Page 1
1
The influence of reproductive timing on white spruce seed escape
and red squirrel hoarding
Devan W. Archibald
Department of Natural Resource Sciences,
McGill University, Macdonald Campus,
Montréal, Québec, Canada
June 2011
A thesis submitted to McGill University in partial fulfillment
of the requirements of the degree of Master of Science
© Devan W. Archibald 2011
Page 2
2
Abstract
This thesis evaluates how reproductive timing influences red squirrel hoarding of white
spruce cones, from the perspective of both the trees and the squirrels. This was
accomplished for white spruce by assessing the degree of intra-annual reproductive
synchrony exhibited by individual trees relative to others in the population and the
amount of cones (i.e. seeds) escaping red squirrel predation. In two years with vastly
different cone production at the population level, individual white spruce trees exhibiting
higher levels of intra-annual reproductive synchrony had enhanced seed escape from red
squirrels, leading to positive directional selection on this trait in the lower cone year. In
red squirrels, we used behavioural observations of radio-collared individuals to assess
how variation in the temporal separation of breeding and hoarding seasons, across four-
years of varying cone production, affects cone hoarding behaviour. Hoarding behaviour
was more affected by cone levels than reproductive timing and under high cone levels
both activities were successfully combined. However, males and females used different
hoarding strategies that were consistent with differences in the timing of reproductive
demands, indicating that although overall hoarding behaviour was driven by resource
levels, the timing of reproduction may be a factor in gender differences. The seasonal
scheduling of reproduction in white spruce trees appears to be an important component of
seed escape from red squirrels, and although the timing of reproduction may promote
gender differences in hoarding behaviour of red squirrels, it is less important than cone
production to overall hoarding behaviour, allowing red squirrels flexibility in their
reproductive timing relative to hoarding.
Page 3
3
Résumé
Cette thèse évalue comment la phénologie reproductive influence la collecte de cône
d‟épinette blanche par l‟écureuil roux, de la perspective de l‟arbre et de l‟écureuil. Chez
l‟épinette blanche, nous avons évalué le degré de synchronie reproductive intra-annuelle
de chaque arbre relatif à la population et le nombre de cône (graines) qui échappe à la
collecte des écureuils roux. Pour deux années avec des productions de cône très
différentes au niveau populationnel, les épinettes blanches davantage synchronisées intra-
annuellement dans leur phénologie reproductive ont davantage de graines qui échappent
aux écureuils, menant vers une sélection directionnelle positive sur ce trait lors d‟une
année de faible production de cône. Chez l‟écureuil roux, au cours de quatre années avec
des productions de cône très différentes, nous avons utilisé des observations
comportementales pour évaluer comment la variation dans la séparation temporelle entre
les saisons de reproduction et de la collecte affecte le comportement de collecte.
L‟activité de collecte de cône est affectée davantage par le niveau de production annuelle
de cône que par la phénologie de reproduction de l‟écureuil et, avec un haut niveau de
production de cône, les deux activités peuvent être combinées avec succès. Cependant,
les males et les femelles utilisent des stratégies différentes selon leur différente
phénologie de demande reproductive, indiquant que malgré que le niveau de ressource
soit déterminant sur l‟activité de collecte, la phénologie de reproduction pourrait être un
facteur dans la différence entre les sexes. La phénologie de reproduction de l‟épinette
blanche semble être une composante importante de la survie des graines face aux
écureuils roux, et malgré que la phénologie de reproduction puisse promouvoir une
différence dans l‟activité de collecte entre les sexes, cela est moins important que la
production de cône pour l‟ensemble de l‟activité de collecte, ce qui permet à l‟écureuil
roux d‟être flexible dans sa phénologie reproductive par rapport à la collecte de cône.
Page 4
4
Contribution of authors
This thesis is presented as two chapters, each intended for publication. For each
chapter the candidate was responsible for developing the research questions,
experimental design, field work, data management and analysis, interpretation and
writing.
Chapter one is co-authored by Andrew McAdam, Stan Boutin, Quinn Fletcher,
and Murray Humphries. Andrew McAdam provided logistical assistance with the
fieldwork, and was responsible for white spruce data collection in 2010. He also
provided input on the study design and comments on the manuscript. Stan Boutin
provided logistical assistance with the field work, provided input on the study design and
early versions of the manuscript. Quinn Fletcher provided input on the study design and
comments on the manuscript. Murray Humphries provided input on the study design,
assistance with logistics, and comments on the manuscript.
Chapter two is co-authored by Andrew McAdam, Stan Boutin, Quinn Fletcher,
and Murray Humphries. Andrew McAdam provided logistical assistance with the
fieldwork, input on the study design and comments on the manuscript. Stan Boutin
provided logistical assistance with the field work, provided input on the study design and
comments on the manuscript. Quinn Fletcher provided input on the study design,
assistance with data management for three of the four years of red squirrel hoarding
behaviour data, and comments on the manuscript. Murray Humphries provided input on
the study design, assistance with logistics, and comments on the manuscript.
Page 5
5
Acknowledgements
I would like to thank my supervisor, Murray Humphries for his patience, support,
guidance, consistent availability and positive attitude. This work received input from the
Humphries lab, and numerous lab-mates helped me gain the skill-set required to complete
this thesis and provided the friendship to flourish (Amy, Emily, Elise, Guillaume,
Jeremy, Jason, Karine, Marianne, Nicholas, Paul, Quinn, and Sébastien). Specifically,
Guillaume was always there to answer my statistical and IT questions, Manuelle was
always willing to help and always knew the answer, Quinn taught me how to use excel,
access, and R after I realized I knew nothing, and was always there for a squirrel chat.
Other members of the Natural Resource Sciences department at Macdonald Campus
provided feedback in the initial stages, as well as more final versions, and the members of
Journal Club helped keep my mind stimulated and thinking about my research from other
perspectives.
Field work was in collaboration with the Kluane Red Squirrel Project (KRSP) and
thus I benefited from the wisdom of Stan Boutin and Andrew McAdam who were
involved in this research from start to finish, and who provided field resources and
assistance which was greatly aided by the coordination of Ainsley Sykes. I benefited
from the collective squirrel lore of all „squirrelers‟ involved with KRSP and thank
everyone for their hard work. In particular my fellow graduate student Ben was
instrumental in assisting me with research design questions on the fly in the field, and
provided excellent comradery and advice during my stint as field crew supervisor.
Meghan and Quinn made learning how to conduct metabolic measurements on red
squirrels fun. I owe an enormous thank you to everyone involved in focal observations,
Emily, Julia, Manuelle, and last but definitely not least Kristin who had squirrels in view
for over 147 hours and likely spent more time entering data! An additional thank you is
required for Tasha for counting cones in 2010 and all the other volunteers involved in this
exciting endeavour. I am grateful for to the Yukon Government and the Alsek
Renewable Resource Council for their permission to conduct field work in the Yukon,
and this project was funded by an NSERC Discovery grant to MMH, as well as a NSERC
CGS scholarship and NSTP grant to DWA.
Page 6
6
In addition to her assistance in the field, my wife Kristin deserves a second thank
you for her patience in never tiring of hearing my squirrel theories. This thesis greatly
benefited from her knowledge of red squirrels, and her moral support in every stage of
the process.
Page 7
7
Table of Contents
Abstract ...................................................................................................................2
Résumé ....................................................................................................................3
Contribution of authors .........................................................................................4
Acknowledgements ................................................................................................5
Table of Contents ...................................................................................................7
List of Tables ..........................................................................................................9
List of Figures .......................................................................................................10
General Introduction and Literature Review ...................................................12
Granivory ...........................................................................................................12
Mast seeding ......................................................................................................13
Food hoarding ....................................................................................................14
North American red squirrels ............................................................................15
White spruce ......................................................................................................16
Kluane Red Squirrel Project ..............................................................................17
Research Objectives ...........................................................................................18
Literature Cited ..................................................................................................20
Chapter 1: Swamping seed predators in number and time: within-season synchrony
of a masting conifer enhances seed escape .........................................................26
Abstract ..............................................................................................................27
Introduction ........................................................................................................28
Methods .............................................................................................................30
Study area .......................................................................................................30
White spruce cone counts ..............................................................................30
White spruce intra-annual reproductive synchrony .......................................31
Red squirrel seed predation and hoarding behavioral observations ...............32
Statistical analysis ..........................................................................................33
Results ................................................................................................................34
Discussion ..........................................................................................................35
Literature Cited ..................................................................................................40
Page 8
8
Tables .................................................................................................................47
Figures ...............................................................................................................48
Connecting Statement ..........................................................................................49
Chapter 2: Reproductive and resource constraints on food hoarding in male and
female red squirrels. ............................................................................................50
Abstract ..............................................................................................................51
Introduction ........................................................................................................52
Methods .............................................................................................................54
Study area .......................................................................................................54
Hoarding behavioral observations .................................................................55
Estimation of total number of cones clipped, hoarded, and the propensity to larder-
hoard ..............................................................................................................56
Reproductive timing .......................................................................................56
Statistical analysis ..........................................................................................57
Results ................................................................................................................59
Reproductive completion and resource levels ...............................................59
Total number of cones clipped and hoarded ..................................................59
Propensity to larder-hoard ..............................................................................60
Hoarding time allocation during concurrent reproductive activity ................60
Discussion ..........................................................................................................62
Literature Cited ..................................................................................................66
Tables .................................................................................................................71
Figures ...............................................................................................................72
General Conclusions ............................................................................................75
Literature Cited ..................................................................................................77
Page 9
9
List of Tables
General Introduction and Literature Review
Table 1. Hypotheses that have been suggested to explain mast seeding or fruiting,
adapted from Kelly (1994)……………………………………………………………25
Chapter 1
Table 1. Standardized directional (β‟) and stabilizing/disruptive (γ’) selection gradients
calculated for intra-annual reproductive synchrony and the number of cones produced by
white spruce during a non-mast (2009; n=212) and mast (2010; n=206) year. Separate
models were used to assess linear and non-linear selection gradients within each year.
Bold font indicates significance estimated with jackknife tests (Mitchell-Olds and Shaw
1987). Stabilizing/disruptive selection gradients have been doubled to accurately
represent their strength (Stinchcombe et al. 2008)……………………………………...47
Chapter 2
Table 1. Range and median date of reproductive completion for adult red squirrels
included in the study in 2002, 2003, 2005, and 2010 with white spruce average cone
index (ln transformed cone count; mean ± SE) calculated from annual cone counts
conducted prior to red squirrel harvesting (n = 167-171 trees per year). Days until
hoarding season was calculated by subtracting the median date last reproductively active
from August 16th
, the approximate date when hoarding activity initiates (Fletcher et al.
2010)……………………………………………………………………………………71
Page 10
10
List of Figures
Chapter 1
Figure 1. Seasonal decline in the number of closed cones on white spruce trees resulting
from red squirrel cone clipping activities as well as cone opening. Average number of
closed cones (closed triangles) and average number of open cones (open circles) per tree
in the study (non-mast; n=607, mast; n=292) plotted with the average observed and
modeled red squirrel cone clipping (dashed line) rate per day (non-mast; n=21, mast;
n=9) throughout the autumn of a non-mast (2009) and (2010) mast year. Cone count
rounds occurred over more than one day but are plotted on median date of each round.
Cone number values represent means ± SE and clipping activity values are means…48
Chapter 2
Figure 1. Total number of cones clipped (A), hoarded (B) and the proportion of hoarded
cones that were larder-hoarded (C) by adult male and female red squirrels over four study
years with varying cone availability and separation of reproduction and hoarding. All
models contained significant sex and year effects with different letters indicating
significant differences found between years using post hoc testing. None of the models
contained significant interactions between year and sex. Values are means ± SE…….72
Figure 2. The proportion of time adult female red squirrels spent feeding (A) and in the
nest (B) varied as a quadratic function of days postpartum, but time spent conducting
cone hoarding-related activities (C) varied linearly with days postpartum during autumn
2010, after accounting for the appropriate fit of Julian date (time spent feeding; linear,
time spent in the nest; linear, time spent hoarding; quadratic). Parturition dates ranged
from Jul 12 to Aug 24 (median; Aug 9). Data were analyzed using generalized linear
mixed models with squirrel identity as a random factor, but plotted values are raw data
representing means ± SE for each day postpartum……………………………………..73
Page 11
11
Figure 3. The proportion of time adult female (closed circles) and male (open squares)
red squirrels spent conducting cone hoarding-related activities throughout autumn 2010
varied non-linearly with Julian date and was not significantly different between the sexes.
The dashed grey line indicates the proportion of study females that were yet to pass the
mid-point of lactation (proportion less than 45 days postpartum). Data were analyzed
using a generalized linear mixed model with squirrel identity as a random factor, but
plotted values are raw data representing means ± SE for each day…………………….74
Page 12
12
General Introduction and Literature Review
The global diversity of animals depends critically on resources provided by
plants, the major primary producers energizing the planet (Price 2002). Plant-animal
interactions likely played an important role in shaping the diversity of both groups
(Ehrlich and Raven 1964, Bascompte and Jordano 2007). Plants and animals interact in
numerous ways, both mutualistic (e.g. pollination and seed dispersal) and antagonistic
(e.g. seed predation) (Herrera and Pellmyr 2002). Herbivorous animals consist of
members of a variety of invertebrate and vertebrate groups and consume a variety of
plant parts, with no plant tissues escaping their attention (Crawley 1983).
Granivory
The plant tissue possessing the highest energetic content per gram is seeds
(Robbins 1983). Seeds are the fertilized ovules of flowering plants and consist of an
embryo with food-storage organs surrounded by a protective seed coat (Hulme and
Benkman 2002). Numerous animals have become specialized to feed mainly or
exclusively on seeds and are termed granivores, or seed-predators (Hulme and Benkman
2002). Granivory differs from other forms of herbivory in that this high quality food
source is provided in discrete packets often with low perish-ability, but with the challenge
that it is only available for brief periods of time that can be highly unpredictable
(Crawley 2000).
Seed predators have been identified as having considerable impact on plant seed
populations due to high predation rates, often as high as 50 to 100 percent of available
seed (Crawley 2000). Seed predation is thought to play a pivotal role in the regeneration,
colonization ability, and spatial distribution of plants (Hulme and Benkman 2002).
Additionally, it has been suggested that seed predators act as agents of natural selection
that influence seed traits (Smith 1970, Hulme 1998, Benkman 1999, Benkman and
Parchman 2009) and production strategies (Silvertown 1980, Ruhren and Dudash 1996,
Curran and Leighton 2000). Seed predator-plant interactions are not always antagonistic,
they can also be mutualistic. Seed predators can be dispersal agents when seeds remain
Page 13
13
viable after ingestion and digestion, or if cached seeds are not recovered (Howe and
Smallwood 1982, Jordano 2000, Hulme 2002, Vander Wall et al. 2005).
Within granivores it is useful to differentiate between pre and post-dispersal seed
predation (Crawley 2000, Hulme 2002, Hulme and Benkman 2002). From the plants
perspective, the costs of defence from pre-dispersal seed predation can be borne by the
parent plant, whereas the costs of defence from post-dispersal seed predation lie solely
with the individual seed (Crawley 2000). Pre-dispersal seed predators can exploit
spatially and temporally aggregated resources and can use searching cues based on the
parent plants, whereas post-dispersal seed predators must search for inconspicuous items
scattered in an often cryptic background at lower densities (Crawley 2000). Most pre-
dispersal seed predators are small, sedentary, specialist feeders and often insects. In
contrast, most post-dispersal seed predators are larger, more mobile, and generalist
herbivorous birds or mammals (Crawley 2000, Hulme 2002, Hulme and Benkman 2002).
Mast seeding
Masting is a forestry term that has taken on a rather precise ecological meaning
(Crawley 2000). Mast seeding is the synchronous intermittent production of large seed
crops in perennial plants (Kelly and Sork 2002). The term originates from a German
word for fattening livestock on abundant seed crops, and thus years with high seed
abundance are called mast years (Kelly and Sork 2002). Kelly (1994) separates mast
seeding into three types based mainly on the level of seed production in inter-mast years:
strict masting, with no seeds produced in non-mast years, normal masting, where plants
produce seed in non-mast years but it is markedly lower than mast levels, and putative
masting, where seed crops vary greatly but there is little evidence of switching between
mast and non-mast years and no evidence it is due to anything other than environmental
variation.
There are at least eight hypotheses suggested to explain mast seeding (Table 1,
Kelly 1994). Of these, three are frequently supported as ultimate causes: predator
satiation, wind pollination, and environmental prediction. The environmental prediction
hypothesis proposes that plants can predict which years will be best for seedlings. This
hypothesis lacks empirical support in relation to climate variation, but is well
Page 14
14
documented in mast events triggered by fire (Kelly 1994). After a fire, enhanced nutrient
availability and reduced competition are favorable for seedling establishment. However,
this explanation applies only to fire prone habitats and thus is less general than the other
two explanations (Kelly 1994). The wind pollination hypothesis proposes that masting
increases the chances of successful pollination in wind pollinated plants. This hypothesis
has received support in a range of species and many of the well known masting species
are wind pollinated (Kelly 1994). The most widely known explanation for masting is the
predator satiation hypothesis (Kelly 1994). Large seed crops satiate seed predators and
thus destroy a lower percentage of the crop. This hypothesis has been well supported in
the literature, but depends on the functional response of the seed predators (Kelly 1994).
As with any form of reproductive synchrony, there may be multiple causes in any given
case (Ims 1990), and masting may result from interactions between the abiotic and biotic
environment.
Food hoarding
Vander Wall (1990) defines food hoarding as “the handling of food to conserve it
for future use” (p. 1). Food hoarding animals have the capacity to control the availability
of food in space and time (Vander Wall 1990). The benefits of this strategy may include
improving the chances of surviving a period of food shortage, allowing an animal to
optimize feeding and foraging time with regard to other activities, improving an animal‟s
competitive status when foraging for limited resources, or ensuring a continuous supply
of food to young in the reproductive season (provisioning, under the above definition is
also considered hoarding; Vander Wall 1990). A variety of arthropods, mammals and
birds hoard food, including members with herbivorous, omnivorous and carnivorous diets
(Vander Wall 1990). Familiar examples include nut storage by tree squirrels (Sciurus
sp.), and honey storage by honey bees (Apis mellifera).
Food hoarding animals distribute hoarded food in a variety of ways, ranging from
highly clumped to highly dispersed (Vander Wall 1990). The end points of this spectrum
of cache-dispersion patterns have been termed larder-hoarding, caching all items in a
central location, and scatter-hoarding, caching one to a few items widely spaced
throughout an area (Vander Wall 1990). Larder-hoards are attractive resources for other
Page 15
15
foragers due to the high concentration of food. As a result, they tend to be placed in
protectable areas where they can be defended by the hoarder, which is usually vigilant
and well equipped to defend its larder-hoard (Vander Wall 1990). Scatter-hoards are
much less attractive and their dispersed nature makes them difficult to defend. Their
protection comes from their inconspicuousness and the hoarder often appears inattentive
(Vander Wall 1990). Inter-specific differences in food hoarding patterns and pilferage
ability have been suggested as mechanisms promoting the coexistence of similar species
(Jenkins and Breck 1998, Leaver and Daly 1998, Price et al. 2000).
Most animals can be categorized as either larder or scatter-hoarders, but some
species use a combination of strategies (Vander Wall 1990, Brodin 2010). Species that
use a combination of larder and scatter-hoarding include some kangaroo rats
(Dipodomys), chipmunks (Tamias), wood-mice (Apodemus), flying squirrels
(Glaucomys), the red fox (Vulpes vulpes), the white footed mouse (Peromyscus
leucopus), the red-headed woodpecker (Melanerpes erythrocephalus) and the red squirrel
(Tamiasciurus hudsonicus) (Vander Wall 1990, Hurly and Lourie 1997). Intra-specific
variation in hoarding behaviour has begun to receive more attention as researchers
attempt to identify factors associated with the use of different food storing strategies
(Daly et al. 1992, Clarke and Kramer 1994, Leaver and Daly 1998, Preston and Jacobs
2001, Leaver 2004, Tsurim and Abramsky 2004, Murray et al. 2006, Jenkins In press).
Factors identified to influence the propensity to larder or scatter-hoard in species that use
mixed strategies include the value of the food (Leaver and Daly 1998, Leaver 2004),
distance from the central-larder that food is encountered (Daly et al. 1992, Tsurim and
Abramsky 2004), the make-up of the competitive environment (Murray et al. 2006), the
pilferage of hoards (Preston and Jacobs 2001), the gender of the hoarder (Jenkins In
press), and the age and reproductive status of the hoarder (Clarke and Kramer 1994).
North American red squirrels
North American red squirrels are small (200-250g), arboreal and diurnal rodents
found throughout the majority of Canada‟s forests and those of the northern United States
of America (USA), and further south along the Rocky Mountains (Steele 1998). In
northern areas of their range red squirrels feed primarily on conifer seeds of white spruce
Page 16
16
(Picea glauca, Boutin et al. 2006). They hoard up to 20,000 conifer cones in autumn,
either burying them in a central-larder or midden, or scattering them throughout their
mutually exclusive territories, appearing to use a deliberate mix of both behaviours
(Smith 1968, Hurly and Lourie 1997, Steele 1998). These food stores aid overwinter
survival and are used to fuel reproduction in spring (Smith 1968, Steele 1998). Large
cone crops in one year lead to earlier parturition dates and higher juvenile growth rates in
the following year (Boutin et al. 2006).
Red squirrels are promiscuous (Lane et al. 2008), and in the northern areas of
their range females usually only attempt one litter per season, except during white spruce
mast years when they may attempt a second litter in late summer just prior to new cone
availability (Boutin et al. 2006). After a gestation period of about 33 days (Steele 1998),
females give birth to about 3 offspring (range; 1-7) usually in late March to early June
(McAdam et al. 2007). Offspring emerge from their nest about 50 days later, and are
weaned at about 70 days postpartum (Humphries and Boutin 1996, McAdam et al. 2007).
Juvenile overwinter survival is largely dependent upon obtaining a territory with a
midden, and occasionally a female will bequeath part or all of her territory to her
offspring (Price 1992, Boutin et al. 1993, Price and Boutin 1993, Berteaux and Boutin
2000, Boutin et al. 2000). Predation on red squirrels is largely by northern goshawk
(Accipiter gentilis), lynx (Lynx canadensis), great-horned owl (Bubo virginianus), and
coyote (Canis lantrans) (Stuart-Smith and Boutin 1995).
White spruce
White spruce has a transcontinental distribution in northern coniferous forests of
North America, being found throughout Canada and regions of the northeastern USA
(Nienstaedt and Zasada 1990). Within these regions it is able to grow under diverse soil
conditions, but it is generally more demanding than other coexisting conifers in achieving
its best development (Nienstaedt and Zasada 1990).
White spruce is monoecious, with female buds generally concentrated in the top
of the crown and the male buds in the middle to lower crown (Eis and Inkster 1972,
Nienstaedt and Zasada 1990). Reproductive buds are differentiated at the time annual
shoot growth ceases, the year before flowering and seed dispersal (Owens and Molder
Page 17
17
1977). Female receptivity and pollen shedding occur at the same time over a three to five
day period generally in May, June or July, varying with region and climate (Owens and
Molder 1979, Nienstaedt and Zasada 1990). White spruce are wind pollinated and
adverse weather such as frost or rain can severely inhibit promising seed crops
(Nienstaedt and Zasada 1990). Self pollination can occur, but as a consequence viable
seed set is greatly reduced (Fowler and Park 1983, Connell et al. 2006).
Fertilization occurs about three weeks after pollination and cones attain maximum
water content and size in late June or July (Nienstaedt and Zasada 1990). However, the
majority of embryo growth occurs after cones attain full size (Nienstaedt and Zasada
1990). Cotyledons appear in mid to late July and embryo development is completed in
late August (Nienstaedt and Zasada 1990). The maturation process continues after
embryo development is completed as cone dry weight continues to increase until about
two weeks prior to cone opening (Cram and Worden 1957, Nienstaedt and Zasada 1990).
Cone maturation stage appears to vary both within and among trees (Cram and Worden
1957). Cone opening coincides with moisture contents of 45 to 70 percent, and specific
gravities of 0.6 to 0.8 (Cram and Worden 1957, Nienstaedt and Zasada 1990). Seed
dispersal may be impacted by the weather. Cool wet weather can delay cone opening, or
even cause open cones to close again, with drier weather re-opening them (Nienstaedt
and Zasada 1990). Seeds are wind dispersed and peak seed fall usually occurs in mid-
September, with minor seed fall earlier in August and later into autumn (Waldron 1965,
Dobbs 1976, Nienstaedt and Zasada 1990).
White spruce is a mast seeding species (Lamontagne and Boutin 2007). The
interval between excellent cone crops varies regionally, and among local sites, from
between two to six years in good areas, to between 10 to 12 years in less favorable ones
(Nienstaedt and Zasada 1990, Lamontagne and Boutin 2007). Mast years may be
triggered by hot, dry summers at the time of bud differentiation (Nienstaedt and Zasada
1990) and are always followed by years with little to no cone production.
Kluane Red Squirrel Project
The Kluane Red Squirrel Project (KRSP) is an ongoing multidisciplinary long-
term ecological research project investigating the ecology, evolution, and energetics of
Page 18
18
red squirrels. Field work associated with this project is conducted near Kluane Lake in
the southwestern Yukon Canada (61 º N, 138 º W). The project started in the late 1980‟s
with Dr. Stan Boutin from the University of Alberta. Since then it has grown into a
collaborative project with three principal investigators (Dr. Stan Boutin, University of
Alberta; Dr. Andrew McAdam, University of Guelph; Dr. Murray Humphries, McGill
University), and several other collaborators at universities in Canada and abroad. To
date, the project has contributed 57 peer reviewed publications.
Populations of red squirrels and the annual white spruce cone production in the
region have been monitored extensively and continuously since 1987. All individual red
squirrels in the study populations are known, and individually marked with ear tags and
colour markers at birth or first capture (for general description see McAdam et al. 2007).
Behavioural observations and live trapping allow individuals to be followed throughout
their lifetime. Annual monitoring of female reproduction, and use of DNA sampling to
determine paternity, has allowed the creation of an extensive pedigree (McFarlane et al.
2010) . White spruce cone production on each of the study areas is assessed annually
using either binoculars or digital photographs to count cones on trees in July or August of
each year (LaMontagne et al. 2005, Lamontagne and Boutin 2007).
Research Objectives
In this thesis I explore how red squirrel hoarding of white spruce cones is
influenced by the reproductive timing of both white spruce trees (Chapter 1) and red
squirrels (Chapter 2).
Despite the obvious inter-annual reproductive synchrony in mast seeding or
fruiting plants, there is a lack of information about the extent and selective implications
of intra-annual reproductive synchrony in these plants (Rathcke and Lacey 1985, Kelly
1994). In chapter one, I explore the intra-annual reproductive synchrony of individual
white spruce trees, how it influences seed escape from red squirrel predation, and the
evolutionary implications for white spruce.
Hoarding and lactation have been shown to be the most energetically demanding
times of the year for red squirrels (Fletcher 2011). When reproduction is not adequately
separated from hoarding, there could be a trade-off between these activities that may or
Page 19
19
may not be alleviated by high resource levels. In chapter two, I explore the cone
clipping, hoarding, and larder-hoarding levels of male and female red squirrels in four
years with varying resource availability and seasonal separation of reproduction from
hoarding.
Page 20
20
Literature Cited
Bascompte, J. and P. Jordano. 2007. Plant-animal mutualistic networks: the architecture
of biodiversity. Annual Review of Ecology, Evolution, and Systematics 38:567-
593.
Benkman, C. W. 1999. The selection mosaic and diversifying coevolution between
crossbills and lodgepole pine. American Naturalist 153:S75-S91.
Benkman, C. W. and T. L. Parchman. 2009. Coevolution between crossbills and black
pine: The importance of competitors, forest area and resource stability. Journal of
Evolutionary Biology 22:942-953.
Berteaux, D. and S. Boutin. 2000. Breeding dispersal in female North American red
squirrels. Ecology 81:1311-1326.
Boutin, S., K. W. Larsen, and D. Berteaux. 2000. Anticipatory parental care: acquiring
resources for offspring prior to conception. Proceedings of the Royal Society of
London Series B-Biological Sciences 267:2081-2085.
Boutin, S., Z. Tooze, and K. Price. 1993. Post-breeding dispersal by female red squirrels
(Tamiasciurus hudsonicus): the effect of local vacancies. Behavioral Ecology
4:151-155.
Boutin, S., L. A. Wauters, A. G. McAdam, M. M. Humphries, G. Tosi, and A. A.
Dhondt. 2006. Anticipatory reproduction and population growth in seed predators.
Science 314:1928-1930.
Brodin, A. 2010. The history of scatter hoarding studies. Philosophical Transactions of
the Royal Society B: Biological Sciences 365:869-881.
Clarke, M. F. and D. L. Kramer. 1994. Scatter-hoarding by a larder-hoarding rodent:
intraspecific variation in the hoarding behaviour of the eastern chipmunk, Tamias
striatus. Animal Behaviour 48:299-308.
Connell, L. M., A. Mosseler, and O. P. Rajora. 2006. Impacts of forest fragmentation on
the reproductive success of white spruce (Picea glauca). Canadian Journal of
Botany 84:956-965.
Cram, W. H. and H. A. Worden. 1957. Maturity of white spruce cones and seed. Forest
Science 3:263-269.
Page 21
21
Crawley, M. J. 1983. Herbivory : the dynamics of animal-plant interactions. Blackwell
Scientific, Oxford.
Crawley, M. J. 2000. Seed predators and plant population dynamics. Pages 167-182 in M.
Fenner, editor. Seeds : the ecology of regeneration in plant communities. CABI
Publishing, Wallingford.
Curran, L. M. and M. Leighton. 2000. Vertebrate responses to spatiotemporal variation in
seed production of mast-fruiting dipterocarpaceae. Ecological Monographs
70:101-128.
Daly, M., L. F. Jacobs, M. I. Wilson, and P. R. Behrends. 1992. Scatter hoarding by
kangaroo rats (Dipodomys merriami) and pilferage from their caches. Behavioral
Ecology 3:102-111.
Dobbs, R. C. 1976. White spruce seed dispersal in central British Columbia. The Forestry
Chronicle 52:225-228.
Ehrlich, P. R. and P. H. Raven. 1964. Butterflies and plants: a study in coevolution.
Evolution 18:586-608.
Eis, S. and J. Inkster. 1972. White spruce cone production and prediction of cone crops.
Canadian Journal of Forest Research 2:460-466.
Fletcher, Q. F. 2011. The role of energy expenditure in resource acquisition and energy
allocation in free-ranging North American red squirrels (Tamiasciurus
hudsonicus). PhD Thesis. McGill University, Montreal.
Fowler, D. P. and Y. S. Park. 1983. Population studies of white spruce. I. Effects of self-
pollination. Canadian Journal of Forest Research 13:1133-1138.
Herrera, C. M. and O. Pellmyr. 2002. Plant-animal interactions: an evolutionary
approach. Blackwell Science, Oxford; Malden, MA.
Howe, H. F. and J. Smallwood. 1982. Ecology of seed dispersal. Annual Review of
Ecology and Systematics 13:201-228.
Hulme, P. E. 1998. Post-dispersal seed predation and seed bank persistence. Seed Science
Research 8:513-519.
Hulme, P. E. 2002. Seed-eaters: seed dispersal, destruction and demography. Pages 257-
273. CABI Publishing, Wallingford.
Page 22
22
Hulme, P. E. and C. W. Benkman. 2002. Granivory. Pages 132-154 in C. M. Herrera and
O. Pellmyr, editors. Plant-animal interactions: an evolutionary approach.
Blackwell Science, Oxford; Malden, MA.
Humphries, M. M. and S. Boutin. 1996. Reproductive demands and mass gains: a
paradox in female red squirrels (Tamiasciurus hudsonicus). Journal of Animal
Ecology 65:332-338.
Hurly, T. A. and S. A. Lourie. 1997. Scatterhoarding and larderhoarding by red squirrels:
size, dispersion, and allocation of hoards. Journal of Mammalogy 78:529-537.
Ims, R. A. 1990. The ecology and evolution of reproductive synchrony. Trends in
Ecology & Evolution 5:135-140.
Jenkins, S. H. In press. Sex differences in repeatability of food-hoarding behaviour of
kangaroo rats. Animal Behaviour In Press, Corrected Proof.
Jenkins, S. H. and S. W. Breck. 1998. Differences in food hoarding among six species of
Heteromyid rodents. Journal of Mammalogy 79:1221-1233.
Jordano, P. 2000. Fruits and frugivory. Pages 125-165 in M. Fenner, editor. Seeds: the
ecology of regeneration in plan communities, 2nd
edition. CABI, Wallingford.
Kelly, D. 1994. The evolutionary ecology of mast seeding. Trends in Ecology &
Evolution 9:465-470.
Kelly, D. and V. L. Sork. 2002. Mast seeding in perennial plants: why, how, where?
Annual Review of Ecology and Systematics 33:427-447.
Lamontagne, J. M. and S. Boutin. 2007. Local-scale synchrony and variability in mast
seed production patterns of Picea glauca. Journal of Ecology 95:991-1000.
LaMontagne, J. M., S. Peters, and S. Boutin. 2005. A visual index for estimating cone
production for individual white spruce trees. Canadian Journal of Forest Research
35:3020-3026.
Lane, J. E., S. Boutin, M. R. Gunn, J. Slate, and D. W. Coltman. 2008. Female multiple
mating and paternity in free-ranging North American red squirrels. Animal
Behaviour 75:1927-1937.
Leaver, L. and M. Daly. 1998. Effects of food preference on scatter-hoarding by
kangaroo rats (Dipodomys merriami). Behaviour 135:823-832.
Page 23
23
Leaver, L. A. 2004. Effects of food value, predation risk, and pilferage on the caching
decisions of Dipodomys merriami. Behavioral Ecology 15:729-734.
McAdam, A. G., S. Boutin, A. K. Sykes, and M. M. Humphries. 2007. Life histories of
female red squirrels and their contributions to population growth and lifetime
fitness. Ecoscience 14:362-369.
McFarlane, S. E., J. E. Lane, R. W. Taylor, J. C. Gorrell, D. W. Coltman, M. M.
Humphries, S. Boutin, and A. G. McAdam. 2010. The heritability of multiple
male mating in a promiscuous mammal. Biology Letters 7:368-371.
Murray, A. L., A. M. Barber, S. H. Jenkins, and W. S. Longland. 2006. Competitive
environment affects food-hoarding behavior of Merriam's kangaroo rats
(Dipodomys merriami). Journal of Mammalogy 87:571-578.
Nienstaedt, H. and J. C. Zasada. 1990. Picea glauca (Moench) Voss white spruce. Pages
165-185 in R. M. Burns and B. H. Honkala, editors. Silvics of North America.
Vol. 1. Conifers. U.S. Department of Agriculture Agricultural Handbook. U.S.
Department of Agriculture, Washington, D.C.
Owens, J. N. and M. Molder. 1977. Bud development in Picea glauca. II. Cone
differentiation and early development. Canadian Journal of Botany 55:2746-2760.
Owens, J. N. and M. Molder. 1979. Sexual reproduction of white spruce (Picea glauca).
Canadian Journal of Botany 57:152-169.
Preston, S. D. and L. F. Jacobs. 2001. Conspecific pilferage but not presence affects
Merriam's kangaroo rat cache strategy. Behavioral Ecology 12:517-523.
Price, K. 1992. Territorial bequeathal by red squirrel mothers: a dynamic model. Bulletin
of Mathematical Biology 54:335-354.
Price, K. and S. Boutin. 1993. Territorial bequeathal by red squirrel mothers. Behavioral
Ecology 4:144-150.
Price, M. V., N. M. Waser, and S. McDonald. 2000. Seed caching by Heteromyid rodents
from two communities: implications for coexistence. Journal of Mammalogy
81:97-106.
Price, P. W. 2002. Species interactions and the evolution of biodiversity. Pages 3-25 in C.
M. Herrera and O. Pellmyr, editors. Plant-animal interactions: an evolutionary
approach. Blackwell Science, Oxford; Malden, MA.
Page 24
24
Rathcke, B. and E. P. Lacey. 1985. Phenological patterns of terrestrial plants. Annual
Review of Ecology and Systematics 16:179-214.
Robbins, C. T. 1983. Wildlife feeding and nutrition. Academic Press, New York.
Ruhren, S. and M. R. Dudash. 1996. Consequences of the timing of seed release of
Erythronium americanum (Liliaceae), a deciduous forest Myrmecochore.
American Journal of Botany 83:633-640.
Silvertown, J. W. 1980. The evolutionary ecology of mast seeding in trees. Biological
Journal of the Linnean Society 14:235-250.
Smith, C. C. 1970. The coevolution of pine squirrels (Tamiasciurus) and conifers.
Ecological Monographs 40:349-371.
Smith, M. C. 1968. Red squirrel responses to spruce cone failure in interior Alaska. The
Journal of Wildlife Management 32:305-317.
Steele, M. A. 1998. Tamiasciurus hudsonicus. Mammalian Species 586:1-9.
Stuart-Smith, A. K. and S. Boutin. 1995. Behavioural differences between surviving and
depredated juvenile red squirrels. Ecoscience 2:34-40.
Tsurim, I. and Z. Abramsky. 2004. The effect of travel costs on food hoarding in gerbils.
Journal of Mammalogy 85:67-71.
Vander Wall, S. B. 1990. Food hoarding in animals. University of Chicago Press,
Chicago.
Vander Wall, S. B., K. M. Kuhn, and M. J. Beck. 2005. Seed removal, seed predation,
and seconday dispersal. Ecology 86:801-806.
Waldron, R. M. 1965. Cone production and seed fall in a mature white spruce stand.
Forestry Chronicle 41:314-329.
Page 25
25
Tables
Table 1. Hypotheses that have been suggested to explain mast seeding or fruiting,
adapted from Kelly (1994).
Hypothesis Explanation
Predator satiation Large seed crops satiate seed predators and
decreases in predator abundance may occur in low
seed years
Wind pollination Increases pollination success in wind pollinated
plants
Environmental prediction Mast years timed to anticipate favorable conditions
for reproduction or establishment
Resource matching Plants vary reproductive effort to match available
resources
Animal pollination Increases pollination success in animal pollinated
plants
Animal dispersal Increases dispersal in animal dispersed plants
Accessory costs High accessory costs of reproduction make small
reproductive efforts less efficient per seed
Large seed size Selection for larger seed size increases contrast
between high and low seed years
Page 26
26
Chapter 1: Swamping seed predators in number and time: within-season
synchrony of a masting conifer enhances seed escape
Authors:
Devan W. Archibald (email: [email protected] )
Natural Resource Sciences, Macdonald campus, McGill University
21 111 Lakeshore Drive, Ste-Anne-de-Bellevue, QC, H9X 3V9, Canada
Andrew G. McAdam (email: [email protected] )
Department of Integrative Biology, University of Guelph,
50 Stone Road East, Guelph, ON, N1G 2W1, Canada
Stan Boutin (email: [email protected] )
Department of Biological Sciences, University of Alberta
Z1109 Biological Sciences Building, Edmonton, AB, T6G 2R3, Canada
Quinn Fletcher (email: [email protected] )
Department of Biological Sciences, University of Alberta
Z1109 Biological Sciences Building, Edmonton, AB, T6G 2R3, Canada
Murray M. Humphries (email: [email protected] )
Natural Resource Sciences, Macdonald campus, McGill University
21 111 Lakeshore Drive, Ste-Anne-de-Bellevue, QC, H9X 3V9, Canada
Page 27
27
Abstract
Predator satiation resulting from inter-annual synchrony has been widely documented in
masting plants, but how reproductive synchrony within a year influences seed escape is
poorly understood. We evaluated whether the intra-annual reproductive synchrony of
individual white spruce trees (Picea glauca) increased seed escape from their primary
pre-dispersal seed predator (North American red squirrels; Tamiasciurus hudsonicus). In
both a mast and non-mast year, seed escape tended to be enhanced by producing many
cones as well as by producing cones that matured synchronously relative to other trees in
the population. This led to significantly positive selection differentials for intra-annual
reproductive synchrony in both years, but after also accounting for the number of cones
produced, natural selection favoring increased synchrony was significantly different from
zero only in the non-mast year. Thus, maximizing number and minimizing time both
contribute to conifer seed escape, but their relative importance varies between mast and
non-mast years.
Page 28
28
Introduction
The reproductive stages of many plants and animals are synchronized among
individuals within populations (reviewed in Ims 1990a). Synchrony is common in
seasonal environments because reproduction is often clustered during times of the year
that are most favorable to offspring survival. However, reproductive synchrony is often
far greater than would be expected from environmental seasonality alone (Hughes and
Richard 1974, Rutberg 1987, Sinsch 1988), suggesting that ecological and socio-
biological processes may influence the temporal pattern of reproduction (Ims 1990a).
Darling (1938) first suggested that reproductive synchrony might serve an anti-predatory
function, and this remains the most general and widely cited adaptive explanation for
reproductive synchrony (Ims 1990b). Among the three mechanisms suggested (see
Rutberg 1987, Ims 1990b), the sudden mass appearance of prey in a vulnerable stage,
which satiates predators due to handling time constraints, has been proposed as an
explanation for reproductive synchrony in birds (Darling 1938, Robertson 1973, Findlay
and Cooke 1982), mammals (Estes 1976, Rutberg 1987, Odonoghue and Boutin 1995),
and to explain mast fruiting or seeding in plants (Janzen 1971, Silvertown 1980, Kelly
1994, Kelly and Sork 2002).
Mast seeding is the synchronous and highly variable inter-annual seed production
by a population of plants (Kelly 1994). This reproductive strategy results in a pattern of
episodic mast years, in which a superabundance of seeds is produced, followed by non-
mast years where little or no seeds are produced. The predator satiation hypothesis is a
prominent, ultimate explanation for this inter-annual reproductive synchrony (Kelly and
Sork 2002). It proposes that more seeds are produced than can be consumed in mast
years and predators starve and may decline in abundance in the intervening non-mast
years. Reduced seed predation in mast years is a basic prediction of the predator satiation
hypothesis, and has been confirmed in numerous plant species with a variety of seed
predators (reviewed in Kelly and Sork 2002).
Inter-annual synchrony is the defining feature of the masting strategy, but, to date,
studies have ignored the potential importance of intra-annual synchrony to seed escape
(Rathcke and Lacey 1985), despite a number of studies in other non-masting plants
Page 29
29
examining the importance of within-season synchrony in flowering or fruiting to
pollination success and seed escape (for a review see Elzinga et al. 2007). In masting
species, high inter-annual synchrony will be most effective in satiating seed predators
when combined with high intra-annual synchrony, such that in a year of abundant seed
production, seeds from all trees are vulnerable to predation during the same, short time
interval. Not only the amount of seeds produced, but also the timing of their seasonal
availability may, therefore, be an important fitness determinant for masting species.
White spruce (Picea glauca) is a mast seeding tree (LaMontagne and Boutin
2007) that satiates its dominant pre-dispersal seed predator, the North American red
squirrel (Tamiasciurus hudsonicus) during mast years (Fletcher et al. 2010). White
spruce is a wind dispersed conifer species with a transcontinental distribution in North
America (Nienstaedt and Zasada 1990), much of which overlaps with that of the red
squirrel (Steele 1998). The red squirrel is the dominant pre-dispersal seed predator of
North American conifers (Benkman et al. 1984, Benkman et al. 2003, Peters et al. 2003,
Benkman and Siepielski 2004). They defend mutually exclusive territories within which
they remove (cut or clip from the trees) and hoard thousands of conifer cones in autumn
(Steele 1998). Neither immature nor open cones are hoarded by red squirrels (Fletcher et
al. 2010). Therefore, in order to escape pre-dispersal seed predation by red squirrels
white spruce cones must remain on trees from maturation to opening, a vulnerable period
that lasts about two weeks (Cram and Worden 1957). The timing of cone maturation for
individual white spruce trees relative to the timing of other trees in the population may
have important consequences for their seed escape from red squirrels, but has not been
previously investigated.
Here we examined the degree of intra-annual reproductive synchrony of
individual white spruce trees and its influence on seed escape to dispersal in both a mast
and non-mast year. We hypothesized that high-levels of intra-annual reproductive
synchrony in white spruce trees could enhance seed escape by satiating red squirrel
hoarding efforts. We, therefore, predicted that white spruce trees whose cones matured
more synchronously with others in the population would have increased seed escape to
dispersal than those with less reproductive synchrony. Since pre-dispersal seed predation
can affect natural selection on a variety of plant traits (Kolb et al. 2007, Parachnowitsch
Page 30
30
and Caruso 2008), we explored the evolutionary implications of our findings by
measuring the strength of directional selection (Lande and Arnold 1983) on intra-annual
reproductive synchrony in white spruce during these two episodes of selection, and the
repeatability of intra-annual reproductive synchrony and cone production between years.
We also concurrently monitored red squirrel cone clipping activities to determine how it
varied with cone maturation phenology.
Methods
Study area
This study was conducted in the autumns of 2009 (non-mast year) and 2010 (mast
year) on three study areas (approximately 40ha each) located near Kluane National Park
in southwestern Yukon, Canada (61°N, 138°W). Two of the study areas were located
across the Alaska Highway from each other. In these areas the red squirrel population
and annual white spruce cone production have been monitored continuously since 1988
(Boutin et al. 2006, McAdam et al. 2007). The third area was less than 500 meters away,
has historically been studied intermittently, and was included in the first year of the study
only. All sites were located in a glacial valley that is composed of boreal forest
dominated by white spruce with a willow (Salix spp.) understory. Red squirrels are
present in this area at an average density of 1.5-2.8 squirrels/ha (Boonstra et al. 2001),
and are the dominant vertebrate pre-dispersal seed predator of white spruce as seed
predation by white-winged crossbills (Loxia leucoptera) is rare and irruptive (Smith and
Folkard 2001). For a more detailed description of the general ecology of the study site
see Krebs, Boutin, and Boonstra (2001).
White spruce cone counts
To assess cone production, pre-dispersal predation, and cone opening we
conducted repeated cone counts of individually marked trees (non-mast; n=607 trees,
mast; n=292 trees) on red squirrel territories (non-mast; n=24 territories, mast; n=14
territories). Red squirrels defend mutually exclusive territories with a central food store
where they cache cones that they have clipped from the surrounding trees (Steele 1998).
Page 31
31
Sampled trees were located within 30m of a red squirrel‟s central food cache, or midden,
which is similar to the average radius of a red squirrel territory in this area (LaMontagne
and Boutin 2007). Specifically, two perpendicular 60m transects were randomly oriented
through the centre of each midden. All trees larger than 5cm diameter at breast height
within 1m of either side of both transects were sampled. This diameter at breast height
approximates the age at which white spruce become cone-bearing in this region
(LaMontagne and Boutin 2007).
The numbers of closed and open cones were measured seven times for each tree
during the cone-hoarding season. In the non-mast year (2009) these counts were
performed between Aug 15 and Oct 2. In the mast year (2010) counts were performed
between Aug 7 and Oct 7. Each of the seven rounds of cone counts lasted between one
and six days (median = 3). The length of time between successive counts of the same
tree ranged from six to 16 days (median = 8). For each count, the total numbers of closed
cones and open cones visible on the top 3 m of one side of each sampled tree were
counted using binoculars. All counts for each tree were made from the same flagged
location. If more than 100 cones were visible, a digital photo was taken (6.0 megapixels)
and cones were counted from the image (LaMontagne et al. 2005). The technique has
been previously shown to correlate strongly with the total number of cones on the entire
tree (LaMontagne et al. 2005).
White spruce intra-annual reproductive synchrony
We assumed that the timing of cone opening reflected phenological patterns of
maturation and used the timing of cone opening as proxy for reproductive timing. White
spruce cone maturation cannot be assessed without removing cones from the trees
(Nienstaedt and Zasada 1990), and since we were interested in cone removal by red
squirrels we did not remove cones ourselves. Cram and Worden (1957) found that a
white spruce cone matures about two weeks prior to opening, indicating opening may be
a signal of maturation. Cone opening is also a function of specific gravity and lowering
water content (Cram and Worden 1957). Therefore, environmental factors may play a
role in opening synchronicity that could be absent in maturation synchrony. But,
different trees and cones likely have varying degrees of susceptibility to opening under
Page 32
32
the same weather conditions due to being in differing phenological stages, so differences
in the timing of cone opening should be reflective of differences in the timing of cone
maturation.
To assess the degree of intra-annual reproductive synchrony exhibited by
individual white spruce trees with others in the population we used Mahoros (2002) index
for flowering synchrony and adapted it to cone opening (Eq‟n 1). There are numerous
methods for assessing fruiting and flowering synchrony in plants (Elzinga et al. 2007).
We chose the method of Mahoro (2002) because it is one of the only methods that
considers the relative numbers of mature fruit or open flowers on each individual
throughout the season, and how that compares to the relative numbers of mature fruit or
flowers of all the other individuals in the population, which are two criteria that are most
important when investigating the consequences of phenological patterns (Freitas and
Bolmgren 2008). Other indices examine the duration of flowering or fruiting, ignoring
the importance of the number of open flowers or mature fruit (Augspurger 1983), or they
only consider the relative number of open flowers or mature fruit of the individual and
ignore the importance of the relative numbers of open flowers or mature fruit on the other
individuals (Marquis 1988, Freitas and Bolmgren 2008).
(1)
In this index the synchronization level (SIi) of an individual i is a function of the ratio of
the total cones opening in the season that newly opened in the jth
interval between counts
(yi,j) and the mean of that value for every other individual in the population during the
same interval between counts (ӯi,j) with n representing the number of intervals between
counts. This results in an index for each tree that ranges between zero and one, with one
indicating complete synchrony with the other trees in the population and zero indicating
complete asynchrony.
Red squirrel seed predation and hoarding behavioral observations
We also concurrently monitored red squirrel cone clipping activities to determine
the extent and timing of pre-dispersal cone predation (2009; n = 21 adult males from Aug
Page 33
33
20 to Oct 1, 2010; 9 adult males from Aug 25 to Sep 30). In both years this encompassed
the beginning, peak, and decline in cone clipping activities. We followed the
methodology presented in detail by Fletcher et al. (2010), using repeated daily seven-
minute focal observations of individuals three times weekly, facilitated by radio collars
and color-marked ear tags. During focal periods we recorded the number of cones
clipped from trees and we used the average clipping rate on a given day to estimate the
daily total number of cones clipped by each squirrel by multiplying this average by the
number of minutes between sunrise and sunset (61°N, 138°W; Herzberg Institute of
Astrophysics; National Research Council of Canada). We used the shape-preserving
piecewise cubic modeling technique (Fritsch and Carlson 1980) adopted by Fletcher et al.
(2010) to estimate clipping activities for each individual on un-observed days between
August 15th
and October 15th
. In order to assess the phenology of cone clipping we
plotted the average of both modeled and observed values for all squirrels on each day
throughout the season.
All animal use procedures were approved by McGill University Faculty of
Agricultural and Environmental Sciences Animal Care Committee.
Statistical analysis
To evaluate the strength and form of phenotypic selection on intra-annual
reproductive synchrony during these two episodes of selection we calculated selection
differentials and the selection gradients (Lande and Arnold 1983) in each year separately.
Standardized selection differentials (S’) were measured as the covariance between
relative fitness and the individual synchrony index. Relative fitness for each tree was
calculated as the absolute fitness of the tree (number of open cones visible during the last
count) divided by the mean fitness of all trees in the population. Standardized linear
selection gradients (β’) were estimated from the partial regression coefficients of a
multiple regression that included individual synchrony index and the number of cones
produced (number of cones visible during the first count) as predictors of relative fitness.
Individual synchrony index and the number of cones produced were each standardized to
a mean of zero and standard deviation of one prior to analysis (Lande and Arnold 1983).
Non-linear (i.e. stabilizing or disruptive) selection gradients (γ’) were estimated as twice
Page 34
34
the regression coefficients from the quadratic terms in the model (Lande and Arnold
1983, Stinchcombe et al. 2008). Since relative fitness was calculated relative to the mean
fitness for each population there were no differences among study areas in relative fitness
so this factor was not included in our selection models. Due to non-normality of the
residuals from the models, we used jackknife re-sampling to generate standard errors and
test significance (Mitchell-Olds and Shaw 1987).
Repeatability is a measure of the proportion of total variance in a trait that is
accountable to differences among individuals (Falconer and Mackay 1996), and therefore
represents an upper limit to heritability (but see Dohm 2002). In balanced designs with
paired measures it is most simply estimated as a correlation, since it is the within-class
correlation of repeated measurements (Nakagawa and Schielzeth 2010). We assessed the
repeatability between years of intra-annual reproductive synchrony and the number of
cones produced in white spruce using Pearson‟s product moment correlation coefficient
(r). Thus, we measured whether the relative, not absolute, measurement of the trait of an
individual was correlated with its relative measurement the following year (Berkum et al.
1989, Chappell et al. 1995). Due to non-normal error distributions the number of cones
produced were ln-transformed (ln +1) prior to this analysis.
All statistical tests were conducted using the statistical software R (R
Development Core Team 2011) with an alpha level of 0.05. We report means ± standard
error throughout the paper.
Results
In the non-mast year (2009) 44.8 percent of all white spruce trees sampled
produced cones, whereas 83.2 percent produced cones in the mast year (2010). On
average, the number of cones produced on sampled trees increased 25-fold from the non-
mast to mast year. The number of closed cones decreased throughout the study in both
years due to the clipping activities of the red squirrels and cone opening (Figure 1).
About two weeks prior to cone opening, red squirrel cone clipping activities increased,
and then peaked as cones were beginning to open in both years.
Page 35
35
In the non-mast year, 34.9 percent of all trees sampled had at least one cone
survive to opening, whereas 72.6 percent of all trees sampled had cones survive to
opening in the mast year. Among individual trees that produced cones, there was a
significant (p<0.001; Wilcoxon rank sum test) 9-fold increase in cone production from
the non-mast to mast year (non-mast; 55 ± 6 cones counted, mast; 480 ± 36 cones
counted), however the percentage of these cones that escaped pre-dispersal seed
predation and opened in the mast (43.9 ± 2.7 percent) and non-mast (45.0 ± 2.8 percent)
were not significantly different (p=0.09; Wilcoxon rank sum test).
Individual intra-annual reproductive synchrony indices were significantly higher
(p<0.001; Wilcoxon rank sum test) in the mast (0.56 ± 0.02; range 0.03 – 0.9) than in the
non-mast year (0.47 ± 0.01; range 0.01-0.89). There were significant positive selection
differentials for increased intra-annual reproductive synchrony in both years (non-mast;
S’= 0.48 ± 0.12, t1,211 = 4.13, p < 0.001, mast; S’ = 0.45 ± 0.09, t1,205 = 5.20, p < 0.001).
However, when accounting for the number of cones produced, there was significant
directional selection for increased intra-annual reproductive synchrony in the non-mast
year but no evidence of selection on synchrony in the mast year (Table 1). There was
significant directional selection for an increase in the number of cones produced in the
non-mast year. In the mast year there were significant linear and non-linear selection on
the number of cones produced, indicating that trees that produced more cones had higher
fitness but also that the strength of selection also increased with increasing cone
production (Table 1). Analysis of white spruce trees included in both years of the study
revealed that intra-annual reproductive synchrony was significantly repeatable between
years (r91 = 0.25, p = 0.015), as was the number of cones produced (r275 = 0.49, p <
0.001).
Discussion
Individual white spruce trees varied widely in the number of cones produced and
their reproductive synchrony. These two traits had important consequences for tree
fitness, but effects differed between a mast year and a non-mast year. It is unsurprising
that the number of cones present at the start of the hoarding period was a significant
Page 36
36
determinant of the number of cones open at the end of the hoarding period, given the
tendency for most trees to have a substantial proportion of their cones escape squirrel
predation. However, our detection of strong selection (double the median value of
known selection gradients reported by Kingsolver et al. 2001) for intra-annual
reproductive synchrony is more surprising and novel. During the mast year, when trees
produced many cones and squirrels are satiated in their cone clipping and hoarding
(Fletcher et al. 2010), we found no evidence for selection on intra-annual reproductive
synchrony. However, in the non-mast year, when mean cone production was much
lower, we found evidence for strong directional selection favoring increased intra-annual
reproductive synchrony. Atlan et al. (2010) recognized two strategies for escape from
seed predation in a shrub (Ulex europaeus), escape in time or numbers, and they
suggested that the plants they studied exhibited polymorphic use of these strategies. Our
results also suggest that white spruce can enhance seed predation either through the
adjustment of seed number or the timing of seed maturation. The “too many cones”
strategy appears to be effective in years of high and low cone production, whereas the
“too little time” strategy seems to be most effective when cone abundance is limited.
To our knowledge, this is the first study to quantify the degree of intra-annual
reproductive synchrony exhibited by individuals in a masting species, and the first to
relate this to seed escape. It is also one of the few studies to find evidence of a positive
influence of intra-annual reproductive synchrony on seed escape in plants (for other
examples see, Augspurger 1981, Honek and Martinkova 2005). Most other studies have
found that fruiting or flowering off-peak is associated with reduced seed predation
(Pettersson 1994, Eriksson 1995, Pilson 2000, Freeman et al. 2003, Lacey et al. 2003).
When using the same synchrony index as this study, Mahoro (2002) found no correlation
between flowering synchrony and flower, fruit or seed predation for individuals of
Vaccinium hirtum. Gomez (1993) found neither consistency in the sign nor significance
of phenotypic selection on flowering synchrony of Hormathophyla spinosa across three
years. Differences between our findings and those of previous studies could be due to
our focus on a system where predation defines the plant-animal interaction, whereas most
previous studies of plant synchrony have focused on mutualistic plant-animal interactions
such as pollination, with seed escape being secondary. Synchrony patterns thought of as
Page 37
37
detrimental for seed predators are often thought of as beneficial to pollinators and
dispersers (Benkman et al. 1984, Elzinga et al. 2007). Most of the studies of intra-annual
reproductive synchrony in plants also have animal pollinators present, and are interpreted
mainly in that context (Kolb et al. 2007). White spruce are wind pollinated and dispersed
(Nienstaedt and Zasada 1990), which eliminates any pollinator or disperser saturation that
could result in a cost of reproductive synchrony.
Other studies of conifer cone phenology have sought to determine population, and
occasionally within individual patterns of maturation or opening (Vander Wall and Balda
1977, Tomback and Kramer 1980, Benkman et al. 1984, Samano and Tomback 2003),
and have compared species traits that facilitate avian dispersal or deter red squirrel
predation (Benkman et al. 1984, Samano and Tomback 2003). But none of these studies
have investigated the implications of inter-individual variation in cone phenology within
a single species. By using a synchrony index we were able to quantify differences
between individuals. Our index accounted for both within and among individual
synchrony by accounting for the relative number of open cones not only on the individual
of interest, but also that of all the other individuals in the population (Mahoro 2002).
However, this index was unable to separate the importance of within and among
individual synchrony, of which variation in maturation or opening pattern could be
important to seed predation (Benkman et al. 1984). Future research should focus on the
relative importance of within and among individual synchrony, which will require
monitoring individual cones on a given tree and following their progression throughout
the autumn.
We further demonstrate that cone production and intra-annual synchrony are both
repeatable traits. Although repeatabilities of these traits (0.25 for synchrony and 0.49 for
production) were lower than repeatabilities of morphological traits in other plants (0.64 to
0.99; Beavis et al. 1991, Jordano 1995, Shykoff and Kaltz 1998, Santos 1999, Di Renzo
et al. 2000), it establishes the potential for heritable variation, which has been shown for
phenological traits in other plants (Chao et al. 2003, Botto and Coluccio 2007, Johnson
2007, Atlan et al. 2010), including a conifer (Matziris 1994). However, differentiating
between the environmental and genetic contributions to the variation in cone production
and intra-annual synchrony of white spruce will be challenging. Nevertheless, a
Page 38
38
comparison of the fitness of more and less synchronous individuals in a population
indicates an adaptive advantage of this strategy (Ims 1990a).
Red squirrels removed all of the available cones before any had opened on about
ten percent of sampled trees in both years. Our inability to estimate intra-annual
synchrony for these highly depredated trees (the „invisible fraction‟, sensu Grafen
(1988)), could have introduced bias into our selection gradient estimates (Grafen 1988,
Bennington and McGraw 1995, Hadfield 2008). However, since the number of trees in
this category is small, consistent across years, and there is still large variation in intra-
annual synchrony index values within the population, the bias is likely to be small.
Our use of the number of open cones as a measure of fitness assumes that the
number of open cones correlated with the number of seeds dispersing. Although quality
and number of seeds per cone can vary depending on cone size or year (Zasada 1970,
Waldron 1965), years with large cone crops tend to also have more seeds per cone
(Zasada and Viereck 1970), indicating our results may underestimate the importance of
these traits to tree fitness in the mast year and overestimate their importance in the non-
mast year. However, the difficulty of assessing seedfall of individual trees requires
additional indicies of fitness such as cone numbers to be used (Parchman et al. 2007).
This study provides additional evidence that red squirrels interact strongly with,
and may act as selective agents on, a variety of conifer species (Smith 1970, Elliott 1974,
Benkman et al. 1984, Benkman and Siepielski 2004). The two years over which we
measured selection on spruce cone number and synchrony represent a small proportion of
the lifetime of these trees, but provide unbiased estimates of the fitness consequences of
these traits during these two episodes of selection. Further estimates of selection on cone
production and synchrony would reveal how representative the two years that we studied
were of mast and non-mast years more generally. Overall selection on these traits could
also be affected by interactions with other potential pre-dispersal seed predators (Smith
and Balda 1979, Nienstaedt and Zasada 1990), but these appear to be less common pre-
dispersal seed predators of spruce in our area. The importance of cone production and
synchrony to post-dispersal seed predation is also unknown. The high proportion of
cones clipped by red squirrels, however, suggests that red squirrel seed predation has
important consequences for the fitness of spruce trees. As a result, the natural selection
Page 39
39
on the timing and amount of cone production in white spruce resulting from squirrel cone
predation that we have documented here likely represents an important component of the
overall selection on these traits.
Page 40
40
Literature Cited
Atlan, A., M. Barat, A. S. Legionnet, L. Parize, and M. Tarayre. 2010. Genetic variation
in flowering phenology and avoidance of seed predation in native populations of
Ulex europaeus. Journal of Evolutionary Biology 23:362-371.
Augspurger, C. K. 1981. Reproductive synchrony of a tropical shrub - experimental
studies on effects of pollinators and deed predators on Hybanthus-prunifolius
(Violaceae). Ecology 62:775-788.
Augspurger, C. K. 1983. Phenology, flowering synchrony, and fruit set of six neotropical
shrubs. Biotropica 15:257-267.
Beavis, W. D., D. Grant, M. Albertsen, and R. Fincher. 1991. Quantitative trait loci for
plant height in four maize populations and their associations with qualitative
genetic loci. Theoretical and Applied Genetics 83:141-145.
Benkman, C. W., R. P. Balda, and C. C. Smith. 1984. Adaptations for seed dispersal and
the compromises due to seed predation in limber pine. Ecology 65:632-642.
Benkman, C. W., T. L. Parchman, A. Favis, and A. M. Siepielski. 2003. Reciprocal
selection causes a coevolutionary arms race between crossbills and lodgepole
pine. American Naturalist 162:182-194.
Benkman, C. W. and A. M. Siepielski. 2004. A keystone selective agent? Pine squirrels
and the frequency of serotiny in lodgepole pine. Ecology 85:2082-2087.
Bennington, C. C. and J. B. McGraw. 1995. Phenotypic selection in an artificial
population of Impatiens pallida: The Importance of the invisible fraction.
Evolution 49:317-324.
Berkum, F. H. V., R. B. Huey, J. S. Tsuji, and T. Garland. 1989. Repeatability of
individual differences in locomotor performance and body size during early
ontogeny of the lizard Sceloporus occidentalis (Baird & Girard). Functional
Ecology 3:97-105.
Boonstra, R., S. Boutin, A. Byrom, T. Karels, A. Hubbs, K. Stuart-Smith, M. Blower, and
S. Antpoehler. 2001. The role of red squirrels and arctic ground squirrels. Pages
179-214 in C.J. Krebs, S. Boutin, and R. Boonstra, editors. Ecosystem dynamics
of the boreal forest: the Kluane project Oxford University Press, Oxford.
Page 41
41
Botto, J. F. and M. P. Coluccio. 2007. Seasonal and plant-density dependency for
quantitative trait loci affecting flowering time in multiple populations of
Arabidopsis thaliana. Plant, Cell & Environment 30:1465-1479.
Chao, C. T., D. E. Parfitt, L. Ferguson, C. Kallsen, and J. Maranto. 2003. Genetic
analyses of phenological traits of pistachio (Pistacia vera, L.). Euphytica
129:345-349.
Chappell, M. A., G. C. Bachman, and J. P. Odell. 1995. Repeatability of maximal aerobic
performance in Belding's ground squirrels, Spermophilus beldingi. Functional
Ecology 9:498-504.
Cram, W. H. and H. A. Worden. 1957. Maturity of white spruce cones and seed. Forest
Science 3:263-269.
Darling, F. F. 1938. Bird flocks and the breeding cycle. Cambridge University Press,
Cambridge.
Di Renzo, M. A., M. A. Ibanex, N. C. Bonamico, and M. M. Poverene. 2000. Estimation
of repeatability and phenotypic correlations in Eragrostis curvula. The Journal of
Agricultural Science 134:207-212.
Dohm, M. R. 2002. Repeatability estimates do not always set an upper limit to
heritability. Functional Ecology 16:273-280.
Elliott, P. F. 1974. Evolutionary responses of plants to seed-eaters: Pine squirrel
predation on lodgepole pine. Evolution 28:221-231.
Elzinga, J. A., A. Atlan, A. Biere, L. Gigord, A. E. Weis, and G. Bernasconi. 2007. Time
after time: flowering phenology and biotic interactions. Trends in Ecology &
Evolution 22:432-439.
Eriksson, O. 1995. Asynchronous flowering reduces seed predation in the perennial forest
herb Actaea spicata. Acta oecologica 16:195.
Estes, R. D. 1976. The significance of breeding synchrony in the wildebeest. African
Journal of Ecology 14:135-152.
Falconer, D. S. and T. F. C. Mackay. 1996. Introduction to quantitative genetics.
Longman, Essex, England.
Page 42
42
Findlay, C. S. and F. Cooke. 1982. Synchrony in the Lesser Snow Goose (Anser
caerulescens caerulescens). II the adaptive value of reproductive synchrony.
Evolution 36:786-799.
Fletcher, Q. E., S. Boutin, J. E. Lane, J. M. LaMontagne, A. G. McAdam, C. J. Krebs,
and M. M. Humphries. 2010. The functional response of a hoarding seed predator
to mast seeding. Ecology 91:2673-2683.
Freeman, R. S., A. K. Brody, and C. D. Neefus. 2003. Flowering phenology and
compensation for herbivory in Ipomopsis aggregata. Oecologia 136:394-401.
Freitas, L. and K. Bolmgren. 2008. Synchrony is more than overlap: measuring
phenological synchronization considering time length and intensity. Revista
Brasileira de Botânica 31:721-724.
Fritsch, F. N. and R. E. Carlson. 1980. Monotone piecewise cubic interpolation. SIAM
Journal on Numerical Analysis 17:238-246.
Gomez, J. M. 1993. Phenotypic selection on flowering synchrony in a high-mountain
plant, Hormathophylla spinosa (Cruciferae). Journal of Ecology 81:605-613.
Grafen, A. 1988. On the uses of data on lifetime reproductive success. Pages 454 – 471 in
T. H. Clutton-Brock, editor. Reproductive success. Univeristy of Chicago Press,
Chicago, IL.
Hadfield, J. D. 2008. Estimating evolutionary parameters when viability selection is
operating. Proceedings of the Royal Society B: Biological Sciences 275:723-734.
Honek, A. and Z. Martinkova. 2005. Pre-dispersal predation of Taraxacum officinale
(dandelion) seed. Journal of Ecology 93:335-344.
Hughes, D. A. and J. D. Richard. 1974. The nesting of the Pacific ridley turtle
Lepidochelys olivacea on Playa Nancite, Costa Rica. Marine Biology 24:97-107.
Ims, R. A. 1990a. The ecology and evolution of reproductive synchrony. Trends in
Ecology & Evolution 5:135-140.
Ims, R. A. 1990b. On the adaptive value of reproductive synchrony as a predator-
swamping strategy. American Naturalist 136:485-498.
Janzen, D. H. 1971. Seed predation by animals. Annual Review of Ecology and
Systematics 2:465-492.
Page 43
43
Johnson, M. T. J. 2007. Genotype-by-environment interactions leads to variable selection
on life-history strategy in Common Evening Primrose (Oenothera biennis).
Journal of Evolutionary Biology 20:190-200.
Jordano, P. 1995. Frugivore-mediated selection on fruit and seed size: Birds and St.
Lucie's Cherry, Prunus mahaleb. Ecology 76:2627-2639.
Kelly, D. 1994. The evolutionary ecology of mast seeding. Trends in Ecology &
Evolution 9:465-470.
Kelly, D. and V. L. Sork. 2002. Mast seeding in perennial plants: Why, how, where?
Annual Review of Ecology and Systematics 33:427-447.
Kingsolver, J. G., H. E. Hoekstra, J. M. Hoekstra, D. Berrigan, S. N. Vignieri, C. E. Hill,
A. Hoang, P. Gibert, and P. Beerli. 2001. The strength of phenotypic selection in
natural populations. The American Naturalist 157:245-261.
Kolb, A., J. Ehrlén, and O. Eriksson. 2007. Ecological and evolutionary consequences of
spatial and temporal variation in pre-dispersal seed predation. Perspectives in
Plant Ecology, Evolution and Systematics 9:79-100.
Krebs, C., S. Boutin, and R. Boonstra, editors. 2001. Ecosystem dynamics of the boreal
forest: the Kluane project Oxford University Press, Oxford.
Lacey, E. P., D. A. Roach, D. Herr, S. Kincaid, and R. Perrott. 2003. Multigenerational
effects of flowering and fruiting phenology in Plantago lanceolata. Ecology
84:2462-2475.
LaMontagne, J. M. and S. Boutin. 2007. Local-scale synchrony and variability in mast
seed production patterns of Picea glauca. Journal of Ecology 95:991-1000.
LaMontagne, J. M., S. Peters, and S. Boutin. 2005. A visual index for estimating cone
production for individual white spruce trees. Canadian Journal of Forest Research
35:3020-3026.
Lande, R. and S. J. Arnold. 1983. The measurement of selection on correlated characters.
Evolution 37:1210-1226.
Mahoro, S. 2002. Individual flowering schedule, fruit set, and flower and seed predation
in Vaccinium hirtum Thunb. (Ericaceae). Canadian Journal of Botany-Revue
Canadienne De Botanique 80:82-92.
Page 44
44
Marquis, R. J. 1988. Phenological variation in the neotropical understory shrub Piper
arielanum: Causes and consequences. Ecology 69:1552-1565.
Matziris, D. I. 1994. Genetic variation in the phenology of flowering in black pine. Silvae
Genetica 43:321-328.
Mitchell-Olds, T. and R. G. Shaw. 1987. Regression analysis of natural selection:
Statistical inference and biological interpretation. Evolution 41:1149-1161.
Nakagawa, S. and H. Schielzeth. 2010. Repeatability for gaussian and non-gaussian data:
a practical guide for biologists. Biological Reviews 85:935-956.
Nienstaedt, H. and J. C. Zasada. 1990. Picea glauca (Moench) Voss white spruce.in R.
M. B. a. B. H. and Honkala, editors. Silvics of North America. Volume 1.
Conifers. USDA Agricultural Handbook. U.S. Department of Agriculture,
Washington, D.C.
Odonoghue, M. and S. Boutin. 1995. Does reproductive synchrony affect juvenile
survival rates of northern mammals. Oikos 74:115-121.
Parachnowitsch, A. L. and C. M. Caruso. 2008. Predispersal seed herbivores, not
pollinators, exert selection on floral traits via female fitness. Ecology 89:1802-
1810.
Parchman, T.L., C.W. Benkman, and E.T. Mezquida. 2007. Coevolution between
Hispaniolan Crossbills and pine: Does more time allow for greater phenotypic
escalation at lower latitude? Evolution 61: 2142-2153.
Peters, S., S. Boutin, and E. Macdonald. 2003. Pre-dispersal seed predation of white
spruce cones in logged boreal mixedwood forest. Canadian Journal of Forest
Research-Revue Canadienne De Recherche Forestiere 33:33-40.
Pettersson, M. W. 1994. Large plant size counteracts early seed predation during the
extended flowering season of a Silene uniflora (Caryophyllaceae) population.
Ecography 17:264-271.
Pilson, D. 2000. Herbivory and natural selection on flowering phenology in wild
sunflower, Helianthus annuus. Oecologia 122:72-82.
R Development Core Team. 2011. R: A language and environment forstatistical
computing in R Foundation for Statistical Computing, editor., Vienna, Austria.
Page 45
45
Rathcke, B. and E. P. Lacey. 1985. Phenological patterns of terrestrial plants. Annual
Review of Ecology and Systematics 16:179-214.
Robertson, R. J. 1973. Optimal niche space of the Redwinged Blackbird: Spatial and
temporal patterns of nesting activity and success. Ecology 54:1085-1093.
Rutberg, A. T. 1987. Adaptive hypotheses of birth synchrony in ruminants: An
interspecific test. The American Naturalist 130:692-710.
Samano, S. and D. F. Tomback. 2003. Cone opening phenology, seed dispersal, and seed
predation in southwestern white pine (Pinus strobiformis) in southern Colorado.
Ecoscience 10:319-326.
Santos, C. A. F. 1999. In situ evaluation of fruit yield and estimation of repeatability
coefficient for major fruit traits of umbu tree [Spondias tuberosa (Anacardiaceae)]
in the semi-arid region of Brazil. Genetic Resources and Crop Evolution 46:455-
460.
Shykoff, J. A. and O. Kaltz. 1998. Phenotypic changes in host plants diseased by
Microbotryum violaceum: Parasite manipulation, side effects, and trade-offs.
International Journal of Plant Sciences 159:236-243.
Silvertown, J. W. 1980. The evolutionary ecology of mast seeding in trees. Biological
Journal of the Linnean Society 14:235-250.
Sinsch, U. 1988. Temporal spacing of breeding activity in the natterjack toad, Bufo
calamita. Oecologia 76:399-407.
Smith, C. C. 1970. The coevolution of Pine squirrels (Tamiasciurus) and conifers.
Ecological Monographs 40:349-371.
Smith, C. C. and R. P. Balda. 1979. Competition among insects, birds and mammals for
conifer seeds. American Zoologist 19:1065-1083.
Smith, J. N. M. and N. F. G. Folkard. 2001. Other herbivores and small predators:
arthropods, birds, and mammals. Pages 261-272 in C.J. Krebs, S. Boutin, and R.
Boonstra, editors. Ecosystem dynamics of the boreal forest: the Kluane project. .
Oxford Univsersity Press, Oxford.
Steele, M. A. 1998. Tamiasciurus hudsonicus. Mammalian Species 586:1-9.
Page 46
46
Stinchcombe, J. R., A. F. Agrawal, P. A. Hohenlohe, S. J. Arnold, and M. W. Blows.
2008. Estimating nonlinear selection gradients using quadratic regression
coefficients: Double or nothing? Evolution 62:2435-2440.
Tomback, D. F. and K. A. Kramer. 1980. Limber pine seed harvest by Clark's Nutcracker
in the Sierra Nevada: Timing and foraging behavior. The Condor 82:467-468.
Vander Wall, S. B. and R. P. Balda. 1977. Coadaptations of the Clark's Nutcracker and
the Pinon pine for efficient seed harvest and dispersal. Ecological Monographs
47:89-111.
Waldron, R.M. 1965. Cone production and seedfall in a mature white spruce stand. The
Forestry Chronicle 41:314-321.
Zasada, J.C. and L.A. Viereck. 1970. White spruce cone and seed production in interior
Alaska, 1957-1968. U.S. Department of Agriculture, Forest Service, Pacific
Northwest Forest and Range Experiment Station Res. Note PNW-129 Portland,
Oregon, USA.
Page 47
47
Tables
Table 1. Standardized directional (β‟) and stabilizing/disruptive (γ’) selection gradients
calculated for intra-annual reproductive synchrony and the number of cones produced by
white spruce during a non-mast (2009; n=212) and mast (2010; n=206) year. Separate
models were used to assess linear and non-linear selection gradients within each year.
Bold font indicates significance estimated with jackknife tests (Mitchell-Olds and Shaw
1987). Stabilizing/disruptive selection gradients have been doubled to accurately
represent their strength (Stinchcombe et al. 2008).
Year Selection
Gradient Model
Parameter
Estimate
SE
P value
Non-mast β' Synchrony index 0.32 0.12 0.007
Non-mast β' Cones produced 1.01 0.45 0.026
Non-mast γ' Synchrony index -0.18 0.10 0.388
Non-mast γ' Cones produced -0.12 0.38 0.864
Mast β' Synchrony index 0.06 0.05 0.216
Mast β' Cones produced 1.30 0.16 <0.001
Mast γ' Synchrony index -0.10 0.08 0.520
Mast γ' Cones produced 0.25 0.08 0.002
Page 48
48
Figures
Figure 1. Seasonal decline in the number of closed cones on white spruce trees resulting
from red squirrel cone clipping activities as well as cone opening. Average number of
closed cones (closed triangles) and average number of open cones (open circles) per tree
in the study (non-mast; n=607, mast; n=292) plotted with the average observed and
modeled red squirrel cone clipping (dashed line) rate per day (non-mast; n=21, mast;
n=9) throughout the autumn of a non-mast (2009) and (2010) mast year. Cone count
rounds occurred over more than one day but are plotted on median date of each round.
Cone number values represent means ± SE and clipping activity values are means.
Page 49
49
Connecting Statement
In this thesis I explore how reproductive timing influences red squirrel hoarding of white
spruce cones, from the perspective of both the trees and the red squirrels. In the first
chapter I showed that the degree of intra-annual reproductive synchrony exhibited by
individual white spruce trees was positively correlated with seed escape from red
squirrels, and was under positive directional selection in a low cone year. This furthers
the evidence of strong interactions between white spruce and their dominant pre-dispersal
seed predator, the red squirrel, and highlights the importance of the brief period of cone
availability to hoarding red squirrels. In the second chapter I explore the implications of
the varying separation of reproduction in red squirrels from this brief period of time that
they have available to hoard resources for overwinter survival and future reproduction.
Page 50
50
Chapter 2: Reproductive and resource constraints on food hoarding in male
and female red squirrels.
Authors:
Devan W. Archibald (email: [email protected] )
Natural Resource Sciences, Macdonald campus, McGill University
21 111 Lakeshore Drive, Ste-Anne-de-Bellevue, QC, H9X 3V9, Canada
Andrew G. McAdam (email: [email protected] )
Department of Integrative Biology, University of Guelph,
50 Stone Road East, Guelph, ON, N1G 2W1, Canada
Stan Boutin (email: [email protected] )
Department of Biological Sciences, University of Alberta
Z1109 Biological Sciences Building, Edmonton, AB, T6G 2R3, Canada
Quinn Fletcher (email: [email protected] )
Department of Biological Sciences, University of Alberta
Z1109 Biological Sciences Building, Edmonton, AB, T6G 2R3, Canada
Murray M. Humphries (email: [email protected] )
Natural Resource Sciences, Macdonald campus, McGill University
21 111 Lakeshore Drive, Ste-Anne-de-Bellevue, QC, H9X 3V9, Canada
Page 51
51
Abstract
When investment in accumulating energy stores is not adequately separated from
reproduction, costs associated with current reproduction could trade-off with the
acquisition of resources for future reproduction and survival. We quantified conifer cone
clipping and hoarding by male and female red squirrels (Tamiasciurus hudsonicus) in
four years with varying resource levels and seasonal separation of reproduction from
hoarding. There was a 50-fold increase in the availability of cones between the lowest
and highest cone years, with higher cone years associated with a later end to the breeding
season that extended into the cone-hoarding season. The total numbers of cones clipped
and hoarded were more related to cone production than reproductive timing. However,
females, who experienced peak reproductive demands approximately three months after
males, were more likely than males to pursue the low cost and low reward strategy of
clipping more cones than they hoarded and scatter-hoarding more cones than they larder-
hoarded. These differences between males and females were most pronounced in the
highest cone year, when females were lactating while hoarding and their time spent
hoarding increased from parturition to post-weaning. Nevertheless, even in this year,
females allocated as much time to hoarding activities as males while successfully
reproducing and accumulated sufficient cone reserves to survive winter. Our findings
demonstrate interacting effects of resource availability and reproductive timing on
hoarding behavior in red squirrels, and show how high resource availability reduces the
temporal separation and trade-offs between reproduction and capital accumulation.
Page 52
52
Introduction
Trade-offs are at the core of ecological and evolutionary theory (Stearns 1989,
Lima and Dill 1990, Werner and Anholt 1993, Roff 2002, Kneitel and Chase 2004,
Ridley 2004). The basis of trade-offs are often presumed to be energy ; energy invested
in one activity cannot simultaneously be invested in another (eg. Brown et al. 2004). But,
trade-offs can also be generated by time allocation, because limited time allocated to one
activity comes at the expense of time allocated to another (Enright 1970, Halle and
Stenseth 2000). One way animals minimize trade-offs in time and energy allocation is to
organize activities seasonally such that investment in one activity does not overlap with
investment in another. For example, many animals accumulate energy reserves at one
time of the year and use these reserves as capital to support reproductive demands at
another time of year (Drent and Daan 1980, Stearns 1992, Jönsson 1997).
Opportunities for capital accumulation are often restricted to one time of the year,
due to annual cycles in resource availability (Jönsson 1997). Year-round residents of
high latitude regions may have particular difficulty in separating reproduction and capital
accumulation if both activities need to be accomplished during short growing seasons.
Furthermore, inadequate separation of the two events in time should generate a trade-off
between current reproduction and future survival and reproductive success.
Alternatively, if animals engaged in reproduction remain capable of capital accumulation,
this trade-off could be avoided or minimized. Most research on capital breeders has
focused on animals using endogenous lipid and protein stores, but many species use
exogenous stores of hoarded food (Jönsson 1997). Food hoarding is important to over-
winter survival in many species, and occasionally is used to fuel reproductive activity in
the following spring (Vander Wall 1990).
Female mammals experience their peak energetic demands during lactation,
whereas males experience their peak energetic demands during mating (Gittleman and
Thompson 1988). As a result, the generally higher energetic costs associated with
reproduction and its later completion in female mammals relative to male mammals
potentially allows females less time and energy for capital accumulation following
reproduction than males. North American red squirrels (Tamiasciurus hudsonicus) have
Page 53
53
a highly promiscuous, scramble competition mating system with energy demands of
reproduction highest in males when most females are sexually receptive (Lane et al.
2010) and highest in females during late lactation (Humphries and Boutin 2000). Given a
33 day gestation period in this species (Steele 1998), and a lactation period lasting about
50 to 70 days (Layne 1954, Humphries and Boutin 1996), females experience peak
energy demands three or more months later than males. In addition to these gender
differences, there is extensive inter-annual variation in the timing of reproduction, with
Boutin et al. (2006) reporting average parturition dates ranging from late March to early
June.
In northern areas of their range red squirrels experience extreme resource-pulses,
where their main food resource, white spruce seed (Picea glauca), exhibits yearly
variation spanning three orders of magnitude (Boutin et al. 2006, LaMontagne and Boutin
2007). Red squirrels hoard up to 15,000 white spruce cones, which are produced in late
summer and autumn, by clipping cones from the tree tops, and then burying these clipped
cones as either scatter-hoards spread throughout their territory or as larder-hoards
concentrated in a central midden (Hurly and Lourie 1997, Fletcher et al. 2010). Burying
cones below the ground surface keeps cones moist, ensuring they remain closed and
retain their seeds (Smith 1968). In contrast, clipped cones that are left on the surface
usually open and lose their seeds, either in autumn before snow accumulation or in spring
after snow melt. Scatter-hoarding is generally presumed to require less in energy and
time than larder-hoarding (Vander Wall 1990, Clarke and Kramer 1994); but see, (Hurly
and Lourie 1997), whereas larder-hoarded cones are most defensible from pilferage
(Donald 2010) and provide the most concentrated (Hurly and Lourie 1997), easily-
accessed and safely consumed energy source during winter and reproduction (Smith
1968, Steele 1998, Boutin et al. 2006, McAdam et al. 2007). Overall, cone hoarding in
autumn has been shown to be as energetically demanding for red squirrels as mating in
males and lactation in females (Lane et al. 2010, Fletcher 2011), suggesting that it might
be incompatible with other energy and time demanding activities.
Here we test the hypothesis that reduced temporal separation between
reproduction and capital accumulation in a food-hoarding rodent compromises hoarding
performance. This hypothesis predicts that in years when hoarding and reproduction are
Page 54
54
least separated, red squirrels will spend less time hoarding, total hoard accumulation will
be reduced, and/or squirrels will pursue low cost, low benefit strategies such as clipping
cones that are not immediately hoarded, and/or scatter-hoarding rather than larder-
hoarding. The alternative hypothesis is that reproduction and hoarding only overlap
under high resource circumstances when they can be simultaneously accomplished
without demonstrable costs. Because females experience peak reproductive demands
much later in the season than males, the alternative hypothesis predicts that females
would spend less time hoarding, accumulate smaller hoard sizes, and be more likely to
clip but not hoard or scatter-hoard than males. To test these predictions, we quantified
the total number of cones clipped, hoarded, and the propensity to larder-hoard in male
and female red squirrels in four years with varying resource levels and seasonal
separation of reproduction from hoarding.
Methods
Study area
The study was conducted on two study areas (approximately 40ha each) located
near Kluane National Park in southwestern Yukon, Canada (61°N, 138°W). The study
areas were located in a glacial valley composed of boreal forest dominated by white
spruce with a willow (Salix spp.) understory. Red squirrels are present at an average
density of 1.5-2.8 squirrels/ha (Boonstra et al. 2001). For a more detailed description of
the general ecology of the study site see Krebs, Boutin, and Boonstra (2001).
Red squirrel populations and white spruce cone production have been monitored
continuously since 1988 (Boutin et al. 2006, LaMontagne and Boutin 2007, McAdam et
al. 2007). Red squirrel population monitoring covered the entire reproductive season,
typically starting in early March and ending in late August, facilitated by approximately
bi-weekly live-trapping and handling of individuals (McAdam et al. 2007). Every
summer the number of cones produced by white spruce trees of cone bearing age was
counted in a consistent manner before red squirrel cone hoarding behavior started
(LaMontagne et al. 2005). Study areas were staked every 30 meters and lines flagged to
Page 55
55
create a grid allowing the recording of spatial locations of field observations with a 3m
resolution (Boutin et al. 2006, McAdam et al. 2007).
Hoarding behavioral observations
Behavioral data from autumn 2010 were combined with data from 2002, 2003,
and 2005, which were included in the analysis of the functional response of red squirrels
to white spruce cone levels presented by Fletcher et al. (2010). The dates of observations
and sample sizes varied among years (autumn 2002, 20 adult females and 8 adult males
monitored Sep 1-Oct 8; autumn 2003, 15 adult females and 17 adult males monitored
Aug 20-Oct 14; autumn 2005, 12 adult females and 9 adult males monitored Aug 28-Sep
24; autumn 2010, 13 adult females and 10 adult males monitored Aug 22-Sep 30).
Red squirrels were live trapped in Tomahawk traps baited with peanut butter and
each animal was marked with ear tags (Monel #1). Colored wires or pipe cleaners were
threaded through the ear tags in unique combinations to allow individual identification in
the field. Animals were fitted with radio collars (Model PD-2C, 4g, Holohil Systems
Limited, Carp, Ontario, Canada) to allow individuals to be reliably located. Each
individual received three focal observation periods per day three days per week. Red
squirrels are diurnal, so observation periods were spaced evenly throughout the day.
Individuals were monitored along the most efficient visitation circuit that minimized
distances between territories, but with starting locations and directions varying randomly
between days. During a focal observation period squirrels were located with radio
telemetry and their behavior was monitored for seven minutes. During the focal
observation period we recorded the number and location of all cones clipped from trees,
and all items hoarded.
In 2010, females were concurrently reproductively active while hoarding,
allowing a direct investigation of the impacts of seasonal overlap of reproduction with
time allocated to hoarding activities. In this year we altered our methodology slightly to
obtain more detailed observations. In addition to the methodology described above
(Fletcher et al. 2010), we also employed an instantaneous sampling approach (Dantzer et
al. 2011) recording the behaviors that squirrels were engaged in every 30 seconds.
Specifically, these behaviors included whether the focal squirrel was in a nest, travelling,
Page 56
56
foraging, feeding, resting, vigilant, grooming, or out of site (Stuart Smith and Boutin
1995). In 2010 we also increased the number of focal observation periods per day from
three to four.
Estimation of total number of cones clipped, hoarded, and the propensity to larder-
hoard
As in Fletcher et al.(2010), we estimated the total number of cones clipped and
hoarded by each squirrel on a given day based on its average clipping or hoarding rate per
minute of focal observation, multiplied by the number of minutes between sunrise and
sunset (61°N, 138°W; Herzberg Institute of Astrophysics; National Research Council of
Canada). In this analysis, hoarding included both scatter- and larder-hoarding. In order
to obtain seasonal estimates of total cones hoarded and total cones clipped, we used the
shape-preserving piecewise cubic modeling technique (Fritsch and Carlson 1980)
employed by Fletcher et al. (2010) to estimate hoarding activities on un-observed days
between August 15th
and October 15th
. The seasonal totals were equal to the sum of the
observed and modeled number of cones hoarded or clipped per day, but we limited our
analysis to only red squirrels monitored throughout the whole sampling period.
Larder-hoarding propensity was quantified as the proportion of cone hoarding
events observed in a season that were brought to the central larder (Jenkins and Breck
1998). Hurly and Lourie (1997) found a strong correspondence between the estimated
proportion of cones scatter or larder-hoarded by red squirrels recorded by behavioral and
direct-sampling methods (quadrats placed throughout territories), indicating behavioral
estimates provide a reliable index of hoarding behavior. We limited this analysis to red
squirrels with at least 10 observed caching events (n= 67 squirrels, median number of
events observed = 65).
Reproductive timing
The seasonal timing of reproduction relative to the food hoarding period was
assessed based on the date each individual was last observed reproductively active in
each year. For females that successfully weaned a litter the date last reproductively
active was defined as 70 days postpartum, the time at which red squirrels wean their
Page 57
57
young (Layne 1954, Humphries and Boutin 1996). Parturition dates were estimated from
the weight of the pups when nests were entered shortly after the female gave birth to her
litter (McAdam et al. 2007), and successful completion of reproduction was confirmed
with post-weaning trapping or observation of juveniles. When biweekly trapping
confirmed the mother had lost her litter before weaning, we estimated the date last
reproductively active as the median date between when a female was last trapped
lactating and first trapped non-lactating. For males, the date last reproductively active
was defined as the date of testes ascension, estimated as the median date between when a
male was last trapped with testes scored as scrotal and the first day trapped with testes
scored as abdominal (empty scrotum).
All animal use procedures were approved by McGill University Faculty of
Agricultural and Environmental Sciences Animal Care Committee.
Statistical analysis
The degree of seasonal separation between reproduction and food hoarding is
likely to be affected by both current and past white spruce cone levels. Female red
squirrels anticipate upcoming large cone crops in the current year by producing a second
litter just prior to new cone availability (Boutin et al. 2006), contributing to overlap in
reproduction and hoarding. Large cone crops in the previous year are correlated with
earlier parturition dates the following spring (Boutin et al. 2006), contributing to reduced
overlap with the hoarding period. Because an autumn with a large cone crop is invariably
followed by an autumn with a small cone crop (LaMontagne and Boutin 2007), high
overlap of reproduction and hoarding typically occurs in years of high cone production,
whereas low overlap between reproduction and hoarding typically occurs in the
subsequent low-cone year. Because of this close correspondence between resource levels
and the amount of separation between reproduction and hoarding, we did not attempt to
separate the unique effects of cone production and reproductive timing on hoarding
behavior. Instead, we focused on differences among years, combined with qualitative
assessments of each year‟s cone levels (lowest, low, medium, high) and date of
reproductive completion (early, middle, late).
Page 58
58
To determine if there were differences in (1) total cones clipped, (2) total cones
hoarded, and (3) the proportion of cones larder-hoarded among years and between
genders we used three separate generalized linear models (GLM). Total cones clipped
and hoarded were each evaluated with quasi-Poisson errors to account for over-dispersion
(Bolker 2008, O‟Hara and Kotze 2010). The proportion of cones larder-hoarded was
evaluated with quasi-binomial errors to account for over-dispersion (Bolker 2008). Each
of these models included the effect of year, sex, and their interaction. In all of our GLMs
we tested for the significance of year, sex, and their interaction using an using an F-test
based on the ratio of deviances (Venables and Ripley 2002, Bolker 2008). When
significant effects of year were found we investigated which years differed by conducting
post-hoc Tukey tests with Bonferroni corrected p-values (Bretz et al. 2011) using the
“glht” function in the “multcomp” package in the statistical software R (Hothorn et al.
2008).
To examine how directly overlapping reproduction and hoarding affected the time
budget of females in autumn of 2010, we modeled: (1) the proportions of time feeding,
(2) in the nest, and (3) devoted to cone hoarding-related activities (clipping cones,
hoarding cones, travelling with cones) in each focal observation period as a function of
days postpartum using three separate generalized linear mixed effects models (GLMM).
Because red squirrel hoarding activity rates increase to a peak and then decline (Fletcher
et al. 2010), we first determined the appropriate effect of the date on which the behavior
was sampled (days since January 1st; hereafter Julian date) without other factors included
in the model. We tested for non-linearity using a quadratic term for Julian date
(hereafter, Julian date2). If there was a significant effect of Julian date, it was included in
the model. We fit the models with quasi-binomial errors (logit link, models fit with
Penalized Quasi-Likelihood) due to over-dispersion (Bolker et al. 2009). Since we
expected maternal behavior to vary non-linearly throughout the lactation period (see
Dantzer et al. 2011), we also tested a quadratic term for days postpartum (hereafter, days
postpartum2) in our models. We then investigated differences between female and male
red squirrels in the proportion of time devoted to cone hoarding-related activities. In this
case, hoarding activity was predicted by sex and Julian date using a fourth GLMM fit as
above, but using all male and female focal data combined. To control for repeated
Page 59
59
observations on the same individuals we included a random intercept term for squirrel
identity in all of our models using the “glmmPQL” function in the “MASS” package
(Venables and Ripley 2002) in the statistical software R (R Development Core Team
2011). The choice to use penalized quasi-likelihood (PQL) was made following Bolker
et al. (2009), since for each focal observation bout our data had an expected number of
successes and failures greater than five (fifteen 30-second intervals). We tested the
significance of our fixed effects using Wald t tests (Bolker et al. 2009). However, since
we used PQL which is not true maximum likelihood, we could not use likelihood ratio
tests to test the significance of our random factors (Bolker et al. 2009) and therefore do
not report these results.
All statistical analysis were conducted using the statistical software R (R
Development Core Team 2011) with an alpha level of 0.05, and means ± standard error
are reported.
Results
Reproductive completion and resource levels
The estimated dates of reproductive completion varied considerably among years,
as did cone availability (Table 1). 2002 and 2003 had low cone levels and well separated
reproduction and hoarding, while 2005 had limited separation with moderate cone levels.
In contrast, 2010 had high cone levels and direct overlap in reproductive activity and
hoarding among females, but little to no overlap among males.
Total number of cones clipped and hoarded
The median and maximum rate of cone clipping (median 2.3 cones/min; max 50
cones/min, n=430 focals in which clipping was observed) far exceeded the median and
maximum rate of cone hoarding (median 0.3 cones/min; max 4.3 cones/min, n=1411
focals in which hoarding was observed).
The total number of cones clipped by red squirrels differed significantly among
years (Figure 1A, F3,93 = 37.5, p<0.001), ranging from 1201 ± 466 cones clipped in 2003
(n=32 squirrels) to 38,046 ± 4914 cones clipped in 2010 (n=21 squirrels). More cones
Page 60
60
were clipped in the small mast (2005) and large mast (2010) years compared to the lower
cone years with early reproductive completion, despite a lower separation of reproduction
and hoarding in the mast years (Figure 1A). On average, female red squirrels clipped 2.2
times as many cones as males (Figure 1A; F1,92 = 9.20, p = 0.003) and this gender
difference was consistent across years (sex x year interaction: F3,89 = 1.84, p=0.15).
The total number of cones hoarded (scatter plus larder-hoarded) by red squirrels
differed significantly among years (Figure 1B, F3, 93 = 24.7, p<0.001), ranging from 860 ±
194 cones hoarded in 2003 (n=32 squirrels) to 8050 ± 977 cones hoarded in 2002 (n=27
squirrels). Red squirrels hoarded significantly fewer cones in the lowest cone year (2003)
than all other years, which did not differ significantly (Figure 1B). On average, male red
squirrels hoarded 1.5 times as many cones as females (Figure 1B, F1,92 = 7.53, p=0.007),
this gender difference was consistent across years (sex x year interaction: F3,89 = 0.58,
p=0.63).
Propensity to larder-hoard
The proportion of cones that we observed being hoarded that were taken to the
central-larder differed significantly among years (Figure 1C, F3, 63 = 17.4, p<0.001),
varying from 25% larder-hoarded in 2010 (i.e. 75% scatter-hoarded) to 75% larder-
hoarded in 2002. Red squirrels larder-hoarded the lowest proportion of cones when cone
levels were highest and reproduction was latest (2010), but also larder-hoarded
proportionately small amounts when cone levels were lowest and reproduction earliest
(2003; 37%). On average, the proportion of cones larder-hoarded by males was 1.5 times
higher than females (Figure 1C, F1,62 = 4.25, p=0.044), and this gender difference was
consistent across years (sex x year interaction: F3,59 = 1.21, p=0.31).
Hoarding time allocation during concurrent reproductive activity
In 2010, the year of extensive overlap of reproduction and hoarding in females
(see Table 1), female time allocated to feeding increased linearly throughout the autumn
(Julian date = 0.02 ± 0.006, Wald t1,648 = 3.62, p<0.001) and was not a quadratic function
of Julian date (Julian date2 = -0.0009 ± 0.0006, Wald t1,647 = -1.41, p=0.16, Julian date =
0.47 ± 0.32, Wald t1,647 = 1.48, p=0.14). Time spent feeding was initially low close to
Page 61
61
parturition, increased, and then declined again to lower levels (Figure 2A ), as indicated
by a significant negative quadratic function of days postpartum (Days postpartum2 = -
0.0005 ± 0.0002, Wald t1,646 = -2.74, p=0.006, Days postpartum = 0.04 ± 0.02, Wald t1,646
= 2.20, p=0.03) with the seasonal effect of Julian date accounted for (Julian date = 0.02 ±
0.01, Wald t1,646 = 1.73, p=0.08).
Female time spent in the nest decreased linearly with throughout the autumn
(Julian date = -0.06 ± 0.01, Wald t1,648 = 0.01, p<0.001) and was not a quadratic function
of Julian date (Julian date2 = 0.001 ± 0.001, Wald t1,647 = 1.02, p=0.31, Julian date = -0.66
± 0.59, Wald t1,647 = -1.12, p=0.26). Time spent in the nest was initially high close to
parturition, decreased, and then increased slightly again (Figure 2B), as indicated by a
significant positive quadratic function of days postpartum (Days postpartum2 = 0.0009 ±
0.0003, Wald t1,646 = 3.24, p=0.001, Days postpartum = -0.10 ± 0.03, Wald t1,646 = -4.12,
p<0.001) with the seasonal effect of Julian date accounted for (Julian date = -0.02 ± 0.02,
Wald t1,646 = -1.07, p=0.29).
Female time allocated to hoarding-related activities increased, peaked, and then
declined throughout the autumn as evidenced by a significant negative quadratic function
of Julian date (Julian date2 = -0.006 ± 0.001, Wald t1,647 = -5.47, p<0.001, Julian date =
3.18 ± 0.58, Wald t1,647 = 5.45, p<0.001). Time allocated to hoarding-related activities
increased further from parturition (Figure 2C), as indicated by a significant linear term
for days postpartum (Days postpartum = 0.02 ± 0.007, Wald t1,646 = 2.48, p=0.013) with
the seasonal effect of Julian date2 accounted for (Julian date
2 = -0.006 ± 0.001, Wald t1,646
= -5.39, p<0.001, Julian date = 3.13 ± 0.59, Wald t1,646 = 5.34, p<0.001), and was not a
quadratic function of days postpartum (Days postpartum2 = 0.0003 ± 0.0003, Wald t1,645
= 0.95, p=0.34, Days postpartum = -0.006 ± 0.026, Wald t1,645 = -0.22, p=0.82) with the
seasonal effect of Julian date2 accounted for (Julian date
2 = -0.007 ± 0.001, Wald t1,645 = -
5.42, p<0.001, Julian date = 3.29 ± 0.61, Wald t1,645 = 5.36, p<0.001).
During this year of extensive overlap of reproduction and hoarding in females but
little to no overlap in males (2010; see Table 1), with both genders included in the model
time allocated to hoarding-related activities was still a negative quadratic function of
Julian date (Figure 3, Julian date2 = -0.007 ± 0.0008, Wald t1,1194 = -8.00, p<0.001, Julian
date = 3.30 ± 0.42, Wald t1,1194 = 7.96, p<0.001). The genders did not differ significantly
Page 62
62
in the proportion of time allocated to cone hoarding-related activities (sex = -0.12 ± 0.21,
Wald t1,1193 = -0.57, p=0.57), with the effect of Julian date2 accounted for (Julian date
2 = -
0.007 ± 0.0008, Wald t1,1193 = -8.01, p<0.001, Julian date = 3.31 ± 0.41, Wald t1,1193 =
7.97, p<0.001).
Discussion
In this study, we examined the effects of varying separation of reproduction and
hoarding, two energetically demanding activities for red squirrels (Lane et al. 2010,
Fletcher 2011), across four years of varying resource availability. The total number of
cones clipped was more affected by cone production than reproductive timing, with more
cones clipped when more cones were available regardless of reproductive timing. The
total number of cones hoarded was less directly related to cone production, with some
evidence of an interacting influence of reproductive timing. The number of cones
hoarded was lowest in the year with the lowest cone availability, but did not differ
significantly among the three remaining years despite substantial variation in cone
production (Figure 1). In these years, cones hoarded tended to vary according to
reproductive timing, with fewer cones hoarded when reproduction occurred later.
Gender differences in hoarding behavior were consistent with the predicted
consequences of females experiencing peak reproductive demands much later in the year
than males (Humphries and Boutin 2000, Lane et al. 2010). In particular, females tended
to pursue lower cost-lower benefit hoarding strategies than males. Females consistently
clipped more cones than males, whereas males consistently hoarded more cones than
females. Furthermore, the ratio of larder-hoards to scatter-hoards was higher in males
than females. This combination of females hoarding less and scatter-hoarding more will
lead to much smaller larder-hoards in females than males, which has been confirmed by
two studies that counted and compared the central-larder food stores of male and female
red squirrels (Gerhardt 2005, Donald 2010). The gender differences in hoarding behavior
that we observed persisted across all years, as indicated by non-significant year by gender
interactions, but the magnitude of the difference in the lower cost-lower benefit hoarding
Page 63
63
strategies tended to increase when reproduction and hoarding over-lapped more (Figure
1).
An additional way in which differences in reproductive timing could promote
gender-differences in hoarding behavior involves the timing of hoard recovery. Recovery
of scatter-hoarded and clipped but not hoarded cones will become easier and less
energetically costly when the snow melts in spring. For example, in spring 2011
following the large mast of 2010, females were observed consuming and hoarding cones
clipped but un-hoarded the previous autumn, which were exposed by melting snow (S.
Boutin, A.G. McAdam, M.M. Humphries, E. Studd, unpublished data). Given that males
experience peak reproductive demands in late winter and early spring when snow cover
remains extensive, whereas females typically experience peak energy demands in late
spring and early summer when snow cover is reduced or gone, clipped and scatter-
hoarded cones are likely to have higher energetic value for reproducing females than
reproducing males. Nevertheless, cones hoarded within the central-larder will be of
much higher energetic value for both genders at all times of the year, given their
centrality, defensibility, and low perishability. The number of cones accumulated in the
central-larder is an important determinant of over-winter survival (Larivée et al. 2010)
and is potentially very important to subsequent reproductive success. Thus, the observed
gender and year differences in larder-hoard accumulation are likely to have important life
history and demographic consequences.
Two other studies of gender-specific hoarding behavior also attributed hoarding
variation to differing reproductive roles (Clarke and Kramer 1994, Jenkins In press).
Clarke and Kramer (1994) argued that female eastern chipmunks (Tamias striatus) were
more likely to scatter-hoard to avoid hoard depletion from young present in their burrow.
Juvenile red squirrels remain near their natal territory until they establish their own
territory (Larsen and Boutin 1994), so it is possible that female red squirrels may scatter-
hoard more to avoid depletion of their central-larder by their young. Bequeathal of part
or all of a territory also occurs in red squirrels (Price 1992, Boutin et al. 1993, Price and
Boutin 1993, Boutin et al. 2000), so having cones scattered throughout the territory, a
portion of which may eventually be acquired by offspring, could be a form of parental
investment. However, juvenile dispersal and territory settlement usually occurs prior to
Page 64
64
the start of the hoarding season (Boutin et al. 1993). Therefore, assuming that the
majority of a female‟s hoard is consumed over the course a single year, these offspring
avoidance-investment arguments only apply to one year of this study (2010) when
hoarding preceded juvenile settlement. Thus, the occurrence of gender differences in
hoarding behavior that persisted across all study years is inconsistent with these
alternative offspring avoidance or parental investment explanations.
Despite finding general support for the hypothesis that low separation of
reproduction from hoarding affects the total number of cones hoarded, our results also
confirms that reproduction and hoarding can be accomplished simultaneously, at least
under high resource conditions. In 2010, when cone production was high and females
were lactating throughout most of the hoarding season, females spent as much time
hoarding as males and differences in the number of cones clipped and hoarded by males
and females were not notably larger than in other years. Although lactating females were
characterized by reduced time spent hoarding shortly after their litters were born, time
allocated to hoarding increased as their litters aged, allowing them to make-up for lost
time later in the hoarding season. All of the females for which we sampled behavior
survived the winter (S. Boutin, A.G. McAdam, M.M. Humphries, unpublished data),
indicating they accumulated the cone hoards necessary for over-winter survival.
Furthermore, their reproductive performance did not appear to be compromised; growth
rates of pups raised during the hoarding season (1.85 ± 0.08 g/day) of the sampled
females, quantified using methodology described in McAdam et al. (2007), were similar
to other years (1.46-2.05 g/day; Humphries and Boutin 2000), and each female had at
least one of their offspring recruit into the population in spring 2011 (S. Boutin, A.G.
McAdam, M.M. Humphries, unpublished data).
The very high cone production in 2010 was likely a key contributor to the ability
of females to reconcile the competing demands of reproduction and hoarding. This
capacity to sustain overlapping demands allows red squirrels flexibility in the separation
of reproduction and hoarding in high cone years, and facilitates their ability to produce
second litters in late summer when they anticipate large cone crops in advance of their
availability (Boutin et al. 2006). Resource availability affects the phenological timing of
energy demanding activities in many animals. For example, the timing of reproduction is
Page 65
65
among the most responsive of traits to experimental food supplementation in vertebrates
(Boutin 1990) and some birds respond to food supplementation by molting sooner after
reproduction (Siikamäki 1998). Together with previous literature, our results highlight
the extent of gender variation in hoarding behavior and how resource-availability can
mediate seasonal separation and trade-offs between reproduction and capital
accumulation (van Noordwijk and de Jong 1986, Dunham et al. 1989, Beilharz and Nitter
1998, Reznick et al. 2000, Boggs 2009).
Page 66
66
Literature Cited
Beilharz, R. G. and G. Nitter. 1998. The missing E: the role of the environment in
evolution and animal breeding. Journal of Animal Breeding and Genetics
115:439-453.
Boggs, C. L. 2009. Understanding insect life histories and senescence through a resource
allocation lens. Functional Ecology 23:27-37.
Bolker, B. M. 2008. Ecological models and data in R. Princeton University Press,
Princeton, N.J.
Bolker, B. M., M. E. Brooks, C. J. Clark, S. W. Geange, J. R. Poulsen, M. H. H. Stevens,
and J.-S. S. White. 2009. Generalized linear mixed models: a practical guide for
ecology and evolution. Trends in Ecology & Evolution 24:127-135.
Boonstra, R., S. Boutin, A. Byrom, T. Karels, A. Hubbs, K. Stuart-Smith, M. Blower, and
S. Antpoehler. 2001. The role of red squirrels and arctic ground squirrels. Pages
179-214 in C.J. Krebs, S. Boutin, and R. Boonstra, editors. Ecosystem dynamics
of the boreal forest: the Kluane project Oxford University Press, Oxford.
Boutin, S. 1990. Food supplementation experiments with terrestrial vertebrates: patterns,
problems, and the future. Canadian Journal of Zoology 68:203-220.
Boutin, S., K. W. Larsen, and D. Berteaux. 2000. Anticipatory parental care: acquiring
resources for offspring prior to conception. Proceedings of the Royal Society of
London Series B-Biological Sciences 267:2081-2085.
Boutin, s., Z. Tooze, and K. Price. 1993. Post-breeding dispersal by female red squirrels
(Tamiasciurus hudsonicus): the effect of local vacancies. Behavioral Ecology
4:151-155.
Boutin, S., L. A. Wauters, A. G. McAdam, M. M. Humphries, G. Tosi, and A. A.
Dhondt. 2006. Anticipatory reproduction and population growth in seed predators.
Science 314:1928-1930.
Bretz, F., T. Hothorn, and P. Westfall. 2011. Multiple Comparisons Using R. Chapman &
Hall/CRC Taylor & Francis Group, Boca Raton, FL, USA.
Brown, J. H., J. F. Gillooly, A. P. Allen, V. M. Savage, and G. B. West. 2004. Toward a
metabolic theory of ecology. Ecology 85:1771-1789.
Page 67
67
Clarke, M. F. and D. L. Kramer. 1994. Scatter-hoarding by a larder-hoarding rodent:
intraspecific variation in the hoarding behaviour of the eastern chipmunk, Tamias
striatus. Animal Behaviour 48:299-308.
Dantzer, B., A. G. McAdam, R. Palme, M. M. Humphries, S. Boutin, and R. Boonstra.
2011. Maternal androgens and behaviour in free-ranging North American red
squirrels. Animal Behaviour In press.
Donald, J. L. 2010. Intraspecific cache pilferage by larder-hoarding red squirrels
(Tamiasciurus hudsonicus) in Kluane, Yukon. M.Sc. Thesis. University of
Alberta, Edmonton.
Drent, R. H. and S. Daan. 1980. The prudent parent: Energetic adjustments in avian
breeding. Ardea 68:225-252.
Dunham, A. E., B. W. Grant, and K. L. Overall. 1989. Interfaces between biophysical
and physiological ecology and the population ecology of terrestrial vertebrate
ectotherms. Physiological Zoology 62:335-355.
Enright, J. T. 1970. Ecological aspects of endogenous rhythmicity. Annual Review of
Ecology and Systematics 1:221-238.
Fletcher, Q. E., S. Boutin, J. E. Lane, J. M. LaMontagne, A. G. McAdam, C. J. Krebs,
and M. M. Humphries. 2010. The functional response of a hoarding seed predator
to mast seeding. Ecology 91:2673-2683.
Fletcher, Q. F. 2011. The role of energy expenditure in resource acquisition and energy
allocation in free-ranging North American red squirrels (Tamiasciurus
hudsonicus). Ph.D. Thesis. McGill University, Montreal.
Fritsch, F. N. and R. E. Carlson. 1980. Monotone piecewise cubic interpolation. SIAM
Journal on Numerical Analysis 17:238-246.
Gerhardt, F. 2005. Food pilfering in larder-hoarding red squirrels (Tamiasciurus
hudsonicus). Journal of Mammalogy 86:108-114.
Gittleman, J. L. and S. D. Thompson. 1988. Energy allocation in mammalian
reproduction. American Zoologist 28:863-875.
Halle, S. and N. C. Stenseth. 2000. Activity patterns in small mammals: An ecological
approach. Springer, Berlin; New York.
Page 68
68
Hothorn, T., F. Bretz, and P. Westfall. 2008. Simultaneous inference in general
parametric models. Biometrical Journal 50:346-363.
Humphries, M. M. and S. Boutin. 1996. Reproductive demands and mass gains: A
paradox in female red squirrels (Tamiasciurus hudsonicus). Journal of Animal
Ecology 65:332-338.
Humphries, M. M. and S. Boutin. 2000. The determinants of optimal litter size in free-
ranging red squirrels. Ecology 81:2867-2877.
Hurly, T. A. and S. A. Lourie. 1997. Scatterhoarding and larderhoarding by red squirrels:
Size, dispersion, and allocation of hoards. Journal of Mammalogy 78:529-537.
Jenkins, S. H. In press. Sex differences in repeatability of food-hoarding behaviour of
kangaroo rats. Animal Behaviour In Press, Corrected Proof.
Jenkins, S. H. and S. W. Breck. 1998. Differences in food hoarding among six species of
Heteromyid rodents. Journal of Mammalogy 79:1221-1233.
Jönsson, K. I. 1997. Capital and income breeding as alternative tactics of resource use in
reproduction. Oikos 78:57-66.
Kneitel, J. M. and J. M. Chase. 2004. Trade-offs in community ecology: linking spatial
scales and species coexistence. Ecology Letters 7:69-80.
Krebs, C., S. Boutin, and R. Boonstra, editors. 2001. Ecosystem dynamics of the boreal
forest: the Kluane project Oxford University Press, Oxford.
LaMontagne, J. M. and S. Boutin. 2007. Local-scale synchrony and variability in mast
seed production patterns of Picea glauca. Journal of Ecology 95:991-1000.
LaMontagne, J. M., S. Peters, and S. Boutin. 2005. A visual index for estimating cone
production for individual white spruce trees. Canadian Journal of Forest Research
35:3020-3026.
Lane, J. E., S. Boutin, J. R. Speakman, and M. M. Humphries. 2010. Energetic costs of
male reproduction in a scramble competition mating system. Journal of Animal
Ecology 79:27-34.
Larivée, M. L., S. Boutin, J. R. Speakman, A. G. McAdam, and M. M. Humphries. 2010.
Associations between over-winter survival and resting metabolic rate in juvenile
North American red squirrels. Functional Ecology 24:597-607.
Page 69
69
Larsen, K. W. and S. Boutin. 1994. Movements, survival, and settlement of red squirrel
(Tamiasciurus hudsonicus) offspring. Ecology 75:214-223.
Layne, J. N. 1954. The biology of the red squirrel, Tamiasciurus hudsonicus-Loquax
(Bangs), in central New York. Ecological Monographs 24:227-267.
Lima, S. L. and L. M. Dill. 1990. Behavioral decisions made under the risk of predation:
a review and prospectus. Canadian Journal of Zoology 68:619-640.
McAdam, A. G., S. Boutin, A. K. Sykes, and M. M. Humphries. 2007. Life histories of
female red squirrels and their contributions to population growth and lifetime
fitness. Ecoscience 14:362-369.
O‟Hara, R. B. and D. J. Kotze. 2010. Do not log-transform count data. Methods in
Ecology and Evolution 1:118-122.
Price, K. 1992. Territorial bequeathal by red squirrel mothers: A dynamic model. Bulletin
of Mathematical Biology 54:335-354.
Price, K. and S. Boutin. 1993. Territorial bequeathal by red squirrel mothers. Behavioral
Ecology 4:144-150.
R Development Core Team. 2011. R: A language and environment forstatistical
computing in R Foundation for Statistical Computing, Vienna, Austria.
Reznick, D., L. Nunney, and A. Tessier. 2000. Big houses, big cars, superfleas and the
costs of reproduction. Trends in Ecology & Evolution 15:421-425.
Ridley, M. 2004. Evolution. Oxford University Press, Oxford; New York.
Roff, D. A. 2002. Life history evolution. Sinauer Associates, Sunderland, Mass.
Siikamäki, P. 1998. Limitation of reproductive success by food availability and breeding
time in Pied Flycatchers. Ecology 79:1789-1796.
Smith, M. C. 1968. Red squirrel responses to Spruce cone failure in interior Alaska. The
Journal of Wildlife Management 32:305-317.
Stearns, S. C. 1989. Trade-Offs in life-history evolution. Functional Ecology 3:259-268.
Stearns, S. C. 1992. The evolution of life histories. Oxford Univ. Press, Oxford.
Steele, M. A. 1998. Tamiasciurus hudsonicus. Mammalian Species 586:1-9.
Stuart-Smith, A. K. and S. Boutin. 1995. Behavioural differences between surviving and
depredated juvenille red squirrels. Ecoscience 2: 34-40.
Page 70
70
van Noordwijk, A. J. and G. de Jong. 1986. Acquisition and allocation of resources: Their
influence on variation in life history tactics. The American Naturalist 128:137-
142.
Vander Wall, S. B. 1990. Food hoarding in animals. University of Chicago Press,
Chicago.
Venables, W. N. and B. D. Ripley. 2002. Modern applied statistics with S. Springer, New
York.
Werner, E. E. and B. R. Anholt. 1993. Ecological consequences of the trade-off between
growth and mortality rates mediated by foraging activity. The American
Naturalist 142:242-272.
Page 71
71
Tables
Table 1. Range and median date of reproductive completion for adult red squirrels included in the study in 2002, 2003, 2005, and 2010
with white spruce average cone index (ln transformed cone count; mean ± SE) calculated from annual cone counts conducted prior to
red squirrel harvesting (n = 167-171 trees per year). Days until hoarding season was calculated by subtracting the median date last
reproductively active from August 16th
, the approximate date when hoarding activity initiates (Fletcher et al. 2010).
Females Males
Year
Cone
Index
Reproductive
Completion
Qualitative
Classification
Days
Until
Hoarding
Season
Date of
Reproductive Completion
Days
Until
Hoarding
Season
Date of
Reproductive Completion
Median
Range
Median
Range
2002 1.9 ± 0.1 Early 37 July 12 May 13 – Aug 15 52 June 29 May 30 – July 30
2003 0.9 ± 0.1 Early 34 July 13 May 6 – Sep 18 30 July 17 June 23 – Aug 8
2005 3.4 ± 0.2 Middle 13 Aug 4 June 30 – Sep 21 11 Aug 5 July 27 – Sep 6
2010 4.9 ± 0.1 Late -63 Oct 18 Sep 13 – Nov 3 -4 Aug 20 July 31 – Sep 3
Page 72
72
Figures
Figure 1. Total number of cones clipped (A), hoarded (B) and the proportion of hoarded
cones that were larder-hoarded (C) by adult male and female red squirrels over four study
years with varying cone availability and separation of reproduction and hoarding. All
models contained significant sex and year effects with different letters indicating
significant differences found between years using post hoc testing. None of the models
contained significant interactions between year and sex. Values are means ± SE.
Page 73
73
Figure 2. The proportion of time adult female red squirrels spent feeding (A) and in the
nest (B) varied as a quadratic function of days postpartum, but time spent conducting
cone hoarding-related activities (C) varied linearly with days postpartum during autumn
2010, after accounting for the appropriate fit of Julian date (time spent feeding; linear,
time spent in the nest; linear, time spent hoarding; quadratic). Parturition dates ranged
from Jul 12 to Aug 24 (median; Aug 9). Data were analyzed using generalized linear
mixed models with squirrel identity as a random factor, but plotted values are raw data
representing means ± SE for each day postpartum.
Page 74
74
Figure 3. The proportion of time adult female (closed circles) and male (open squares)
red squirrels spent conducting cone hoarding-related activities throughout autumn 2010
varied non-linearly with Julian date and was not significantly different between the sexes.
The dashed grey line indicates the proportion of study females that were yet to pass the
mid-point of lactation (proportion less than 45 days postpartum). Data were analyzed
using a generalized linear mixed model with squirrel identity as a random factor, but
plotted values are raw data representing means ± SE for each day.
Page 75
75
General Conclusions
This thesis evaluated how the seasonal timing of reproduction influenced red
squirrel hoarding of white spruce cones, from the perspective of both the trees (Chapter
1) and the squirrels (Chapter 2).
In the trees we found evidence that the degree of intra-annual reproductive
synchrony exhibited by individual white spruce trees relative to others in the population
enhanced seed escape from red squirrels in both a mast and non-mast year, and that this
trait was under strong selection in the lower cone year. This indicates masting species
can employ a mixture of strategies to avoid seed predation, both escape in numbers and
time, but that the relative importance of these strategies varies with levels of seed
production. This second aspect of reproductive synchrony, intra-annual synchrony, has
rarely been addressed in masting species (Rathcke and Lacey 1985). Selection on intra-
annual reproductive synchrony is likely imposed by a combination of many members of
the seed predator community. Post-dispersal seed predation can also be severe (Janzen
1971), and if intra-annual reproductive synchrony is under selection from seed predators
they will likely have an influence as well. Other studies should examine the importance
of intra-annual reproductive synchrony in other masting plants, how it varies with
variation in seed production, and the community of seed predators.
In red squirrels, our findings suggest that cone hoarding behaviour is more
affected by annual cone levels than reproductive timing, and in years with high cone
production both activities can successfully be combined despite their associated high
energetic costs (Lane et al. 2010, Fletcher 2011). Since both reproduction and hoarding
can be accomplished by red squirrels, their ability to anticipate large cone crops in
advance of their availability by producing second litters in late summer (Boutin et al.
2006) does not appear to compromise their ability to simultaneously establish a food
hoard necessary for future survival and reproduction. The compatibility of reproduction
and hoarding under high resource conditions provides red squirrels considerable
flexibility in reproductive timing relative to the autumn hoarding period, ranging from
multi-month separation in a year of low cone production preceded by high cone
production to almost complete overlap in a year of high cone production preceded by low
Page 76
76
cone production. However, we also found that males and females used different hoarding
strategies that were consistent with differences in the timing of reproductive completion,
indicating that although overall hoarding behaviour was driven by resource levels, the
timing of reproduction may be a factor in gender differences. These findings highlight
the importance of considering the timing of environmental and physiological events
throughout the year in understanding animal behaviour (McNamara and Houston 2008).
The timing of reproduction is important in predator-prey dynamics, but often only
the reproduction of the prey is considered (Ims 1990b, a). Combined, the findings of this
thesis illustrate the importance of considering the timing of reproduction of both the prey,
and the predator, which will lead to a more complete understanding of how phenological
patterns in reproduction influence predator-prey dynamics. Future research should
investigate the relationship of hormone levels of both the predator and the prey and how
their variation throughout the year influences predator-prey dynamics. Hormones are
important components in mediating animal behaviour (von Holst 1998, Lonstein and De
Vries 2000) and plant growth and timing of reproduction are also under hormonal control
(Hooley 1994). Red squirrel maternal behavioural patterns are correlated with androgen
levels suggesting hormones may play a role in mediating red squirrel behaviour during
reproduction (Dantzer et al. 2011). In other rodents removal of the testis or ovaries has
influenced hoarding (Nyby et al. 1973, Borker and Gogate 1984) as have metabolic
hormones (Keen-Rhinehart et al. 2010), indicating that hoarding is under hormonal
control. Additionally, variation in hoarding strategies are often explained in terms of
varying levels of territoriality (Vander Wall 1990), and aggression is often linked with
testosterone levels (Soma 2006). Hormone levels of both predator and prey could be
linked via ingestion of the prey by the predator. Red squirrel anticipation of large cone
crops in advance of their availability (Boutin et al. 2006) may be influenced by white
spruce hormone levels in the differentiating cone buds. The plant hormone gibberellins is
associated with increased cone production in white spruce (Pharis et al. 1986), and in the
low cone year prior to the mast of 2010 we noted red squirrels were eating many buds
during the time that cone bud differentiation occurs. Understanding the hormonal
changes throughout the year of both the predator and the prey, in addition to the timing of
reproduction, may lead to a greater understanding of predator-prey dynamics.
Page 77
77
Literature Cited
Borker, A. S. and M. G. Gogate. 1984. Role of ovarian hormones in hoarding pattern.
Indian Journal of Physiology and Pharmacology 28:115-120.
Boutin, S., L. A. Wauters, A. G. McAdam, M. M. Humphries, G. Tosi, and A. A.
Dhondt. 2006. Anticipatory reproduction and population growth in seed predators.
Science 314:1928-1930.
Dantzer, B., A. G. McAdam, R. Palme, M. M. Humphries, S. Boutin, and R. Boonstra.
2011. Maternal androgens and behaviour in free-ranging North American red
squirrels. Animal Behaviour In press.
Fletcher, Q. F. 2011. The role of energy expenditure in resource acquisition and energy
allocation in free-ranging North American red squirrels (Tamiasciurus
hudsonicus). PhD Thesis. McGill University, Montreal.
Hooley, R. 1994. Gibberellins: perception, transduction and responses. Plant Molecular
Biology 26:1529-1555.
Ims, R. A. 1990a. The ecology and evolution of reproductive synchrony. Trends in
Ecology & Evolution 5:135-140.
Ims, R. A. 1990b. On the adaptive value of reproductive synchrony as a predator-
swamping strategy. American Naturalist 136:485-498.
Janzen, D. H. 1971. Seed predation by animals. Annual Review of Ecology and
Systematics 2:465-492.
Keen-Rhinehart, E., M. J. Dailey, and T. Bartness. 2010. Physiological mechanisms for
food-hoarding motivation in animals. Philosophical Transactions of the Royal
Society B: Biological Sciences 365:961-975.
Lane, J. E., S. Boutin, J. R. Speakman, and M. M. Humphries. 2010. Energetic costs of
male reproduction in a scramble competition mating system. Journal of Animal
Ecology 79:27-34.
Lonstein, J. S. and G. J. De Vries. 2000. Sex differences in the parental behavior of
rodents. Neuroscience & Biobehavioral Reviews 24:669-686.
Page 78
78
McNamara, J. M. and A. I. Houston. 2008. Optimal annual routines: behaviour in the
context of physiology and ecology. Philosophical Transactions of the Royal
Society B: Biological Sciences 363:301-319.
Nyby, J., P. Wallace, K. Owen, and D. D. Thiessen. 1973. An influence of hormones on
hoarding behavior in the Mongolian gerbil (Meriones unguiculatus). Hormones
and Behavior 4:283-288.
Pharis, R. P., D. Tomchuk, F. D. Beall, R. M. Rauter, and G. Kiss. 1986. Promotion of
flowering in white spruce (Picea glauca) by gibberellin A4/7, auxin
(naphthaleneacetic acid), and the adjunct cultural treatments of girdling and
Ca(NO3)2 fertilization. Canadian Journal of Forest Research 16:340-345.
Rathcke, B. and E. P. Lacey. 1985. Phenological patterns of terrestrial plants. Annual
Review of Ecology and Systematics 16:179-214.
Soma, K. K. 2006. Testosterone and aggression: Berthold, birds and beyond. Journal of
Neuroendocrinology 18:543-551.
Vander Wall, S. B. 1990. Food hoarding in animals. University of Chicago Press,
Chicago.
von Holst, D. 1998. The concept of stress and its relevance for animal behavior. Pages 1-
131 in M. M. Anders Pape Møller and J. B. S. Peter, editors. Advances in the
study of behavior. Academic Press.