ArticlePDF Available

Abstract and Figures

Understanding whether organisms will be able to adapt to human-induced stressors currently endangering their existence is an urgent priority. Globally, multiple species moult from a dark summer to white winter coat to maintain camouflage against snowy landscapes. Decreasing snow cover duration owing to climate change is increasing mismatch in seasonal camouflage. To directly test for adaptive responses to recent changes in snow cover, we repeated historical (1950s) field studies of moult phenology in mountain hares ( Lepus timidus ) in Scotland. We found little evidence that population moult phenology has shifted to align seasonal coat colour with shorter snow seasons, or that phenotypic plasticity prevented increases in camouflage mismatch. The lack of responses resulted in 35 additional days of mismatch between 1950 and 2016. We emphasize the potential role of weak directional selection pressure and low genetic variability in shaping the scope for adaptive responses to anthropogenic stressors.
Content may be subject to copyright.
royalsocietypublishing.org/journal/rspb
Research
Cite this article: Zimova M, Giery ST, Newey
S, Nowak JJ, Spencer M, Mills LS. 2020 Lack of
phenological shift leads to increased
camouflage mismatch in mountain hares.
Proc. R. Soc. B 287: 20201786.
https://doi.org/10.1098/rspb.2020.1786
Received: 24 July 2020
Accepted: 19 November 2020
Subject Category:
Global change and conservation
Subject Areas:
ecology, evolution
Keywords:
adaptation, climate change, historical resurvey,
phenological mismatch, phenotypic plasticity,
snow
Author for correspondence:
Marketa Zimova
e-mail: mzimovaa@umich.edu
Electronic supplementary material is available
online at https://doi.org/10.6084/m9.figshare.
c.5227963.
Lack of phenological shift leads to
increased camouflage mismatch in
mountain hares
Marketa Zimova1,2, Sean T. Giery4, Scott Newey5,
J. Joshua Nowak2, Michael Spencer6and L. Scott Mills3
1
School for Environment and Sustainability, University of Michigan, Ann Arbor, MI 48104, USA
2
Wildlife Biology Program, and
3
Wildlife Biology Program and Office of Research and Creative Scholarship,
University of Montana, Missoula, MT 59812, USA
4
Department of Biology, The Pennsylvania State University, University Park, PA 16801, USA
5
The James Hutton Institute, Craigiebuckler, Aberdeen AB15 8QH, UK
6
Scotlands Rural College, Kings Buildings, Edinburgh EH9 3JG, UK
MZ, 0000-0002-8264-9879; STG, 0000-0003-3774-5295; SN, 0000-0002-2264-964X;
JJN, 0000-0002-8553-7450; LSM, 0000-0001-8771-509X
Understanding whether organisms will be able to adapt to human-induced
stressors currently endangering their existence is an urgent priority. Globally,
multiple species moult from a dark summer to white winter coat to maintain
camouflage against snowy landscapes. Decreasing snow cover duration
owing to climate change is increasing mismatch in seasonal camouflage. To
directly test for adaptive responses to recent changes in snow cover, we
repeated historical (1950s) field studies of moult phenology in mountain
hares (Lepus timidus) in Scotland. We found little evidence that population
moult phenology has shifted to align seasonal coat colour with shorter
snow seasons, or that phenotypic plasticity prevented increases in camou-
flage mismatch. The lack of responses resulted in 35 additional days of
mismatch between 1950 and 2016. We emphasize the potential role of weak
directional selection pressure and low genetic variability in shaping the
scope for adaptive responses to anthropogenic stressors.
1. Introduction
Recent climate change has already subjected wild populations to large changes
in environmental conditions [1]. Failure of populations to sufficiently track
these changes will result in local declines and extinctions [2,3]. Some popu-
lations have responded adaptively via phenotypic plasticity and/or evolution
[47]. However, others have failed to track climate change or seem to have
responded in non-adaptive ways [810]. Predicting population responses to
climate change remains challenging, in part because many interacting factors
determine future trajectories of inherently complex natural systems [11]. Yet,
understanding whether and how populations will respond to climate change
is one of the most urgent challenges facing biologists [12].
In response to climate change, snow cover duration is decreasing in many
parts of the Northern Hemisphere [13,14], and consequently imposing chan-
ging and potentially novel selection pressures on organisms adapted to
seasonally changing environments [15]. A diverse group of birds and mammals
moult from summer dark to winter white coat annually to increase crypsis
against snow [16,17]. While weather, especially temperature, can fine tune the
phenology of those moults each year, changing daylength is the principal
driver of the moults across taxa [16,18]. As snow duration declines owing to
climate change, colour moulting species become increasingly mismatched
with their background [19]. Field studies indicate that mismatch in seasonal
coat colour and snow presence or absence has negative individual and popu-
lation consequences via increased predator-induced mortality [2025]. For
example, snowshoe hares (Lepus americanus) experience 714% decreased
© 2020 The Author(s) Published by the Royal Society. All rights reserved.
weekly survival when mismatched against their background
[23,24]. Given this strong selection against mismatch, persist-
ence of colour moulting species will require adaptation to
future changes in global snow cover [26,27].
Climate-mediated phenotypic plasticity, described in
several colour moulting species, can, theoretically, buffer
against camouflage mismatch [16,19]. However, previous
studies that investigated plasticity in response to climate
change showed that current levels of plasticity are insufficient
to prevent mismatch; snowshoe hares and least weasels
(Mustela nivalis) became mismatched during years with fewer
days of snow cover [18,19,22,28]. This suggests that adaptive
evolution of moult phenology and/or moult plasticityevol-
utionary rescue [27,29]may be crucial for the persistence of
colour moulting species. Whether evolutionary shifts in
moult phenology (e.g. shifts in the photoperiodic response)
or plasticity (e.g. shifts in the sensitivity to temperature) can
occur is unknown. However, existing intrapopulation variation
in moult phenology and strong selection favouring cryptic
coloration suggest evolutionary rescue is possible [23,30].
Historical phenological studies provide some of the
only opportunities to test whether organisms have already
responded to climate change. Unfortunately, such datasets
are extremely rare for moult phenology. Fortunately, Watson
[31] and Flux [32] described seasonal moult of wild mountain
hares (Lepus timidus scoticus) in the northeast and central high-
lands of Scotland over spring and autumn seasons during the
1950s and 1960s. To our knowledge, this effort represents
the longest-running systematic historical survey of moult
phenology in any species. These studies documented intrapo-
pulation variation in haresmoult phenology each year and
population-level phenotypic plasticity in response to ambient
temperature, especially in the spring [31,32]. The adaptive
capacity of mountain hares to mitigate camouflage mismatch
via phenotypic plasticity is unknown, however. Similarly, the
selective costs of camouflage mismatch in the highly managed
Scottish Highlands have not been investigated; but based
on insights gained from the study of other populations [33]
and other colour moulting species [22,23], extant Scottish
mountain hares may have avoided increases in camouflage
mismatch via adaptive shifts in response to widespread
reductions in snow cover [34].
In this study, we assessed the potential of a wild popu-
lation of a common, seasonally colour moulting species
to adaptively track climate change. We took advantage of
the detailed historical surveys of mountain hare moult
phenology in the Scottish Highlands to examine population
responses to decreasing snow cover over the past 65 years.
First, we quantified current population mean moult phenolo-
gies and tested whether they have shifted since the 1950s.
We hypothesized that mountain hares should have shifted
moult phenologies in ways that reduce camouflage mismatch
by moulting to white winter fur later in the autumn and to
brown summer fur earlier in the spring. Second, we quantified
population-level phenotypic plasticity to examine whether
it contributes to any potential shifts in mean phenology.
Third, we quantified historical and present-day frequency of
mismatch as measures of species vulnerability to future
environmental change. We end with general conclusions on
some key considerations when predicting adaptive responses
of wild populations to climate change.
2. Methods
(a) Study areas
Historical surveys were carried out at six sites in the northeast and
central highlands of Scotland, UK, from 1951 to the end of 1961
[31,32]. We were unable to resurvey the same sites owing to
changes in land management, access restrictions to private land
and loss of mountain hares from some historic sites [35]. We,
therefore, surveyed different sites with comparable topography,
land management practices and vegetation type. The current
sites were located in the upland areas of the northeast and central
highlands of Scotland and spanned a similar elevational range as
the historical sites (table 1; electronic supplementary material,
table S1). All historic and current sites were dominated by
dwarf heath and subalpine plant communities, common veg-
etation type of the Scottish uplands and represent the habitat
type preferred by mountain hares in the geographical area [36].
(b) Field surveys
We followed the original historical field survey methods [31,32];
one surveyor walked along a predetermined route (ca 36km
Table 1. Historical and present study sites in northeast and central Scotland, average elevation at the sites in metres above sea level, years when surveys were
carried out, latitude and longitude.
region site elev. survey years lat. long.
historic surveys
Angus Glens Glen Esk high 610 19571961 56.957 2.839
Angus Glens Glen Esk low 270 19571961 56.943 2.835
Deeside/Strathdon Corndavon 450 1951, 1955, 19571959 57.068 3.234
Deeside/Strathdon Glen Muick 380 19581959 57.022 3.046
Deeside/Strathdon Punchbowl 310 19571959 56.860 2.730
Deeside/Strathdon Roar Hill 450 19581959 57.129 2.999
current surveys
Highland Findhorn high 640 2016 57.235 4.136
Highland Findhorn low 430 2016 57.206 4.102
Deeside/Strathdon Lecht 730 20152016 57.193 3.240
royalsocietypublishing.org/journal/rspb Proc. R. Soc. B 287: 20201786
2
long). Hares were detected as they were either flushed (moved)
in response to the surveyor or in reaction to other hares, or less
frequently, as the surveyor thoroughly scanned the surroundings
with binoculars. Hares are largely inactive during the day and
the majority of hares were detected as they flushed. Our experi-
ence and that reported in the literature is that hares tend not to
flush until an observer is very close to them, or unless disturbed
by other fleeing hares; therefore, the large majority of detections
were within less than 50 m of the observer [37]. For all hares
detected within 200 m of the observer which provided a clear
view that allowed coat colour to be assessed, we recorded coat
colour (described below). Surveys were repeated twice a month
(OctoberJanuary and MarchJune) for a total of 511 surveys
per season in 2016 at the two Findhorn sites and 2015 and
2016 at the Lecht site, giving eight yearseasonsite combinations
(electronic supplementary material, table S1). Surveys were
always undertaken in clear and dry conditions so as to reduce
the possible effects of weather on detecting hares. The risk of
repeat observations of the same individuals within a survey
was minimized by visually monitoring flushed individuals
as far as possible. Each survey was completed within a single
45 h session.
(c) Moult phenology
We recorded, and where possible photographed, the coat colour
for each observed hare using the moult score protocol developed
by Watson [31]. Each hare was ranked in one of five colour cat-
egories; DD (completely dark), D (mostly dark), LD (half-dark
and half-white), L (mostly white) or LL (completely white) by
the surveyor. Observations accompanied by photographs
(greater than 80%) were later verified by a single observer (elec-
tronic supplementary material, figure S1). Historical surveys at
one site used a slightly different method [32] to determine the
five colour categories (=colour was scored independently for
seven body parts and averaged), but interchangeability of the
two scoring methods was confirmed by the sites observer
(J. Flux 2015, personal communication) and by agreement with
records from similar dates and sites [31]. Finally, to reduce poten-
tial bias between observers and to simplify parameter estimation,
we reduced the initial five categories into three: white (LL, L),
moulting (LD) and brown (D, DD) for all analyses (electronic
supplementary material, figure S1).
(d) Statistical analyses
We used R (v. 3.5.2; R Core Team 2016) for all statistical analyses.
(i) Climate variables and analysis
To characterize climate in the study region, we calculated temp-
erature and snow cover variables over the past 65 years. The
mean seasonal temperature, tavg was calculated for each year
from 1950 to 2016 using gridded 5 × 5 km resolution monthly
average temperature (Met Office UKCP09) [38] at each study
site. The seasons were defined as spring (1 March31 May) and
autumn (1 September30 November) and encompassed the
main periods when hares underwent moults. Days with snow
cover (snow days) were summed for each season of each year;
for 19602011, snow days were those days when snow cover
was present at a site (snow water equivalent greater than
0 mm) based on daily gridded 5 × 5 km resolution data [39].
Because this dataset became unavailable after 2011, for 2012
2016, snow days were defined by days when snow cover was
present at a site (grid cells were greater than 50% snow covered)
based on daily 1.5 × 1.5 km resolution data [40]. We combined
the two snow datasets to span the entire period of interest and
verified the compatibility of the two datasets by comparing the
period of overlap (20002011; electronic supplementary material,
figure S2). Next, we calculated for each year and site 25th percen-
tile of snow days in the autumn and 75th percentile of snow days
in the spring, as indices for early autumn and late spring snow
days, respectively. Lastly, we calculated the number of transitions
as the number of changes between snow cover presence and
absence at each site by summing the number of times snow
days were followed by days without snow and vice versa each
year and season. The resulting number of transitions is a measure
of environmental stochasticity with snow cover repeatedly
falling and melting multiple times during each season. All snow
variables were calculated for the main snowfall periods in
Scotland; spring snowfall (1 March31 May), autumn snowfall
(1 October31 December) and autumn-to-spring snowfall period
(snow season, 1 October31 May).
Although high-resolution snow data do not exist for Great
Britain prior to 1960, we assumed that the 1950s data were com-
parable to the 1960s and used the 1960s data as a proxy to
calculate historical mismatch. We validated this assumption by
comparing number of snow days during 19511960 and 1961
1970 which were collected during the Snow Survey of Great Brit-
ain [41]. Only records from stations lying within 40 km of any of
our study sites and that recorded daily snow cover for at least 6
years during both decades (n= 7) were included in the compari-
son. We found no difference between the number of snow days
during the entire snow season (here referring to the period 1 Octo-
ber31 May) between the two decades using a Wilcoxon rank-sum
test with continuity correction ( p= 0.45, W= 1684).
Changes in mean temperature, number ofsnow days, number
of snow transitions and timing of early autumn and late spring
snow were quantified using mixed effects models. The mean sea-
sonal temperature, number of snow days, number of transitions
and the 25th or 75th percentile snow dates were used as response
variables, year as a fixed effect and site as a random effect using
the lmer function from the lme4 [42] package in R [43].
(ii) Moult phenology
We developed a hierarchical multinomial logistic regression analy-
sis within a Bayesian framework to describe moult phenology and
its phenotypic plasticity [18]. For all models, we estimated the
probability of a hare colour ybeing in colour category iat site j
on a Julian day das
Pr (y¼i)¼e
a
iþ
b
1idþsi,j
1þPi1
k¼1e
a
iþ
b
1idþsi,j
:
Coat colour was treated as a categorical variable, such that a hare
on day dwas either brown ( p
brown
), white ( p
white
) or moulting
(p
moult
) and Σ(p
1:3, j,d
) = 1. Site was coded as a random covariate
s
i,j
to reflect the hierarchical structure of the dataset and admit
repeat measures. α
i
was the intercept and β1
i
was the effect of
Julian day on the probability of being either brown, white or
moulting. Autumn and spring moults were modelled separately.
Hereafter, we refer to this model without additional covariates as
the basic model.
To compare moult phenology between the time periods, we
combined colour observations from all years and sites in one
dataset and added a fixed effect of time period β2
i
(1950s or
2010s) on the probability of being in a certain colour category
to the basic model. We used the estimated probabilities to
derive approximate dates when hares initiated and completed
the moults as initiationand completiondates during each
time period. Autumn initiation was specified as the first Julian
day when the mean p
brown
< 0.9 and completion date when the
mean p
white
> 0.9; the opposite condition was used to estimate
the spring dates (i.e. initiation
d
p
white
< 0.9 and completion
d
p
brown
> 0.9).
Next, we investigated the role of phenotypic plasticity
in moult phenology in response to ambient temperature.
royalsocietypublishing.org/journal/rspb Proc. R. Soc. B 287: 20201786
3
Because ambient temperature is thought to moderate mountain
hare moult phenology [31,32], and thereby improve a models
ability to detect differences between time periods, we con-
structed an additional set of models with temperature as an
additional covariate (tavg
j,e
). tavg
j,e
was the average seasonal
temperature at site jduring year eand was added as a fixed
effect β3
i
to the basic model containing time period β2
i
described
above. We standardized tavg to have a mean of 0 and s.d. of
1. Additionally, to explicitly test the effect of tavg, we constructed
a univariate model with a single fixed effect β3
i
(tavg
j,e
). The
resulting β3
i
coefficients were the slopes of reaction norms of
the probabilities of being brown (β3
brown
) or white (β3
white
)
on tavg.
For all models, we obtained posterior distributions of all
parameters along with their 95% credible intervals (CRI) using
Markov chain Monte Carlo implemented in JAGS (v. 4.0.1),
which we called using the R2jags package [44]. Model conver-
gence was assessed using the GelmanRubin statistic, where
values less than 1.1 indicated convergence [45]. We generated
three chains of 300 000 iterations after a burn-in of 150 000 iter-
ations and thinned by three. Parameters α
i
,β1
i
and β3
i
received
a vague prior of N(0, 0.001), while β2
i
and the standard deviation
of random effect s
i,j
received uniform priors of U(10, 10) and
U(0, 100), respectively.
(iii) Phenotypic mismatch
To examine the occurrence of mismatch between hare winter coat
colour and snow-free ground between 1951 and 2016, we calcu-
lated the number of days of mismatch at each site each year and
season. Mismatch occurred on days when hares were white and
snow was absent at each site. We defined white hares when the
mean p
white
> 0.6 based on the basic model, as this threshold
would include completely (LL) and mostly white (L) hares (elec-
tronic supplementary material, figure S1) and is consistent with
previous studies [18,19]. To test for increase in the mismatch
over the 60+ years of climate change, we ran a univariate
linear mixed model with mismatch days as the response variable,
year as a fixed effect and site as a random effect. Finally, to
explore the sensitivity of the definition of white threshold, we
repeated the analysis with an alternative threshold at mean
p
white
> 0.9.
3. Results
(a) Climate change
Temperature increased and snow cover duration decreased for
all sites and seasons over the 19502016 period, while snow sto-
chasticity did not change. Seasonal average temperature (tavg)
increased by a mean (±s.d.) of 0.17 (±0.018)°C decade
1
during
spring and 0.13 (±0.016)°C decade
1
during autumn ( p
0.001; electronic supplementary material, figure S3). This led
to increases in average seasonal temperature of 1.15°C in the
spring and 0.84°C in the autumn between 1950 and 2016.
The number of snow days decreased during both seasons by
ameanof2.79 (±0.33) days decade
1
in spring, and 1.72
(±0.30) days decade
1
in autumn ( p0.001; electronic sup-
plementary material, figure S4) and by a mean of 6.52
(±0.68) snow days decade
1
for the entire snow season ( p
0.001; figure 1). This led to an average decline of 37.14 days
of annual snow cover at our sites between 1960 and 2016.
Next, we found that the mean date of early autumn snow
occurs about 4 days later (0.069 ± 0.033 days, p= 0.038) and
the late spring snow now occurs about 7 days earlier (0.12
± 0.039, p= 0.0031) since the 1960s (electronic supplementary
material, figure S6). Finally, we found no change in stochasti-
city of snow, measured by the number of transitions between
bare ground and snow cover during the entire snow season
(β=0.018, s.e. = 0.014, p= 0.17) or autumn seasons
(β=0.0075, s.e. = 0.0078, p= 0.33). In the spring, there was a
significant decrease in the number of transitions (β=0.029,
s.e. = 0.0063, p0.001), although this effect size is small (1.63
fewer transitions between 1960 and 2016), probably owing to
the confounding effect of the decreasing number of springtime
snow days (electronic supplementary material, figures S4 and
S5).
(b) Moult phenology
We did not detect any significant shifts in spring or autumn
moult phenology between 1951 and 2016 (table 2a,b). The
60
autumn spring
site
Findhorn high
Findhorn low
Corndavon
Glen Esk high
Glen Esk low
Lecht
Glen Muick
Punchbowl
Roar Hill
40
20
no. snowdays
0
1960 1980 2000 2016 1960
year
1980 2000 2016
Figure 1. Number of snow days during the autumn- and spring snowfall season at the study sites between 1960 and 2016. Coloured lines show linear regression
slopes for each site with 95% confidence intervals depicted in grey. Solid (dashed) lines indicate sites used in current (historical) surveys. (Online version in colour.)
royalsocietypublishing.org/journal/rspb Proc. R. Soc. B 287: 20201786
4
effect of the time period covariate on the probabilities of
being brown (β2
brown
) or white (β2
white
) overlapped zero for
both seasons in models with (table 2b), or without seasonal
temperature (tavg; table 2a, figure 2). Next, mean population
moult initiation dates did not differ significantly between
moult phenology of the 1950s and 2010s in spring or
autumn as indicated by the overlapping 95% CRI (figure 2);
hares initiate autumn moults in late October (mean Julian
date = 296) and spring moults around mid-March (mean
Julian day = 78). Similarly, the estimated completion dates
have not changed between the two time periods for either
season with spring moults completing in mid-May (mean
Julian day = 135) and autumn moults in mid-December
(mean Julian day = 345; figure 2).
We found evidence for phenotypic plasticity in response to
annual variation in temperature tavg (table 2b,c). In the spring,
the effect of tavg (β3
i
) was significant, indicating that moults
were delayed during colder springs. This resulted in up to a
20 day difference in the mean population completion dates
between some springs. In the autumn, tavg had a significant
effect only on the probability of being white (β3
white
, table 2c),
with non-significant shifts towards earlier initiation and
completion of the moult during colder autumns.
(c) Phenotypic mismatch
Estimated mismatch in coat colour increased between 1950
and 2016 at all sites and seasons. The increases were steepest
over the entire snow season (1 October31 May) (β
Year
= 0.52,
p0.001) and evident when moulting seasons were con-
sidered separately (autumn β
Year
= 0.14, p0.001; spring
β
Year
= 0.18, p0.001; figure 3). Since the 1960s, from when
1.00
period
1950s
2010s
autumn initiation
24 Oct (17 Oct, 31 Oct)
21 Oct (10 Oct, 31 Oct)
Pwhite = 1
Pwhite = 0
autumn completion
11 Dec (6 Dec, 17 Dec)
12 Dec (2 Dec, 19 Dec)
period
1950s
2010s
spring initiation
18 Mar (8 Mar, 26 Mar)
19 Mar (7 Mar, 29 Mar)
spring completion
13 May (5 May, 20 May)
16 May (6 May, 29 May)
0.75
0.50
0.25
probability of white pelage
0
15 Oct 5 Nov 26 Nov
date
17 Dec 7 Jan 12 Feb 5 Mar 26 Mar 16 Apr
date
7 May 28 May
Figure 2. Similar mean mountain hare moult phenologies during 1950s and 2010s in the highlands of Scotland. Solid lines depict predicted probabilities of being
white over time based on the basic model including seasonal average temperature tavg. The shaded areas and dashed lines show 95% credible intervals (CRI) and
the perpendicular hash marks along the x-axis depict survey dates, colour coded for each time period. Photographs show mountain hares when probability of being
in white pelage is 100% (left) and 0% (right). Dates above plots indicate the mean initiation and completion dates and CRIs. (Online version in colour.)
Table 2. Absence of shifts in moult phenology from 1951 to 2016 and some phenotypic plasticity in mountain hares in the highlands of Scotland in autumn
and spring. (Mean effect sizes and 95% credible intervals (CRI) estimates for slopes for models including (a) time period only, (b) time period and seasonal
average temperature (tavg), and (c) tavg only. β2
i
indicates the effect of time period and β3
i
indicates the effect of seasonal temperature tavg on the
probability of brown (β3
brown
) and white (β3
white
). Asterisks indicate CRIs not overlapping zero.)
(a) Pr(y=i)=α
i
+β1
i
×day+β2
i
× time period
j
+s
i,j
moult season β2
brown
β2
white
autumn 0.62 (2.12, 0.89) 0.62 (0.26, 1.56)
spring 0.23 (0.83, 0.35) 0.01 (1.00, 1.00)
(b) Pr(y=i)=α
i
+β1
i
×day+β2
i
× time period
j
+β3
i
×tavg
j,e
+s
i,j
moult season β2
brown
β2
white
β3
brown
β3
white
autumn 0.34 (1.68, 1.00) 0.06 (0.68, 0.92) 0.25 (0.05, 0.54) 0.46* (0.71, 0.20)
spring 0.37 (1.56, 0.60) 0.05 (1.30, 0.93) 1.00* (0.87, 1.14) 0.78* (0.92, 0.64)
(c) Pr(y=i)=α
i
+β1
i
×day+β3
i
×tavg
j,e
+s
i,j
moult season β3
brown
β3
white
autumn 0.270 (0.018, 0.554) 0.456* (0.676, 0.288)
spring 1.005* (0.870, 1.144) 0.766* (0.906, 0.627)
royalsocietypublishing.org/journal/rspb Proc. R. Soc. B 287: 20201786
5
gridded snow data are available, the regression slopes trans-
late to 29.7 more days with white hares (mean p
white
> 60%)
against snowless background than in the 2010s, with an
additional 5.2 days when data are extrapolated to the 1950s
(figure 3b). Across all the sites, the mean number mismatch
days increased from 44.3 (s.d. = 24.8) days during 1950s and
1960s to 69.9 (s.d. = 30.1) during the 2010s ( figure 3a). The
results were similar when an alternative mismatch threshold
(=mean p
white
> 90%) was used (electronic supplementary
material, table S2).
4. Discussion
Across the northeast and central highlands of Scotland seaso-
nal temperatures have increased and the number of snow
days has declined since the 1950s; a trend seen across the
Northern Hemisphere [13,14]. Despite the directionality and
large magnitude of the observed climate shift documented
here, our results suggest that moult phenology did not track
the shortening snow seasons to prevent camouflage mismatch.
Further, temperature-mediated phenotypic plasticity in moult
phenology was detectable, but insufficient to prevent camou-
flage mismatch. Altogether, this resulted in 35 additional
days of phenotypic mismatch whereby mostly white hares
inhabited snowless backgrounds. As snow cover is expected
to decline by up to additional 50% by 2100 across Scotland
[34], mountain hares in the Scottish uplands are very likely to
experience further phenotypic mismatch in the future.
The lack of sufficient adaptive phenological responses in
mountain hares was unexpected for two main reasons.
First, phenotypic plasticity has been commonly documented
across taxa in a range of traits [46,47] and some plasticity in
moult phenology has been observed in several seasonally
colour moulting species [16,18,28]. Yet, the observed levels
of plasticity were apparently insufficient to prevent increases
in camouflage mismatcha finding consistent with field
studies of snowshoe hares over shorter time periods
[18,28,48]. Second, strong natural selection against camou-
flage mismatch has been documented in other colour
moulting species [2224] and negative population conse-
quences of mismatch were found in mountain hares in
Norway [33]. Therefore, evolutionary shifts in moult phenol-
ogies are a plausible, if not expected, response to reduced
snow cover. Given that these two components of adaptive
capacity are so widely observed, our results provide a strik-
ing contrast to evidence for adaptive shifts observed in
other systems [22,23,27]. Multiple factors may have contribu-
ted to the lack of shifts in moult phenology in mountain
hares. In the next paragraphs, we discuss the potential contri-
butions of environmental stochasticity, potentially low
genetic variance and attenuation of selection pressure against
camouflage mismatch in Scotland. We also discuss how the
increasing duration of camouflage mismatch in mountain
hares might influence these populations in the future.
Adaptive tracking of decreasing snow cover could be
slowed or stalled if temporally varying selection pressures pre-
vent the generation of stable optimal phenotypes via
phenotypic plasticity or evolutionary adaptation [49,50]. The
climate of Scotlands highlands is extremely variable and
unpredictable in time and space, subjecting mountain hares
to high environmental stochasticity. Although temperature
exerts major control over snow cover and depth in Scotland,
snowfall is often associated with frontal systems and a cold
winter does not necessarily mean a snowy one [51]. Indeed,
hares experience high variability in snow cover during each
winter, with an average of 14.2 snow cover transitions per
winter during our study period. However, the high stochasticity
in climate has not increased over the past 60 or more years (elec-
tronic supplementary material, figure S5), so environmental
site
Findhorn high
Findhorn low
Corndavon
Glen Esk high
Glen Esk low
Lecht
Glen Muick
Punchbowl
Roar Hill
19601950
120
80
40
no. minsmatch days
0
120
(a)(b)
80
40
no. mismatch days
0
1950 1970
year
1980 1990 2000 2010
2010
time period
Figure 3. Estimated number of days when white mountain hares would be found mismatched against snowless background from 1950 to 2016 in the highlands of
Scotland. The number of mismatch days is calculated over the entire snow season (1 October31 May) for each year. (a) Boxplots show the number of mismatch
days in the 1950s and 2010s. Horizontal lines within the boxes denote the medians, boxes the first and third quartiles, whiskers extend to the largest and smallest
value within 1.5 × the interquartile range and the point represents an outlier. (b) Coloured lines show linear regression slopes for each site with 95% confidence
intervals depicted in grey. Solid (dashed) lines indicate sites used in current (historical) surveys. (Online version in colour.)
royalsocietypublishing.org/journal/rspb Proc. R. Soc. B 287: 20201786
6
stochasticity seems unlikely to be a primary inhibitor of recent
adaptive responses.
For moult phenology or its plasticity to evolve by natural
selection, sufficient heritable genetic variation must exist in
the trait and population must be large enough that selection
is not overwhelmed by genetic drift [52,53]. Circannual phe-
nological traits often have a heritable basis [54,55], and the
genetic basis for winter colour per se (i.e. winter dark
versus white morphs) has been determined for some popu-
lations of snowshoe hares [56] and mountain hares [57].
The genetic basis, and response to selection, of moult
timing and rate (phenology) has not yet been described,
but is similarly likely to be affected by genetic architecture,
gene expression and the disruptive effect of genetic drift in
small populations [5860]. Genetic drift owing to small popu-
lation size may be relevant in this case because recent (i.e.
since 1990s) population reductions have been reported for
some areas in the northeast and central highlands of Scotland
[35,61]. Furthermore, although genetic variation in Scotland
populations is unknown, some evidence suggests it is lower
than in other mountain hare populations in Europe [62,63].
However, without better information on genetic variance
in Scottish hares, we cannot infer whether it might have
contributed to the lack of response in moult phenology.
We believe a primary contributing factor for the apparent
lack of phenological shifts in mountain hares in Scotland is
attenuation of selection pressure. Natural selection for cryptic
coloration is one of the strongest drivers of adaptive evol-
ution [64,65], with examples including peppered moths
(Biston betularia) in Great Britain [66], mice inhabiting light-
coloured substrates [67], and seasonal colour moults in
birds and mammals [16]. However, relaxed selection (i.e.
reduced effect of phenotypic trait on fitness) can lead to a
loss of functional traits or diminished phenotypic plasticity
[68,69]. The main adaptive advantage of the winter white
moult is predator avoidance (thermoregulatory properties
are overwhelmingly controlled by changes in hair length
and density, not hair colour; Zimova et al. [16]). Therefore,
evolutionary shifts in moult timing would require directional
selection imposed by predation.
In Scotland, mountain hares are prey for a range of
species including red fox (Vulpes vulpes), wild cat (Felis silves-
tris), otter (Lutra lutra) and golden eagle (Aquila chrysaetos)
[70,71]. However, in the northeast and central highlands of
Scotland, mountain hares are associated with heather-
dominated moorlands managed for commercial shooting of
red grouse (Lagopus lagopus scoticus) [36,72]. Predator num-
bers and diversity are severely depressed across these lands
owing to legal and illegal predator control over the last cen-
tury [61,73,74]. Thus, the relatively low predator-induced
mismatch costs would be expected to relax natural selection
against mismatch in these areas relative to regions with
more intact predator communities such as in Norway [33]
or Montana, USA [23]. Given the highly altered selection
regime on intensively managed moorlands, camouflage
mismatch might have little-to-no fitness costs for mountain
hares in our study system, now and in the recent past.
If attenuated predation risk is the main contributing factor
to what we suggest is a relatively static moult phenology, we
expect that a return of predation pressure could lead to nega-
tive population consequences. For example, if generalist
predators were to increase in response to land use or policy
changes, the accumulated duration of camouflage mismatch
could threaten hare population persistence. This latent mala-
daptationis, therefore, worth considering when assessing
the species vulnerability to climate and land use change
[35,75]. Irrespective of any potential increase in predator
numbers, we recommend management efforts that favour
evolutionary rescue (i.e. large connected populations that
harbour high genetic diversity; [27,29]) to achieve evolution-
ary resilience and long-term persistence in the face of future
biotic and abiotic changes [76]. This recommendation is
especially relevant for the mountain hare populations in the
northeast and central highlands of Scotland, where there is
evidence of local population declines [35,61] and additional
stressors related to game bird management and woodland/
forestry expansion [77,78].
Two potential limitations of our study are worth noting.
First, for analyses, we collapsed moult observations into
three categories, which may decrease resolution of initiation
and completion dates. Second, we only had 2 years of cur-
renthare moult phenology, making it difficult to eliminate
the possibility of plasticity in moult phenology that may
manifest at other temperatures. Future studies that include
additional years and sites of observations will help elucidate
Scottish mountain harescapacity to respond to a wider range
of conditions under current and future climate.
For at least 21 species across the Northern Hemisphere,
seasonal coat colour has been shaped directly by climate
[27]. The general consensus is that as decreasing snow dur-
ation continues to cause winter white animals to be found
against dark snowless backgrounds, evolutionary change
will be necessary to mitigate the negative effects of increasing
camouflage mismatch [19,23,27]. However, here we found
little evidence that moult phenology in mountain hares in
Scotland has changed despite directional climate change
over the past 60 or more years. While more study is necessary
to understand the full extent of phenotypic plasticity and
why it appears moult phenology has not shifted in response
to environmental change, we suggest that relaxed selection
for camouflage, potentially coupled with low genetic
variance, would be consistent with our findings. If true, we
expect that the fitness consequences of climate change will
ultimately depend on the strength of selection pressures
such as predation. Altogether, our findings underscore that
wildlife adaptive responses to anthropogenic stressors will
ultimately depend on both abiotic and biotic conditions.
Ethics. This study meets the terms of the ethics committee at the
University of Montana.
Data accessibility. Data are available from the Dryad Digital Repository:
https://doi.org/10.5061/dryad.cc2fqz64m [79].
Authorscontributions. M.Z. and L.S.M. designed the study; M.Z. and
S.N. collected the data; M.Z. and J.J.N. analysed the data; M.Z.,
S.T.G., S.N., J.J.N., M.S. and L.S.M. wrote the paper.
Competing interests. We declare we have no competing interests.
Funding. This work was supported by the Department of the Interior
Southeast Climate Adaptation Science Center Global Change Fellow-
ship through Cooperative Agreement no. G10AC00624 to M.Z. and
S.T.G.; North Carolina State University, University of Montana; The
Explorers Club Exploration Fund to M.Z.; the National Science Foun-
dation Division of Environmental Biology grants 1743871 and
1907022 to L.S.M. and the National Science Foundation EPSCoR
Award no. 1736249 to University of Montana. S.N. and M.S. were
supported by the Rural & Environment Science & Analytical Services
Division of the Scottish Government.
Acknowledgements. We would like to thank J. Flux and A. Watson for
providing the historical data and making this research possible. We
thank Glenn Iason and Paulo Celio Alves for assistance with
royalsocietypublishing.org/journal/rspb Proc. R. Soc. B 287: 20201786
7
identifying survey locations and discussions about this work,
Andreas Dietz for providing downscaled snow cover data, and
Denny Becks for help with field surveys. We also thank Z. Cheviron,
C. Nadeau, T.L. Morelli, A. Sirén and M. Urban for helpful comments
on earlier versions of this manuscript. Next, we thank Allargue and
other estate owners and staff for access, and especially to Andrew
Dempster, Kenny Graham and Lewis Rose for logistical support
with data collection.
References
1. IPCC. 2013 Climate change 2013: the physical
science basis. In Contribution of working group I to
the fifth assessment report of the
Intergovernmental Panel on climate change (eds
TF Stocker, D Qin, G-K Plattner, M Tignor, SK Allen,
J Boschung, A Nauels, Y Xia, V Bex, PM Midgley),
p. 1535. Cambridge, UK and New York, NY:
Cambridge University Press.
2. Bellard C, Bertelsmeier C, Leadley P, Thuiller W,
Courchamp F. 2012 Impacts of climate change on
the future of biodiversity. Ecol. Lett. 15, 365377.
(doi:10.1111/j.1461-0248.2011.01736.x)
3. Scheffers BR et al. 2016 The broad footprint of
climate change from genes to biomes to people.
Science 354, aaf7671. (doi:10.1126/science.aaf7671)
4. Kovach RP, Gharrett AJ, Tallmon DA. 2012 Genetic
change for earlier migration timing in a pink
salmon population. Proc. R. Soc. B 279, 38703878.
(doi:10.1098/rspb.2012.1158)
5. Buckley J, Butlin RK, Bridle JR. 2012 Evidence for
evolutionary change associated with the recent
range expansion of the British butterfly, Aricia
agestis, in response to climate change. Mol.
Ecol. 21, 267280. (doi:10.1111/j.1365-294X.2011.
05388.x)
6. Phillips BL, Muñoz MM, Hatcher A, Macdonald SL,
Llewelyn J, Lucy V, Moritz C. 2016 Heat hardening
in a tropical lizard: geographic variation explained
by the predictability and variance in environmental
temperatures. Funct. Ecol. 30, 11611168. (doi:10.
1111/1365-2435.12609)
7. Bonnet T, Morrissey MB, Morris A, Morris S, Clutton-
Brock TH, Pemberton JM, Kruuk LEB. 2019 The role
of selection and evolution in changing parturition
date in a red deer population. PLoS Biol. 17,
e3000493. (doi:10.1371/journal.pbio.3000493)
8. Mayor SJ et al. 2017 Increasing phenological
asynchrony between spring green-up and arrival of
migratory birds. Sci. Rep. 7, 1902. (doi:10.1038/
s41598-017-02045-z)
9. Evans AE, Forester BR, Jockusch EL, Urban MC. 2018
Salamander morph frequencies do not evolve as
predicted in response to 40 years of climate change.
Ecography (Cop.)41, 16871697. (doi:10.1111/
ecog.03588)
10. Lane JE, Kruuk LEB, Charmantier A, Murie JO,
Dobson FS. 2012 Delayed phenology and reduced
fitness associated with climate change in a wild
hibernator. Nature 489, 554557. (doi:10.1038/
nature11335)
11. Chevin L-M, Lande R, Mace GM. 2010 Adaptation,
plasticity, and extinction in a changing
environment: towards a predictive theory. PLoS Biol.
8, e1000357. (doi:10.1371/journal.pbio.1000357)
12. Urban MC et al. 2016 Improving the forecast for
biodiversity under climate change. Science 353,
aad8466. (doi:10.1126/science.aad8466)
13. Kunkel KE, Robinson DA, Champion S, Yin X, Estilow
T, Frankson RM. 2016 Trends and extremes in
Northern Hemisphere snow characteristics. Curr.
Clim. Change Rep. 2,6573. (doi:10.1007/s40641-
016-0036-8)
14. IPCC. In press. IPCC Special Report on the Ocean and
Cryosphere in a Changing Climate. (eds H-O Pörtner
et al.).
15. Williams CM, Henry HAL, Sinclair BJ. 2015 Cold
truths: how winter drives responses of terrestrial
organisms to climate change. Biol. Rev. 90,
214235. (doi:10.1111/brv.12105)
16. Zimova M, Hackländer K, Good JMJM, Melo-Ferreira
J, Alves PCPC, Mills LS. 2018 Function and
underlying mechanisms of seasonal colour moulting
in mammals and birds: what keeps them changing
in a warming world? Biol. Rev. 93, 14781498.
(doi:10.1111/brv.12405)
17. Beltran RS, Burns JM, Breed GA. 2018 Convergence
of biannual moulting strategies across birds and
mammals. Proc. R. Soc. B 285, 20180318. (doi:10.
1098/rspb.2018.0318)
18. Zimova M, Sirén APK, Nowak JJ, Bryan AM, Ivan JS,
Morelli TL, Suhrer SL, Whittington J, Mills LS. 2020
Local climate determines vulnerability to
camouflage mismatch in snowshoe hares. Glob.
Ecol. Biogeogr. 29, 503515. (doi:10.1111/geb.
13049)
19. Mills LS, Zimova M, Oyler J, Running S, Abatzoglou
JTJT, Lukacs PMPM. 2013 Camouflage mismatch in
seasonal coat color due to decreased snow duration.
Proc. Natl Acad. Sci. USA 110, 73607365. (doi:10.
1073/pnas.1222724110)
20. Sultaire SM, Pauli JN, Martin KJ, Meyer MW,
Notaro M, Zuckerberg B. 2016 Climate change
surpasses land-use change in the contracting range
boundary of a winter-adapted mammal.
Proc. R. Soc. B 283, 20153104. (doi:10.1098/rspb.
2015.3104)
21. Imperio S, Bionda R, Viterbi R, Provenzale A.
2013 Climate change and human disturbance
can lead to local extinction of Alpine rock
ptarmigan: new insight from the western Italian
Alps. PLoS ONE 8, e81598. (doi:10.1371/journal.
pone.0081598)
22. Atmeh K, Andruszkiewicz A, Zub K. 2018 Climate
change is affecting mortality of weasels due to
camouflage mismatch. Sci. Rep. 8, 7648. (doi:10.
1038/s41598-018-26057-5)
23. Zimova M, Mills LS, Nowak JJ. 2016 High fitness
costs of climate change-induced camouflage
mismatch. Ecol. Lett. 19, 299307. (doi:10.1111/
ele.12568)
24. Wilson EC, Shipley AA, Zuckerberg B, Peery MZ,
Pauli JN. 2018 An experimental translocation
identifies habitat features that buffer camouflage
mismatch in snowshoe hares. Conserv. Lett. 12,
e12614. (doi:10.1111/conl.12614)
25. Melin M, Mehtätalo L, Helle P, Ikonen K, Packalen T.
2020 Decline of the boreal willow grouse (Lagopus
lagopus) has been accelerated by more frequent
snow-free springs. Sci. Rep. 10,110. (doi:10.1038/
s41598-020-63993-7)
26. Danco JF, DeAngelis AM, Raney BK, Broccoli AJ.
2016 Effects of a warming climate on daily
snowfall events in the Northern Hemisphere.
J. Clim. 29, 62956318. (doi:10.1175/jcli-d-
15-0687.1)
27. Mills LS et al. 2018 Winter color polymorphisms
identify global hot spots for evolutionary rescue
from climate change. Science 359, 10331036.
(doi:10.1126/science.aan8097)
28. Kumar AV, Zimova M, Sparks JR, Mills LS. 2020
Snow-mediated plasticity does not prevent
camouflage mismatch. Oecologia 194, 301310.
(doi:10.1007/s00442-020-04680-2)
29. Carlson SM, Cunningham CJ, Westley PAH.
2014 Evolutionary rescue in a changing world.
Trends Ecol. Evol. 29, 521530. (doi:10.1016/j.tree.
2014.06.005)
30. Caro T. 2005 The adaptive significance of coloration
in mammals. Bioscience 55, 125136. (doi:10.1641/
0006-3568(2005)0550125:tasoci]2.0.co;2)
31. Watson A. 1963 The effect of climate on the colour
changes of mountain hares in Scotland. Proc. Zool.
Soc. Lond. 141, 823835.
32. Flux JEC. 1970 Colour change of mountain hares
(Lepus timidus scoticus) in north-east Scotland.
J. Zool. 162, 345358. (doi:10.1111/j.1469-7998.
1970.tb01270.x)
33. Pedersen S, Odden M, Pedersen HC. 2017 Climate
change induced molting mismatch? Mountain hare
abundance reduced by duration of snow cover and
predator abundance. Ecosphere 8, e01722. (doi:10.
1002/ecs2.1722)
34. Brown I. 2019 Snow cover duration and extent for
Great Britain in a changing climate: altitudinal
variations and synoptic-scale influences.
Int. J. Climatol. 39, 46114626. (doi:10.1002/
joc.6090)
35. Watson A, Wilson JD. 2018 Seven decades of
mountain hare counts show severe declines where
high-yield recreational game bird hunting is
practised. J. Appl. Ecol. 55, 26632672. (doi:10.
1111/1365-2664.13235)
royalsocietypublishing.org/journal/rspb Proc. R. Soc. B 287: 20201786
8
36. Patton V, Ewald JA, Smith AA, Newey S, Iason GR,
Thirgood SJ, Raynor R. 2010 Distribution of
mountain hares Lepus timidus in Scotland: results
from a questionnaire. Mamm. Rev. 40, 313326.
(doi:10.1111/j.1365-2907.2010.00162.x)
37. Newey S, Bell M, Enthoven S, Thirgood S. 2003 Can
distance sampling and dung plots be used to assess
the density of mountain hares Lepus timidus?Wildl.
Biol. 9, 185192. (doi:10.2981/wlb.2003.049)
38. Perry M, Hollis D. 2005 The generation of monthly
gridded datasets for a range of climatic variables
over the UK. Int. J. Climatol. A J. R. Meteorol. Soc.
25, 10411054. (doi:10.1002/joc.1161)
39. Spencer M. 2016 Reanalysis of Scottish mountain
snow conditions. Edinburgh, UK: The University of
Edinburgh.
40. Dietz AJ, Kuenzer C, Dech S. 2015 Global SnowPack:
a new set of snow cover parameters for studying
status and dynamics of the planetary snow cover
extent. Remote Sens. Lett. 6, 844853. (doi:10.
1080/2150704X.2015.1084551)
41. Spencer M, Essery R, Chambers L, Hogg S. 2014 The
historical snow survey of Great Britain: digitised
data for Scotland. Scott. Geogr. J. 130, 252265.
(doi:10.1080/14702541.2014.900184)
42. Bates D, Mächler M, Bolker B, Walker S. 2015 Fitting
linear mixed-effects models using lme4. J. Stat.
Softw. 67, 48. (doi:10.18637/jss.v067.i01)
43. R Core Team. 2018 R: a language and environment
for statistical computing. Vienna, Austria: R
Foundation for Statistical Computing.
44. Su YS, Yajima M. 2012 R2jags: a package for
running jags from R. See http://CRAN.R-project.org/
package=R2jags.
45. Gelman A, Rubin DB. 1992 Inference from iterative
simulation using multiple sequences. Stat. Sci. 7,
457511. (doi:10.1214/ss/1177011136)
46. Fox RJ, Donelson JM, Schunter C, Ravasi T, Gaitán-
Espitia JD. 2019 Beyond buying time: the role of
plasticity in phenotypic adaptation to rapid
environmental change. Phil. Trans. R. Soc. B 374,
20180174. (doi:10.1098/rstb.2018.0174)
47. Merilä J, Hendry AP. 2014 Climate change,
adaptation, and phenotypic plasticity: the problem
and the evidence. Evol. Appl. 7,114. (doi:10.1111/
eva.12137)
48. Zimova M, Mills LS, Lukacs PMPM, Mitchell MSMS.
2014 Snowshoe hares display limited phenotypic
plasticity to mismatch in seasonal camouflage.
Proc. R. Soc. B 281, 20140029. (doi:10.1098/rspb.
2014.0029)
49. Senner NR, Stager M, Cheviron ZA. 2017 Spatial and
temporal heterogeneity in climate change limits
speciesdispersal capabilities and adaptive
potential. Ecography (Cop.)41, 14281440. (doi:10.
1111/ecog.03234)
50. Nadeau CP, Urban MC, Bridle JR. 2017 Climates
past, present, and yet-to-come shape climate
change vulnerabilities. Trends Ecol. Evol. 32,
786800. (doi:10.1016/j.tree.2017.07.012)
51. Spencer M, Essery R. 2016 Scottish snow
cover dependence on the North Atlantic Oscillation
index. Hydrol. Res. 47, 619629. (doi:10.2166/nh.
2016.085)
52. Wright S. 1969 Evolution and the genetics of
populations. Chicago, IL: University of Chicago Press.
53. Hoffmann AA, Sgrò CM, Kristensen TN. 2017
Revisiting adaptive potential, population size, and
conservation. Trends Ecol. Evol. 32, 506517.
(doi:10.1016/j.tree.2017.03.012)
54. van Asch M, van Tienderen PH, Holleman LJM,
Visser ME. 2007 Predicting adaptation of phenology
in response to climate change, an insect herbivore
example. Glob. Chang. Biol. 13, 15961604. (doi:10.
1111/j.1365-2486.2007.01400.x)
55. Tarka M, Hansson B, Hasselquist D. 2015 Selection
and evolutionary potential of spring arrival
phenology in males and females of a migratory
songbird. J. Evol. Biol. 28, 10241038. (doi:10.
1111/jeb.12638)
56. Jones MR et al. 2018 Adaptive introgression
underlies polymorphic seasonal camouflage in
snowshoe hares. Science 360, 13551358. (doi:10.
1126/science.aar5273)
57. Giska I et al. 2019 Introgression drives repeated
evolution of winter coat color polymorphism in
hares. Proc. Natl Acad. Sci. USA 116, 24 150
24 156. (doi:10.1073/pnas.1910471116)
58. Jones MR, Mills LS, Jensen JD, Good JM. 2020
Convergent evolution of seasonal camouflage in
response to reduced snow cover across the
snowshoe hare range. Evolution (NY)74,
20332045. (doi:10.1111/evo.13976)
59. Ferreira MS, Alves PC, Callahan CM, Marques JP,
Mills LS, Good JM, Melo-Ferreira J. 2017 The
transcriptional landscape of seasonal coat color molt
in the snowshoe hare. Mol. Ecol. 26, 41734185.
(doi:10.1111/mec.14177)
60. Jones MR, Mills LS, Jensen JD, Good JM. 2020 The
origin and spread of locally adaptive seasonal
camouflage in snowshoe hares. Am. Nat. 196,
316332. (doi:10.1086/710022)
61. Massimino D, Harris SJ, Gillings S. 2018 Evaluating
spatiotemporal trends in terrestrial mammal
abundance using data collected during bird surveys.
Biol. Conserv. 226, 153167. (doi:10.1016/j.biocon.
2018.07.026)
62. Melo-Ferreira J, Boursot P, Randi E, Kryukov A,
Suchentrunk F, Ferrand N, Alves PC. 2007 The rise and
fall of the mountain hare during Pleistocene
glaciations: expansion and retreat with hybridization
in the Iberian Peninsula. Mol. Ecol. 16, 605618.
(doi:10.1111/j.1365-294X.2006.03166.x)
63. Zachos FE, Ben Slimen H, Hackländer K, Giacometti
M, Suchentrunk F. 2010 Regional genetic in situ
differentiation despite phylogenetic heterogeneity
in Alpine mountain hares. J. Zool. 282,4753.
(doi:10.1111/j.1469-7998.2010.00710.x)
64. Duarte RC, Flores AA V, Stevens M. 2017 Camouflage
through colour change: mechanisms, adaptive value
and ecological significance. Phil. Trans. R. Soc. B
372, 20160342. (doi:10.1098/rstb.2016.0342)
65. Wallace AR. 1879 The protective colours of animals.
Science 2, 128137.
66. Saccheri IJ, Rousset F, Watts PC, Brakefield PM, Cook
LM. 2008 Selection and gene flow on a diminishing
cline of melanic peppered moths. Proc. Natl Acad.
Sci. USA 105, 16 21216 217. (doi:10.1073/pnas.
0803785105)
67. Barrett RDH et al. 2019 Linking a mutation to
survival in wild mice. Science 363, 499504.
(doi:10.1126/science.aav3824)
68. Darwin C. 1859 The origin of species. London, UK:
John Murray.
69. Lahti DC, Johnson NA, Ajie BC, Otto SP, Hendry AP,
Blumstein DT, Coss RG, Donohue K, Foster SA. 2009
Relaxed selection in the wild. Trends Ecol. Evol. 24,
487496. (doi:10.1016/j.tree.2009.03.010)
70. Flux JEC. 1970 Life history of mountain hare (Lepus
timidus scoticus) in north-east Scotland. J. Zool.
161,75123. (doi:10.1111/j.1469-7998.1970.
tb02171.x)
71. Watson J, Leitch AF, Rae SR. 1993 The diet of
golden eagles Aquila chrysaetos in Scotland. Ibis
(Lond. 1859)135, 387393. (doi:10.1111/j.1474-
919x.1993.tb02110.x)
72. Hesford N, Baines D, Smith AA, Ewald JA. 2020
Distribution of mountain hares Lepus timidus in
Scotland in 2016/2017 and changes relative to
earlier surveys in 1995/1996 and 2006/2007. Wildl.
Biol. 2020. (doi:10.2981/wlb.00650)
73. Thirgood S, Redpath S, Newton I, Hudson P.
2000 Raptors and red grouse: conservation
conflicts and management solutions. Conserv. Biol.
14,95104. (doi:10.1046/j.1523-1739.2000.
99013.x)
74. Thompson PS, Douglas DJT, Hoccom DG, Knott J,
Roos S, Wilson JD. 2016 Environmental impacts of
high-output driven shooting of red grouse Lagopus
lagopus scotica.Ibis (Lond. 1859)158, 446452.
(doi:10.1111/ibi.12356)
75. Rehnus M, Bollmann K, Schmatz DR, Hackländer K,
Braunisch V. 2018 Alpine glacial relict species losing
out to climate change: the case of the fragmented
mountain hare population (Lepus timidus) in the
Alps. Glob. Chang. Biol. 24, 32363253. (doi:10.
1111/gcb.14087)
76. Sgrò CM, Lowe AJ, Hoffmann AA. 2011 Building
evolutionary resilience for conserving biodiversity
under climate change. Evol. Appl. 4, 326337.
(doi:10.1111/j.1752-4571.2010.00157.x)
77. Harrison A, Newey S, Gilbert L, Haydon DT, Thirgood
S. 2010 Culling wildlife hosts to control disease:
mountain hares, red grouse and louping ill virus.
J. Appl. Ecol. 47, 926930. (doi:10.1111/j.1365-
2664.2010.01834.x)
78. Mustin K, Arroyo B, Beja P, Newey S, Irivine RJ,
Kestler J, Redpath SM. 2018 Consequences of game
bird management for non-game species in Europe.
J. Appl. Ecol. 55, 22852295. (doi:10.1111/1365-
2664.13131)
79. Zimova M, Giery ST, Newey S, Nowak JJ, Spencer M,
Mills LS. 2020 Data from: Lack of phenological shift
leads to increased camouflage mismatch in
mountain hares. Dryad Digital Repository. (https://
doi.org/10.5061/dryad.cc2fqz64m)
royalsocietypublishing.org/journal/rspb Proc. R. Soc. B 287: 20201786
9
... As with other phenological mismatches, colour moulting species might be able to reduce camouflage mismatch via phenotypic plasticity in moult phenology [12]. Temperature-and snow-mediated plasticity in moult phenology has been observed in several colour moulting taxa including snowshoe hares (Lepus americanus) and mountain hares (Lepus timidus) [4,9,13,14]. Although this plasticity was adaptive, in the sense that moults tracked the temperature and snow each particular season, it was insufficient to meaningfully decrease camouflage mismatch during seasons with low snow cover [13,[15][16][17]. ...
... Temperature-and snow-mediated plasticity in moult phenology has been observed in several colour moulting taxa including snowshoe hares (Lepus americanus) and mountain hares (Lepus timidus) [4,9,13,14]. Although this plasticity was adaptive, in the sense that moults tracked the temperature and snow each particular season, it was insufficient to meaningfully decrease camouflage mismatch during seasons with low snow cover [13,[15][16][17]. ...
... Pelage colour phenology has been shown to covary with climate [15][16][17], implying that the timing and rate of seasonal pelage colour shape has been adaptively shaped by local snow duration and timing. However, this generality may be compromised in cases where small population size or relaxed selection pressure may compromise scope for evolutionary shift in a changing climate [13]. ...
Article
Species that seasonally moult from brown to white to match snowy backgrounds become conspicuous and experience increased predation risk as snow cover duration declines. Long-term adaptation to camouflage mismatch in a changing climate might occur through phenotypic plasticity in colour moult phenology and or evolutionary shifts in moult rate or timing. Also, adaptation may include evolutionary shifts towards winter brown phenotypes that forgo the winter white moult. Most studies of these processes have occurred in winter white populations, with little attention to polymorphic populations with sympatric winter brown and winter white morphs. Here, we used remote camera traps to record moult phenology and mismatch in two polymorphic populations of Arctic foxes in Sweden over 2 years. We found that the colder, more northern population moulted earlier in the autumn and later in the spring. Next, foxes moulted earlier in the autumn and later in the spring during colder and snowier years. Finally, white foxes experienced relatively low camouflage mismatch while blue foxes were mismatched against snowy backgrounds most of the autumn through the spring. Because the brown-on-white mismatch imposes no evident costs, we predict that as snow duration decreases, increasing blue morph frequencies might help facilitate species persistence.
... The geographic distribution of the nominal subspecies has gradually receded northwards in Sweden. One possible reason for this is competition from the introduced brown hare (Lepus europaeus Pallas, 1778) in the mid-19 th century 3 , but the retreat may also be driven by climate change induced camouflage mismatch or other habitat changes [4][5][6] . While the heath hare is better camouflaged in areas void of snow, it too might face local extinction due to habitat alteration and competition from brown hare, as indicated from a persistent decline in hunting harvest data 7 . ...
Article
Full-text available
We provide the first whole genome sequences from three specimens of the mountain hare subspecies the heath hare (Lepus timidus sylvaticus), along with samples from two mountain hares (Lepus timidus timidus) and two brown hares (Lepus europaeus) from Sweden. The heath hare has a unique grey winter pelage as compared to other mountain hares (white) and brown hares (mostly brown), and face regional extinction, likely due to competitive exclusion from the non-native brown hare. Whole genome resequencing from the seven hare specimens were mapped to the Lepus timidus pseudoreference genome and used for detection of 11,363,883 polymorphic nucleotide positions. The data presented here could be useful for addressing local adaptations and conservation status of mountain hares and brown hares in Sweden, including unique subspecies.
... Species are 44 commonly defined by emphasizing reproductive isolation between them, also known as the 45 biological species concept (Dobzhansky, 1937;Mayr, 1942). Although the concept is not fully Mountain hare and brown hare have followed radically different evolutionary paths, with 124 multiple selective pressures driving the evolution of the observed morphological, physiological 125 and behavioral differences (Fickel et al., 2008;Reid, 2011;Zimova et al., 2020). While some 126 of these differences are obvious, such as the color of the winter pelage, whose evolution can 127 be understood using modern genomic approaches (Ferreira et In the present study, we have performed a detailed analysis of fibroblasts isolated from a 178 sympatric population of brown hares and mountain hares, including global gene expression, 179 cell growth and migration, cell cycle control, mitochondrial mass, and oxidative metabolism. ...
Article
Full-text available
Speciation is a fundamental evolutionary process, which results in genetic differentiation of populations and manifests as discrete morphological, physiological and behavioral differences. Each species has travelled its own evolutionary trajectory, influenced by random drift and driven by various types of natural selection, making the association of genetic differences between the species with the phenotypic differences extremely complex to dissect. In the present study, we have used an in vitro model to analyze in depth the genetic and gene regulation differences between fibroblasts of two closely related mammals, the arctic/subarctic mountain hare (Lepus timidus Linnaeus) and the temperate steppe‐climate adapted brown hare (Lepus europaeus Pallas). We discovered the existence of a species‐specific expression pattern of 1,623 genes, manifesting in differences in cell growth, cell cycle control, respiration, and metabolism. Interspecific differences in the housekeeping functions of fibroblast cells suggest that speciation acts on fundamental cellular processes, even in these two interfertile species. Our results help to understand the molecular constituents of a species difference on a cellular level, which could contribute to the maintenance of the species boundary.
... Snow melt-out can be an essential parameter initiating vegetation end of dormancy for some species (Jonas et al., 2008;Petraglia et al., 2014;Körner et al., 2019;Jabis et al., 2020;Marumo et al., 2020), or animal breeding period (Bison et al., 2020). Moreover, snow cover duration is an important factor affecting species that change their coat color to white in winter, such as ptarmigan (Lagopus spp.) and mountain hare (Lepus timidus; Melin et al. (2020); Zimova et al. (2020)), or to access food resources (Espunyes et al., 2022). Finally, measuring the duration and intensity of summer heat waves, calculated from air sensors, is critical in order to assess the ability of plants and animals to cope with extreme heat events. ...
Article
Full-text available
Linking climate variability and change to the phenological response of species is particularly challenging in the context of mountainous terrain. In these environments, elevation and topography lead to a diversity of bioclimatic conditions at fine scales affecting species distribution and phenology. In order to quantify in situ climate conditions for mountain plants, the CREA (Research Center for Alpine Ecosystems) installed 82 temperature stations throughout the southwestern Alps, at different elevations and aspects. Dataloggers at each station provide local measurements of temperature at four heights (5 cm below the soil surface, at the soil surface, 30 cm above the soil surface, and 2 m above ground). Given the significant amount of effort required for station installation and maintenance, we tested whether meteorological data based on the S2M reanalysis could be used instead of station data. Comparison of the two datasets showed that some climate indices, including snow melt-out date and a heat wave index, can vary significantly according to data origin. More general indices such as daily temperature averages were more consistent across datasets, while threshold-based temperature indices showed somewhat lower agreement. Over a 12 year period, the phenological responses of four mountain tree species (ash (Fraxinus excelsior), spruce (Picea abies), hazel (Corylus avellana), birch (Betula pendula)), coal tits (Periparus ater) and common frogs (Rana temporaria) to climate variability were better explained, from both a statistical and ecological standpoint, by indices derived from field stations. Reanalysis data out-performed station data, however, for predicting larch (Larix decidua) budburst date. Overall, our study indicates that the choice of dataset for phenological monitoring ultimately depends on target bioclimatic variables and species, and also on the spatial and temporal scale of the study.
... Snowshoe hares undergo seasonal molts to match their background, between a uniform brown coat in summer and a uniform white in winter, with intermediate phases of brown and white patches during the seasonal transitions in autumn and spring. In a series of studies on phenotypic plasticity of snow-shoe hares (Lepus americanus) in response to decreased snow cover and mismatches with coat color, researchers used data collected over 3 years on snowpack and color polyphenism to illustrate the extent of seasonal mismatch and associated mortality (Mills et al., 2013;Zimova et al., 2016Zimova et al., , 2020a. A large portion of mortality is directly attributed to predation, making their predator interactions a primary selective pressure for local adaptations. ...
Article
Full-text available
Natural habitats are increasingly affected by anthropogenically driven environmental changes resulting from habitat destruction, chemical and light pollution, and climate change. Organisms inhabiting such habitats are faced with novel disturbances that can alter their modes of signaling. Coloration is one such sensory modality whose production, perception and function is being affected by human-induced disturbances. Animals that acquire pigment derivatives through diet are adversely impacted by the introduction of chemical pollutants into their environments as well as by general loss of natural habitat due to urbanization or logging leading to declines in pigment sources. Those species that do manage to produce color-based signals and displays may face disruptions to their signaling medium in the form of light pollution and turbidity. Furthermore, forest fragmentation and the resulting breaks in canopy cover can expose animals to predation due to the influx of light into previously dark environments. Global climate warming has been decreasing snow cover in arctic regions, causing birds and mammals that undergo seasonal molts to appear conspicuous against a snowless background. Ectotherms that rely on color for thermoregulation are under pressure to change their appearances. Rapid changes in habitat type through severe fire events or coral bleaching also challenge animals to match their backgrounds. Through this review, we aim to describe the wide-ranging impacts of anthropogenic environmental changes on visual ecology and suggest directions for the use of coloration both as an indicator of ecological change and as a tool for conservation.
... Without considering such ecological mismatches, restoring predators to disturbed communities can have unanticipated consequences (Reznick et al. 2008). For example, mountain hares in northeast Scotland did not adapt their molt phenology to decreasing snow cover and currently experience a nearly two-fold increase in camouflage mismatch since the 1950s (Zimova et al. 2020). However, hares persist in these areas likely because their predators, including golden eagles Aquila chrysaetos and red foxes Vulpes vulpes, are functionally extinct (Thompson et al. 2016). ...
Article
In recent decades, anthropogenic and natural disturbances have increased in rate and intensity around the world, leaving few ecosystems unaffected. As a result of the interactions among these multiple disturbances, many biological communities now occur in a degraded state as collections of fragmented ecological pieces. Restoration strategies are traditionally driven by assumptions that a community or ecosystem can be restored back to a pre-disturbance state through ecological remediation. Yet despite our best efforts, attempts to restore fragmented communities are often unsuccessful. One explanation, the humpty-dumpty effect, suggests that once a community is disassembled , it is difficult to reassemble it even in the presence of all the original pieces. This hypothesis, while potentially useful, often fails to incorporate the multitude of other critical mechanisms that affect our abilities to put fragmented communities back together. Here, we extend the original humpty-dumpty analogy to incorporate eco-evolutionary changes that can hinder successful restoration. A systematic literature review uncovered few studies that have explicitly considered how the original humpty-dumpty effect has affected restoration success in the 30 years since its inception. Using case studies, we demonstrate how the application of our extended eco-evolutionary humpty-dumpty framework may determine the success of restoration actions via ecological and evolutionary changes in fragments of communities. Lastly, given continued anthropogenic disturbances and projected climatic changes, we make five recommendations to facilitate more successful restoration efforts given our revised eco-evolutionary humpty-dumpty effects framework. These guidelines, combined with clearly defined management goals are aimed at both keeping ecological communities as intact as possible while ensuring that future ecosystem restorations might more successfully put the ecological community pieces back together.
... Due to the shorter snow-covered season, camouflage mismatch with white winter pelage on bare ground has been shown to carry fitness costs 19 , which cannot be compensated through adaptive behavioral plasticity 20 . Interestingly, it seems that natural selection can be too weak to adaptively shift the phenology of color molt in mountain hares 21 , underscoring the importance of obtaining adapted color morph alleles. 6 . ...
Article
Full-text available
Brown hares (Lepus europaeus Pallas) are able to hybridize with mountain hares (L. timidus Linnaeus) and produce fertile offspring, which results in cross-species gene flow. However, not much is known about the functional significance of this genetic introgression. Using targeted sequencing of candidate loci combined with mtDNA genotyping, we found the ancestral genetic diversity in the Finnish brown hare to be small, likely due to founder effect and range expansion, while gene flow from mountain hares constitutes an important source of functional genetic variability. Some of this variability, such as the alleles of the mountain hare thermogenin (uncoupling protein 1, UCP1), might have adaptive advantage for brown hares, whereas immunity-related MHC alleles are reciprocally exchanged and maintained via balancing selection. Our study offers a rare example where an expanding species can increase its allelic variability through hybridization with a congeneric native species, offering a route to shortcut evolutionary adaptation to the local environmental conditions.
... There are large areas of intervening agriculture and roads: a difficult migration for a species whose natal dispersal range is less than 1 km (Angerbjörn and Flux, 1995). Notwithstanding these challenges, the warming climate also reduces snow cover, thereby increasing the vulnerability of L. timidus to predators, because of the camouflage mismatch arising from its white winter pelage (Zimova et al., 2020). The increasing number of wildfires inevitably also threatens hares on the uplands (Albertson et al., 2010). ...
Article
Full-text available
The congeneric lagomorphs Lepus timidus and L. europaeus share allopatric distributions in many areas of Europe characterised by competitive exclusion and hybridisation. We investigated prospects for these species under climate change in northern England uplands. We created ensemble models predicting niche realisation for these species, influenced by abiotic and biotic factors, estimating niche overlap in geo-environmental space. The two species occupy distinctly different niches, influenced more by vegetation preferences than climatic differences. The current climate niche for L. timidus featured higher elevations with cooler temperatures and 168 km 2 range extent. Its current habitat niche scale was larger at 269 km 2 , comprised entirely of upland dwarf shrubs: heather, cotton grass, moorland grasses. By contrast, the current climate niche predicted L. europaeus occupying lowland areas with a milder climate and range extent of 252 km 2. Its current habitat niche was also greater, 401 km 2 , being mostly improved grassland. Competition was presently limited. The current niche predictions showed very little geographic overlap between the species. Niche overlap measured by Schoener Index was low: current climate niche 0.16; current habitat niche 0.07. The future climate niches for 2050 (IPCC RCP2.6), predicted L. timidus range contracting to 19 km 2 , on hilltops and L. europaeus range expanding to 765 km 2. Consequently L. timidus range would be wholly within the L. europaeus range. In many contact zones throughout Europe, L. europaeus outcompetes L. timidus; however, in the Peak District their distributions are largely distinct. Future replacement of L. timidus by L. europaeus may be engendered by dietary convergence, should a warmer climate cause a transition of upland dwarf shrub vegetation to grasses.
... Concurrently, historical resurveys have emerged as an invaluable approach to understanding ecology and evolution [9][10][11][12]. Resurveys can expose dominant ecological processes underlying variation in population abundances [13,14], community composition [15,16] and geographic range [17] and can reveal phenotypic and genetic shifts in wild populations [18][19][20][21][22]. Knowing how and why traits change can provide insights into the fate of wild populations facing rapid global change [23,24]. ...
Article
Full-text available
Understanding how genetic variation is maintained in a metapopulation is a longstanding problem in evolutionary biology. Historical resurveys of polymorphisms have offered efficient insights about evolutionary mechanisms, but are often conducted on single, large populations, neglecting the more comprehensive view afforded by considering all populations in a metapopulation. Here, we resurveyed a metapopulation of spotted salamanders ( Ambystoma maculatum ) to understand the evolutionary drivers of frequency variation in an egg mass colour polymorphism. We found that this metapopulation was demographically, phenotypically and environmentally stable over the last three decades. However, further analysis revealed evidence for two modes of evolution in this metapopulation—genetic drift and balancing selection. Although we cannot identify the balancing mechanism from these data, our findings present a clear view of contemporary evolution in colour morph frequency and demonstrate the importance of metapopulation-scale studies for capturing a broad range of evolutionary dynamics.
Article
Full-text available
Global reduction in snow cover duration is one of the most consistent and widespread climate change outcomes. Declining snow duration has severe negative consequences for diverse taxa including seasonally color molting species, which rely on snow for camouflage. However, phenotypic plasticity may facilitate adaptation to reduced snow duration. Plastic responses could occur in the color molt phenology or through behavior that minimizes coat color mismatch or its consequences. We quantified molt phenology of 200 wild snowshoe hares (Lepus americanus), and measured microhabitat choice and local snow cover. Similar to other studies, we found that hares did not show behavioral plasticity to minimize coat color mismatch via background matching; instead they preferred colder, snow free areas regardless of their coat color. Furthermore, hares did not behaviorally mitigate the negative consequences of mismatch by choosing resting sites with denser vegetation cover when mismatched. Importantly, we demonstrated plasticity in the initiation and the rate of the molt and established the direct effect of snow on molt phenology; greater snow cover was associated with whiter hares and this association was not due to whiter hares preferring snowier areas. However, despite the observed snow-mediated plasticity in molt phenology, camouflage mismatch with white hares on brown snowless ground persisted and was more frequent during early snowmelt. Thus, we find no evidence that phenotypic plasticity in snowshoe hares is sufficient to facilitate adaptive rescue to camouflage mismatch under climate change.
Article
Full-text available
Mountain hares Lepus timidus, incorporating the subspecies L. t. varronis, L. t. hibernicus and L. t. scoticus, have declined in range throughout continental Europe where they face pressures from climate-change, competition and land-management. In Scotland, the absence of a national monitoring scheme, including mandatory reporting of hunting records, means that producing robust estimates of mountain hare population trends is difficult. We repeated questionnaire surveys conducted in 1995/1996 and 2006/2007 to assess the 2016/2017 distribution and hunting records of mountain hares in Scotland and describe regional changes in their distribution over a 20-year period in relation to management for red grouse Lagopus lagopus scotica shooting. Comparisons of areas covered in all surveys indicated no net change in the area of Scotland occupied by mountain hares, but within that we found changes in range between regions and sites of differing grouse management intensity. Between 1995/1996 and 2016/2017, range contractions in southern Scotland contrasted with no changes in north-east Scotland. In north-west Scotland range expanded by 61% in areas practicing driven grouse shooting, declined by 57% in areas practicing walked-up grouse shooting and remained low and stable in areas which did not shoot grouse. A total of 33 582 mountain hares were killed in 2016/2017 representing a 71% and 48% increase from 1995/1996 and 2006/2007 respectively. However, the average kill density in 2016/2017 (12.4 3.3 hares km2) was comparable to 2006/2007 (10.8 3.0 km2) and we found no relationship between kill density and contractions in range. Despite increases in numbers of mountain hares killed over the last 20 years, it appears that range contraction may be attributed to factors other than culling, such as changes in habitat and management. Disentangling these factors should be the focus of future research.
Article
Full-text available
Climate change has influenced a range of species across the globe. Yet, to state a noted decline in the abundance of a given species as a consequence of a specific environmental change, for instance, spatially explicit long-term data are a prerequisite. This study assessed the extent to which prolonged snow-free periods in autumn and spring have contributed to the decline of the willow grouse, the only forest grouse changing into a white winter plumage. Time-series data of willow grouse numbers from summer surveys across the study area were integrated with local data on weather (snow cover), mammalian predator abundance and hunting intensity. Modelling was conducted with a hierarchical Bayesian Poisson model, acknowledging year-, area- and location-specific variability. The results show that while willow grouse numbers had decreased continuously across the study landscapes, the decrease was accelerated at the sites where, and during the years when the preceding April was the most snow-free. This indicates a mismatch between the change into a white winter plumage and the presence of snow, turning the bird into an ill-camouflaged prey. The results thus also confirm past hypotheses where local declines of the species have been attributed to prolonged snow-free periods. Across our study area, autumns and springs have become more snow-free, and the trend has been predicted to continue. Thus, in addition to conservation actions, the future of a species such as the willow grouse is also dependent on its ability to adapt to the changed environmental conditions.
Article
Full-text available
Changing environmental conditions cause changes in the distributions of phenotypic traits in natural populations. However, determining the mechanisms responsible for these changes—and, in particular, the relative contributions of phenotypic plasticity versus evolutionary responses—is difficult. To our knowledge, no study has yet reported evidence that evolutionary change underlies the most widely reported phenotypic response to climate change: the advancement of breeding times. In a wild population of red deer, average parturition date has advanced by nearly 2 weeks in 4 decades. Here, we quantify the contribution of plastic, demographic, and genetic components to this change. In particular, we quantify the role of direct phenotypic plasticity in response to increasing temperatures and the role of changes in the population structure. Importantly, we show that adaptive evolution likely played a role in the shift towards earlier parturition dates. The observed rate of evolution was consistent with a response to selection and was less likely to be due to genetic drift. Our study provides a rare example of observed rates of genetic change being consistent with theoretical predictions, although the consistency would not have been detected with a solely phenotypic analysis. It also provides, to our knowledge, the first evidence of both evolution and phenotypic plasticity contributing to advances in phenology in a changing climate.
Article
Full-text available
Snow cover is an important indicator of climate change but constraints on observational data quality can limit interpretation of spatial and temporal variability, especially in mountain areas. This issue was addressed using archived data from the Snow Survey of Great Britain to infer key climate relationships which were then used to reference larger‐scale patterns of change. Data analysis using non‐linear (logistic) regression showed average changes in yearly snow cover were strongly related to mean temperature rather than precipitation values. Inferred change shows long‐term decline in average yearly snow cover with greatest declines in some mountain areas, notably in northern England, that can be related to their position on the most temperature‐sensitive segment of the logistic curve. Further declines in snow cover were projected in the future: a central ensemble projection from HadRM3 climate model showed average yearly snow cover predominantly confined to Great Britain mountain areas by the 2050s. However, interannual variability means some years can deviate significantly from average snow cover patterns. Site‐based analysis showed this variability has distinctive geographical variations and different influences for mountains compared to adjacent valleys. Comparison of inter‐annual variability with Lamb weather type frequency and NAO index shows the influence of large‐scale airflow patterns on snow cover duration. Most notable is the role of northwesterly and northerly flows in explaining snowy years on mountains exposed to that direction, compared to influence of easterly flows at lower levels. Future changes will therefore depend on dominant annual/decadal circulation patterns in addition to long‐term declines from climate warming.
Article
Adaptation is central to population persistence in the face of environmental change, yet we seldom precisely understand the origin and spread of adaptive variation in natural populations. Snowshoe hares (Lepus americanus) along the Pacific Northwest coast have evolved brown winter camouflage through positive selection on recessive variation at the Agouti pigmentation gene introgressed from black-tailed jackrabbits (Lepus californicus). Here, we combine new and published whole-genome and exome sequences with targeted genotyping of Agouti to investigate the evolutionary history of local seasonal camouflage adaptation in the Pacific Northwest. We find evidence of significantly elevated inbreeding and mutational load in coastal winter-brown hares, consistent with a recent range expansion into temperate coastal environments that incurred indirect fitness costs. The genome-wide distribution of introgression tract lengths supports a pulse of hybridization near the end of the last glacial maximum, which may have facilitated range expansion via introgression of winter-brown camouflage variation. However, signatures of a selective sweep at Agouti indicate a much more recent spread of winter-brown camouflage. Through simulations, we show that the delay between the hybrid origin and subsequent selective sweep of the recessive winter-brown allele can be largely attributed to the limits of natural selection imposed by simple allelic dominance. We argue that while hybridization during periods of environmental change may provide a critical reservoir of adaptive variation at range edges, the probability and pace of local adaptation will strongly depend on population demography and the genetic architecture of introgressed variation.
Article
Determining how different populations adapt to similar environments is fundamental to understanding the limits of adaptation under changing environments. Snowshoe hares (Lepus americanus) typically molt into white winter coats to remain camouflaged against snow. In some warmer climates, hares have evolved brown winter camouflage – an adaptation that may spread in response to climate change. We used extensive range‐wide genomic data to 1) resolve broad patterns of population structure and gene flow and 2) investigate the factors shaping the origins and distribution of winter‐brown camouflage variation. In coastal Pacific Northwest (PNW) populations, winter‐brown camouflage is known to be determined by a recessive haplotype at the Agouti pigmentation gene. Our phylogeographic analyses revealed deep structure and limited gene flow between PNW and more northern Boreal populations, where winter‐brown camouflage is rare along the range edge. Genome sequencing of a winter‐brown snowshoe hare from Alaska shows that it lacks the winter‐brown PNW haplotype, reflecting a history of convergent phenotypic evolution. However, the PNW haplotype does occur at low frequency in a winter‐white population from Montana, consistent with the spread of a locally deleterious recessive variant that is masked from selection when rare. Simulations of this population further show that this masking effect would greatly slow the selective increase of the winter‐brown Agouti allele should it suddenly become beneficial (e.g., owing to dramatic declines in snow cover). Our findings underscore how allelic dominance can shape the geographic extent and rate of convergent adaptation in response to rapidly changing environments. This article is protected by copyright. All rights reserved
Article
Phenological mismatches, when life‐events become mistimed with optimal environmental conditions, have become increasingly common under climate change. Population‐level susceptibility to mismatches depends on how phenology and phenotypic plasticity vary across a species’ distributional range. Here, we quantify the environmental drivers of colour moult phenology, phenotypic plasticity, and the extent of phenological mismatch in seasonal camouflage to assess vulnerability to mismatch in a common North American mammal. North America. 2010–2017. Snowshoe hare (Lepus americanus). We used > 5,500 by‐catch photographs of snowshoe hares from 448 remote camera trap sites at three independent study areas. To quantify moult phenology and phenotypic plasticity, we used multinomial logistic regression models that incorporated geospatial and high‐resolution climate data. We estimated occurrence of camouflage mismatch between hares’ coat colour and the presence and absence of snow over 7 years of monitoring. Spatial and temporal variation in moult phenology depended on local climate conditions more so than on latitude. First, hares in colder, snowier areas moulted earlier in the fall and later in the spring. Next, hares exhibited phenotypic plasticity in moult phenology in response to annual variation in temperature and snow duration, especially in the spring. Finally, the occurrence of camouflage mismatch varied in space and time; white hares on dark, snowless background occurred primarily during low‐snow years in regions characterized by shallow, short‐lasting snowpack. Long‐term climate and annual variation in snow and temperature determine coat colour moult phenology in snowshoe hares. In most areas, climate change leads to shorter snow seasons, but the occurrence of camouflage mismatch varies across the species’ range. Our results underscore the population‐specific susceptibility to climate change‐induced stressors and the necessity to understand this variation to prioritize the populations most vulnerable under global environmental change.
Article
Changing from summer-brown to winter-white pelage or plumage is a crucial adaptation to seasonal snow in more than 20 mammal and bird species. Many of these species maintain nonwhite winter morphs, locally adapted to less snowy conditions, which may have evolved independently. Mountain hares ( Lepus timidus ) from Fennoscandia were introduced into the Faroe Islands in 1855. While they were initially winter-white, within ∼65 y all Faroese hares became winter-gray, a morph that occurs in the source population at low frequency. The documented population history makes this a valuable model for understanding the genetic basis and evolution of the seasonal trait polymorphism. Through whole-genome scans of differentiation and single-nucleotide polymorphism (SNP) genotyping, we associated winter coat color polymorphism to the genomic region of the pigmentation gene Agouti , previously linked to introgression-driven winter coat color variation in the snowshoe hare ( Lepus americanus ). Lower Agouti expression in the skin of winter-gray individuals during the autumn molt suggests that regulatory changes may underlie the color polymorphism. Variation in the associated genomic region shows signatures of a selective sweep in the Faroese population, suggesting that positive selection drove the fixation of the variant after the introduction. Whole-genome analyses of several hare species revealed that the winter-gray variant originated through introgression from a noncolor changing species, in keeping with the history of ancient hybridization between the species. Our findings show the recurrent role of introgression in generating winter coat color variation by repeatedly recruiting the regulatory region of Agouti to modulate seasonal coat color change.
Article
Adaptive evolution in new or changing environments can be difficult to predict because the functional connections between genotype, phenotype, and fitness are complex. Here, we make these explicit connections by combining field and laboratory experiments in wild mice. We first directly estimate natural selection on pigmentation traits and an underlying pigment locus, Agouti, by using experimental enclosures of mice on different soil colors. Next, we show how a mutation in Agouti associated with survival causes lighter coat color through changes in its protein binding properties. Together, our findings demonstrate how a sequence variant alters phenotype and then reveal the ensuing ecological consequences that drive changes in population allele frequency, thereby illuminating the process of evolution by natural selection.