Inter-specific synchrony of two contrasting ungulates: Wild boar (Sus scrofa) and roe deer (Capreolus capreolus)

Abstract

Very few studies on ungulates address issues of inter-specific synchrony in population responses to environmental variation such as climate. Depending on whether annual variation in performance of ungulate populations is driven by direct or indirect (trophic) interactions, very different predictions regarding the pattern of inter-specific synchrony can be derived. We compared annual autumn body mass variation in roe deer (Capreolus capreolus) and wild boar (Sus scrofa) from Poland over the period 1982-2002, and related this to variation in winter and summer climate and plant phenological development [the Normalized Difference Vegetation Index (NDVI), derived from satellites]. Roe deer fawns (approximately 1.3 kg increase from year 1982 to 2002) and yearlings both increased markedly in mass over years. There was also an increase for wild boar mass over years (approximately 4.2 kg increase for piglets from 1982 to 2002). Despite our failure to link annual body mass to spring or winter conditions or the NDVI, the body mass of roe deer and wild boar fluctuated in synchrony. As this was a field roe deer population, and since wild boar is an omnivore, we suggest this may be linked to annual variation and trends in crop structure (mainly rye). We urge future studies to take advantage of studying multiple species in order to gain further insight into processes of how climate affect ungulate populations.

Full text

Oecologia (2007) 151:232–239
DOI 10.1007/s00442-006-0584-z
123
POPULATION ECOLOGY
Inter-speciWc synchrony of two contrasting ungulates: wild boar
(Sus scrofa) and roe deer (Capreolus capreolus)
Atle Mysterud · Piotr Tryjanowski · Marek Panek ·
Nathalie Pettorelli · Nils C. Stenseth
Received: 19 May 2006 / Accepted: 11 October 2006 / Published online: 11 November 2006
© Springer-Verlag 2006
Abstract Very few studies on ungulates address
issues of inter-speciWc synchrony in population
responses to environmental variation such as climate.
Depending on whether annual variation in perfor-
mance of ungulate populations is driven by direct or
indirect (trophic) interactions, very diVerent predic-
tions regarding the pattern of inter-speciWc synchrony
can be derived. We compared annual autumn body
mass variation in roe deer (Capreolus capreolus) and
wild boar (Sus scrofa) from Poland over the period
1982–2002, and related this to variation in winter and
summer climate and plant phenological development
[the Normalized DiVerence Vegetation Index (NDVI),
derived from satellites]. Roe deer fawns (»1.3 kg
increase from year 1982 to 2002) and yearlings both
increased markedly in mass over years. There was also
an increase for wild boar mass over years (»4.2 kg
increase for piglets from 1982 to 2002). Despite our
failure to link annual body mass to spring or winter
conditions or the NDVI, the body mass of roe deer and
wild boar Xuctuated in synchrony. As this was a Weld
roe deer population, and since wild boar is an omni-
vore, we suggest this may be linked to annual variation
and trends in crop structure (mainly rye). We urge
future studies to take advantage of studying multiple
species in order to gain further insight into processes of
how climate aVect ungulate populations.
Keywords Climate · Crops · NDVI · Poland ·
Trophic interactions
Introduction
Climate inXuence dynamics of ungulate populations
both directly and indirectly through trophic interac-
tions (reviews in Weladji et al. 2002; Mysterud et al.
2003). Population sizes of northern ungulates typically
decline after severe winters with much snow (e.g.
Jacobson et al. 2004; Grøtan et al. 2005; Mysterud and
Østbye 2006b), and there may also be negative eVects
of severe winter conditions on body mass (Cederlund
et al. 1991). Early summer conditions (temperature
and precipitation) are regarded as particularly impor-
tant for individual growth and operate through climate
eVects on plants (Klein 1965; Sæther and Heim 1993;
Langvatn et al. 1996; Mysterud et al. 2001). This has
recently been veriWed directly through the use of satel-
lite-derived vegetation indices [Normalized DiVerence
Communicated by Jean-Michel Gaillard.
Electronic supplementary material Supplementary material is
available in the online version of this article at http://dx.doi.org/
10.1007/s00442-006-0584-z and is accessible for authorized users.
A. Mysterud (&) · N. Pettorelli · N. C. Stenseth
Centre for Ecological and Evolutionary Synthesis (CEES),
Department of Biology, University of Oslo,
P.O. Box 1066, Blindern, 0316 Oslo, Norway
e-mail: atle.mysterud@bio.uio.no
P. Tryjanowski
CEH Monks Wood, Abbots Ripton,
Huntingdon, Cambs PE28 2LS, UK
P. Tryjanowski
Department of Behavioural Ecology,
Adam Mickiewicz University, Umultowska 89,
61-614 Poznaj, Poland
M. Panek
Polish Hunting Association, Research Station,
64-020 Czempin, Poland

Oecologia (2007) 151:232–239 233
123
Vegetation Index (NDVI)] at large spatial scales
(Pettorelli et al. 2005a, c).
In the literature on rodents, inter-speciWc patterns of
synchrony have been a main theme to separate hypoth-
eses related to predation and food (e.g. Stenseth and
Ims 1993; Hanski and Henttonen 1996; Korpimäki
et al. 2005). However, very few studies on ungulates
address issues of inter-speciWc synchrony in population
responses to climate. In Norway, body mass of domes-
tic sheep (Ovis aries) has been shown to vary annually
in synchrony with mass of red deer (Cervus elaphus)
(Mysterud et al. 2001), moose (Alces alces) (Sæther
1985) and semi-domestic reindeer (Rangifer tarandus)
(Weladji et al. 2003), while on Greenland, some evi-
dence has suggested population synchrony between
muskox (Ovibos moshcatus) and reindeer (e.g. Post
and Forchhammer 2002, but see also Vik et al. 2004).
Depending on whether annual variation in perfor-
mance of ungulate populations is driven by direct or
indirect (trophic) interactions, very diVerent predic-
tions regarding the pattern of inter-speciWc synchrony
can be derived. If ungulates are aVected directly by fac-
tors such as snow depth, patterns of annual variation in
performance should be irrespective of ungulate diet
(and dietary overlap). However, we would expect
large-sized species to be less strongly aVected and, pos-
sibly, grazers to be more strongly aVected than brows-
ers since the Weld layer is more restricted by snow than
deciduous browse. In contrast, if annual variation in
performance is driven by climate operating through
plants, we would expect synchrony only among species
with a similar diet, unless diVerent forage plants are
synchronous as well.
In this study, we compare patterns of annual varia-
tion in body mass of two very diVerent ungulates, the
wild boar (Sus scrofa) and roe deer (Capreolus capreo-
lus), using a long-term data set (1982–2002) from
Poland. Wild boar is a medium-sized (»80–150 kg)
omnivore (Jedrzejewska and Jedrzejewski 1998), while
the roe deer is a small (»20–30 kg) browser (Andersen
et al. 1998). We aim to test whether these species are
aVected directly by conditions during winter (duration
of snow cover and temperature) and/or early summer
(temperature and precipitation, as well as the satellite
derived NDVI), and to test whether there is inter-
speciW
c synchrony in annual body mass variation. Our
prediction is that both species may be aVected by snow,
but that roe deer should be more inXuenced by plants
(NDVI) being a strict herbivore. We predict inter-
speciWc synchrony in annual variation if body mass of
both species is driven by direct eVects of climate
(through snow depth), but not if they are aVected indi-
rectly through plants.
Materials and methods
Study area
The study was carried out in the experimental area,
approximately 150 km
2
in extent, of the Polish Hunting
Association Research Station at Czempij, western
Poland (52°08⬘N, 16°44⬘E). This is a typical farmland
region, with arable Welds occupying nearly 70% of the
area. The climate of the region is typical for central
Europe, where oceanic and continental climate types
meet. The mean annual temperature is ca. 8°C (sub-
zero mean monthly temperatures occur in December–
February) with mean annual precipitation ca. 550 mm
(Ryszkowski et al. 1996). The two study species are the
most common and most numerous large mammal spe-
cies both in the study area, and in western Poland in
general (Bresinski and Jedryczkowski 1999). While
both species in the past were mainly connected with
forests, roe deer in the 1930s and wild boar in the 1970s
started to establish local Weld populations in this area
of Poland (Andrzejewski and Jezierski 1978; Pielowski
and Bresinski 1982; Kaluzinski 1982b). Roe deer live
here mainly on arable Welds, sporadically using small
woodland patches as resting sites (Pielowski and Bres-
inski 1982; Bresinski 1982; Kaluzinski 1982a, b). Food
of roe deer in the study area consists mainly of crops
and grasses (over 75%), while in some years, big Xocks
of deer foraged on oil-seed rape Welds (Kaluzinski
1982a).
Roe deer data
Data on roe deer derive from hunting in the Czempij
area of Poland. We obtained data on carcass mass (i.e.,
live mass minus viscera, bleedable blood and metapo-
dials), hereafter referred to as body mass. Carcass
mass correlates closely with total body mass (Wallin
et al. 1996, for moose). Roe deer were aged by tooth
wear (e.g. Hewison et al. 1999; Mysterud and Østbye
2006a). As this method is not highly reliable especially
for older ages, we considered only fawns (»0.5 year)
and yearlings (»1.5 years). We restricted the analysis
to data from October through January. From this
period, there was limited data on males, and we there-
fore limited the analysis to data on females. Sample
size was therefore 444 individuals from 1982 to 2002
(Appendix).
Wild boar data
Data on wild boar also derive from hunting in the
Czempij area of Poland. In Poland, wild boar are shot

234 Oecologia (2007) 151:232–239
123
not only for trophies but also as a farmland pest (And-
rzejewski and Jezierski 1978). Body mass data are car-
cass mass as for roe deer (see above). In this area, wild
boar only have one litter each year. Age groups can be
classiWed to (1) piglets, which are individuals less than
1 year (from 1 April y
t
to 31 March y
t+1
), (2) yearlings,
which are between 1 and 2 years of age, and (3) adults
that are more than 2 years of age. Due to low sample
sizes from parts of the year, we restricted the analysis
to data from September through January. Further, as
body mass is not stable from 2 years onwards, we
excluded adults and analysed variation in mass of pig-
lets and yearlings. Sample size was therefore 311 indi-
viduals from 1982 to 2002 (Appendix). For wild boar,
there were also available data on counts during drive
hunts in spring each year. We used this as an index for
density, though we have no assessment of how well this
tracks actual population size, therefore results must be
interpreted with some caution. The trend in counts
over time was marked. There was a decrease from the
Wrst half of 1980s (»150 counted annually), until mid-
1990 (»50 counted annually), and then an increase
again until 2002 (»150 counted annually).
Hunting in the area is done using three methods:
drive hunts, stalking and posting (i.e. waiting for prey
in one location). There have been no special prefer-
ences for diVerent methods over the study period nor
are we aware of any changes over the study period.
The problem in most cases working with hunters’ data
are that no independent data are available, so that
hunter selectivity cannot be addressed. However, as we
work only with wild boar piglets and yearlings and roe
deer females, it is highly unlikely that hunter selectivity
is important. The methods of eviscerating roe deer and
wild boar have not changed over time.
Climate data
We obtained monthly averages of snow depth, temper-
ature and precipitation from a local weather station at
Turew, in the western sector of the study area.
We also used the seasonal indices of the North Atlan-
tic Oscillations (NAO) (Hurrell et al. 2003; available
from http://www.cgd.ucar.edu/»jhurrell/nao.pc.html)
(winter, December–February; spring, March–May; sum-
mer, June–August; autumn, September–November).
The best indices are regarded as those based on princi-
pal component analysis (Hurrell et al. 2003), and they
were therefore used in addition to the station-based
winter index, which is the one most commonly used
(Stenseth et al. 2003). In general, high NAO values are
correlated with much precipitation and high tempera-
tures in the study area in Poland, in particular during
winter. A more detailed description of these indices is
given elsewhere (Hurrell et al. 2003), and also concern-
ing their major impact on ecological systems (Stenseth
et al. 2002; Mysterud et al. 2003).
NDVI data
We used the NDVI from the pathWnder Advanced
Very High Resolution Radiometer (AVHRR) col-
lected by the National Oceanic and Atmospheric
Administration (NOAA). We used NDVI data from
the GIMMS group (kindly provided by Compton
Tucker) at a spatial scale of 8£8 km and aggregated for
every second week. The NDVI is regarded as the most
robust index and correlates strongly with aboveground
net primary productivity and absorbed photosyntheti-
cally active radiation (Myneni et al. 1997; Kerr and
Ostrovsky 2003; review in Pettorelli et al. 2005b). The
NDVI index is derived from the ratio of red to near-
infrared reXectance [NDVI = (NIR ¡ RED)/(NIR +
RED)], where NIR and RED are the amounts of near-
infrared and red light reXected by the vegetation. This
corresponds to the absorption of red light by chloro-
phyll and the scattering of near-infrared light by meso-
phyll leaf structure. NDVI values typically range from
¡0.2 to 1 (theoretically from ¡1 to +1, but values less
than ¡0.2 correspond to water), negative values corre-
sponding to an absence of vegetation (Justice et al.
1985; Pettorelli et al. 2005b). Data on the NDVI was
available from 1982 to 2002. The seasonal and annual
pattern of variation in the NDVI is given in appendix
Fig. 1. We also used the integrated NDVI value
(INDVI), which corresponds to the sum of the NDVI
values over the growing season (Pettorelli et al. 2005b).
Statistical analyses
We explored the correlation between environmental
covariates with simple Pearson correlation coeY
cients,
and trends over years with simple linear regressions.
To analyse variation in body mass, we used mainly
linear models (LM) after some initial use of additive
models (AM; Hastie and Tibshirani 1990) with
smoothing splines to ensure that predictors were line-
arly related to response variables. We used the stan-
dard logarithmic transformation [ln(weight)] of body
mass to get residuals with constant variance. We also
used linear-mixed eVects (LME) models with year as a
random eVect (Lindsey 1999), which is a more conser-
vative approach than using each year as the level of
replication (rather than number of individuals).
As we had a fairly high number of climatic variables
potentially inXuencing the dynamics, we used the

Oecologia (2007) 151:232–239 235
123
Akaike Information Criterion (AIC; Burnham and
Anderson 1998; Johnson and Omland 2004) for select-
ing an appropriate model for hypothesis testing. The
model with the lowest AIC value is regarded as the
best compromise between explaining most of the
variation and simultaneously using as few parameters
as possible. We used the small-sample correction
AIC
c
=AIC + 2K(K +1)/(N ¡ K + 1), where N is the
number of observations and K is the number of regres-
sion coeYcients including intercept. The detailed strat-
egy when selecting models is given in tables (see
Appendix). Model selection was always done on LM
and not LME (cf. Crawley 2003).
Separate models were run for roe deer and wild
boar. When comparing the pattern of synchrony, we
used a simpler model without any environmental vari-
ables, but with “cohort year” entered as a categorical
term so as to estimate body mass for each year. We
then calculated mass change from 1 year to the next
and used linear regression, with predicted values for
mass change of wild boar (most common category, i.e.
male piglets in November) regressed on predicted val-
ues for mass change of roe deer (female fawns in
November) with the (square root) number of observa-
tions for wild boar as “weights” (but years with n<5
were excluded).
All analyses were done in S-Plus versus 6.2 (Ven-
ables and Ripley 1994; Crawley 2003).
Results
Temporal trends and correlations in environmental
variables
There was no trend in April temperature (r
2
=0.007,
T=0.355, P=0.726), May temperature (r
2
=0.002,
T=¡0.181, P=0.858), number of snow days (r
2
=0.031,
T=0.776, P=0.447) or in the winter index of the NAO
(r
2
=0.022, T=¡0.658, P=0.518) over the study period.
There tended to be a positive trend in the NDVI over
time, as seen for the INDVI (r
2
=0.160, INDVI=
¡55.020 (§35.137) + 0.0336 (§0.0176) years, T=1.905,
P=0.072) and to a lesser extent for NDVI in spring
(r
2
=0.126, NDVI, 15 April=¡4.495 (§3.050) + 0.0025
(§0.0015) years, T=1.653, P=0.115). There was no cor-
relation between NDVI and temperature in April
(NDVI, 15 April r=0.055; INDVI r=¡0.145). The NAO
in winter was positively correlated with temperature
(January r=0.642; February r=0.453; March r=0.703),
but not with precipitation (January r=0.259; February
r=0.296; March r=0.269) or duration of snow cover
(r=0.298).
Roe deer
Carcass mass of fawns averaged 11.3 kg (from 7.0 to
16.0) and yearlings 15.7 kg (from 10.2 to 20.2) over the
whole period 1982–2002. The most parsimonious
model as assessed with the AICc explained 62.7% of
the variation in female roe deer body mass (slaugh-
tered) between October and January (Table 1). As
would be expected, fawns were smaller than yearlings.
Body masses were smaller in October than from
November to January, especially for fawns. Body mass
increased over years (Fig. 1). A fawn in October had a
mass of 10.25 kg in 1982 and 11.55 kg in year 2002.
Temperature in April entered the most parsimonious
model being marginally signiWcant (Table 1). In a more
restrictive LME with year as a random variable, the
eVect of temperature in April was not signiWcant (l.s.
estimate=0.00906, SE=0.00615, df=19, T=1.474,
P=0.157). The diVerent indices based on the NDVI did
not enter the most parsimonious model.
Wild boar
There was huge variation in body mass of wild boar,
ranging from 7 to 51 kg in piglets and from 27 to 86 kg
in yearlings. The most parsimonious model as assessed
with the AICc explained 68.6% of the variation in
body mass (slaughtered) of piglets and yearlings
between September and January (Table 2). Naturally,
piglets were smaller than yearlings, and males were
larger than females, and the diVerence between males
and females was larger for yearlings than for piglets.
There was also some variation in mass due to month,
and this interacted with age. There was a positive trend
in body mass over years (Table 2); piglets and yearlings
weighed on average, respectively, 27.1 and 55.8 kg in
1982 and 31.3 and 64.5 kg in 2002 (Fig. 1). There was
also a positive eVect of population density as indexed
Table 1 Results from the most parsimonious model for annual
variation in (ln) body mass (kg) of roe deer (Capreolus capreolus)
from 1982 to 2002 in Poland
Variable L.s. estimate SE TP
Intercept ¡9.1562 2.2745 ¡4.026 0.000
Month
January versus
December
0.0114 0.0177 0.647 0.518
November versus
December
0.0192 0.0199 0.965 0.335
October versus
December
¡0.0194 0.0208 ¡0.931 0.352
Age (1.5 vs 0.5 year) 0.3362 0.0128 26.266 0.000
Cohort year 0.0058 0.0011 5.036 0.000
Temperature, April 0.0089 0.0049 1.835 0.067

236 Oecologia (2007) 151:232–239
123
from the spring counts (Table 2), which also remained
when running a more conservative LME with year as a
random eVect (l.s. mean=0.00209, SE=0.000608, df=20,
T=3.431, P=0.003). Temperature in May entered the
most parsimonious model, but was not signiWcant
(Table 2).
Inter-speciWc synchrony
Based on estimated body mass separately for both spe-
cies (read out for wild boar piglets and yearling roe
deer), there was signiWcant correlation between body
mass change of roe deer and wild boar from one year
to the next (r
2
=0.412, Fig. 2). This result was similar
when using bootstrap (bootstrap estimate=0.130, 95%
CI=0.0296, 0.232), and remained when adding a linear
year term to remove possible trends (Table 3). This
result was robust to the choice of age class. For roe
deer, when restricting analysis to years with data on
both age classes (Appendix), adding an interaction
term between age and cohort year (categorical)
resulted in a less parsimonious model (AIC=7.000).
For wild boar, when restricting analysis to years with
data on both age and sex classes (Appendix), adding an
interaction term between cohort year (categorical) and
age (AIC=12.655) or sex (AIC=14.882) resulted in
less parsimonious models.
Discussion
For both roe deer and wild boar, body mass increased
considerably over years (Fig. 1). There tended to be a
parallel increase in the NDVI. However, we failed to
link variation in NDVI to body mass of either roe deer
or wild boar, and if this non-signiWcant term was never-
theless added to the model, the estimate of the NDVI
eVect was positive in roe deer and negative in wild
boar. Therefore, the rather strong trend in mass of
Fig. 1 Temporal development of average body mass of a roe deer
(Capreolus capreolus) and b wild boar (Sus scrofa) in Poland
1982–2002. Dotted lines indicate 95% conWdence intervals. Note
that data are unadjusted for monthly variation. Size of circles is di-
rectly proportional to the (sqrt) sample size (within species), but
overall sample sizes are smaller for wild boar than for roe deer
Year
10
12
14
16
18
ydob reed eoR
Roe deer Yearlings
Fawns
1985 1990 1995 2000
1985 1990 1995 2000
Year
0
20
40
60
80
100
)gk( ssam
)gk( ssam
ydob raob dliW
Wild boar
Yearlings - males
Yearlings - females
Piglets - males
Piglets - females
A
B
Table 2 Results from the
most parsimonious model for
annual variation in (ln) body
mass (kg) of wild boar from
1982 to 2002 in Poland
Variable L.s. estimate SE TP
Intercept ¡19.9221 5.3726 ¡3.708 0.000
Sex (male vs female) 0.1789 0.0511 3.503 0.001
Age (piglets vs yearlings) ¡0.5369 0.0913 ¡5.881 0.000
Month
January versus December ¡0.0659 0.0861 ¡0.766 0.445
November versus December ¡0.0465 0.0861 ¡0.540 0.590
October versus December 0.0221 0.0917 0.241 0.810
September versus December ¡0.1263 0.0869 ¡1.453 0.147
Cohort year 0.0120 0.0027 4.452 0.000
Density (harvest size) 0.0020 0.0005 3.718 0.000
Temperature, May ¡0.0177 0.0114 ¡1.550 0.122
Sex £ age ¡0.1647 0.0651 ¡2.528 0.012
Age £ month
January versus December 0.0307 0.1036 0.296 0.767
November versus Decembr ¡0.0517 0.1062 ¡0.487 0.627
October versus December ¡0.2715 0.1188 ¡2.285 0.023
September versus December ¡0.4039 0.1079 ¡3.744 0.000

Oecologia (2007) 151:232–239 237
123
both species may likely be caused by some other factor
in the environment. As there was no eVect of winter
conditions in either wild boar or roe deer, it was sur-
prising that annual body mass variation of the two spe-
cies was synchronous, as we predicted synchronous
pattern only if direct eVects of climate operating
through plants were present.
The lack of strong responses of climatic variation on
performance of roe deer and wild boar may be either
biological or methodological in origin. Litter size was
unknown neither in roe deer nor in wild boar.
Although heavier animals might be born in smaller lit-
ters (wild boar, Fernándex-Llario et al. 2003; roe deer,
Andersen et al. 2000), whether the diVerence persists
over time is unclear. Further, we also did not have data
on population density of roe deer, and only the spring
counts as a measure of density in wild boar. Lack of
data on density for roe deer may also have made it
more diYcult to Wnd eVects of variation in climate,
especially if the eVect of population density interacts
with the climate eVect (Sauer and Boyce 1983; Portier
et al. 1998; Coulson et al. 2001; Stenseth et al. 2004).
The eVect of density on wild boar body mass was small
and positive, suggesting no competition within the den-
sity range experienced. The general impression is that
population density for both species has increased dur-
ing the study period (Bresinski and Jedryczkowski
1999), thus we would have predicted decreased body
mass over time. As there was both a common trend
and inter-speciWc synchrony in body mass, and since
for both species we would have predicted decreasing
rather than increasing body mass if density increased
over time, there is likely some other factor that we
have not measured that is responsible for these pat-
terns.
The causes of inter-speciWc synchrony have been
highlighted as complex and diYcult to identify (Lieb-
hold et al. 2004). In our case, we can only provide
implicit evidence for a mechanism, by excluding both
direct eVects of winter weather and indirect eVects of
plants (NDVI). A possible explanation is that both the
trend and synchrony is related to crop structure, which
may be important for both species. While rye, lucerne
and oil-seed rape are the most important food sources
for roe deer (over 60% of diet except in spring; Kalu-
zinski 1982a), cereals, potatoes and maize are the most
important food sources for the wild boar (for our area,
Genov 1981; general review in Schley and Roper 2003).
Even though wild boar seem to prefer maize, they also
eat rye as do roe deer. In the study area (around 1995),
agricultural crops were 50% cereals (mainly rye), 20%
row crops (including oil-seed rape), 10% perennial fod-
der crops and 20% others (Ryszkowski et al. 1996).
Both maize and cereals may be links between the two
species, and crop rotation is similar to the Norfolk sys-
tem (cf. Berzsenyi et al. 2000), therefore cover structure
changes annually and likely may show a trend as well.
We suggest further studies comparing sympatric
populations of ungulates to be rewarding—both inter-
speciWc synchrony and lack of such will help us under-
stand better the mechanisms by which climate impacts
on ungulates (Mysterud et al. 2001). Part of the lack of
such studies likely arises due to data limitations. Most
monitoring is focussed on single species, and data from
diVerent species are often gathered by diVerent people
or institutions not usually collaborating. A better over-
view of the rewards of such cooperation may indeed
facilitate more interspeci
Wc comparisons in the future.
Acknowledgments We gratefully acknowledge the Wnancial
support of the Research Council of Norway to A.M. (YFF Pro-
ject). We are grateful to W. Bresijski, R. Kamieniarz and many
of local hunters for help in collecting Weld data, J. Karg for assis-
tance with obtaining temperature data. P.T.’s sabbatical at Monks
Wood was funded by the Foundation for Polish Science. We
Fig. 2 The relationship between annual autumn body mass
change of roe deer and wild boar in Poland for the period 1982–
2002. Each circle represent mass change from 1 year to the next
[i.e. (mass year
t
)/(mass year
t¡1
)]. Size of circles is directly propor-
tional to the (sqrt) sample size for wild boar (more data were
available for roe deer). Dotted lines indicate 95% conWdence
intervals
0.90 0.95 1.00 1.05 1.10
Roe deer body mass change (relative)
0.96
0.98
1.00
1.02
1.04
)evit
al
e
r( egnah
c ssam
ydob raob
dl
iW
Table 3 Resu
l
ts from t
h
e
li
near mo
d
e
l
for annua
l
c
h
ange
i
n
(l
n
)
body mass (kg) of wild boar as a function of annual change in roe
deer mass and cohort year as a continuous term (to remove
trends)
Variable L.s. estimate SE TP
Intercept ¡0.3713 0.7853 ¡0.473 0.644
Ln
(roe deer mass change)
0.1359 0.0542 2.510 0.025
Cohort year
(continuous)
0.0006 0.0004 1.571 0.139

238 Oecologia (2007) 151:232–239
123
greatly appreciated the comments of Jean-Michel Gaillard,
Nicolas Morellet, and one anonymous referee to previous drafts
of this paper.
References
Andersen R, Duncan P, Linnell JDC (1998) The European roe
deer: the biology of success. Scandinavian University Press,
Oslo
Andersen R, Gaillard J-M, Linnell JDC, Duncan P (2000) Factors
aVecting maternal care in an income breeder, the European
roe deer. J Anim Ecol 69:672–682
Andrzejewski R, Jezierski J (1978) Management of a wild boar
population and its eVects on commercial land. Acta Theriol
23:309–339
Berzsenyi Z, GyorVy B, Lap D (2000) EVect of crop rotation and
fertilisation on maize and wheat yields and yield stability in
a long-term experiment. Eur J Agron 13:225–244
Bresinski W (1982) Grouping tendencies in roe deer under agr-
ocenosis conditions. Acta Theriol 27:427–447
Bresinski W, Jedryczkowski WB (1999) Situation of hunting
game and some protected species in Dezydery Chlapowski
landscape park and its surroundings. Biul Parków Krajo-
brazowych Wielkopolski 5:81–101
Burnham KP, Anderson DR (1998) Model selection and infer-
ence: a practical information-theoretic approach. Springer,
Berlin Heidelberg New York
Cederlund G, Sand H, Pehrson Å (1991) Body mass dynamics of
moose calves in relation to winter severity. J Wildl Manage
55:675–681
Coulson T, Catchpole EA, Albon SD, Morgan BJT, Pemberton
JM, Clutton-Brock TH, Crawley MJ, Grenfell BT (2001)
Age, sex, density, winter weather, and population crashes in
Soay sheep. Science 292:1528–1531
Crawley MJ (2003) Statistical computing. An introduction to data
analysis using S-Plus. Wiley, Chichester
Fernándex-Llario P, Mateos-Quesada P, Silvério A, Santos P
(2003) Habitat eVects and shooting techniques on two wild
boar (Sus scrofa) populations in Spain and Portugal. Z Jagd-
wiss 49:120–129
Genov P (1981) Food composition of wild boar in north-eastern
and western Poland. Acta Theriol 26:185–205
Grøtan V, Sæther B-E, Engen S, Solberg EJ, Linnell JDC,
Andersen R, Brøseth H, Lund E (2005) Climate causes
large-scale spatial synchrony in population Xuctuations of a
temperate herbivore. Ecology 86:1472–1482
Hanski I, Henttonen H (1996) Predation on competing rodent
species: a simple explanation of complex patterns. J Anim
Ecol 65:220–232
Hastie T, Tibshirani R (1990) Generalized additive models.
Chapman & Hall, London
Hewison AJM, Vincent JP, Angibault JM, Delorme D, Van La-
ere G, Gaillard J-M (1999) Tests of estimation of age from
tooth wear on roe deer of known age: variation within and
among populations. Can J Zool 77:58–67
Hurrell JW, Kushnir Y, Ottersen G, Visbeck M (2003) The North
Atlantic Oscillation. Climatic signiWcance and environmen-
tal impact. American Geophysical Union, Washington DC
Jacobson AR, Provenzale A, von Hardenberg A, Bassano B, Fes-
ta-Bianchet M (2004) Climate forcing and density dependence
in a mountain ungulate population. Ecology 85:1598–1610
Jedrzejewska B, Jedrzejewski W (1998) Predation in vertebrate
communities. The Bialowieza Primeval forest as a case study.
Springer, Berlin Heidelberg New York
Johnson JB, Omland KS (2004) Model selection in ecology and
evolution. Trends Ecol Evol 19:101–108
Justice CO, Townshend JRG, Holben BN, Tucker CJ (1985)
Analysis of the phenology of global vegetation using meteo-
rological satellite data. J Remote Sens 6:1271–1318
Kaluzinski J (1982a) Composition of the food of roe deer living in
Welds and the eVects of their feeding on plant production.
Acta Theriol 27:457–470
Kaluzinski J (1982b) Dynamics and structure of a Weld roe deer
population. Acta Theriol 27:385–408
Kerr JT, Ostrovsky M (2003) From space to species: ecological
applications for remote sensing. Trends Ecol Evol 18:299–
305
Klein DR (1965) Ecology of deer range in Alaska. Ecol Monogr
35:259–284
Korpimäki E, Norrdahl K, Huitu O, Klemola T (2005) Predator-
induced synchrony in population oscillations of coexisting
small mammal species. Proc R Soc Lond B 272:193–202
Langvatn R, Albon SD, Burkey T, Clutton-Brock TH (1996) Cli-
mate, plant phenology and variation in age at Wrst reproduc-
tion in a temperate herbivore. J Anim Ecol 65:653–670
Liebhold A, Koenig WD, Bjørnstad ON (2004) Spatial synchrony
in population dynamics. Annu Rev Ecol Syst 35:467–490
Lindsey JK (1999) Models for repeated measurements. Oxford
University Press, Oxford
Myneni RB, Keeling CD, Tucker CJ, Asrar G, Nemani RR
(1997) Increased plant growth in the northern high latitudes
from 1981 to 1991. Nature 386:698–702
Mysterud A, Østbye E (2006a) Comparing simple methods for
ageing roe deer Capreolus capreolus: can any of them be use-
ful in management? Wildl Biol 12:101–107
Mysterud A, Østbye E (2006b) The eVect of climate and density
on individual and population growth of roe deer Capreolus
capreolus at northern latitudes—the Lier valley, Norway.
Wildl Biol 12:321–329
Mysterud A, Stenseth NC, Yoccoz NG, Langvatn R, Steinheim G
(2001) Nonlinear eVects of large-scale climatic variability on
wild and domestic herbivores. Nature 410:1096–1099
Mysterud A, Stenseth NC, Yoccoz NG, Ottersen G, Langvatn R
(2003) The response of the terrestrial ecosystems to climate
variability associated with the North Atlantic Oscillation. In:
Hurrell JW, Kushnir Y, Ottersen G, Visbeck M (eds) The
North Atlantic Oscillation. American Geophysical Union,
Washington, pp 235–262
Pettorelli N, Mysterud A, Yoccoz NG, Langvatn R, Stenseth NC
(2005a) Importance of climatological downscaling and plant
phenology for red deer in heterogeneous landscapes. Proc R
Soc Lond B 272:2357–2364
Pettorelli N, Vik JO, Mysterud A, Gaillard J-M, Tucker C, Stens-
eth NC (2005b) Using the satellite-derived normalized diVer-
ence vegetation index (NDVI) to assess ecological responses
to environmental change. Trends Ecol Evol 20:503–510
Pettorelli N, Weladji RB, Holand Ø, Mysterud A, Breie H, Stens-
eth NC (2005c) The relative role of winter and spring condi-
tions: linking climate and landscape-scale plant phenology to
alpine reindeer performance. Biol Lett 1:24–26
Pielowski Z, Bresinski W (1982) Population characteristics of roe
deer inhabiting a small forest. Acta Theriol 27(28):409–425
Portier C, Festa-Bianchet M, Gaillard J-M, Jorgenson JT, Yoccoz
NG (1998) EVects of density and weather on survival of big-
horn sheep lambs (Ovis canadensis). J Zool 245:271–278
Post E, Forchhammer MC (2002) Synchronization of animal
population dynamics by large-scale climate. Nature 420:168–
171
Ryszkowski L, French NR, Kedziora A (1996) Dynamics of an
agricultural landscape. PWRiL, Poznañ

Oecologia (2007) 151:232–239 239
123
Sæther B-E (1985) Annual variation in carcass weight of Norwe-
gian moose in relation to climate along a latitudinal gradient.
J Wildl Manage 49:977–983
Sæther B-E, Heim M (1993) Ecological correlates of individual
variation in age at maturity in female moose (Alces alces): the
eVects of environmental variability. J Anim Ecol 62:482–489
Sauer JR, Boyce MS (1983) Density dependence and survival of
elk in northwestern Wyoming. J Wildl Manage 47:31–37
Schley L, Roper TJ (2003) Diet of wild boar Sus scrofa in Western
Europe, with reference to consumption of agricultural crops.
Mammal Rev 33:43–56
Stenseth NC, Chan K-S, Tavecchia G, Coulson T, Mysterud A,
Clutton-Brock T, Grenfell BT (2004) Modelling non-addi-
tive and nonlinear signals from climatic noise in ecological
time series: Soay sheep as an example. Proc R Soc Lond B
271:1985–1993
Stenseth NC, Ims RA (1993) The biology of lemmings. Academic
Press, London
Stenseth NC, Mysterud A, Ottersen G, Hurrell JW, Chan K-S,
Lima M (2002) Ecological eVects of climate Xuctuations.
Science 297:1292–1296
Stenseth NC, Ottersen G, Hurrell JW, Mysterud A, Lima M,
Chan K-S, Yoccoz NG, Ådlandsvik B (2003) Studying cli-
mate eVects on ecology through the use of climate indices:
the North Atlantic Oscillation, El Niño Southern Oscillation
and beyond. Proc R Soc Lond B 270:2087–2096
Venables WN, Ripley BD (1994) Modern applied statistics with
S-plus. Springer, Berlin Heidelberg New York
Vik JO, Stenseth NC, Tavecchia G, Mysterud A, Lingjerde OC
(2004) Living in synchrony on Greenland coasts? Nature
427:697–698
Wallin K, Cederlund G, Pehrson Å (1996) Predicting body mass
from chest circumference in moose Alces alces. Wildl Biol
2:53–58
Weladji RB, Klein DR, Holand Ø, Mysterud A (2002) Compara-
tive response of Rangifer tarandus and other northern ungu-
lates to climatic variability. Rangifer 22:33–50
Weladji RB, Steinheim G, Holand Ø, Moe SR, Almøy T, Ådnøy
T (2003) Temporal patterns of juvenile body weight variabil-
ity in sympatric reindeer and sheep. Ann Zool Fenn 40:17–26