Environmental, biological and anthropogenic effects on grizzly bear body size:
temporal and spatial considerations
BMC Ecology 2013, 13:31 doi:10.1186/1472-6785-13-31
Scott E Nielsen (scottn@ualberta.ca)
Background
Individual body growth is controlled in large part by the spatial and temporal heterogeneity
of, and competition for, resources. Grizzly bears (Ursus arctosL.) are an excellent species for
studying the effects of resource heterogeneity and maternal effects (i.e. silver spoon) on life
history traits such as body size because their habitats are highly variable in space and time.
Here, we evaluated influences on body size of grizzly bears in Alberta, Canada by testing six
factors that accounted for spatial and temporal heterogeneity in environments during
maternal, natal and ‘capture’ (recent) environments. After accounting for intrinsic biological
factors (age, sex), we examined how body size, measured in mass, length, and body
condition, was influenced by: (a) population density; (b) regional habitat productivity; (c)
inter-annual variability in productivity (including silver spoon effects); (d) local habitat
quality; (e) human footprint (disturbances); and (f) landscape change.
Results
We found sex and age explained the most variance in body mass, condition and length (R
2
from 0.48–0.64). Inter-annual variability in climate the year before and of birth (silver spoon
effects) had detectable effects on the three-body size metrics (R
2
from 0.04–0.07); both
maternal (year before birth) and natal (year of birth) effects of precipitation and temperature
were related with body size. Local heterogeneity in habitat quality also explained variance in
body mass and condition (R
2
from 0.01–0.08), while annual rate of landscape change
explained additional variance in body length (R
2
= 0.03). Human footprint and population
density had no observed effect on body size.
Conclusions
These results illustrated that body size patterns of grizzly bears, while largely affected by
basic biological characteristics (age and sex), were also influenced by regional environmental
gradients the year before, and of, the individual’s birth thus illustrating silver spoon effects.
The magnitude of the silver spoon effects was on par with the influence of contemporary
regional habitat productivity, which showed that both temporal and spatial influences explain
in part body size patterns in grizzly bears. Because smaller bears were found in colder and
less-productive environments, we hypothesize that warming global temperatures may
positively affect body mass of interior bears, provided there is sufficient snow-cover during
the denning periods.
Discussion
Biological factors and body size
Measurements of body mass and length of grizzly bears in Alberta were strongly dependent
on intrinsic biological factors: age (positive, non-linear relationship) and sex (males >
females). Age, sex and offspring dependence were important factors affecting body
condition, which is a short-term measure of growth. Adult females, and especially adult
females with cubs of the year, were likely to be in poorer condition than male bears. A
negative effect of capture history (number of captures) was also observed for body condition
measures which is consistent with previous observations [61]. Although population density
(density dependence) is known to inversely affect body-size patterns in animals [80-82], no
density dependent effects on body size patterns of grizzly bears were observed in our study.
Grizzly bear populations in Alberta are likely to be below carrying capacity given locally
high rates of human-caused mortality [83,84], and were recently classified by the province as
‘threatened’ given the low observed population densities [75]. This is in contrast to brown
bears in Sweden that are considered healthy [85], but where body sizes of adult female bears
are inversely related to population density [48].
Temporal and spatial environmental heterogeneity
Environmental heterogeneity is an important mechanism by which animal populations are
regulated [86]. Here, we found that regional heterogeneity in habitat productivity was a
moderate predictor of body size patterns of grizzly bears in Alberta. The smallest bears by
mass and length occurred in the least-productive and coldest environments as measured by
alpine habitat use and home ranges occupying both cool average spring temperatures and
high average March precipitation (snowfall). In the Canadian Rocky Mountains, all three of
these factors are associated with late timing of spring snowmelt and plant emergence, which
are known to affect population dynamics of other alpine mammals [87]. Since den emergence
in grizzly bears in our area typically occurs in April to early May [88], the amount and timing
of spring snowpack is likely a factor affecting the availability of early season food resources
such as roots [89], and generally might restrict access to early spring food resources.
Inter-annual variations in climate during the years’ prior, duringand/or just following birth
(maternal, in-uteroand natal environments, respectively) also affected adult body size. Such
silver-spoon effects by which animals that are born into ‘rich’ conditions are favoured
throughout life are consistent with observations in other mammals including polar bears [90],
Soay sheep [1], red squirrels [37] and caribou [91]. Common among these studies is the
importance of winter and spring climate during (natal environments) or just prior (maternal or
in uteroenvironments) to the year of birth, which we also observed in this study. Winter and
spring climate is related to summer drought conditions in the Canadian Rocky Mountains
[92], which suggests that the effect of winter and spring climate may not necessarily be
directly associated with denning period, but affects summer environments where water is
limiting. We are unsure, however, how late summer precipitation affects cubs-of-the-year. It
may be related to late summer food resources, such as fruit production, or affect foodresource abundance in the following year when bears are yearlings. Further, winter
precipitation (December-March) anomalies during the natal birth year were positively related
to body mass. We interpreted this as snow cover during winter denning providing energetic
benefits (e.g. insulation) in the den for cubs of the year.
During the year prior to birth, late summer (July-August) temperature anomalies were
negatively associated with body mass but positively associated with body length in grizzly
bears. This late-summer environment might have directly affected maternal body condition
prior to denning and thus subsequent condition of offspring [e.g. 53] or conversely, it may
have affected the following years’ food supply during the cub-of-year period, since lag
effects in fruit production are caused by weather conditions favourableto flower primordia in
the mid-to-late summer period the year prior to fruiting [93]. Although we cannot be certain
which factor is more important, the fact that body mass is negatively associated with late-
summer temperature anomalies, where as body length is positively associated with latesummer temperature anomalies suggests to us that maternal condition is less likely (as we
would expect similar responses in body mass and length if it were solely a maternal effect).
Further investigations of mid and late-summer weather on pulsing in food resource
abundance the following year are needed, especially in regard to the apparent opposite effects
on bear mass and length.
One important consideration to our purported silver spoon effect should be discussed: that is,
we have no information on our study animals prior to their first capture. This has two
important implications: 1) we cannot account for litter size effects, and 2) the centroid data
used to determine natal climatic conditions may not be reflective of the actual natal location.
In regards to the former, not accounting for litter size should inflate the variance around our
estimates. For the centroid data, this would likely only influence dispersing males, as females
are philopatric [94]. For males, average dispersal distances in the province are under 50
kilometers [94], thus still largely reflective of the climate in the centroid of the current home
range (differences in climates among bears are mainly regional in effect, not within
populations). Further, for this limitation to bias our results, males would consistently have to
disperse to poorer environments, again something we deem unlikely. Thus, weargue that the
silver spoon pattern is unlikely to be altered by these factorsin such away that the statistical
pattern would disappear.
Anthropogenic considerations
Human footprint did not directly relate to body size patterns of grizzly bears, but human
activity indirectly affected body size by influencing habitats. The two most important
measures of habitat quality were canopy closure and the age structure of forests. Bears that
used habitats associated with higher canopy variability, such as forest/non-forest landscapes
in the mountains or expanses of old growth forests with a recent, single-harvestsequence, had
lower body masses. Conversely, bears that used forests with highervariability in regenerating
forest age had higher body condition. Likewise, body length was positively related to annual
landscape change. Taken together, these results suggest that human activities that fragment
forests are positively associated with body size measures, although survival of bears in these
environments is compromised due to high rates of human-caused mortalities [57,84]. Early
successional and highly variable forests are therefore important indicators of improved
habitat quality for bears given the relationship to body size patterns reported here, habitat use
studies [72] and measures of food resource abundance [73,74]. We hypothesize that positive
associations between body size patterns and variability in regenerating forest age are due in
part to local landscape patterns in protein availability. For instance, both ungulate and ant
resource use in Alberta are associated with disturbed forests [46,74].
Conclusions
While bear body size is largely dictated by age and sex, it still only accounted for little more
than 50% of the variation. More consideration of the spatial and temporal patterns of resource
availability, including the conditions early in life, is needed to better understand individual
performance of animals and population dynamics. For grizzly bears in Alberta,
environmental effects on body size are most affected by regional environmental gradients
(space) and the environmental conditions animals are born into (time). Local-habitat
heterogeneity (particularly young, patchily disturbed forests), and landscape dynamics also
had a small influence on body size. It is important to emphasize that while patchily disturbed
forests positively affected body size, these areas also have high rates of mortality, which
could negate any positive population-level effect.
Worldwide, relationships between carnivore body size and climate warming show ambiguous
trends [95]; however, polar bears body sizes have recently declined, which has been
attributed primarily to loss in habitat (i.e., sea ice as a platform for hunting; [96,97]). Despite
unequivocal global patterns [95], a 50 year examination regional study showed that carnivore
body sizes have generally increased over the past half century [98]. Given the short season
associated with high-alpine environments, such as the Rocky Mountains in Alberta, we
hypothesize that individuals with a limited growing season and temperature-limited
ecosystems, such as interior grizzly bears, might actually benefit from increases in season
length associated with climate change. This prediction is largely consistent with observed
body size and seasonality patterns in grizzly bears across North America [40], but is
dependent on sufficient snow cover during the denning period. In conclusion, we have
demonstrated a complex interplay of biological, spatial and temporal factors on body size that
collectively explained between 60 and 84% of the variation seen in Alberta’s grizzly bears.
www.biomedcentral.com/content/pdf/1472-6785-13-31.pdfMarc RL Cattet (marc.cattet@usask.ca)
John Boulanger (boulange@ecological.bc.ca)
Jerome Cranston (arctos@telus.net)
Greg J McDermid (mcdermid@ucalgary.ca)
Aaron BA Shafer (aaron.shafer@ebc.uu.se)
Gordon B Stenhouse (gstenhouse@foothillsri.ca)