0S2, Canada.

+ Terr. Plants * Terr. lnverts O RHAU Eggs @ CAAU Eggs * intertidal

Figure 3-1. Stable carbon and nitrogen isotope values (760) for prey types available to Keen's mice (Perornysczis keeni) and Townsend's voles (Microtzis townsendii cozunni), Triangle Island, British Columbia.

TERR. INVERTS. (2)

PLANTS (1 0) T

Figure 3-2. Mean stable carbon and nitrogen isotope values (%O f SD) for sample prey types. Numbers in parentheses are sample sizes.

signihcantly from all 0 t h prey types (Wilks' h = 0.073, P c 0.0001, Tdcey's test, P c

0.05).

Isotope Anal ses of Rodents

Liver tissues: Both 815N and 613C values of liver tissues, indicative of short-

term diets of about one week, differed significantly among groups of rodents

(Wilks' ;C = 0.230, P < 0.001; Figure 3-314. Voles had liver tissues with 613C values

similar to those of terrestriai plants, and which differed sisnificantly from liver

tissues of all mice groups (Tukey's test, P < 0.05). 613C values from liver tissues of

mice were between 613C values of seabird eggs and terrestrial invertebrates and did

not differ among mouse groups, with the following exception: 613C values of Cabin

mice liver tissues more closely resembled 613C values of ùitertidal organisms and

had the greatest "C enrichment of mouse groups. Liver 613C values from Cabin

mice differed significantly from rnice from Calamity Cove and the pooled August

1998 sarnples from South Bay (Figure 3-3; Tukey's test, P < 0.05). Highest 615N

values were found in vole liver tissues; these differed significantly from al1 mouse

groups except Calamity Cove. Calamity Cove mice had liver tissues with the

highest 615N values of all mouse groups, and differed significantly from 615N

values of liver tissues from Cabin mice (Tukey's test, P < 0.05), which had the

Mean stable carbon and nitrogen isotope values for rodents' liver and muscle tissues at Triangle Island are archived with the NRC. See Footnote 3 for further detaiis.

LIVER TISSUES vs

1 TERR. 1 INVERTS.

PREY TYPES

--

O SOUTHBAY O WESTBAY 0 CALAMITY A CABIN + VOLES l

Figure 3-3. Stable carbon and nitrogen concentrations in liver tissues of Keen's mice (Peromyscns keeni) and Townsend's voles (Microtiis townsendii cownni) collected at Triangle Island, British Columbia, 1997-1998. Filled cirdes indicate mean values of possible prey types adjusted for fractionation effects, where PLANTS = terrestrial plants, TERR. INVERTS. = terrestrial invertebrates, CAAU = Cassin's Auklet eggs, RHAU = Rhinoceros Auklet eggs, and INTERTTDAL = inter tidal organisms.

lowest 615N values of al l mouse groups. The remaining mouse groups ail had b e r

tissues with similar 6"N values (Figure 3-3; Tukey's test, P > 0.05).

Mouse groups had coefficients of variance for liver 6 T values that ranged

from -2.3 to -6.5, with the exception of the Cabin group that had a coefficient of

vanance of -11.9, nearly double the largest value of other mice groups. Coefficients

of variance for liver 615N values for South Bay and West Bay mice ranged from 2.9

to 4.9 and were lower than coefficients of variance for liver 615N values for voles

and mice from the Cabin and Calamity Cove, which ranged from 9.0 to 12.0.

Muscle tissues: 6"N and 6l3C values differed among groups of rodents

(Wilks' h = 0.193, P < 0.001; Figure 3-4), and followed a similar pattern to 615N and

613C values in liver tissues. Voles had muscle tissues with lowest 613C values,

which differed significantly from muscle tissues of all mouse groups (Tukey's test,

P < 0.05), and were similar to terrestrial plants (Figure 3-4). Cabin mice had musde

tissues with 613C values more closely resembling 613C values of intertidal organisms

(Figure 3-4). These values were the highest of all rodents', and differed significantly

from mice from Calamity Cove and the Plot 1 Late (1998) sample (Tukey's test, P <

0.05). 6I3C values did not differ among musde tissues of mice from West Bay,

Calamity Cove and the 1998 South Bay samples (Tukey's test, P > 0.05), all of which

had 613C values between those of seabird eggs and terrestrial invertebrates.

PLANTS i MUSCLE TISSUES

T vs PREY TYPES TERR. INVERTS.

INTER

O SOUTHBAY O WESTBAY o CALAMITY A CABIN 4k VOLES

Figure 3-4. Stable carbon and nitrogen concentrations in muscle tissues of Keen's mice (Perornyscns keen i) and Townsend's voles (Microt us townsendii cowani) collected on Triangle Island, British Columbia, 1997-1998. Filled circles indicate mean values of possible prey types adjusted for fractionation effects, where PLANTS = terrestrial plants, TERR. INVERTS. = terrestrial invertebrates, CAAU = Cassin's Auklet eggs, RHAU = Rhinoceros Auklet eggs, and R\ITERTIDAL = intertidal organisms.

Vole musde tissues had the highest 615N values, and differed significantly

from muscle tissues of all mice groups (Tukey's test, P < 0.05). Of the mice, Calamity

Cove mice had highest 6I5N values in musde tissues and differed significantly

from West Bay and Cabin mice (Tukey's test, P c 0.05), whidi had the lowest 615N

values. The 1998 South Bay samples had musde tissues with intermediate 6I5N

values and did not differ among each other and from the 1997 samples.

Coefficients of variance for muscle 815N and 613C values followed the same

pattern as the liver tissues. Cabin mice had the largest variance of 613C values

compared to other rodent groups. Coefficients of variance for muscle S"N values

were similar among Çouth Bay and West Bay mice, but were Lower than analogous

values for voles and mice from the Cabin and Calamity Cove.

Mzrltivle-sozrrce rnixing mode1

The multiple-source mixing model indicated that the primary sources of

protein for Keen's mice on Triangle Island were terrestrial invertebrates and CAAU

and RHAU eggs, whereas terrestrial plants and intertidal organisms provided

relatively less protein (Table 3-2). When considered together, CAAU and RHAU

eggs provided the b u k of the protein (estimate overall, 51.0% for muscle, 43.0% for

liver) in mouse tissues. RHAU and CAAU eggs provided similar amounts of

protein in tissues for mice at sites where both bird species occur together (19.0% and

20.0% for CAAU and RHAU eggs, respectively, at Calamity Cove; 21.0% and 20.0%

at South Bay). The Cabin group was an exception, where CAAU eggs appeared to be

45

Table 3-2. Relative coiitributions ('%) of prey types to diet of Keen's mice (Peron~yscscits keerii) and Townsend's voles

(Micro1 irs toruiiseiidii coruniri) on Triangle Island, British Columbia (1997-1998), estiinated using a multiple-source

mixing mode1 with isotopic data from two tissues.

CAAU RH AU lntertidal Terrestrial Terrestrial

c66s eg6s Organisnis Plants lnvertcbratcs

Liver lissucs - M i c e

South Bay - Cabin

Calamity Covc

West Bay m

Soutli Bay - Plot 1 Early

Soutli h y - Plot 2 Early

Soiith Ray - Plot 1 Latc

Livcr tissues - Voles 3

Table 3-2 contiiiued on ncxt page

more important than RHAU eggs in diets of mice (Table 3-2). This was probably

due to the early date of collection of these animals, as Cassin's Auklets begin laying

before Rhinoceros Auklets. In addition, the CAAU eggs prey type had a larger

variance relative to the other prey types for mice in the Cabin group, partïdarly in

muscle tissues (Table 3-2). Little difference existed in the relative contribution of

seabird eggs in diets of rnice between the Early South Bay samples and the Late

South Bay samples. In contrast to the mice, voles on Triangle Island derived their

protein primarily from terrestrial plants (32.0% and 39.0% for muscle and liver

tissues, respectively), followed by terrestrial invertebrates (30.0% and 25.0%; Table 3-

2). The multiple source mixing model estimated that seabird eggs and intertidal

organisms provided relatively less protein for voles, but as the model forces ail prey

types to be considered, it presumably overestimated the contribution of these two

prey types.

Discussion

Isotope Analvses o f Prew Tvves

Triangle Island is a varied isotopic environment, in which terrestrial,

intertidal, and pelagic elements provide omnivorous rodents with a wide range of

isotopic inputs. The 615N values for terrestrial plants and invertebrates on Triangle

Island certainly exceeded levels found in many terrestrial biomes, and are typical of

places with high inputs of marine-derived nitrogen. These indude other seabird

colonies (Mizutani and Wada 1988, Cocks et al. 1998, Wainright et al. 1998), otter

latrines (Ben-David et al. 1998), and salmon spawnuig grounds (Kline et al. 1990,

1993). Salrnonberry provides a drarnatic example: on Triangle Island, sahnonberry

had a mean S"N value of 15.3Ym, whereas in southeast Alaska, salmonberry had a

615N value of 1.1 %O at sites with no contribution of marine-derived nitrogen (Ben-

David et al. 1998). Terrestrial plants that do not fix nitrogen typically have 615N

values that range from -8 to 10%0, depending on the isotopic composition of their

soils (Peterson and Fry 1987). The highly enriched 615N values of plants on

Triangle Island likely resulted from nitrogen uptake from soils enriched in "N due

to incoming marine-derived nitrogen in seabird guano, and by the process of

volatilization of isotopicaIly light ammonia from the guano (Mizutani et al. 1986,

Mizutani and Wada 1988). In this situation, where two very different biomes (i.e.

marine and non-marine) contribute to nutrient input, I5N serves as a usefd tracer

of nutrient source but loses its utility as a straightfonvard indicator of trophic level.

On Triangle Island, mice and voles feeding on primary producers (terrestrial plants)

had higher 615N values than rodents feeding at higher trophic levels. In studies

using stable isotopes, characterising 61SN values of available food sources is

required before inferences about trophic level of a consumer can be made (Hobson

1999).

Intertidal organisms were more enriched in 13C than seabird eggs, a pattern

that reflects the gradient of decreasing abundance of 13C in the ocean from the

intertidal to the pelagic zone (France 1995). Similarly, Hobson (1993) found that

benthic-feeding seabirds had tissues more enriched in 13C than those from pelagic

49

seabirds. Although we could not use 6I5N values to predict rodent trophic level at

Triangle Island, stable nïtrogen isotopes reliably predict seabird trophic positions

since their diets are strictly marine in ongin (Hobson et al. 1994); the difference we

observed between 615N values of RHAU eggs and CAAU eggs reflect dietary

differences between the two seabird species (Hobson 1991). Rhinoceros Auklets

primarily eat fish, while Cassin's AukIets typically feed on plankton (Gaston et al.

1998) and thus have lower 615N values in their eggs.

Seabird e w s and rodenfs on Triande Island

The 6I3C values and the multiple-source mixing model indicated that mice

on Triangle Island derive the bulk of their protein firstly fiom seabird eggs and

secondly from terrestrial invertebrates. However, as plant prey (low protein)

typically enters energy metabolic pathways, and animal prey (high protein) is a

primary source of tissue synthesis in consumers (Gannes et al. 1997, Hobson and

Stirling 1997), our estimates of protein contribution are greater than the actual

proportion of a prey item in a rodent's diet. In contrast to mice, vole tissues have

613C and 6lSN values indicative of a largely terrestrial diet, with Little marine input.

This indication is supported by the model and by observations of voles on Triangle

Island feeding exclusively on terrestrial plants (L.K. Blight, pers. obs.). Voles

(Microtzls and/or Clethrionomys) on St. Lawrence Island, Bering Sea, have been

observed eating eggs and nestlings of nesüng aukleh (Sealy 1982), although voles

elsewhere are largely herbivorous and consume only small amounts (c 10 %) of

animal protein, in the forrn of arthropods (Batzli 1985).

The importance of seabird eggs in diets of mice varied with tirne of year and

locality at Triangle Island. 613C and 61sN values suggest that when eggs become

available, predation on seabird eggs is not the work of specialist mice but rather is a

general phenomenon in the mouse population, despite local variation in diet. The

Cabin mice were captured in early Spring before the majorïty of seabirds begin

breeding, although some early nesting activity by Cassin's Aukiets at that time is

likely. These mice then represent mice that are least affected by breeding bird

activity in the sampled years. 613C and 615N values for the Cabin group have larger

variance than other mice groups, indicative of a more diverse diet that indudes

marine invertebrates. However, some of these mice may also have had access to

particles of human food from dishwater disposed of on the beach. Stable-carbon

isotope ratios of mice from Calamity Cove, West Bay and South Bay (Early), trapped

in the middle of the seabird breeding season, had a smaller variance than the mice

trapped in March, suggesting that at this time of year mice have a smaller dietary

breadth and may focus on seabird eggs when they become available. Such a

phenomenon occurred at Langara Island, where isotopic signals of rats feeding on

eggs and birds in the Ancient Murrelet colony had much smaller variances than

signals of rats collected elsewhere on the island. Since rat 6 " ~ and 613C values

were similar to SISN and 613C values of Ancient Murrelet tissues, the small

variance indicated that rats fed on murrelets almost exdusively (Drever and

Harestad 1998, Hobson et al. 1999). Sirnilarly, rats raised on monotonous diets in

the laboratory showed minimal variance in bone collagen 613C values (Ambrose

and Norr 1993; although collagen is expected to have lower variance as it represents

a much longer-term integration of diet than muscle or liver.). Mice sampled by us

in South Bay showed little difference between Early (May/June) and Late (August)

samples. Although seabirds have finished incubating by the latter date, this la& of

difference suggests that mice continue to feed on seabird prey, as abandoned eggs,

dead chicks, or adult carcasses remain avaîlable into the late summer.

Irnalications for rodent vovulations

The acquisition of protein is crucial for growth, survival, maintenance and

reproduction. Seabird eggs Iikely provide the bulk of the input of protein for

mouse tissues on Triangle Island, yet are available only as a seasonal influx of

nutrients. It has been proposed that reduced food availability over the winter

months is the limiting factor for mouse populations at Triangle Island (Car1 et al.

1951). We suggest that mice there synduonize their breeding efforts to this influx

of avian food, particularly as availability of seabird eggs precedes that of seeds and

invertebrates. Island populations of rodents typically have distinct, short

reproductive seasons (Gliwicz 1980, Adler and Levins 1994), and rodent populations

often eat more animal prey when breeding relative to other times (Clark 1981).

We also suggest that seabird presence may account for the exceptionally high

density of mice on Triangle Island. A study that trapped rodents at 46 islands in

southem British Columbia had an average trapping success of 28.2% for islands

where mice (P. maniculatus) were present (Redfield 1976). No islands under 25 ha.

(n=16) had a population of mice, with the exception of Mandarte Island (5 ha.), the

only seabird colony sampled: trapping success there was 73.8% (Redfield 1976). The

Mandarte Island values compare with a mean trapping success of 63.5% at Triangle

Island. Availability of seabird prey may also account for the sympatry of mice and

voles on our study island; in sampling 46 islands from 1 - 18,000 ha. in size,

Redfield (1976) found only one where Microtus occurred, and that was at an island

where Peromysczis was absent, likely due to cornpetitive exclusion.

In 1997, the March/April sarnple of mice from the Çouth Bay - Cabin site had

the smallest proportion of breeding individuals, and the later samples from West

Bay and Calamity Cove contained larger proportions of breeding individuals. In

1998, juvenile mice were virtually absent from the May/June trapping but

comprised the bulk of the animals trapped in June/July (L.K. Blight, unpubl. data).

While our sampling scheme partially confounds tune with location, this

concurrence of breeding effort and high seabird egg contribution to mouse diet

suggests that mice at Triangle Island time their breeding to coincide with the laying

season of Cassin's and Rhinoceros Auklets, when a large influx of protein-rich food

becomes available.

Acknowledgements: We thad John Ryder, Boyd Pyper, Ginny Collins and

fellow researchers on Triangle Island Research Station for aiding in collecting

samples on the island. British Columbia Parks kindly gave us permission to work

at the Anne Vallée Ecological Resewe (Triangle Island). This research was funded

in part by grants from the Nestucca Oil Spill Trust Fund to DFB and Gary W. Kaiser

(Canadian Wildlife Service). It was aiso funded by a Natural Sciences and

Engineering Research Council (NSERC) post-graduate scholarship and John K.

Cooper Foundation award to LKB, and by NSERC operating grants to Drs. Fred

Cooke and Tony D. Williams. Stable isotope sarnples were prepared by P. Healy and

analyzed by G. Parry at the University of Saskatchewan Soi1 Science Laboratory.

Daniel Ricard provided assistance with statistical analyses and Tom Olvet helped

with sample preparation. We also benefited from valuable discussions with Dr.

Merav Ben-David on the use of multiple-source mixï.ng models. Comments by

Tony Williams improved the manuscript.

Chapter I V 5

Egg neglect in a member of the Alcidae: an experimental approach

Abstract

Many birds interrupt incubation in order to spend time away from the nest.

For passerines, these excursions usually last only minutes. However, for some

marine birds, egg neglect may last a day or more, likely allowing parents to inaease

foraging time at sea. Egg neglect in alcids is reportedly most common early in

incubation. To test whether developmental constraints of oldes embsyos have

selected against neglect late in incubation, we experimentally chilled two separate

groups of Rhoceros Auklet eggs. Eggs were chilled to ambient temperature

(513.0" C) for 48 hours at 7 or 30 days of age, and survivorship for the two groups

recorded. We also examined the effect of experimental neglect on hatdiling size

and incubation period length. We found no difference in morphometrics or

hatchability among chicks from expesimental and control eggs. However,

experimentally chilled eggs hatched an average of two days later than controls. Our

results suggest that for some seabirds, embryo suMval is not affected by egg age at

neglect; parent birds may neglect at any stage of incubation. Lengthy neglect early in

incubation instead may be explained by parental responses to the energetic costs of

the early breeding period, or by increased parental investment in the embryo.

Key words: Cerorhinca monocerata; cost of reproduction; egg neglect; embryo

development; incubation behaviour; thermal tolerance.

5 This chapter is in the process of being submitted to the JournaI of Avian Biology as Blight, L.K. and T.D. Williams. Egg neglect in a member of the Alcidae: an experimental approach-

Introduction

Many species of birds rouünely interrupt their incubation shifts in order to

spend time away from the nest For passerines, these excursions may be only a few

minutes long (Haftom 1988). However, for seabirds such as procellariifomis and

alcids, breaks in incubation may occur over periods of a day or more, probably in

order to allow parents to increase foraging time at sea. Interruptions of greater than

24 hours are known as egg neglect (Gaston and Powell 1989). As tolerance to

chilling is a feature of many seabird embryos (Boersrna and Wheelwright 1979,

Warham 1990), temporary egg desertion may be less costly to these species than it

would first appear. La& (1967) hypothesized that resistance to embryonic chilling

in marine birds has allowed egg neglect to emerge as a life history strategy for a

group of organiçms that feeds on ephemeral oceanic resources, often at large

distances from the colony. In other words, the possibility of neglecting theV egg is a

strategy that allows parent birds to regulate the trade-off between risk of breeding

failure and risk of their own mortality.

Although egg age (i.e., stage of incubation) is thought to exert a strong effect

on ernbryo survivorship in the face of exposure to temperature extremes during

neglect, the degree of tolerance to cold stress has been shown to vary markedly for

the few avian species studied to date (Webb 1987). Many birds appear to require the

maintenance of relatively constant temperatures throughout incubation (Haftorn

1988), but in general little is known about the effects of temperature on embryo

s u ~ v a l and development for wild birds (Stoleson and Beissinger 1999). In alcids,

neglect seems to occur most frequently early in incubation. For example,

56

Astheirner (1991) found that for Cassin's Auklets (Ptychorumphus aleu ticus),

incidents of egg neglect were least frequent during the period of incubation that

corresponded to most rapid embryonic development, Le., in the latter two-thirds of

incubation. Wilson (1977) also found that the majority of neglect occurs early in

incubation for Rhinoceros Auklets (Cerorhinca monocera ta). While it is tempting

to assume that developmental constraints have restricted the evolution of

resistance to chillïng in older embryos (Webb 1987), decreasing neglect with embryo

age might also be explained by the parental investment hypothesis (Webb 1987),

since the amount of parental resources committed to a developing embryo

increases as incubation progresses, due to the limited tirne available for relaying.

In Our study, we experimentally chilled two separate groups of Rhinoceros

Auklet eggs removed from their nests either early or late in incubation. Our goal

was to examine whether the parents' tendency to interrupt incubation is

constrained by the changing developmental requirements of an older embryo, or

whether it might be affected by some other factor. We also wished to test whether

sumival, growth, or development of the avian embryo is affected by egg age at

neglect, as reported in Webb's (1987) review and in more recent research (e-g.,

Stoleson and Beissinger 1999). Among alcids, older embryos of Thick-billed Murres

(Uria lomvia) were killed by cooling, but in contrast, embryo age at neglect did not

affect hatchability for a s m d number of h u e n t Murrelet (Synthliboramphus

antiquus) eggs (Gaston and Powell 1989). Although it has long been suggested that

temporary chilling slows embryonic metabolism and increases the number of days

to hatch for alcids and other taxa (Murray et al 1989, Moreby 1995, Tombre and

Erikstad 1996)' the implications of intermittent egg neglect for embryo and neonate

viability have not been well shidied.

Study site and methods

Fieldwork took place at Triangle Island, British Columbia (50'52' N, 129°050

W) . Triangle Island is western Canada's largest seabird colony, and approximately

42,000 pairs of Rhinoceros Auklets breed there (Rodway 1991). In late April, before

laying by Rhinoceros Auklets commenced. we prepared 80 burrows by excavating

access holes at arm's-length intervals. These holes were then fitted with cedar

shingles and covered with earth. The entire burrow could thus be periodically

checked for the presence of an incubating bird.

In order to establish a sample of eggs with known lay dates, burrows were

checked a minimum of four (usually five) days per week. As Rhinoceros Auklets

lay only a single egg, each burrow represented no more than one experimental egg.

Once an egg was found, it was assigned to one of two treahnent groups for

experimental cldling. Eggs were altemately assigned to treatments in the order

they were found in order to avoid confounding effects of lay date. In order to avoid

only replacing abandoned 7-day chill eggs (found abandoned at time of treatment),

we also checked those due to be diilled at day 30 approxirnately one week after their

lay date. Any deserted burrows in the 30-day dull group were also replaced at this

time. Abandoned burrows were replaced by new ones, containhg known-age eggs,

in the order that they were found deserted.

Our treatments were similar to those used by Gaston and Powell (1989).

Experimental eggs were M e d to approximate burrow temperature (6-12°C; L.K.

Blight, unpublished data) for 48 hours at ca. seven or Ca. 30 days after laying. A s

empty burrows were checked for newly laid eggs nearly every day, eggs could be

aged to within one or two days. We chose 48 hours as our diilling period as we

wished to mimic natural durations of egg neglect. This thne penod is based on

actual, extreme durations of egg neglect recorded by us for this species (L.K. Blight,

unpublished data [see Chapter 51). Experimental eggs were removed from their

burrow at the assigned date and temporarily replaced with a previously warmed

substitute auklet egg. These substitute eggs had earlier been removed from other

burrows and the majority of them replaced by artificial eggs being used for

collection of data on nest attendance. Some substitute eggs were also removed

from foster burrows (see below). Experimental eggs were placed in an egg carton

and stored inside a Styrofoam cooler for the duration of the chilling period. We

kept the cooler outdoors in an open-sided, North-facing shed. In July, the warmest

month of Our study, daytime intemal cooler temperature did not exceed 13.0°C,

similar to ambient maximum burrow temperature of approxirnately 12°C for

Triangle Island Rhhoceros Auklet burrows &.K. BLight, unpublished data [see

Chap ter 51).

After 48 h, chilled eggs were slowly rewarmed to incubation temperature and

replaced in their original burrow. Occasionally, parent birds were absent from the

nest when the egg was replaced. Chilled eggs were nonetheless left in the nest

chamber, and small sticks were placed upright in the burrow entrante. The burrow

was then checked on the following day, and the nest was searched if displaced sticks

showed that it had been entered. If the burrow had been entered but no adult was

present, we assumed that the egg had been incubated on the previous night. We

repeated this monitoring on the second day fo110wing egg replacement, but if the

parent was stiU not present on the second day, we rewarmed the experimental egg

and fostered it to another, occupied burrow. If the sticks indicated that the burrow

had not been entered on the day following initial replacement, the egg was placed

in a foster burrow immediately- The foster parents' own eggs were removed and

were also used as replacement eggs (see above).

The incubation period for Rhinoceros Auklets ranges from 39-52 days

(Gaston and Dechesne 1996). We therefore began checks for hatdung chi& in

experimental burrows on the 36th day after the egg was found. Burrows were

checked every second day until the egg was found fully pipped, at which time it was

checked daily. AU chi& were found and rneasured within 24 hours of hatching.

As mass alone is not an accurate measurement for inter-individual cornparisons of

structural size (Piersma and Davidson 1991), we recorded a suite of body size

variables. Mass, tarsus and culmen length, wing chord, and "head-bill," i.e.,

distance from the occipital condyle to distal tip of the bill, were measured for al1

chicks. Chicks were processed as soon as they were found, with the exception of a

group of four found newly-hatched in stormy weather. These were weighed

immediately but we collected ail other data on the following day to avoid chiiling

of neonates.

As controls, we used a subset of eggs from a group of burrows being

monitored weekly for rate of egg depredation by rodents (Blight et al. 1999). These

burrows were manually checked every seven days throughout the incubation

period. Lay date for most of these burrows was known to w i t h seven to eight

days, although for a s m d fraction of our controls, greater than eight days elapsed

between the initial preparation of the empty burrow and the check when the egg

was first found (see below). In order to use these burrows as controls for our

experiment, the experimentd protocol of checking for and measuring chicks was

used once control eggs reached the fifth week of incubation. Hatching success and

chick size and mass were also recorded for all control burrows.

Swival of experimentai and control eggs to hatching was compared using

&-square analysis. In order to test our nd i hypothesis of no difference in mass

and size among experimental and control chi& at hatch, we performed a

multivariate analysis of variance (MANOVA), using chick rnass and

morphometrics as our variables. We also compared mass as a single variable

among the three groups using a one-way analysis of variance (ANOVA). Length of

incubation period among the three groups of embryos were compared using a one-

way ANOVA, and contrasted usïng Tukey's multiple cornparison tests to contrast

levels where factors were significant at a 0.05. As egg age of controls was only

known to within seven to eight days, we calculated their incubation period length

by assurning lay date as the mid-way point between the day on which the egg was

found and the previous nest check date. Where egg age of controls was known to

within >8 d (see above), these eggs were exduded from cornparisons of incubation

period length.

As some adult birds deserted substitute eggs, while others accepted (non-

viable) substitute eggs but refused to re-incubate their own, we had no reason to

suspect that failure by parents to accept an experimental egg back into the nest

indicated that the embryo was killed by diilluig. Instead, we assumed that if

parental desertion of an experimental egg happened immediately post-diilling it

was due to handling stress on the part of the parent Thus, experimental eggs that

failed due to not being reaccepted by their own or by foster parents were exduded

from o u analyses of hatching success. Values reported below are mean I SD.

Resul ts

Inequities in the number of eggs per treatrnent group that were abandoned by

parents pre-treatrnent and not accepted by foster parents, or that were depredated by

rodents, led to an unequal number of eggs in each treatrnent. In addition, two of

the eggs in the 30-day chil l group were broken during rewarming. In total, nine

eggs were chilled at 30 days of age, while 17 were chilled at seven days of age. We

chose 26 burrows to act as controls.

Twelve of the 17 eggs (71%) hatched from the 7-day chill group, and six of

nine eggs (67%) hatched from those chilled at 30 days of age. In the control group,

21 of 26 eggs (81%) hatched. Despite the disparity in amount and timing of chilling,

there was no difference in embryo survival among experimentally chïlled and

control groups ( x ~ ~ = 0.14, P = 0.93). Neither did we find a difference in size or mass

62

among neonates hatched from experirnentaliy dulled eggs or from non-dulled

controls (Wïlk's A= 0.71, F = 1.17,,,, P = 0.30). Mean neonate masses for the three

groups were the same (51.2 i 5.7 g, 52.1 2 5.9 g, and 51.3 i+ 3.2 g for 7-day and 30-day

chills and controls, respectively; one-way ANOVA, F = 0.09, P = 0.92). The two

lowest weights (36.0 g and 42.5 g) occurred in a neonate from each of the two

experimental groups; however, the single highest weight (60.0 g) also occurred in a

chick from the 30 d chi11 treatment.

Five of 21 control chi& hatched from burrows where egg age was not

known to within eight days, and these were exduded from comparisons of

incubation period length. Although there was no apparent difference among the

three groups in terms of sunrivorship or viability (as indicated by size and mass at

hatch), we did find a difference in incubation period between experimental eggs and

controls (one-way ANOVA, F = 4.53, , P = 0.02; pooled experimental data, F =

8.86133f P = 0.006). Eggs experirnentdly chilled during incubation at either seven or

30 days hatched two days later than controls on average. Incubation period was 44.7

+ 1.8 d and 42.5 + 2.5 d for experimental and control eggs, respectively-

Discussion

Life history theory predicts a trade-off between curent and future

reproductive potential, particularly for long-lived organisms such as seabirds

(Williams

reduction

1966, Steams 1992). For birds laying multiple egg clutches, brood

is available as a means of reducing parental energy expenditure when

breeding conditions are sub-optimal. Brood reduction is not an option for those

birds that have evolved a single egg clutch, however. During poor years, parents of

single egg broods rnay reduce the costs of reproduction either by choosing not to

breed, or by maintaining theh own body condition at the expense of the amount of

time spent feeding a du& (Sather et al. 1993) or incubatirtg an egg (Chaurand and

Weimerskirch 1994). Given the patchy and fluctuating nature of food availability at

sea, offshore-foraging seabirds rnay have an optimal dutch size of less than one egg

(Chastel et al. 1995). Thus, egg neglect mîght be an effective way of regulating the

trade-off between adult survival and annual production of Young.

The results of our experirnent appear to indicate that for Rhinoceros Auklets,

lengthy periods of neglect either early or late in incubation affect only hatch date,

and rnay have no effect on embryo viability. (In contrast, in at least one surface-

nesting alcid, the Thidc-billed Murre, resistance to egg chilLing is poorly developed

(Gaston and Powell 1989); variability in incubation period for this species (31-34 d;

Campbell et al. 1990) is minimal.) Webb (1987) stated that embryo age at exposure to

ambient temperature affects avian egg su~ivorship, but our results suggest that for

some seabird taxa this is not the case. Egg neglect rnay conceivably be employed

thrnughout incubation (and this appears to be the case for Rhinoceros Auklets,

particularly in poor food years, L.K. Blight, unpublished data [see Chapter 51) if

limited food availability and/or parental body condition warrant it. However,

there rnay be other costs of neglect to parents or embryos: the extended incubation

period of a chilled egg means that egg-neglecting parents rnay increase their risk of

predation through decreasing the benefit of breeding synchrony and predator

swamping. Chicks that fledge late may face an increased predation risk for the same

reason. At Triangle Island, neglect also incurs the potential cost of egg predation by

rodents (Blight et al. 1999). Breeduig auklets thus likely interrupt incubation only

when necessary for their own survival, such as early in incubation when body

condition has been comprornised by breeding effort.

If a protracted period of egg neglect is utilized by Rhinoceros Auklets early in

incubation, it may be a necessary response to the energetic demands of courtship

and egg production. Patterns of nest attendance tend to stabilize as incubation

progresses despite the apparently minimal cost of moderate neglect at any stage of

incubation. This occurs both in seabirds such as Ancient Murrelets and Rhinoceros

and Cassin's Auklets (Gaston and Powell 1989, Astheimer 1991, L.K. Blight,

unpublished data) and in avian species in general (Webb 1987). Rhinoceros

Auklets frequently neglect their single egg for the first 8 days of incubation, though

it is not known whether or not eggs are incubated at night (Wilson 1977). Other

alcid species also utilize egg neglect as a strategy early in incubation. Unlike most

alcids, Synthliboramphus murrelets lay a two egg clutch, and do not commence

incubation until about eight days after clutch initiation, one or more days after theV

second egg is laid (Murray et al. 1983, Gaston 1992). The eggs of the hvo

Synthliboramphzis murrelet speàes are exceptionally large (e.g., 21.9% of adult body

weight for Ancient Murrelets; Sealy 1975), and it is likely that neglect of the first egg

allows the female to increase foraging time and regain the capital required for the

production of the second egg (Sealy 1975, Murray et al. 1983). We propose that the

lengthy penod of initial neglect obsewed in the Rhinoceros Auklet similarly allows

parent birds to return to prebreeding condition prior to initiating continuous

incubation.

In light of our results, we also suggest that the hypothesis of inaeasing

parental investment better explains obsemed nest attendance patterns than do

explanations invoking increasing sensitivity to cooling in older embryos.

However, if it is the case that increasing parental investment leads to decreased

neglect, this does not d e out an eventual cost of egg neglect, and indeed implies it.

Our results indicate that sensitivity to cooling does not increase with age for

Rhuioceros Auklet embryos, but it is possible that indirect or cumulative costs of

egg neglect could result in the incubation patterns obsewed for this and other alud

species. Although we found no apparent impact of embryo cooling on neonate

viability, our study did not follow chicks hatched fi-om experimental eggs through

to fledging. Lindstrorn (1999) states that conditions experienced during early

development may affect later sumival and reproductive performance in mammals

and birds. Whether the slowing of embryo growth due to egg neglect results in

slower post-hatching growth - as it does in reptiles (Deerning and Ferguson 1991) -

and hence in delayed fledging and maturity of chicks (and ultimately in lower-

quality adult birds) remains to be tested.

Acknowledgements: This research was funded by a gan t from the Nestucca

Oil Spill Trust Fund to Douglas F. Bertram, by Natural Sciences and Engineering

Research Council (NSERC) operating grants to TDW and F. Cooke, and by an

NSERC post-graduate scholarship and a John K. Cooper Foundation research award

to LKB. We would like to thank the field crew at Triangle Island, particularly John

66

Ryder, Carina Gjerdrum and Boyd Pyper, for their companionship and assistance.

We are also grateful to the Canadian Coast Guard for providing us with ship and

helicopter support, and to the British Columbia Ministry of Environment, Lands

and Park for permission to work on the Triangle Island Ecological Reserve.

Chapter V6

Interannual variation in egg neglect and incubation routine of the

Rhinoceros Auklet

Introduction

Although there are dear aerodynamic advantages to the extemal

development of offspring in birds (ZRe et al. 1996), these are traded off against a

decreased availability of foraging time during incubation (Martin 1983).

Norietheless, the majority of bird species appear to maintain relatively stable

incubation temperatures after clutch completion (White and Kinney 1974), and do

not absent themselves from the nest for long enough to a h w egg temperatures to

fa11 below that at which no embryonic development takes place ("physiological

zero;" Haftom 1988). Unlike terreshiai birds, many species of seabird forage at

substantial distances from theV breeding colonies, and nest attendance by both

parents allows altemating foraging trips to sea during the often lengthy incubation

period. Limited food availability or poor weather conditions can sometirnes delay

seabirds in their return to the nest, however, and their eggs may be allowed to drop

to ambient temperature for periods of a day or more (Wheelwright and Boersma

1979, Gaston and Powell 1989, Warham 1990), even at localities where ambient

summer temperature may be as low as 0°C (Roberts 1940, Pefaur 1974). Repeated

This chapter is in preparation as a journal submission (BLight, L-K., D.F. Bertram and T.D. Williams. Interannual variation in egg neglect and incubation routine of the Rhinoceros Auklet) and has been written accordingly-

periods of egg neglect do not necessdy compromise embryo viability (Boersma

and Wheelwright 1979, Gaston and Powell 1989, Blight and Williams, in review),

and this inconsistent nest attendance during incubation has often been suggested as

an explanation for the wide intra-specific variation in incubation periods for some

marine birds (Boersma and Wheelwright 1979, Roby and Riddefs 1984, Sealy 1984,

Astheimer 1991), as well as for egg-laying vertebrates in general (Deeming and

Ferguson 1991).

Nest attentiveness inueases as the incubation period progresses for a

nurnber of avian taxa, particularly in relation to the time around ciutch completion

(Webb 1987, Stoleson and Beissinger 1999, Poussart et al. 2000). Similarly, egg

neglect in some seabirds is less common later in incubation (Gaston and Powell

1989, Astheimer 1991), although this does not appear to be the case for all marine

birds (Boersma and Wheelwright 1979, Chaurand and Wiemerskirch 1994). It has

been suggested that this pattern of decreasing neglect is related to more stringent

developmental requirements of older embryos (Webb 1987, Astheimer 1991), but it

may also be explained by other factors (Webb 1987, Blight and Williams, in review).

Variation in incubation behaviour should have consequences for embryonic

growth and development, and perhaps even for chïck growth and fledgling

performance, but few studies have continuously monitored the egg temperatures of

marine birds throughout incubation. In this study, I used artificial eggs containing

miniature temperature loggers to quantify nest attendance patterns of breeding

Rhinoceros Auklets (Cerorhinca monocerata) throughout the incubation period

over two seasons. My first study season (1998) coincided with a strong El Niiio

event and poor forage fish availability, while seabird breeding parameters for the

second year of research indicated above-average availability of food (Triangle Island

Research Station, unpubl. data). Since it has been suggested that egg neglect in

marine birds is primariiy determined by food availability (La& 1967), 1 predicted

that patterns of nest attendance at the colony would show a pronounced difference

between the two years of my study. 1 also predicted that egg neglect would vary

among individuals within years, probably due to a range of factors su& as age and

breeding experience. In addition, 1 wanted to test whether incidents of egg neglect

did actually decrease as incubation progressed, following the pattern reported for

other aluds (Murray et al. 1983, Gaston and Powell 1989, Astheirner 1991).

Study site and methods

Field work took place at the seabird colony at Triangle Island, British

Columbia (50°52' N, 129'05' W), from April to July of 1998 and 1999. Data were

collected from each of two study plots (located within 300 m of each other) in 1998,

and hom one of the two original plots in 1999. In 1998,I made artifiual eggs by

placing miniature temperature loggers (Hobo Tidbits, Onset Cornputer Corp., MA,

USA) inside hollow plastic hobby eggs. The two halves of each egg were filled with

an agar solution in order to approximate the thermal conductivity of a real egg, as

weLl as to hold the temperature logger in place. The eggs were glued shut and

painted white once the agar solution solidified. Mthough these artificial eggs were

about the minimm size of that reported for Rhinoceros Auklets (63.2 x 42.8 mm;

Gaston and Dechesne 1996), the rejection rate of artificial eggs in 1998 was greater

70

than the acceptance rate. For this reason, in 1999 1 manufactured eggs that were

slightly larger, and more elliptical in shape than those used the previous year.

These eggs measured 68.6 x 48.4 mm on average. Data loggers were encased in an

agar medium inside a plastic hobby egg, as they were in 1998, but each egg's shape

was amended using Dry-hard modelling clay and a thin coating of plaster of paris.

Eggs were then painted to guard against possible disintegration in the auklets' damp

nesting burrows. Data Loggers were set to read temperatures at intervals of 10

minutes in 1998, and 30 minutes in 1999, and recorded temperature to the nearest

0.0I0C. 1 compared the temperature sensitivity of the two different egg models by

placing two of each type in a drymg oven set at 35' C for 48 h.

In 1998,I checked auklet burrows once per week for newly-laid eggs,

beguuiing in late April. Most artificial eggs were substituted for the birds' own egg

within three days of their being found (i.e., within 5 10 d of lay date), but four

artificial eggs were inïtially field tested, and were not available to be placed until

mid-May, about a week after the eggs they replaced were found. In 1999, burrows

were monitored for a newly-laid egg at least twice a week and most had an artificial

egg placed within one week of the actual lay date. In 1998, artificial eggs were left in

place until the period of peak auklet chick hatch at the colony, while in 1999 they

were removed slightly earlier in the season, when the parents' real egg would have

been a minimum of 42 d old. This length of time approximates the average

incubation period of 45 d (range: 39-52 d; Wilson 1977) for the speùes. Leaving the

artificial eggs in place until they were abandoned by the parents would have

resulted in a more complete temperature profile for the incubation period in 1999,

71

but Rhinoceros Auklets may kick their egg out of the burrow after abandoning it

(pers. obs.). Two of the data loggers went missing in 1998, possibly due to this

behaviour. In all, burrows were monitored with the temperature loggers for at

least 40 d in 1998, while in 1999 all burrows were monitored for at least 36 d with aU

but two of these being monitored for 40 d or more. Field testing the artificial eggs in

1998 meant that they were generally placed later relative to individual lay dates

than they were in 1999, but in both years 1 attempted to nile out confounding effects

of parental experience by placing artificial eggs with birds whose lay dates spanned

the laying period. As Rhinoceros Auklets lay only a single egg, it was not possible

for me to collect simultaneous data on hatdiing success for rnonitored burrows.

1 defineci a period of egg neglect as one where artificial eggs were at ambient

burrow temperature for >3.5 h. (This value was not arbitrary as the shortest period

of actual neglect was 9 h.) This cut-off value was based upon apparent daytime

behaviour by incubating birds. Occasionally, egg temperature dropped steadily from

incubation temperature for short periods of time (53.5 h), either to ambient levels

or to fluctuate around a lower mean. These short-term events occurred during

daylight hours, when breeding birds do not corne and go from the colony, as well as

at night. 1 therefore assumed that bnef nighttime temperature drops did not always

indicate the adult's departue; they could also be due to a bird taking a temporary

recess within the burrow or to its not closely incubating the egg. When artificial

eggs recorded short-term temperature drops at night, they were likely the result of

asynchronous incubation exchmges, but since 1 was unable to ascertain their precise

nature, 1 did not consider them to be incidents of neglect per se and 1 excluded them

from my analyses. Artificial eggs that were incubated for only a day or two were

considered to be rejected, and their data were not incorporated into my results.

Statistical analvses

In order to examine differences in inter-annual kequency of egg neglect, I

used a two-taiied t-test to compare the mean nurnber of neglect events per burrow

for 1998 and 1999. In 1998, two of our monitored burrows were deserted

approximately halfway through incubation, and to increase the sample size for that

year 1 performed the analysis a second time and induded these two partial data sets.

The inter-annual variation in mean duration of neglect periods was also compared

using a two-tailed t-test. To test whether most egg neglect took place early in

incubation, as has been shown with other alcids, 1 used &-square analysis to

contrast incidents of neglect for each third of the incubation period. 1 considered

monitored days 0-10 to be the first third of incubation, as loggers were not placed

immediately after egg lay. My 1999 data were most representative of nest

attendance behaviour for the entire incubation period, since artificial eggs were

placed later relative to lay date in 1998. Due to this, and to the small number of

burrows sampled through to the end of the incubation period in 1998 (n=5), 1 used

only 1999 data for this latter analysis.

To look at inter-individual variance in incubation temperatures over entire

incubation periods, I calculated mean temperatures for all burrows, with neglect

periods included. As an estimate of natural Rhuioceros Auklet incubation

temperatures, 1 calcdated a mean value using combined 1998-1999 data from

artifiual eggs that were never neglected. 1 used data from these burrows o d y in

order to avoid including the intermediate values that were recorded as a neglected

egg was coolhg down, or being rewarmed by a parent. h order to determine

whether there were differences in temperature sensitivity between my two egg

models, I used two-tailed t-tests to contrast the mean &es required by each egg

type to stabilize at oven temperature, as well as the mean maximum temperatures

a ttained.

Values below are given as mean I SD.

Results

In 1998, at least 7 of 18 artificial eggs were accepted by incubating birds as

substitutes for their own egg. (As one artificial egg was "depredated" by ravens or

crows, the acceptance rate may have been slightly higher.) In 1999,I was successfd

in obtaining data with 13 of 20 Hobo eggs. When artificial eggs were left unattended

by the incubating adult, intemal temperatures rapidly dropped to the low ambient

temperature levels (range: 5.71°C - 12.23"C, depending on the individual burrow)

that were consistent with parental absence.

The temperature logger data showed pronounced differences in incubation

patterns between the two years of the study (Table 5-1; Figures 5-1 and 5-2), with

1998, an El NiIio year, being characterised by highly variable incubation behaviour,

including nest abandonment approximately halfway through the incubation period

in 28.5% (2 of 7) of the monitored burrows. In 1999, al1 artificiai eggs were incubated

until the day I removed them, with the exception of one that appeared to have been

74

Table 5-1. Iiicu bation rcsults of s~icccssf~illy-ylaced artificial Rfiinoceros Auket c g p , 1998-1999,

Uiirrow Datc Dn tc Date No. iieglect Date nbandoiied/ iirimbcr cggfoiind* deployed incubated events Abandoned? rcniavcd

14 May 14 May 11 May 07 May 04 May 14 May

?A

13 May ? ? ?

09 May 04 May 16 May 16 May O9 May 09 May 08 May 04 May 02 May

16 May 16 May 13 May 16 May 10 May 16 May 14 May

20 May 14 May 17 May 21 May 10 May 04 May 17 May 17 May 10 May 10 May 10 Miiy 04 May 03 May

16 May 16 May 13 May 16 May 11 May 16 May 14 May

20 May 14 May 17 May 21 May 10 May 07 May 17 May 17 May 10 May 10 May 10 May 04 May 03 May

25 Junc 25 Juiie 24 June 25 J uiie 24 Junc

30 May/ 12 Jiiiie

31 May/24 June

01 July 25 Jiinc 26 June

02 JiiLy/04 July 16 June 14 Junc 29 Junc 28 jiiiic 24 Juiie 25 Juiie 15 June 13 Jiine 13 Junc

'Approxiinates lay date in 1999, when burrows checked at least twice per week; burrows were checked every 7 d in 1998. "Lay date of ? wlicre burrows grubbed later in lay period and egg found on grubbing date,

"incubation" begins

-24 h neglect period

0 5 / 3 0 date

Figure 5-1. Temperature profile of an incubated artificial auklet egg, 1998.

76

"@cubationn begins

artificial egg removed

-24 h neglect period

date

Figure 5-2. Temperature profile of an incubated artificial auklet egg, 1999. The data logger had been launched several days before the egg rvas placed in the auklet nest.

77

abandoned for 2.5 d before 1 retrieved it from the nest. The number of periods of

neglect per burrow in 1998 ranged from O - 17. Each period of negled lasted about

one day on average (see below). In 1999, the total number of days of neglect per

burrow only ranged from O - 3. The mean number of incidents of neglect per

monitored burrow (6.14 f 7.01 and 1.38 + 1.12 for 1998 and 1999, respectively)

differed signihcantly between years, whether the two 1998 burrows abandoned

h a h a y through incubation were induded (P = 0.025, t l s = 2.45) or not (P = 0.019, tl6

= 2.61). The artificial eggs were not deployed as soon as the auklets' real eggs were

laid, so 1 inevitably did not record all instances of parental absence from the burrow.

Although on average, srtificial eggs were placed in burrows somewhat earlier

(relative to the individual lay date of birds' real eggs) in 1999 than in 1998,I am

confident that the data loggers recorded real inter-annual differences in incubation

routine. As most egg neglect appears to take place earlier in incubation (see below),

1 would expect any bias in the data to somewhat underestimate neglect frequency in

1998 relative to 1999: if anything, frequency of egg neglect was even more common

in 1998 than is reported here.

Despite the greater frequency of neglect in 1998, the duration of individual

periods of parental absence was the same for each year (21.8 f 8.7 h and 22.0 + 11.7 h

for 1998 and 1999, respectively; P = 0.93, tse =-0.09). In addition, a similar percentage

of burrows each year (29% in 1998 and 31% in 1999) experienced no egg neglect

during the monitored period. The maximum length of time that an egg was left

unattended in 1998 was 49 h, while in the second year of my study, both parents

were absent from one burrow for 69 h. The minimum duration of egg neglect over

78

the two years of the study was 9 h. In 1999,I placed most artifiaal eggs (10 of 13, or

77%) within one week of individual lay dates. 1999 data, more indicative of

incubation behaviour throughout the entire nesting period than were those

collected the previous year, indicated that the majority of incidents of neglect

occurred during the first thud (monitored days O - 10) of the incubation period (P =

0.013, x2, = 8.67; Figure 53).

Mean incubation temperatures for alI artificial eggs ranged from a low of

25.72 + 9.70°C, for an egg neglected 17 times over the incubation period, to a high of

34.06 tr 0.93"C for an artificial egg that recorded no neglect. The mean incubation

temperature for ail non-neglected artificial eggs in my study (n=6) was 33.11 f 1.36"C

(range 30.47 + 1.06"C - 34.06 t 0.93"C), but removing the single outiying value

(30.47"C) from the data set increased the average temperature slightly to 33.64 + 0.45"C. In 1999, the temperature profiles of four artifiual eggs showed an increase

in "incubation" temperature over time (Figure 5-4), but the remairting profiles,

including those from 1998, showed no such trend.

The data recorded while testing both egg models in a drying oven revealed

no differences in either their ability to record ambient temperatures, or the time

required by them to reach equilibrium. Both the plastic and the Dry-hard clay egg

models attained the same maximum temperature (P = 0.88, t z = 0.17), and attained it

in the same amount of tirne (P = 0.63,12 = 0.56).

- -

Proportion of Burrows Showing Neglect

Figure 5-3. Distribution of neglect events in relation to stage of incubation, 1999.

artificial eg 1 removed -

-

-

I I 1 1 I 1 0 4 / 2 4 0 5 / 0 4 OS/ 14 OS/24 06/03 06/13 06/23 1999

date

Figure 5-4. Temperature profile of an incubated artificial auklet egg, showhg increase in incubation temperature over time.

Discussion

The temperature logger data show that there were pronounced differences in

individual Rhirtoceros Auklet incubation patterns between the two years of my

study. In 1998, an El Niiio year, frequent and repeated egg neglect by nesting auklets

was commonplace, likely due to the difficulty that some breeding birds had in

locating forage fish during their time away from the nest. Lengthy absences from

the nest preceded nest abandonment on two occasions. This behaviour is

consistent with colony-wide data on chi& growth and diet for 1998 (Figure 5-5;

Bertram et al., in review) indicating that it was a year of extremely poor food

availability (also from Triangle Island Research Station, unpubl. data). In 1999, a

year of above-average breeding success (Figure 5-5; Bertram et al., in review), egg

neglect was still recorded for a similar percentage of burrows, albeit with greatly

reduced frequency. Periodic egg neglect appears to be a consistent feature of

Rhinoceros Auklet breeding biology, but its frequency is greatly reduced when food

is readily available.

The sample sizes from the first year of my study were not large enough to

allow me to analyze intra-seasonal nest attendance patterns for 1998, but in 1999 the

majority of egg neglect took place during the first third of incubation (Figure 5-3).

Previous studies have noted that alcids are more likely to leave their eggs

unattended early in incubation, and this has been attributed to poor initial

coordination of incubation exchanges on the part of inexperienced breeders

(Wilson 1977); more stringent developmental requirements of older embryos

(Webb 1987, Astheimer 1991); facilitating the synchronous hatdiing of young (Sealy

YEAR

Figure 5-5. Growth rate anomalies for Rhinoceros Auklet (Cerorhinca monocerata) chicks, Triangle Island, British Columbia, 1994-1999. From data in Bertram e t al. (in review) .

1984); or allowing the female to regain the capital required to lay a dutch's second

egg (Murray et al. 1983). Based on colony-wide trends in breeding success (Bertram

et al., in review), it seems likely that the Rhinoceros Auklets monitored in this

study appeared to adjust their inter-annual nest attendance patterns in accordance

with at-sea food availability. I suggest that greater neglect early in incubation is

likely correlated with intra-seasonal cycling of forage fish availabiüty near the

colony, or the body condition of breeding birds early in incubation. Rhinuceros

Auklets produce an egg that is proportionately larger than that of most other aluds

(Sealy 1975). Egg production rnay be an energetically costly period for this species, as

it likely is for S ynthliborarnphus murrelets (Sealy 1975, Murray et al. 1983), and may

sometimes require extra foraging time at sea.

1 predicted that the amount of egg neglect recorded would Vary among

individuais within years, likely due to factors such as age and breeding experience.

My sample size for 1998 was small, but inter-individual variation in incubation

patterns appeared to be extreme: some auklet pairs left their egg unattended every

two to three days, while others consistently incubated theirs. Although 1 was

unable to assess the age or experience of any of the monitored birds, my data suggest

that a small proportion of breeding pairs at a colony do not need to neglect their

eggs in order to increase foraging time, even in years of dramatically reduced food

availability. In 1999, while inter-individual variation in incubation patterns was

small, it was still apparent. Even in a "good" food year, most breeding pairs were

unable to maintain constant egg temperatures throughout the incubation period.

Rhinoceros Auklets lay only a single egg. As 1 replaced real eggs with

artificial substitutes in my shidy nests, I could not determine whether monitored

birds wodd have hatched a chi& despite their erratic presence at the nest.

However, in 1999 we experimentally diilled auklet eggs for 48 h at different stages

of incubation (Blight and Williams, in review [Chapter 4]), and hatchability did not

appear to be affected. Survival of up to four days of continuous neglect has been

reported for other members of the Alcidae (Murray et al. 1979, Gaston and Powell

1989) and Rhùioceros Auklet eggs have also been obsemed hatdung after continual

periods of natural neglect lasthg up to four or five days (Sumrners and Drent 1979).

Embryonic tolerance of periodic chilling can be highly adaptive for seabird species

that feed on ephemeral resources at a distance from the colony. Had the monitored

birds been incubating real eggs, the embryos would Likely have survived the

prolonged absences recorded by the data loggers.

According to the temperatures I recorded for non-neglected artifiüal eggs,

mean incubation temperature for Rhinoceros AuWets is approximately 33 - 34°C.

This temperature falls within the range of incubation temperatures (32.3 - 37.4"C)

reported for other seabird species (Riddefs 19&4), but 1 suggest that the artifiual eggs

slightly underestimated actual incubation temperatures. The thermistor on the

data loggers was situated so that egg huning by the incubating parent wodd have

led to its periodic orientation away from the auklet's brood patch and towards the

lower temperature of the burrow floor. Thus, 1 expect that the real temperattue of

an embryo, which lies dose to the upper surface of the incubated egg (Drent 1975),

would be somewhat higher than the mean values that 1 recorded. One artificial egg

recorded an anomalously low (about 30.S°C) mean incubation temperature for one

of the non-neglected eggs. This temperature suggests the existence of fa*ly extreme

inter-individual variation in Rhinoceros Auklet incubation temperahues. It is

noteworthy that some of the temperature profiles showed a progressive increase in

"incubation" temperature over t h e . This trend has been observed for the

incubation periods of other seabird species (Grant et al. 1982, W e f s 1984), but has

been at least partially attributed to the metabolic heat contributed by more

developed embryos (Drent 1970, Grant et al. 1982). The data obtained from the

artificial eggs used in this study indicates that tighter incubation of older embryos

(or increasing vascularization of the brood patch) could also contibute to this

phenornenon.

The maximum duration of continuous neglect for Rhinoceros Auklets at

Triangle Island was about three days, much shorter than the maximum of seven

consecutive days recorded for a viable egg of the highly pelagic Fork-tailed Storm

Petrel (Oceanodrorna fzircata; Boersma et al. 1980), but similar to that recorded for

other alcids (e.g., Murray et al. 1979). However, the mean duration that 1 report for

each recess is comparable to that recorded for Wilson's Storm Petrels (Oceanites

ocennicus) at Robert Island, Antarctica, where the mean duration of neglect for two

successful nests was 22.5 and 28.5 h (Pefaur 1974). This similarity likely reflects the

fact that both storm petrels and Rhinoceros Auklets only return to the colony at

night, so that an incubating parent departing from the nest before its mate returns

is most likely to be leaving its egg unattended for an entire day, until sometime

during the following evening.

Selection pressures such as time-dependent mortality and sibling rivalry

have favoured the evolution of more rapid avian incubation (Ricklefs and Starck

1998)' but for marine birds a disproportionately lengthy incubation period îs the

n o m (Ricklefs 1984). While this phenornenon is apparent even ui the absence of

egg neglect (Ricklefs 1984), periodic chilling of the embryo to low ambient

temperatures does dernonstrably inuease the incubation period by a length of time

cornmensurate with the amount of cooling (Boersrna and Wheelwright 1979,

Blight and Williams, in review). Relaxed selection pressures, e.g., absence of

terrestrial predators on islands, has Likely allowed the evolution of embryonic

resistance to periodic bouts of neglect (or its retention as an ancestral reptilian trait).

This hypothesis is supported by the reduced resistance to M i n g in older embryos

of surface-nesting seabirds such as Thick-billed Murres (Uria lomuia) (Gaston and

Powell 1989) and Southem Giant Petrels (Macronectes gigantetis; Roby and Ricklefs

1984). Native terrestrial predators are largely absent at most seabird colonies, but

aerial ones may be comrnon, and while the nests of burrow-nesting species such as

Rhinoceros Auklets are not at risk from avian predators, birds that nest in the open

cannot leave their nests unattended without risking egg loss.

Previous work on pelagic seabirds suggests that the deusion to temporarily

desert an egg is assouated with the parent readiing a lower mass threshold, below

which the adult's own survival is compromised (Chaurand and Weimerskirch

1994). The inter-annual variation in egg neglect frequency employed by Rhinoceros

Auklets breeding at Triangle Island over two years of contrasting food availability

suggests that auklets' decision to neglect their egg may also be mediated by dianges

in adult body condition.

Chapter VI

Conclusion

"Even when predators kill a large proportion of their prey each year, they do

not necessarily reduce breeding numbers" (Newton 1998). One of the objectives of

my thesis was to quanti.& the effects of egg depredation on breeding efforts by

Rhinoceros Auklets at the Triangle Island colony. Predation by Peromyscus -

confirmed in this study both through the use of an infrared monitoring camera and

stable isotope analysis - removes a substantial proportion of the eggs laid there each

year. Despite thïs, various population estimates for the colony show stable (D.F.

Bertram, pers. comm.) or increasing Rhinoceros Auklet numbers (Rodway et al.

1992). These population data indicate that egg loss through predation, though high,

likely replaces other possible sources of mortality (Le., it is compensatory; Newton

1998), such as post-fledging starvation, for young-of-the-year, rather than being an

additional source of mortality unto itself.

Although my a priori expectation was that the rate of egg predation would

Vary in accordance with vegetation cover at Triangle Island, as per the situation

described by Murray (1980), this was not the case. The egg depredation rate was the

same between the two differently-vegetated study plots in both 1998 (Chapter 1) and

1999 (19.2% vs. 23.1% depredation for two study plots, missing eggs exduded; x2, =

0.14, P = 0.71). Neither did my study show any inter-annual variation in the rate of

predation among the two study plots over two years (13.7% and 21.5% depredation

overall for 1998 and 1999, respectively, missing eggs exduded; xZ3 = 3.65, P = 0.30),

89

despite the fact that egg neglect (and hence predation opportwiity) decreased

signihcantly in the second year. Although the density of the mouse population did

not change over the months of the study (for Plot 1, xZ, = 1.33, P = 0.25; for Plot 2, X:

= 3.37, P = 0.07; 1998 data), the majority of eggs were depredated early in incubation

(Chapter 2), indicatuig that something other than the density of rodent populations

per se drives egg predation at Triangle Island. Mice may eat fewer auklet eggs later

in incubation due to decreased egg neglect and hence egg availability, as suggest in

Chapter 1, or, aitematively, they may simply adopt an optimal foraging approach

and partially switch to energy-rich, easily-accessible plant foods as the season

progresses and these items become available. The results of our stable isotope

analyses indicated that this diet switch is not complete, however, as they showed

that rnice rely on seabirds (and likely on dead chi& as well as eggs) as a source of

food throughout the summer.

A further objective of my research was to determine whether Rhinoceros

Aukle ts facultatively adjusted egg neglect in response to perceived (egg) predation

risk. Testing this hypothesis in the field proved unworkable, but results of other

aspects of my research led me to infer that breeding aduits neglect their eggs

independent of predation risk and instead may do so in response to changes in

their own body condition and/or the availability of food at sea. In the early part of

the incubation period, when egg loss to rodent predation was the highest, iive

trapping results did not indicate significant inter-annual differences in mouse

density (z = 1.64, P = 0.10; results for late May trapping, 1998 and 1999). Despite this,

nest attendance patterns varied considerably between years, in apparent accordance

90

with offshore oceanic conditions and forage fish availability. My suggestion that

breeding birds neglect their egg independent of predation risk but in response to

their own body condition is consistent with the maxirnization of individual fitness

that is expected in a longer-iïved species of bird (Williams 1966; Charnov and Krebs

1973; Stearns 1992), and with the results of other studies that correlate mass loss

with temporary nest desertions in pelagic seabirds (Chaurand and Weimerskïrch

1994). Similarly, it is to be expected that the potentiai loss of a yearfs production of

young is less costly to a population of long-lived birds than it would be to a species

whose lifetime reproductive output OCCLUS over only two or three seasons.

My use of miniature temperature loggers inside artificial eggs showed wide

inter-seasonal and inter-individual variation in Rhinoceros Auklet nest attendance

patterns. Some pairs of birds appeared capable of incubating their egg without a

single recess during the incubation period, while other pairs neglected their egg

fr equently, par ticularly during the El Niiio conditions that prevailed in 1998.

Temperature loggers inside artificial eggs proved to be an effective way of

continually monitoring parental nest attendance throughout the entire incubation

period, and could be recommended as a technique for nest monitoring in other

situations.

Understanding the factors that influence the varying reproductive success of

seabirds is critical to their management and conservation, particularly given recent

climate change scenarios. 1 have inferred here that depredation by Keen's mice

does not appear to be adversely affecting the Rhinoceros Auklet population at

Triangle Island, but making a more definitive statement is beyond the scope of this

91

research. Nonetheless, the data collected for this study have the potential to be

used in a predictive mode1 designed to quantify impacts of dimate change on

Rhinoceros Auklet populations. In a broader conceptual setting, information on

how seabirds respond to rodent predation in a "natural" situation, i.e., where

predator and prey have CO-evolved, could also be incorporated into m o d e k g

seabird responses to introduced rodent predators. This could prove useful in

decision-making on timing or pnority-setting around rodent eradication efforts at

seabird colonies where depredation by introduced rodents has led to population

declines-

Literature Cited

Adler, G.H. and R. Levins. 1994. The island syndrome in rodent populations. Q. Rev. Biol. 69:473-490,

Ambrose, S.H. and L. Nom. 1993. Experimentai evidence for the relationship of carbon isotope ratios of whole diet and dietary protein to those of bone collagen and carbonate, pp. 1-38. In J.B. Lambert and G. Gmpe (eds.), Prehistoric human bone: archaeology at the molecular level. Sp ringer-Verlag, Berlin.

Astheimer, L.B. 1991. Embryo metabolism and egg neglect in Cassin's auklets. Condor 93:486-495.

Batzli, (2.0. 1985. Nutrition, p. 779-811. Zn R.H. Tamarin (ed.), Biology of new world Microtus. Spec. Publ. No. 8, Am. Soc. of Mamrnalogists, Boston.

Ben-David, M., R.T. Bowyer, L.K. Duffy, D.D. Roby and D.M. Sdieli. 1998. Social behavior and ecosystem processes: river otter latrines and nutrient d ynamics of terres trial vegetation. Ecology 79:2567-2571.

Ben-David, M., R.W. Flynn and D.M. çchell. 1997a. Annual and seasonal changes in diets of martens: evidence from stable isotope analysis. Oecologia 1 ll:280-291.

Ben-David, M., T.A. Hanley, D.R. nein and D.M. Schell. 1997b. Seasonal changes in diets of coastal and riverine mink: the role of spawning Pacific salmon. Cm. J. 2001. 752303-811.

Bertram, D.F., D.L. Mackas and S. McKinnell. The seasonal cycle revisited: interannual variation and ecosystem consequences. Progr. Oceanography. (In review).

Bligh, E.G. and W.J. Dyer. 1959. A rapid method of total iipid extraction and purification. Cm. J. Biochem. Physiol. 373911-917.

Blight, L.K., J.L. Ryder and D.F. Bertram. 1999. Depredation on Rhinoceros Auklet eggs by a native population of Perornyscus. Condor 101:871-876.

Blight, L.K. and T.D. Williams. Egg neglect in a member of the Aludae: an experimental approach. (In submission).

Boersrna, P.D. and N.T. Wheelwright. 1979. Egg neglect in the procellariiformes: reproductive adaptations in the fork-tailed storm- petrel. Condor 81:157-165.

Boersrna, P.D., NT- Wheelwright, M.K. Nerini and ES. Wheelwright. 1980. The breeding biology of the Fork-tailed Storm Petrel (Oceanodrorna fiirca ta). Auk 97:268-282.

Bures, S. 1997. High cornmon vole (Microtus aroulis) predation on ground- nesting bird eggs and nesthgs. Ibis 139:173-174.

Byrd, GY., J.L. Trapp and C.F. Zeillemaker. 1994. Removd of introduced foxes: a case study in restoration of native birds. Tram. N. Amer. Wildl. Nat. Res. Conf. 59:317-321-

Campbell, R. W., N.K. Dawe, 1. McTaggart-Cowan, J.M. Cooper, G.W. Kaiser and M.C.E. MCNall- 1990. The birds of British Columbia, Volume 2. Nonpasserines: Diurnal birds of prey through woodpedcers. UBC Press, Vancouver, British Columbia.

Carl, G.C., C.J. Guiguet and GA. Hardy. 1951. Biology of the Scott Island Group, British Columbia. Rep. Provincial Mus., Victoria, British Columbia.

Charnov, E.L. and J.R. Krebs. 1973. On clutch-size and fitness. Ibis 116:217-219.

Chastel, O., H. Weimerskirch and P. Jouventin. 1995. Body condition and seabird reproductive performance: a study of three petrel species. Ecology 762240-2246.

Chaurand, T. and Weimerskirch, H. 1994. Incubation routine, body mass regulation and egg neglect in the Blue Petrel Halobnena caerulea. Ibis 136:285-290.

Clark, D.A. 1981. Foraging patterns of black rats across a desert-montane forest gradient in the Galapagos Islands. Biotropica 13: 182-194.

Cocks, M.P., I.P. Newton and W.D. Stock. 1998. Bird effects on organic processes in soils from five microhabitats on a nunatak with and without breeding snow petrels in Dronning Maud Land, Antarctica. Polar Biol. 20:112-120.

Cowan, LM. and C.J. Guiguet. 1975. The marnmals of British Columbia. Handbook No. 11 British Columbia Provincial Mus., Victoria, British Columbia.

Craig, D. P. 1998. Chipmunks use leverage to eat oversized eggs: support for the use of quail eggs in artificial nest studies. Auk 115:486-489.

Deeming, D. C. and M.W.J. Ferguson. 1991. Physiological effects of incubation temperature on embryonic development in reptiles and birds, pp. 147- 171. In D. C. Deeming and M.W.J. Ferguson (eds.), Egg incubation: its effects on embryonic development in birds and reptiles. Cambridge University Press, Cambridge, England.

DeGraaf, R. M., and T. J. Maier. 1996. Effect of egg size on predation by white- footed mice. Wilson Bull. 108:535-539.

DeNiro, M.J. and S. Epstein. 1978. Influence of diet on the distribution of carbon isotopes in animals. GeodUm. Cosmodum. Acta 42:495-506.

DeNiro, M.J. and S. Epstein. 1981. Influence of diet on the distribution of nitrogen isotopes in anirnals. Geochim. Cosmochim. Acta 45:341-351.

Drennan, J.E., P. Beier and N.L. Dodd. 1998. Use of track stations to index abundance of sciurids. J. Mammal. 79:353-359.

Drent, R.H. 1970. Functional aspects of incubation in the herring gull (Larus nrgentatzis Pont.). Behav. (Suppl.) 17:l-132.

Drent, R.H. 1975. Incubation, pp. 333-420. In D.S. Farner and J.R. King (eds.), Avian biology. Vol. 5. Academic Press, New York.

Drever, M.C. and A.S. Harestad. 1998. Diets of Norway rats, Rattzis norvegict~s, on Langara Island, Queen Charlotte Islands, British Columbia: implications for conservation of breeding seabirds. Can. Field Nat. 112:676-683.

Duffy, D.C. and S. Jackson. 1986. Diet studies on seabirds: a review of methods. Col. Waterbirds 9:l-17.

France, R.L. 1995. lakes using

Differentiation between littoral stable carbon isotopes. Limnol.

and pelagic food webs in Oceanogr. 40:1310-1313.

Fumess, R.W. and P. Monaghan. 1987. Seabird ecology. Blackie and Son, Glasgow.

Gannes, L.Z., D.M. O'Brien and C. Martuiez del Rio. 1997. Stable isotopes in animal ecology: assumptiow, caveats, and a cal1 for more laboratory experiments. Ecology 78:1271-1276.

Garrettson, P R , F.C. Rohwer, J.M. Zimmer, B-J. Mense and N. Dion. 1996. Effects of mammalian predator rernoval on waterfowl and non-game birds in North Dakota. Trans. N. Amer- Wildl. Nat. Res. Conf. 65:94 101-

Gaston, A.J. 1992. The AnCient Murrelet: a natural history in the Queen Charlotte Islands. T. and A.D. Poyser, London.

Gaston, A.J. and S.C.B. Dechesne. 1996. Rhinoceros auklet (Cerorhinca monocerata). In A. Poole and F. GU (eds.), The Birds of North America, Xo. 212. The Academy of Natural Sciences, Philadelphia, and The American Ornithologists' Union, Washington, DC.

Gaston, A.J., I.L. Jones and 1. Lewington. 1998. The auks (Alcidae: bird families of the World, 4). Oxford University Press, Oxford.

Gaston, A. J. and D.W. Powell. 1989. Natural incubation, egg neglect, and hatchability in the anuent murrelet. Auk 106:433-438.

Gliwicz, J. 1980. Island populations of rodents: their organization and functioning. Biol. Rev. 55:109-138.

Grant, G.S., T.N. Pettit, H. Rahn, G. C. W t t o w and C. V. Paganelli. 1982. Water loss from Laysan and Black-footed Albatross eggs. Physiol. Zool. 55:405-414.

Guillory, H.G. 1987. Cavity cornpetition and suspected predation on Protonothary Warblers by Peromyscns. J. Field Ornithol. 58:425-427.

Haftorn, S. 1988. Incubating female passerines do not let the egg temperature fa11 below the 'physiological zero temperature' during their absences from the nest. Omis Scand. 19:97-110.

Haskell, D.G. 1995. Forest fragmentation and nest predation: are experiments with Japanese Quail eggs rnisleading? Auk 112:767-770.

Hüderbrand, GY., S.D. Farley, C.T. Robbins, T.A. Hanley, K. Titus and C. Servheen. 1996. Use of stable isotopes to determine diets of Living and extinct bears. Cm. J. 2001. 742080-2088.

Hobson, K.A. 1986. Use of stable-carbon isotope analysis to estimate marine and terreshial protein content in gull diets. Cm. J. 2001. 659397-903.

Hobson, K.A. 1991. Stable isotopic determinations of the trophic relationships of seabirds: prelllninary investigations of alcids from coastal British Columbia, pp. 16-20. In W.A. Montevecchi and A. J. Gaston (eds.), Studies of high-latitude seabirds. 1. Behavioural, energetic and oceanographic aspects of seabird feeding ecology. Can. Wild. Sem. Occas. Paper 68:16-20, Ottawa.

Hobson, K.A. 1993. Trophic relationships among high Arctic seabirds: inçights from tissue-dependent stable-isotope models. Mar. Ecol. Prog. Ser. 95:7- 18.

Hobson, K.A. 1999. Stable-carbon and nitrogen isotope ratios of songbird feathers grown in two terrestrial biomes: implications for evaluating trophic relationships and breeding origins. Condor 101:799-805.

Hobson, K.A. and R.G. Clark. 1992. Assessing avian diets using stable isotopes 1: turnover of 13C in tissues. Condor 94:181-188.

Hobson, KA., MC. Drever and G.W. Kaiser. 1999. Nonvay rats as predators of burrow-nesting seabirds: insights fiom stable isotope analyses. J. Wildl. Man. 63:14-25.

Hobson, K.A., J.F. Piatt and J. Pitoccheli. 1994. Using stable isotopes to determine seabird trophic relationships. J. Anim. Ecol. 63:786-798.

Hobson, K.A. and S.G. Sealy. 1991. Marine protein contributions to the diet of northern Saw-whet Owls on the Queen Charlotte Islands: a stable- isotope approach. Auk 108:437-440.

Hobson, K.A., and 1. Stirling. 1997. Low variation in blood 613C among Hudson Bay polar bears: implications for metabolism and tracing terrestrial foraging. Mar. Mam. S i . 13:359-367.

Hogan, KM., M C Hedin, HS. Koh, S.K. Davis and I.F. Greenbaum. 1993. Systematic and taxonomie implications of karyotypic, electrophoretic, and mitochondrial-DNA variation in Peromyscus from the Pacific Northwes t. J. Mammal. 74:819-831.

Johnstone, G.W. 1985. Threats to birds on sub-antarctic islands, p. 101-121. In Int Council Bird Presem. Tech. Publ. 3, Cambridge, England.

Kaiser, G.W. and L.S. Forbes. 1992. C h a t i c and oceanographic influences on island use in four burrow-nestïng aluds. Omis Scand. 23:l-6.

Kline, TC. Jr., J.J. Goering, O.A. Mathisen, P.H. Poe and P.L. Parker. 1990. Recyding of elements transported upstream by runs of Pacific salmon: 1. 615N and tjUC evidence in Sashin Creek, Southeastem Alaska. Cm. J. Fish. Aquat. SU. 47:136-144.

Kline, T.C. Jr., J. J. Goering, O. A. Mathwn, P.H. Poe, P.L. Parker and R.S. Scalan. 1993. Recycling of elements transported upstrearn by runs of Pacific salmon: II. 615N and 613C evidence in the Kvichak River watershed, Bristol Bay, Southwestern Alaska. Can. J. Fish. Aquat. Su. 50:2350-2365.

Lack, D. 1967. Interrelationships in breeding adaptations as shown by marine birds. Proc. Int. Omithol. Congr. 14:3-42.

Larivière, S. 1999. Reasons why predators cannot be inferred from nest remains. Condor 101:718-721.

Lee, S.J., M.S. Witter, I.C. Cuthill and A.R. Goldsmith. 1996. Reduction in escape performance as a cost of reproduction in gravid starlings, Sturnus vulgaris. Proc. R. SOC. B 2633619424.

Lindstrom, J. 1999. Early developrnent and fitness in birds and mammals. Trends Ecol. Evol. 14543-348.

Lutemauer, J.L., J.J. Clague, K.W. Conway, J.V. Barrie, B. Blaise and R.W. Mathewes. 1989. Late Pleistocene terrestrial deposits on the continental shelf of western Canada: evidence for rapid sea-level change at the end of the last glaciation. Geology 17:357-360.

Maier, T.J. and R.M. DeGraaf. 2000. Predation on Japanese Quail vs. House Sparrow eggs in artificial nests: small eggs reveal srnall predators. Condor 102:325-332.

Martin, T.E. 1984. Competition in breeding birds: on the importance of considering processes at the level of the individual, pp. 181-210. In R.F. Johnston (ed.), Current Ornithology, Volume 4. Plenum Press, New York.

Maxson, S.J. and L.W. Oring. 1978. Mice as a source of egg loss among ground- nesting birds. Auk 95:582-584.

Michener, R.H. and D.M. Schell. 1994. Stable isotope ratios as tracers in marine aquatic food webs, pp.138-157. In K. Lajtha and R.H. Michexter (eds.), Stable isotopes in ecology and environmental science. Bladcwell Scientific Publications, Oxford.

Minagawa, M., and E. Wada. 1984. Stepwise enrichment of 15N along food chains: further evidence and the relation between 615N and animal age. Geochim. Cosmochim. Acta 48:1135-1140.

Mizutani, H., H. Hasewaga and E. Wada. 1986. High nitrogen isotope ratio for soils of seabird rookenes. Biogeochem. 2221-247.

Mizutani, H and E. Wada. 1988. Nitrogen and carbon isotope ratios in seabird rookeries and their ecological implications. Ecology 69:340-349.

Moors, P.J. and I.A.E. Atkinson. 1984. Predation on seabirds by introduced animals, and factors affecting its severity, pp.667-690. In J.P. Croxall, P.G.H. Evans and R.H. Schreiber (eds.), Status and consemation of the world's seabirds. Int. Council Bird Preserv. Tech. Publ. No. 2, Cambridge, England.

Morbey, Y.E. 1995. Fledging variability and the application of fledging models to the behaviour of Cassin's auklets (Ptychoramphris aleüticzis) at Triangle Island, British Columbia. M . S . thesis, Simon Fraser University, Bumaby, British Columbia.

Mueller-Dombois, D. and H. Eilenberg. 1974. Aims and methods in vegetation ecology. John Wiley and Sons, New York.

Munro, J.A. and 1. McTaggart-Cowan. 1947. A review of the bird fauna of British Columbia. British Columbia Provincial Mus. Spec. Publ. No. 2, Victoria, British Columbia.

Murray, K.G. 1980. Predation by deer mice on Xantus' Murrelet eggs on Santa Barbara Island, California. M.*. thesis, California State University, Northridge, California.

Murray, KG., K. WUuiett-Murray, Z.A. Eppley, G.L. Hunt, Jr. and D.B. Schwartz. 1983. Breeding biology of the Xantus' murrelet. Condor 85:12- 21.

Murray, KG., K. Winnett-Murray and G.L. Hunt, Jr. 1979. Egg neglect in Xantus' murrelet. Proc. Colonial Waterbird Gr. 3:186-195.

Newton, 1. 1998. Population limitation in birds. Academic Press, London.

Parr, R. 1993. Nest predation and numbers of Golden Plovers Pluvialis apricaria and other moorhd waders. Bird Study 403223-231.

Piersma, T. and N.C. Davidson. 1991. Confusions of mass and size. Auk 108:441-444.

Poussart, C., J. Larochelle and G. Gauthier. 2000. Thermal regime of eggs in Greater Snow Geese. Condor 102:292-300.

Pefaur, J.E. 1974. Egg-neglect in the Wilson's Storm Petrel. Wilson Bull. 86:16- 22.

Peterson, B.T. and B. Fry. 1987. Stable isotopes in ecosystem studies. Ann. Rev. Ecol. Syst. 18:293-320.

Peterson, B.J., R.W. Howarth and RH. Garritt. 1985. Multiple stable isotopes used to trace the flow of organic matter in estuarine food webs. Science 227:1361-1363.

Redfield, J.A. 1975. Distribution, abundance, size, and genetic variation of Peromysczis maniculatus on the Gulf Islands of British Columbia. Cm. J. 2001. 54:463-474.

Ricklefs, R.E. 1984. Prolonged incubation in pelagic seabirds: a comment on Boersrna's paper. Am. Nat 123:710-720.

Ricklefs, R.E. and J.M. Starck. 1998. Embryonic growth and development, pp. 31-58. In J.M. Starck and R.E. Ricklefs (eds.), Avian growth and development: evolution within the altriual-precocial spectnun. Oxford University Press, New York.

Roberts, B. 1940. The life cycle of Wilson's Petrel Oceanites oceanicus (Kuhl). British Graham Land Expedition 1934-37 Scientific Reports 1(2):141-194.

Roby, D.D. and R.E. Riddefs. 1984. Observations on the cooling tolerance of embryos of the Diving Petrel Pelecanoides georgiczrs. Auk 101:160-161.

Rodway, M. 1991. Status and conservation of breeding seabirds in British Columbia, p. 43-102.In J. P. Croxall (ed.), Çeabird status and conservation: a supplement. Int. Council Bird Presew. Tech. Publ. 11, Cambridge, England.

Rodway, M.S., M.J.F. Lemon and KR. Summers. 1990. British Columbia Seabird Colony Inventory: Rep. No. 4 - Scott Islands. Census results from 1982 to 1989 with reference to the Nestucca Oil Spill. Tech. Rep. Ser. No. 86. C m . Wildl. Sew., Pacific and Yukon Region, British Columbia.

Rodway, M-S, M.J.F. Lemon and K. Summers. 1992. Seabird breeding populations in the Çcott Islands on the west coast of Vancouver Island, 1982-1989, pp. 52-59. In K. Vermeer, R.W. Butler and K.H. Morgan (eds.), The ecology, status and conservation of marine and shoreline birds on the west coast of Vancouver Island. Can. Wiid. Serv. Occ. Paper No. 75, Ottawa.

Rogers, CM., M.J. Taitt, J.N.M. Smith and G. Jongejan. 1997. Nest predation and cowbird parasitism create a demographic sink in wetland-breeding Song sparrows. Condor 99:622633.

Roper, J.J. 1992. Nest predation experiments with quail eggs: too much to swallow? Oikos 65528-530.

Rosïng, M.N., M. Ben-David and R.P. Barry. 1998. Analysis of stable isotope data: a K nearest neighbors randomization test. J. Wüdl. Man. 62:380- 388.

Sæther, B-E., R Andersen and H.C. Pedersen. 1993. Regdation of parental effort in a long-lived seabird: an experimental manipulation of the cost of reproduction in the antarctic petrel, Thnlassoica antarctica. Behav. Ecol. Sociobiol. 33:147-150.

Schoeninger, M.1. and M.J. DeNiro. 1984. Nitrogen and carbon isotopic composition of bone collagen from marine and terrestrial animals. Geochim. Cosmochirn. Acta 48:625-639.

Sealy, S.G. 1975. Egg size in murrelets. Condor 77500-501.

Sealy, S.G. 1982. Voles as a source of egg and nestling loss among nesting auklets. Murrelet 63:9-14.

Sealy, S.G. 1984. Inter~ptions extend incubation by AnCient Murreleh, Crested Auklets, and Least Auklets. Murrelet 65:53-56-

Stearns, S C . 1992. The evolution of life histories. Oxford University Press, Oxford.

Stoleson, S.H. and S.R. Beissinger. 1999. Egg viability as a consbaint on hatching s ynchrony at high ambient temperatures. J. Animal Ecol. 68:951-962.

Sullivan, T.P. 1997. Sampling methodology for small mammals. Faculty of Forestry, University of British Columbia, Vancouver.

Summers, KR. and R.H. Drent. 1979. Breeding biology and twinning experiments of Rhinoceros Auklets on Cleland Island, British Columbia. Murrelet 60:16-22.

Tieszen, L.L., T.W. Boutton, KG. Tesdahl and N.A. Slade. 1983. Fractionation and turnover of stable carbon isotopes in animal tissues: implications

for 613c analysis of diet. Oecologia 5732-37.

Tombre, LM. and K.E. Erikstad. 1996. An experimental study of incubation effort in high-Arctic bamacle geese. J. Anim. Ecol. 65:325-331.

Verbeek, N.A.M. 1970. Breeding ecology of the Water Pipit. Auk 87425-451.

Vickery, W.L. , J.L. Daoust, A. El Wartiti and J. Peltier. 1994. The effect of energy and protein content on food choice by deer mice, Peromyscus man ici1 la t zis (Rodentia). Anim. Behav. 47:55-64.

Wada, E. and A. Hattori. 1991. Nitrogen in the sea: forrns, abundances and rate processes. CRC Press. Florida.

Wainright, S.C., J.C. Haney, C. Kerr, A.N. Golovkin and M.V. Flint. 1998. Utilization of nitrogen derived from seabird guano by terrestrial and marine plants at St. Paul, Pribilof Islands, Berïng Sea, Alaska. Mar. Biol. 131:63-71.

Warharn, J. 1990. The Petrels: their ecology and breeding systems. Academic Press, London.

Webb, D.R. 1987. Thermal tolerance of avian embryos: a review. Condor 892374-898.

Wheelwright, N.T. and P.D. Boersrna. 1979. Egg chilling and the thermal environment of the Fork-tailed Storm Petrel (Oceanodroma fi rcata) nest. Physiol. Zool.52:231-239.

White, F.N. and J.L. b e y . 1974. Avian incubation. Science 186:107-115.

Williams, G.C. 1966. Natural selection, the costs of reproduction, and a refinement of Lack's principle. Am. Nat 100:687-690.

Wilson, U.W. 1977. A study of the biology of the Rhinoceros Auklet on Protection Island, Washington. M.Sc. thesis, University of Washington, Seattle.

Wilson, U.W. and D.A. Manuwal. 1986. Breeding biology of the Rhhoceros Auklet in Washington. Condor 88: 143-155-

Zar, J. H. 1984. Biostatistical analysis. Second Edition. Prentice-Hall, Englewood Cliffs, New Jersey.

Zar, J.H. 1996. Biostatistical analysis. Third Edition. Prentice Hall, Upper Saddle River, New Jersey.