hemispheric-scale patterns of climate-related shifts in ...post.queensu.ca/~pearl/planktonic/ruhland...
TRANSCRIPT
Hemispheric-scale patterns of climate-related shiftsin planktonic diatoms from North American andEuropean lakes
K AT H L E E N R U H L A N D *, A N D R E W M . PA T E R S O N w and J O H N P. S M O L *
*Paleoecological Environmental Assessment and Research Laboratory (PEARL), Department of Biology, Queen’s University,
Kingston, ON, Canada K7L 3N6, wDorset Environmental Science Centre, Ontario Ministry of the Environment,
1026 Bellwood Acres Road, Dorset, ON, Canada P0A 1E0
Abstract
A synthesis of over 200 diatom-based paleolimnological records from nonacidified/
nonenriched lakes reveals remarkably similar taxon-specific shifts across the Northern
Hemisphere since the 19th century. Our data indicate that these diatom shifts occurred in
conjunction with changes in freshwater habitat structure and quality, which, in turn, we
link to hemispheric warming trends. Significant increases in the relative abundances of
planktonic Cyclotella taxa (Po0.01) were concurrent with sharp declines in both heavily
silicified Aulacoseira taxa (Po0.01) and benthic Fragilaria taxa (Po0.01). We demonstrate
that this trend is not limited to Arctic and alpine environments, but that lakes at
temperate latitudes are now showing similar ecological changes. As expected, the onset
of biological responses to warming occurred significantly earlier (Po0.05) in climatically
sensitive Arctic regions (median age 5 AD 1870) compared with temperate regions
(median age 5 AD 1970). In a detailed paleolimnological case study, we report strong
relationships (Po0.005) between sedimentary diatom data from Whitefish Bay, Lake of
the Woods (Ontario, Canada), and long-term changes in air temperature and ice-out
records. Other potential environmental factors, such as atmospheric nitrogen deposition,
could not explain our observations. These data provide clear evidence that unparalleled
warming over the last few decades resulted in substantial increases in the length of the
ice-free period that, similar to 19th century changes in high-latitude lakes, likely
triggered a reorganization of diatom community composition. We show that many
nonacidified, nutrient-poor, freshwater ecosystems throughout the Northern Hemisphere
have crossed important climatically induced ecological thresholds. These findings are
worrisome, as the ecological changes that we report at both mid- and high-latitude sites
have occurred with increases in mean annual air temperature that are less than half of
what is projected for these regions over the next half century.
Keywords: climate, Cyclotella, ice-out, meta-analysis, Northern Hemisphere lakes, paleolimnology,
planktonic diatoms
Received 8 January 2008; revised version received 16 April 2008 and accepted 23 April 2008
Introduction
Unequivocal evidence of pronounced warming in high
latitude and high altitude regions beginning in the 19th
century is well documented (Moritz et al., 2002; ACIA,
2004). Similarly, air temperatures in temperate latitudes
of the Northern Hemisphere have increased over the
last century, with an amplification of this warming
trend over the past 30–40 years that is unprecedented
in the last �1300 years (Jansen et al., 2007).
The effects of recent temperature increases on fresh-
water ecosystems may be blurred at temperate lati-
tudes, as these regions are typically subjected to
multiple stressors that can mask or override climatic
signals (Smol, 2008). However, in remote regions where
extensive anthropogenic disturbances are diminished,
the effects of climatic fluctuations on physical, biologi-
cal and chemical processes of freshwater ecosystems are
more clearly evident (Schindler et al., 1996; Gerten &Correspondence: John P. Smol, tel. 1 1 613 533 6147, fax 1 1 613
533 6617, e-mail: [email protected]
Global Change Biology (2008) 14, 2740–2754, doi: 10.1111/j.1365-2486.2008.01670.x
r 2008 The Authors2740 Journal compilation r 2008 Blackwell Publishing Ltd
Adrian, 2002). Although the nature and magnitude of
climate-driven responses may vary among freshwater
systems, it is clear that a warmer climate will affect
important lake properties and thus aquatic biota in a
myriad of complex ways (Keller, 2007).
Analyses of spatial synchrony (i.e. temporal coher-
ence) in lake ecosystem responses to climatic changes
have rarely been examined at a regional or global scale.
Given the wide range of differences in geographic set-
tings, atmospheric circulation patterns and various feed-
back mechanisms, the temporal and spatial consistency
of biological responses to these climatic changes is not
expected. Furthermore, regional and site-specific differ-
ences in lakes (e.g. lake morphology, depth, bedrock
geology) can result in a variety of responses, even under
large-scale climatic changes. However, recent studies
have reported strong correlations between plankton dy-
namics and climate variables (North Atlantic Oscillation,
El Nino-Southern Oscillation Index) in lakes in North
America and Europe (Straile & Adrian, 2000; Weyhen-
meyer et al., 2002; Blenckner et al., 2007; Rusak et al.,
2008). Thus, despite differences among lake settings and
characteristics, regional climatic drivers may result in
significant coherence among lakes within bioregions.
In many dimictic lakes, siliceous algae, particularly
diatoms (Bacillariophyceae), are an important compo-
nent of seasonal phytoplankton maxima, the timing of
which may play a key role in the success of zooplankton
populations (e.g. Winder & Schindler, 2004a; Adrian
et al., 2006). Numerous climate-related physical proper-
ties in lakes, such as the length of the open-water season,
the timing, duration and strength of thermal stratifica-
tion, the timing and strength of the spring freshet, and
the duration of spring overturn play critical roles in algal
dynamics and community structure. These lake proper-
ties, in turn, affect water-column mixing processes, light
availability and stability, and nutrient distribution (Smol,
1988). For example, warmer temperatures (quite notably
winter temperatures) have resulted in earlier ice-out
dates, earlier onset of thermal stratification, and earlier
development of diatom blooms in both European (e.g.
Adrian et al., 2006) and North American (e.g. Winder &
Schindler, 2004b) lakes. Changes in the abundance and
composition of primary producers in lakes can directly
impact abundances of slower growing zooplankton taxa
(Adrian et al., 2006), with important implications for
cascading effects throughout aquatic foodwebs (Winder
& Schindler, 2004b).
Recent increases in the relative abundances of small
Cyclotella species, concurrent with decreases in heavily
silicified Aulacoseira species and/or small, benthic
Fragilaria taxa, were originally linked to warming-
related changes in lakes from subarctic regions of
Finland (Sorvari & Korhola, 1998; Sorvari et al., 2002)
and Canada (Ruhland et al., 2003; Ruhland & Smol, 2005).
Recently, paleolimnological studies from temperate re-
gions across Canada (e.g. Werner et al., 2005; Harris et al.,
2006) have likewise reported increases in planktonic
Cyclotella species that we found to be strikingly similar
to the above subarctic studies. However, a global synth-
esis of these trends has hitherto not been undertaken. The
emergence of this recurring pattern in taxon-specific
changes in lakes from diverse ecological settings
prompted us to examine these trends more closely.
Plankton-based studies (e.g. sediment traps, water-
column sampling, monitoring data, surface sediments)
have added valuable insights into the seasonal dynamics
of various plankton groups including species-level
diatom trends. Although generally short in duration
(typically 2 years or less), these empirical studies have
found climate-related patterns in Cyclotella–Aulacoseira–
Fragilaria diatom assemblages that may be related to the
length of the ice-free season, the timing, duration and
strength of thermal stratification, the depth of the mixed
epilimnion, the duration of the mixing period, and/or
subsurface habitat development in relation to these
factors (e.g. Fahnenstiel & Glime, 1983; Kilham et al.,
1996; Raubitschek et al., 1999; Lotter & Bigler, 2000;
Rautio et al., 2000; Ptacnik et al., 2003; Chu et al., 2005;
Forsstrom et al., 2005; Pannard et al., 2008). In general,
these studies have concluded that warming-induced
lakewater properties favor small Cyclotella species (Cyclo-
tella comensis/gordonensis, Cyclotella stelligera/pseudostelli-
gera/glomerata) over larger and heavier Aulacoseira
species and/or small benthic Fragilaria diatom species.
Longer-term monitoring data (�20–30 years) are rare,
and when they exist rarely measure algal assemblages
beyond coarse taxonomic groupings. While these studies
clearly show that climate has played an important role in
the composition of phytoplankton in freshwater lakes
over the last few decades (e.g. Findlay et al., 2001), there
is unfortunately little or no information on diatom sea-
sonal dynamics at the species level.
In this paper, we provide a meta-analysis of more
than 200 diatom-based paleolimnological studies that
explores the geographic extent, spatial synchrony, and
possible mechanisms for recent taxon-specific shifts (i.e.
Cyclotella–Aulacoseira–Fragilaria species) recorded in a
wide variety of Northern Hemisphere lakes. Specifi-
cally, we ask: (1) How many diatom-based paleolimno-
logical profiles contain significant populations of
Cyclotella species? (2) How many of these profiles record
an increase or decrease in the relative abundance of
Cyclotella species in recent sediments, and what do
these lakes have in common? (3) Do studies that record
increases in Cyclotella show concurrent decreases in
Aulacoseira and/or Fragilaria species? (4) What is the
geographic extent and timing of this increasing Cyclo-
P L A N K T O N I C D I AT O M R E S P O N S E S T O R E C E N T WA R M I N G 2741
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
tella trend? (5) What are the most plausible mechanisms
for this taxon-specific diatom shift? Furthermore,
we examined a high-resolution diatom record from
Whitefish Bay (Lake of the Woods, Ontario, Canada) in
relation to long-term climate records, lake ice records,
and atmospheric nitrogen deposition to elucidate the
primary driver of this taxon-specific diatom shift (e.g.
climate, habitat change, atmospheric contaminants). We
conclude that nonacidified/nonenriched lakes from
both circum-Arctic and temperate regions are showing
remarkably similar warming-induced taxon-specific
shifts since the 19th century. Diatom responses to
warming were most sensitive in Arctic lakes but as
temperature-related thresholds were exceeded at lower
latitudes during the mid-20th century, this ecological
reorganization is now evident throughout the Northern
Hemisphere.
Methods
Paleolimnological studies containing Cyclotella species:meta-analysis
Biological proxy data preserved in lake sediments pro-
vide a reliable means for assessing ecological response to
long-term environmental change (Smol, 2008). In parti-
cular, subfossil diatoms have been a mainstay for paleo-
limnological assessments because of their ability to
respond rapidly to environmental change, their siliceous
frustules preserve well, and their wide distribution
among a diverse array of aquatic habitats (Stoermer &
Smol, 1999). Dated, continuous lake-sediment sequences
(i.e. lake-sediment records examined at close, consecutive
intervals for the full core length) provide valuable re-
cords of the timing, magnitude, and direction of environ-
mental change based on changes in diatom assemblages
preserved in the sediment (Stoermer & Smol, 1999).
We recognize that not all lake types will support the
particular diatom taxa examined in this study (i.e. Cyclo-
tella–Fragilaria–Aulacoseira: Appendix S1) and, in addi-
tion to the genera considered here, other diatom species
may also be sensitive to climatically induced limnologi-
cal changes (e.g. see Smol et al., 2005). For example,
recent studies from both the Russian Arctic, as well as
in the Experimental Lakes Area (ELA; Ontario, Canada),
have also reported increases in the relative abundances
of planktonic diatoms, but these include genera such as
Asterionella and Tabellaria with concurrent decreases in
Aulacoseira taxa over the past few decades [Smol et al.,
2005; Solovieva et al., 2005, 2008; M. Enache (National
Academy of Sciences, Philadelphia, PA), personal com-
munication]. These shifts may also be the result of recent
warming with Asterionella and Tabellaria dominating the
water column when stratification is strongest at the end
of summer, and Aulacoseira taxa dominating in early
spring with blooms occurring under ice (Solovieva
et al., 2005). However, it was not practical for us to
examine all available diatom records that elicit a re-
sponse to climate. Rather, we have restricted our study
to paleolimnological data that specifically contain the
word ‘Cyclotella’ in published manuscripts, theses, or
consultant reports, and that contain Cyclotella in recent
sedimentary profiles (last ca. 200 years).
An important criterion for our meta-analysis was that
lakes included in this study must not be profoundly
disturbed by human activities so as to eliminate some of
the more important confounding factors, such as high
nutrient concentrations and low pH. For example, lakes
in which concentrations of total phosphorus (TP) are
420mg L�1 are more apt to experience cyanobacterial
algal blooms than more oligotrophic lakes (MOE, 1994).
Similarly, a pH level below 6 is found to be the point at
which an ecological threshold is passed whereby acid-
related aquatic community changes become measurable
(e.g. Holt et al., 2003). Therefore, we categorized lakes as
not being profoundly affected by cultural disturbances
as having concentrations of TP 20 mg L�1 or less and that
pH levels must be 6.0 or higher. These criteria also
included lakes that are currently showing a recovery
from these impacts. For example, many lakes in both
Europe and North America have undergone intensive
remediation programs initiated in the early 1970s in
response to eutrophication of freshwater systems from
the addition of phosphorus and nitrogen compounds.
As a result of these remediation efforts, many of these
lakes are starting to show some recovery over the last
decade or more. We rationalize that once these pollution
stressors have been sufficiently removed, ambient cli-
matic signals would become prevalent on the diatom
assemblages from these systems.
All manuscripts included in our analysis were tabu-
lated, irrespective of whether or not these studies
registered increases, decreases or no trend in the rela-
tive abundances of small, planktonic Cyclotella taxa
through time. Diatom-based paleolimnological studies
that were included in this meta-analysis conformed to
the following criteria. (1) All detailed diatom profiles
had an established chronology (e.g. 210Pb dating) for the
last ca. 200 years; (2) all continuous diatom records
contained at least 2% relative abundance of Cyclotella
species in at least one sedimentary interval; (3) all detailed
cores, as well as ‘top-bottom’ regional paleolimnological
analyses, included lakes with present TP measurements
below 20 mg L�1 and pH above 6 (i.e. lakes unimpacted
by anthropogenic nutrient enrichment or acid deposi-
tion). This criterion eliminated lakes with confounding
factors as described above. (4) Diatom profiles were
categorized as showing a rise in Cyclotella species if this
2742 K . R U H L A N D et al.
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
increase was greater than 5% above background relative
abundances.
We examined differences in the timing of the diatom
changes in lakes among regions [i.e. high latitude, high
altitude, low latitude (temperate)] using an ANOVA on
ranks followed by Dunn’s post hoc test to examine pair-
wise differences. Study lakes were categorized by region
based on the original authors’ classifications; Arctic and
subarctic lakes were considered to be high latitude. The
approximate timing (year) of the change was determined
from the authors’ interpretations, or by directly examin-
ing the published diatom stratigraphies and estimating
the point at which the relative abundance of Cyclotella
exceeded 5% above background levels (i.e. beyond the
estimated percentage of possible error associated with
diatom counting or chance alone; see Wolfe, 1997).
To examine trends over a relatively wide geographic
range, we compared several regional datasets that used
a technique known as the ‘top-bottom’ or ‘before and
after’ paleolimnological approach, which has been suc-
cessfully used to explore a large variety of environmen-
tal change issues (Smol, 2008). In brief, this approach
compares environmental indicators from discrete sedi-
mentary layers [top sediments: modern environment,
and bottom sediments: preindustrial environment
(before AD 1850)]. These broad-scale studies were in-
cluded in our analysis to provide additional data to
support our collection of detailed sediment core ana-
lyses, thereby adding to the geographic extent and
reproducibility of the taxon-specific trend within a
given region under examination. We acknowledge that
the depth in the core of the bottom sample varies from
lake to lake, as does the depth at which baseline (pre-
industrial) conditions would be reached. However, each
regional dataset used in this paper consists of several210Pb-dated and fully analyzed (whole core) sedimen-
tary profiles that serve to substantiate that the bottom
sediments represent preindustrial (pre-1850) diatom
assemblages (Smol, 2008) for each region. For example,
based on numerous dated cores, sediments from tem-
perate North American lakes deposited below 20–30 cm
typically represent pre-1850 lake conditions (e.g. Forrest
et al., 2002; Harris et al., 2006; Smol, 2008; as well as the
discussions in the original papers we cite).
For the ‘top-bottom’ portion of our synthesis, we have
collected the results of studies from the Northwest
Territories and Nunavut (Ruhland et al., 2003), from
New Brunswick (Harris et al., 2006), and from Ontario.
The Ontario lake-set consists of three separate studies
that we have combined into one representation of
Ontario [South-central Ontario: collected 1992 (Hall
& Smol, 1996); South-eastern Ontario: collected 1998
(Reavie et al., 2002); South-western Ontario: collected
1999 (Werner, 2003)]. To test whether the mean relative
abundances of diatom species (i.e. Cyclotella, Fragilaria,
Aulacoseira species) in the recent samples were signifi-
cantly different from the preindustrial sediment sam-
ples, we used a nonparametric Wilcoxon Signed-Rank
test for each regional lake set.
Detailed paleolimnological case study
To explore the primary mechanisms influencing the
plankton increases, we analyzed a high-resolution fossil
diatom record from Whitefish Bay, Lake of the Woods,
Canada (49.381N, 94.141W), in conjunction with long-
term climate-related instrumental and historical data.
Lakes in northwestern Ontario, such as Whitefish Bay in
the Lake of the Woods, are of particular interest as most
have remained relatively isolated from long-term influ-
ences of direct human perturbations and are often
considered ‘reference’ sites for local and regional im-
pact studies (e.g. ELA; Schindler, 1997). In addition,
climate records from northwestern Ontario have re-
ported some of the highest rates of temperature in-
creases in North America over the last few decades,
particularly during the winter (Schindler, 1997; Chiotti
& Lavender, 2008). Winter and spring snow fall has
decreased significantly in this region with significant
declines in streamflow since the 1970s (Chiotti & Laven-
der, 2008; Warren & Egginton, 2008). Although total
annual precipitation has increased, it is likely insufficient
to compensate for the rise in evaporation rates due to
substantially warmer temperatures (Schindler, 1997;
Warren & Egginton, 2008).
From a paleolimnological perspective, the exception-
ally high temporal resolution of the Whitefish Bay
sedimentary profile, the availability of long-term con-
tinuous climate data from the nearby town of Kenora,
and the existence of a rare long-term lake ice record
from this same site (Whitefish Bay) provided us with an
ideal opportunity to directly examine relationships
between diatom compositional changes and climate-
related instrumental and historical data. In addition,
to test the possibility that atmospheric nitrogen deposi-
tion may also have contributed to the recent increases in
planktonic Cyclotella taxa, we examined the relationship
between long-term nitrogen deposition data available
from Lake 239 in nearby ELA and our Whitefish Bay
diatom data. An examination of these relationships
seeks to provide insights into some of the plausible
triggers for this recent taxon-specific shift.
Kenora climate records dating back to 1899 were
attained from the Historical Adjusted Climate Database
for Canada available at the Environment Canada web-
site (http://www.cccma.ec.gc.ca/hccd/). We examined
monthly, seasonal, and annual temperature trends. The
continuous Whitefish Bay Lake ice record (starting in
P L A N K T O N I C D I AT O M R E S P O N S E S T O R E C E N T WA R M I N G 2743
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
1964) was obtained from the Ontario Ministry of Nat-
ural Resources, Kenora, Ontario, Canada. We examined
changes in the day in which ice formed on the lake (ice-
on), the day in which ice left the lake (ice-out), as well as
the number of ice-free days (ice-on day of year minus
ice-out day of same year – in Julian days). Atmospheric
inorganic nitrogen deposition data from 1970 to 2004
was collected at Lake 239 in the ELA and was measured
as total annual delivery in precipitation to the lake
surface (mg N m�2) (data provided by Department of
Fisheries and Oceans, Canada). Given the proximity of
Lake 239 to Whitefish Bay (�40 km), similar nitrogen
deposition rates are expected. Comparisons were made
between each of these measured datasets and trends in
the percent relative abundances of Cyclotella and Aula-
coseira species. Running means were applied to the
monitoring data, where appropriate, to ensure that
these data were compared to the diatom data at a
similar chronological interval.
Results
Meta-analysis: geographic extent of Cyclotella trend
Our meta-analysis included detailed diatom profiles
from 170 lakes that contained at least 2% relative
abundance of Cyclotella taxa in their sediments (Appen-
dix S2). Twenty-six of these lakes (15%) did not have
established chronologies (e.g. 210Pb dates) for the recent
sediments and thus could not be included in our
analysis. The remaining 144 diatom profiles met our
first two criteria, of which 41 lakes showed a decrease in
the relative abundance of Cyclotella species in the recent
sediments. However, closer inspection revealed that
95% of the profiles that reported decreases in Cyclotella
species were from lakes that had undergone cultural
eutrophication or acidification in their recent history
(e.g. Ek & Korsman, 2001; Lotter, 2001), and thus failed
our third criterion. Therefore, 105 lakes met our first
three criteria of which 84 lakes (80%) showed a 45%
increase in the relative abundance of Cyclotella species
since the mid-19th century (Fig. 1), 19 lakes (18%) did
not register an increase above 5% relative abundance,
and two lakes (2%) showed a decrease (Appendix S2).
Studies recording recent increases in Cyclotella species
were almost exclusively found in naturally occurring,
freshwater lakes from northern parts of North America
(i.e. Canada and the northern USA) and western Europe,
including both temperate and Arctic/alpine regions.
Notably, increases in small Cyclotella taxa were commonly
found in relatively pristine, nutrient-poor, nonacidified
lakes (e.g. lakes that have averted, or recovered from, the
effects of cultural eutrophication and acidification). An
interesting result that became evident from our meta-
analysis of studies containing diatoms from the species
Cyclotella was the identification of gaps in the geographi-
cal distribution of paleolimnological research. In some
parts of the Northern Hemisphere and in most of the
Southern Hemisphere, there was a distinct lack of paleo-
limnological records focusing on recent sediments (with
or without Cyclotella). For example, large sections of
Russia (particularly Siberia) and Northern Eurasia re-
main largely unsampled for diatom-based paleolimno-
logical analyses spanning the recent sediments
(MacDonald et al., 2004) Perhaps, not surprisingly,
much of the Southern Hemisphere did not yield paleo-
limnological studies suitable for the purposes of this
study, primarily because naturally occurring freshwater
lakes are not always abundant, human-made reservoirs
are common, and nonimpacted lakes are generally rare.
A comparison of general trends in the timing of initial
increases in Cyclotella species among the various bior-
egions (e.g. high latitude, high altitude and low lati-
tude) clearly demonstrated that high latitude sites,
although variable, recorded significantly earlier in-
creases than low latitude sites (Fig. 2a–c; ANOVA on
ranks, H 5 15.002, Po0.001, df 5 2). The median years
of change for high and low latitude sites were AD 1870
and 1970, respectively. The median age at which these
changes occurred in high altitude sites (located both in
temperate and in northern latitudes) was AD 1920. The
least amount of variability in timing was recorded in
temperate lakes, with 75% of the changes occurring
post-1940 (Fig. 2b).
Although the results from our detailed, dated cores
provided us with a large number of lakes (just under
100) that registered strikingly similar directional taxon-
specific diatom shifts, the addition of several regional
(i.e. top-bottom) diatom analyses gave us a more geo-
graphically extensive survey of the reproducibility of
this trend within a given region. Similar to the detailed
diatom profiles, top-bottom studies throughout Canada
(see Fig. 1 for locations) recorded statistically significant
(Po0.01; Wilcoxon Signed-Rank test) taxon-specific
changes in 120 out of 147 lakes (Fig. 3). Lakes from both
the Canadian subarctic and from temperate regions of
Canada, including New Brunswick and Ontario, showed
similar, clear patterns of increasing Cyclotella species (Fig.
3a, c and e). Concurrent with these planktonic increases,
small, benthic Fragilaria species, commonly found in lakes
with short open-water seasons throughout circumpolar
regions (Lotter & Bigler, 2000; Smith, 2002), recorded
higher relative abundances in the preindustrial sediments
of the subarctic lakeset (Fig. 3b), whereas heavily silicified
Aulacoseira species were more commonly found at higher
relative abundances in the preindustrial sediments of
temperate lakes (Fig. 3d and f). However, both Fragilaria
and Aulacoseira taxa were common to all latitudes.
2744 K . R U H L A N D et al.
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
Detailed paleolimnological case study
The Whitefish Bay diatom record from the Lake of the
Woods, Ontario (Fig. 4) shows a clear increase in small
planktonic Cyclotella species beginning ca. AD 1980 with
concurrent decreases in Aulacoseira subarctica (Fig. 1,
inset). This marked diatom floristic change closely
tracks an equally sharp rise in temperature recorded
at the Kenora climate station over the same time period
(Fig. 5a and b). A strong and significant positive corre-
Fig. 1 Geographical distribution of diatom-based paleolimnological studies that record an increase in the relative abundances of
planktonic Cyclotella species since the 19th century. Circles represent dated sediment cores where detailed diatom analyses have been
undertaken. Selected examples of these diatom stratigraphies are provided by the inset figures, which typify both the characteristics and
the timing of this trend for Arctic/alpine regions and for temperate lakes. Regional-scale coring sites [i.e. top-bottom or before–after
paleolimnological approach (Smol, 2008)] that record an increase in Cyclotella species in the modern sediments are represented by
squares. (A) Northwest Territories and Nunavut, Canada, n 5 40 (Ruhland et al., 2003), (B) Ontario, Canada, combined datasets, n 5 91
(Hall & Smol, 1996; Reavie et al., 2002; Werner, 2003), and (C) New Brunswick, Canada, n 5 16 (Harris et al., 2006).
P L A N K T O N I C D I AT O M R E S P O N S E S T O R E C E N T WA R M I N G 2745
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
lation was observed between Cyclotella relative abun-
dances and Kenora annual temperature data (r 5 0.73,
Po0.001), as well as a strong and significant negative
correlation between Aulacoseira abundances and annual
temperature trends (r 5�0.65, Po0.001).
Lake-ice records reveal that the ice-free period on
Whitefish Bay has lengthened by 28 days since 1964
(Fig. 6) and is, not surprisingly, strongly related to
increases in annual temperature recorded during this
period at the Kenora climate station (r 5 0.77, Po0.001).
The ice-out data show a strong inverse relationship to
increases in Cyclotella taxa (r 5�0.57, P 5 0.003) (Fig.
7a), and equally striking positive relationship to de-
creases in Aulacoseira taxa (r 5 0.66, Po0.001) (Fig. 7b).
Trends in long-term inorganic nitrogen deposition in
Lake 239 (ELA) revealed no significant relationship to
our taxon-specific diatom trends. We found that total
inorganic nitrogen deposition (TIN 5 NH4 1 NO3,
mg m�2 yr�1) at ELA was not significantly correlated to
Cyclotella (r 5 0.28, P 5 0.20) or to Aulacoseira (r 5�0.28,
P 5 0.20) relative abundance data in Whitefish Bay.
Discussion
The compilation of detailed diatom biostratigraphies
throughout the Northern Hemisphere (Fig. 1), supple-
mented with regional paleolimnological (i.e. ‘top-
bottom’) studies from various parts of Canada (Fig. 3),
shows a strikingly coherent, yet geographically asyn-
chronous large-scale pattern of ecological change. These
data indicate that extensive reorganizations of algal
community structure are evident in lakes across the
Northern Hemisphere since the mid-19th century. Spe-
cifically, abrupt increases in planktonic Cyclotella spe-
cies and concomitant decreases in heavily silicified
Aulacoseira and/or small, benthic Fragilaria species are
evident in regions where pronounced warming has
been recorded (e.g. Chapman & Walsh, 1993).
The onset of the shift to higher relative abundances of
Cyclotella taxa in Arctic lakes occurred significantly
(Po0.05) earlier (ca. AD 1870) than in lakes from tem-
perate latitudes (ca. 1970) (Fig. 2c). The higher varia-
bility in the timing of this shift among high latitude
lakes may be explained by the generally lower temporal
resolution of Arctic sediment cores, and differences in
the relative sensitivity of subarctic and high Arctic
lakes, both of which were grouped together in this
category. High altitude lakes, regardless of latitude
(Arctic–subarctic–temperate), were intermediate in
terms of the timing of this species change (ca. AD
1920). Despite the larger number of diatom profiles
collected from temperate lakes than from high latitude/
high altitude lakes for this meta-analysis, and large
differences in lake characteristics across the temperate
Northern Hemisphere, there was remarkable consis-
tency in the timing of Cyclotella increases recorded in
the biostratigraphic profiles from temperate lakes (i.e.
low variance) with 75% of these studies recording this
shift post-1940.
Cyc
lote
lla r
elat
ive
abun
danc
e (%
)
0
10
20
30
40
50
60 Whitefish Bay Emmett LakeWolfe LakeFairy Lake Indian LakeBig Rideau Lake
Year
1630
0
10
20
30
40
50Slipper LakeLake SaanajarviLake TK20Lake TK54Lake BirgervatnetLake CF-11
(a)
(b)
(c)
187
0
197
0
n = 60n = 7n = 15
High latitude
Yea
r
1750
1800
1850
1900
1950
2000
1870
19201970
1980193018801830178017301680
TemperateHigh altitude
Fig. 2 Comparison between the timing of the onset of planktonic
diatom increases showing the percentage relative abundances of
Cyclotella taxa in a selection of lakes from (a) high latitude and (b)
temperate regions. The median age of change is illustrated as a
straight line. Slipper Lake and Lake TK-20 (Ruhland, 2001); Lake
TK-54 (Paul, 2005); Lake Saanajarvi (from Sorvari et al., 2002); Lake
Birgervatnet (Jones & Birks, 2004); Lake CF-11 (Wolfe et al., 2006);
Emmett Lake (Werner et al., 2005); Wolfe Lake (Harris et al., 2006);
Fairy Lake (Quinlan et al., 2008); Indian and Big Rideau lakes
(Forrest et al., 2002). The trends portrayed in parts (a) and (b)
consist of combining small Cyclotella taxa, mainly C. stelligera,
C. pseudostelligera, C. glomerata, C. comensis, and C. gordonensis. (c)
A comparison in the timing of Cyclotella increases among different
bioregions (i.e. high latitude, high altitude, temperate) from
detailed diatom profiles collected in this meta-analysis. The earlier
timing of high latitude sites (median AD 1870) was significantly
different (Po0.05) than low latitude sites (median AD 1970):
ANOVA on rank; H 5 15.002; Po0.001; df 5 2. The timing of change
in high-altitude sites (median AD 1920) was not significantly
different from either high or low latitude lakes.
2746 K . R U H L A N D et al.
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
Preindustrial sedimentary diatoms
Mo
der
n s
edim
enta
ry d
iato
ms
Northwest Territories–Nunavut regional dataset (n=40)
% Cyclotella species0
% C
yclo
tella
spe
cies
0
10
20
30
40
50
60
70
% Fragilaria species
0
% F
ragi
laria
spe
cies
0
20
40
60
80
100
% Cyclotella species
0
% C
yclo
tella
spe
cies
0
10
20
30
40
50
60
70
% Aulacoseira species
0
% A
ulac
osei
ra s
peci
es
0
10
20
30
40
50
60New Brunswick regional dataset (n=16)
1:1
1:1
1:1 1:1
(a)
(c)
(e) (f)
(d)
(b)
% Cyclotella species
0
% C
yclo
tella
spe
cies
0
20
40
60
80
100Ontario combined regional datasets (n = 91)
% Aulacoseira species0
% A
ulac
osei
ra s
peci
es
0
20
40
60
80
1:11:1
P<0.010 P<0.012
P<0.003 P<0.011
P<0.001
P<0.001
70605040302010
10080604020 80604020
70605040302010 10080604020
605040302010
Fig. 3 The percent relative abundances of Cyclotella species, benthic Fragilaria species, and Aulacoseira species recorded in the top
sedimentary intervals (modern) vs. the bottom sedimentary intervals (pre-1850) for regional datasets from (a) Northwest Territories and
Nunavut in the Canadian subarctic, (Ruhland et al., 2003), (b) New Brunswick, Canada, (Harris et al., 2006), and (c) combined datasets
from Ontario including south-western Ontario, (Werner, 2003), south-central Ontario, (Hall & Smol, 1996), and south-eastern Ontario,
(Reavie et al., 2002). All lakes in figure had a total phosphorus (TP) measurement o20mg L�1 and/or did not record an increase in
diatom-inferred TP. Changes between the diatoms preserved in modern (top intervals) and preindustrial (bottom intervals) for each
lakeset were deemed significant using a Wilcoxon Signed-Rank test. P-values indicate significance after correcting for the False Discovery
Rate (Benjamini & Hochberg, 1995).
P L A N K T O N I C D I AT O M R E S P O N S E S T O R E C E N T WA R M I N G 2747
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
The response of aquatic organisms to climatically
induced changes is typically abrupt and pronounced
in the paleolimnological record when an ecological
threshold is exceeded. Such a regime shift has been
clear in high-latitude aquatic environments where a
sharp increase in species turnover was reported among
circumpolar lakes over the last ca. 200 years (Smol et al.,
2005). Most of the Arctic and alpine sedimentary pro-
files we present in Fig. 1 (e.g. Catalan et al., 2002; Sorvari
et al., 2002; Ruhland & Smol, 2005; Karst-Riddoch et al.,
2005) consist of diatom assemblages that have been
relatively stable for millennia with the most pro-
nounced changes starting in the 19th century. In all
these studies, a marked and unprecedented shift to a
diatom assemblage that is characterized by increases in
the relative abundances of planktonic Cyclotella species
suggests an abrupt transformation of aquatic habitat
availability and quality.
In comparison to the well-documented sensitivity of
high latitude/altitude environments to even moderate
changes in climate, freshwater ecosystems in more
southern (temperate) latitudes that experience longer
ice-free periods and growing seasons will, understand-
ably, take longer and/or require a greater increase in
temperature before passing through climate-related
ecological thresholds. This is consistent with the find-
ings from our meta-analysis where the median age of
increasing Cyclotella abundances recorded in temperate
lakes (ca. AD 1970) lagged this same taxonomic shift in
Arctic lakes by �100 years (Fig. 2a–c). The timing of
this change in temperate lakes of the Northern Hemi-
sphere is concurrent with instrumental records that
show a rise in air temperature post-1950 that is unpre-
cedented in the last millennium (Jansen et al., 2007).
The nature and timing of the recent diatom changes
recorded in our meta-analysis is consistent with warm-
ing that is associated with a shorter ice-cover period, a
longer diatom growing season, changes in the aquatic
light regime, increased nutrient cycling, and/or
changes to the thermal and mixing properties of lakes.
Changes in these lake water properties may provide
favorable habitats for small, fast-growing planktonic
Cyclotella species (Raubitschek et al., 1999; Rautio et al.,
2000; Pannard et al., 2008). C. stelligera and C. comensis
complexes, in particular, may exploit longer ice-free
periods and/or deeper, subsurface habitats where nu-
trient concentrations are somewhat elevated and where
light properties become more stabilized as thermal
stratification develops (Fahnenstiel & Glime, 1983).
For example, results from seasonal empirical studies
suggest that, after the onset of thermal stratification,
Cyclotella taxa (C. comensis, C. stelligera/pseudostelligera/
glomerata) may bloom in the metalimnion (the layer
of water below the mixed upper layer) where the
–96° –95° –94° –93°
49°
50°
Man
itoba
Ontario
Minnesota
Whitefish Bay
KenoraLake 239
0 10
km
–95°
45°
50°
55°M
a ni t
o
ba
Ontario
Minnesota
Man
itoba
–75°–80°–85°–90°
Fig. 4 Location of Whitefish Bay (Lake of the Woods, Ontario, Canada) showing the relative location of the Kenora climate station and
nearby Lake 239 in the Experimental Lakes Area (ELA).
2748 K . R U H L A N D et al.
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
temperature gradient is steepest (Fahnenstiel & Glime,
1983; Raubitschek et al., 1999), and are the dominant
taxa when the growing season is lengthened, spring
overturn period is extended, hypolimnetic waters are
warmer, and/or when water stability is at its peak
during autumn (Rautio et al., 2000). In contrast, earlier
sedimentary intervals from Arctic/alpine and temperate
profiles were dominated by diatoms with life strategies
linked to longer ice cover and/or greater turbulence. For
example, large, thickly silicified diatoms, such as Aula-
coseira taxa, require periods of resuspension through
turbulence to maintain their position in the water col-
umn (Kilham et al., 1996), and thus commonly decrease
in abundance during periods of strong stratification
(Pannard et al., 2008). During colder periods with ex-
tended ice cover (particularly in high latitude/altitude
lakes), opportunistic and pioneering diatoms, such as
small, benthic Fragilaria species, appear to be more
competitive (Lotter & Bigler, 2000), whereas planktonic
Cyclotella species fare poorly as extended ice cover
inhibits growth in the planktonic zone.
Tem
pera
ture
(°C
)
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0
10
20
30
40
(b) Aulacoseira spp.
Date (year )
19001.0
1.5
2.0
2.5
3.0
3.5
4.0
% r
elat
ive
abun
danc
e
20
30
40
50
(a) Cyclotella spp.
1980–2002 mean annual temperature
r = 0.73, P<0.001
r = –0.65, P<0.001
2000199019801970196019501940193019201910
Fig. 5 Comparison of temperature data [Kenora (Ontario,
Canada) climate station] and diatom percent relative abundance
for Whitefish Bay (Lake of the Woods, ON, Canada). (a) Mean
annual temperature vs. Cyclotella species, (b) mean annual
temperature vs. Aulacoseira species. A 7-year running mean
was applied to the 100-year temperature record to enable com-
parisons to the diatom data. The Kenora temperature data is
available from http://www.ccma.bc.ec.gc.ca/hccd.
Year1960 1990
# of
ice-
free
day
s
180
190
200
210
220
230
240
250
260
Tem
pera
ture
(°C
)
0
1
2
3
4
5
6
# of ice-free daysAnnual temperature
y = 0.6556x + 203.64
r = 0.77
19801970 20102000
Fig. 6 Change in the number of ice-free days since 1964 from
the lake ice record from Whitefish Bay (Lake of the Woods,
Ontario, Canada). The number of ice-free days has increased
by � 28 days over the last four decades and is highly correlated
to the warming trend recorded at the Kenora (Ontario) climate
station for this same time period (Po0.001).
0
10
20
30
40
Ice-
out d
ay o
f yea
r
110
115
120
125
130
135
(b) Aulacoseira spp.
Date (year )
1965
% r
elat
ive
abun
danc
e
20
30
40
50
110
115
120
125
130
135
(a) Cyclotella spp.
r = –0.57, P = 0.003
r = 0.66, P<0.001
20052000199519901985198019751970
Fig. 7 Comparison of historical ice-out records and diatom
percent relative abundance for Whitefish Bay (Lake of the
Woods, Ontario, Canada). (a) Day of ice-out vs. Cyclotella species,
and (b) day of ice-out vs. Aulacoseira species. A 3-year running
mean was applied to the 35-year ice-out record to enable
comparisons to the diatom data. The Whitefish Bay ice-out data
was provided by the Ministry of Natural Resources, Kenora,
Ontario, Canada.
P L A N K T O N I C D I AT O M R E S P O N S E S T O R E C E N T WA R M I N G 2749
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
In general, the results of our meta-analysis indicate
that increases in small Cyclotella species were commonly
found in relatively pristine, nutrient-poor, nonacidified
lakes. This is consistent with previous studies that have
found Cyclotella species to be notably absent from lakes
that were directly impacted by anthropogenic activity,
such as acids from industrial emissions (Battarbee et al.,
1999), nutrients from agricultural and sewage runoff
(Stoermer et al., 1985), metal toxicity from industrial
effluents (Ruggiu et al., 1998), and the overdevelopment
of shorelines (Little et al., 2000). Understandably, cli-
mate-related changes in temperate regions may be
masked by direct anthropogenic stressors acting simul-
taneously on freshwater ecosystems. Therefore, in tem-
perate regions, reliable test sites for this climate-related
threshold response would logically be lakes that have
been minimally impacted by direct human disturbance
and/or other major stressors (e.g. eutrophication, acid
deposition) (Schindler, 1997). Our detailed case study
location, Whitefish Bay in the Lake of the Woods
(Fig. 4), provides such a setting.
The sharp increase in the relative abundance of
Cyclotella taxa and concurrent decrease in Aulacoseira
taxa recorded in our detailed analysis of the Whitefish
Bay sediment core (inset Fig. 1) are very similar in
nature (although asynchronous, as expected) to diatom
changes reported in both temperate (e.g. Fritz et al.,
1993; Alefs & Muller, 1999; Marchetto et al., 2004; Harris
et al., 2006) and circumpolar and alpine (e.g. Catalan
et al., 2002; Sorvari et al., 2002; Ruhland et al., 2003)
regions of the Northern Hemisphere. The striking re-
lationships observed between the Whitefish Bay diatom
data and the Kenora temperature record (both visually
and statistically) provide strong evidence that warming
trends over the past few decades have played an
important role in observed shifts in diatom community
structure (Fig. 5a and b).
A decrease in the duration of ice cover with warmer
temperatures, and associated changes in lakewater
properties, have often been surmised as triggering the
recent increases in planktonic Cyclotella species in sub-
arctic/alpine lakes (e.g. Lotter & Bigler, 2000; Catalan
et al., 2002; Sorvari et al., 2002; Ruhland et al., 2003;
Forsstrom et al., 2005; Smol et al., 2005). It follows that
long-term ice cover records could provide an ideal
means of verifying the relationship between recent
warming and increases in planktonic diatoms. For ex-
ample, a rare, long-term ice cover record available from
a river in Finland was used to indirectly corroborate
recent changes in diatom assemblage composition re-
corded in several neighboring lakes (Sorvari et al., 2002).
However, testing direct linkages between lake ice-cover
duration and these taxon-specific changes (i.e. changes
in the relative abundances of Cyclotella and Aulacoseira/
Fragilaria taxa) has not, as of yet, been undertaken,
understandably because continuous ice-out data for
the same lakes where diatoms have been examined
are extremely rare, particularly in remote Arctic and
alpine regions.
An increase in the ice-free period by almost 30 days
over the past ca. 40 years on Whitefish Bay is substantial
and has undoubtedly played a critical role in determin-
ing the structure and distribution of plankton commu-
nities throughout the water column. For example, the
strong and highly significant correlations between
Whitefish Bay ice-out and diatom trends (Fig. 7a and
b) suggest that recent increases in planktonic Cyclotella
taxa (decreases in Aulacoseira taxa) are predominantly
triggered by the lengthening of the ice-free period and
related changes in lakewater properties. These White-
fish Bay data corroborate the premise that abrupt in-
creases in Cyclotella species, concurrent with decreases
in Aulacoseira and Fragilaria species reported through-
out the Northern Hemisphere, can be primarily ex-
plained by climatically driven changes in lakewater
properties. In particular, these taxon-specific diatom
shifts were likely triggered by a warmer climate that
commenced in the mid-1800s, and escalated to unpre-
cedented highs over the last half century.
Other potential stressors and synergistic effects
The shift to planktonic diatom species, which we show
is correlated with recent warming in temperate lakes,
partly overlaps with a period of accelerated atmo-
spheric deposition of anthropogenically derived con-
taminants and nutrients. Increased deposition of
inorganic nitrogen, for example, has been cited as a
possible trigger for these abrupt compositional turn-
overs (e.g. Stoermer et al., 1985; Fritz et al., 1993; Wolfe
et al., 2003). However, we found no significant relation-
ship between inorganic nitrogen deposition from near-
by ELA and our diatom trends from Whitefish Bay. In
fact, Cyclotella relative abundance and ice-out data from
Whitefish Bay over this same time frame (i.e. 1974–2004)
were significantly and negatively correlated (r 5�0.45,
P 5 0.03), adding further support that climate-related
factors are primarily responsible for recent taxon-
specific diatom shifts. The relationships found in our
case study are consistent with findings in nearby
regions. For example, recent increases in the relative
abundance of Cyclotella species have been reported in
lakes in south-central Ontario since the early 1980s
(Clerk et al., 2000; Quinlan et al., 2008) where significant
increases in the duration of the ice-free period have
been recorded in these lakes since the mid-1970s (Futter,
2003). However, no statistically significant trend has
2750 K . R U H L A N D et al.
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
been noted in nitrate or ammonium deposition over the
past three decades (Schindler et al., 2006).
The lack of a strong relationship between recent
diatom community reorganization and nitrogen deposi-
tion in our case study is also consistent with recent
paleolimnological findings in the Arctic (Smol &
Douglas, 2007; Keatley et al., 2008) where the effects of
climatic change are first experienced. Aerially trans-
ported nutrients and contaminants would reach Arctic
and alpine lakes decades after these pollutants would
reach temperate lakes; however, the changes in diatom
species occur decades earlier in the former. Therefore,
the timing of these taxon-specific diatom shifts predat-
ing the onset of anthropogenic pollutants in Arctic lakes
underscore that aerially transported nutrients and/or
contaminants cannot explain the marked hemispheric-
scale diatom shifts we describe.
Although we have found no support to link increases
in Cyclotella species to nitrogen deposition, it is pos-
sible that a rise in inorganic nitrogen deposition may
be linked to recent climatic changes. For example,
in regions proximal to nitrogen sources, changes in
precipitation could be important for the delivery of
atmospheric inorganic nitrogen to nearby lakes (Macdo-
nald, 2005). A comparison of Kenora precipitation data
with the ELA TIN data resulted in an nonsignificant
positive correlation (r 5 0.33, P 5 0.06). Although not
conclusive, the possibility of a synergistic relationship
exists. While it is clear that diatoms respond to a multi-
tude of complex factors, it is equally clear that the nature,
magnitude, and timing of these taxon-specific diatom
shifts are best explained by recent warming and asso-
ciated limnological changes.
Conclusions
The diatom data included in our meta-analysis of over
200 lakes provide a spatially coherent picture that
climate-driven, taxon-specific changes are now evident
across vast regions of the Northern Hemisphere repre-
senting a wide spectrum of lake ecosystems. Although
the degree of ecological change may vary between lakes
(as would be expected, based on morphometry and
other limnological variables), the similarity in this dia-
tom trend across a regional scale is striking. Freshwater
ecosystems in the Northern Hemisphere have crossed
ecological thresholds with changing climate that
were initiated in the 19th century in Arctic and alpine
regions, but typically only occurred in the mid-20th
century in lakes from mid-latitude regions of North
America and western Europe. The recurring and wide-
spread trend of recent increases in the relative abun-
dances of planktonic Cyclotella species is strongly
correlated to recent increases in temperature and sub-
stantially longer ice-free periods. Synergistic effects
between climate warming and other human-induced
stressors also occur (Wolfe et al., 2006). However, based
on the timing, magnitude, and nature of the diatom
changes, we conclude that climatically induced change
in lake-ice cover (and associated limnological changes)
was the primary explanatory metric for hemispheric-
scale increases in planktonic Cyclotella species over the
last ca. 200 years. Our findings are particularly trou-
bling given that the ecological changes we report are
widespread, and have occurred with increases in mean
annual temperatures that are substantially lower than
levels projected by climate models for both high- and
mid-latitude regions of the Northern Hemisphere. If
this rate and magnitude of temperature increase con-
tinues, it is likely that new ecological thresholds will be
crossed, many of which may be unexpected.
Acknowledgements
We thank R. Hall, A. Harris, E. Reavie, and P. Werner for kindlyallowing the use of their data for producing Fig. 3, M. Stainton(Department of Fisheries and Oceans, Canada) for providing thesediments and dating the Whitefish Bay core, the Lake of theWoods District Property Owners Association for supporting fieldwork for the Whitefish Bay core, and M. Turner (Department ofFisheries and Oceans, Canada) for providing nitrogen depositiondata from ELA. We thank two anonymous reviewers for theirthoughtful and constructive comments that helped to strengthenand clarify this paper. This work was supported by the NaturalSciences and Engineering Research Council of Canada (to J. P.Smol), and the Ministry of the Environment, Ontario.
References
ACIA (2004) Impacts of a Warming Arctic: Arctic Climate Impact
Assessment. Cambridge University Press, Cambridge, UK.
Adrian R, Wilhelm S, Gerten D (2006) Life-history traits of lake
plankton species may govern their phenological response to
climate warming. Global Change Biology, 12, 652–661.
Alefs J, Muller J (1999) Differences in the eutrophication dy-
namics of Ammersee and Starnberger See (Southern Ger-
many), reflected by the diatom succession in varve-dated
sediments. Journal of Paleolimnology, 21, 395–407.
Battarbee R-W, Charles D-F, Dixit S-S, Renberg I (1999) Diatom as
indicators of surface water acidity. In: The Diatoms: Application
for the Environmental and Earth Sciences (eds Stoermer EF, Smol
JP), pp. 85–127. Cambridge University Press, Cambridge, UK.
Benjamini Y, Hochberg Y (1995) Controlling the false discovery
rate: a practical and powerful approach to multiple testing.
Journal of the Royal Statistical Society. Series B (Methodological),
57, 289–300.
Blenckner T, Adrian R, Livingstone DM et al. (2007) Large-scale
climatic signatures in lakes across Europe: a meta-analysis.
Global Change Biology, 13, 1314–1326.
Catalan J, Pla S, Rieradevall M et al. (2002) Lake Redo ecosystem
response to an increasing warming in the Pyrenees during the
twentieth century. Journal of Paleolimnology, 28, 129–145.
P L A N K T O N I C D I AT O M R E S P O N S E S T O R E C E N T WA R M I N G 2751
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
Chapman WL, Walsh JE (1993) Recent variations of sea ice and
air temperature in high latitudes. Bulletin American Meteorolo-
gical Society, 74, 33–47.
Chiotti Q, Lavender B (2008) Ontario. In: From Impacts to Adapta-
tion: Canada in a Changing Climate 2007 (eds Lemmen DS,
Warren FJ, Lacroix J, Bush E), pp. 227–274. Government of
Canada, Ottawa, ON.
Chu G, Liu J, Schettler G et al. (2005) Sediment fluxes and varve
formation in Sihailongwan, a maar lake from northeastern
China. Journal of Paleolimnology, 34, 311–324.
Clerk S, Hall R, Quinlan R, Smol JP (2000) Quantitative inferences
of past hypolimnetic anoxia and nutrient levels from a Canadian
Precambrian Shield lake. Journal of Paleolimnology, 23, 319–336.
Ek AS, Korsman T (2001) A paleolimnological assessment of the
effects of post-1970 reductions of sulfur deposition in Sweden.
Canadian Journal of Fisheries and Aquatic Sciences, 58, 1692–1700.
Fahnenstiel GL, Glime JM (1983) Subsurface chlorophyll max-
imum and associated Cyclotella pulse in Lake Superior. Inter-
nationale Revue der Gesamten Hydrobiology und Hydrographie, 68,
605–618.
Findlay DL, Kasian SEM, Stainton MP, Beaty K, Lyng M (2001)
Climatic influences on algal populations of boreal forest lakes
in the Experimental Lakes Area. Limnology and Oceanography,
46, 1784–1793.
Forrest F, Reavie ED, Smol JP (2002) Comparing limnological
changes associated with 19th century canal construction and
other catchment disturbances in four lakes within the Rideau
Canal system, Ontario, Canada. Journal of Limnology, 61,
183–197.
Forsstrom L, Sorvari S, Korhola A, Rautio M (2005) Seasonality of
phytoplankton in subarctic Lake Saanajarvi in NW Finnish
Lapland. Polar Biology, 28, 846–861.
Fritz SC, Kingston JC, Engstrom DR (1993) Quantitative trophic
reconstruction from sedimentary diatom assemblages: a cau-
tionary tale. Freshwater Biology, 30, 1–23.
Futter MN (2003) Patterns and trends in southern Ontario Lake
ice phenology. Environmental Monitoring and Assessment, 88,
431–444.
Gerten D, Adrian R (2002) Effects of climate, North Atlantic
Oscillation, and El Nino-Southern Oscillation on thermal con-
ditions and plankton dynamics in Northern Hemisphere lakes.
The Scientific World, 2, 586–606.
Hall RI, Smol JP (1996) Paleolimnological assessment of long-
term water quality changes in south-central Ontario lakes
affected by cottage development and acidification. Canadian
Journal of Fisheries and Aquatic Sciences, 53, 1–17.
Harris MA, Cumming BF, Smol JP (2006) Assessment of recent
environmental changes in New Brunswick (Canada) lakes
based on paleolimnological shifts in diatom species assem-
blages. Canadian Journal of Botany, 84, 151–163.
Holt CA, Yan ND, Somers KM (2003) pH 6 as the threshold to use
in critical load modeling for zooplankton community change
with acidification in lakes of south-central Ontario: accounting
for morphometry and geography. Canadian Journal of Fisheries
and Aquatic Sciences, 60, 151–158.
Jansen E, Overpeck J, Briffa KR et al. (2007) Paleoclimate. In:
Climate Change 2007: The Physical Science Basis. Contribution of
Working Group I to the Fourth Assessment Report of the Intergovern-
mental Panel on Climate Change (eds Solomon S, Qin D, Manning
M, Chen Z, Marquis M, Averyt KB, Tignor M, Miller HL), pp.
433–497. Cambridge University Press, Cambridge, UK.
Jones VJ, Birks HJB (2004) Lake-sediment records of recent
environmental change on Svalbard: results of diatom analysis.
Journal of Paleolimnology, 31, 445–466.
Karst-Riddoch TL, Pisaric MFJ, Youngblut DK, Smol JP (2005)
Postglacial record of diatom assemblage changes related to
climate in an alpine lake in the northern Rocky Mountains,
Canada. Canadian Journal of Botany, 83, 968–982.
Keatley BE, Douglas MSV, Smol JP (2008) Prolonged ice cover
dampens diatom community responses to recent climatic
change in high Arctic lakes. Arctic, Antarctic, and Alpine
Research, 40, 364–372.
Keller W (2007) Implications of climate warming for Boreal
Shield lakes: a review and synthesis. Environmental Reviews,
15, 99–112.
Kilham SS, Theriot EC, Fritz SC (1996) Linking planktonic
diatoms and climate change in the large lakes of the Yellow-
stone ecosystem using resource theory. Limnology and Oceano-
graphy, 41, 1052–1062.
Little J, Hall R, Quinlan R, Smol JP (2000) Past trophic status and
hypolimnetic anoxia during eutrophication and remediation
of Gravenhurst Bay, Ontario: comparison of diatoms, chirono-
mids, and historical records. Canadian Journal of Fisheries and
Aquatic Sciences, 57, 333–341.
Lotter AF (2001) The palaeolimnology of Soppensee (Central
Switzerland), as evidenced by diatom, pollen, and fossil-pig-
ment analyses. Journal of Paleolimnology, 25, 65–79.
Lotter AF, Bigler C (2000) Do diatoms in the Swiss Alps reflect
the length of ice-cover? Aquatic Sciences, 62, 125–141.
MacDonald GM, Edwards TWD, Gervais B et al. (2004) Paleo-
limnological research from northern Russian Eurasia. In: De-
velopments in Paleoenvironmental Research (Vol. 8): Long-term
Environmental Change in Arctic and Antarctic Lakes (eds Pienitz
R, Douglas MSV, Smol JP), pp. 349–380. Springer, Dordrecht,
the Netherlands.
Macdonald RW (2005) Climate change, risks and contaminants: a
perspective from studying the Arctic. Human and Ecological
Risk Assessment, 11, 1099–1104.
Marchetto A, Lami A, Musazzi S, Massaferro J, Langone L,
Guilizzoni P (2004) Lake Maggiore (N. Italy) trophic history:
fossil diatom, plant pigments, and chironomids, and compar-
ison with long-term limnological data. Quaternary Interna-
tional, 113, 97–110.
MOE – Ministry of the Environment (1994) Water management
policies guidelines: provincial water quality objectives of the
Ministry of Environment and Energy. Queen’s Printer for
Ontario, 67 pp.
Moritz RE, Bitz CM, Steig EJ (2002) Dynamics of recent climate
change in the Arctic. Science, 297, 1497–1502.
Pannard A, Bormans M, Lagadeuc Y (2008) Phytoplankton species
turnover controlled by physical forcing at different time scales.
Canadian Journal of Fisheries and Aquatic Sciences, 65, 47–60.
Paul A (2005) A paleolimnological assessment of climatic and envir-
onmental changes in a subarctic treeline lake. Unpublished Hon-
our’s thesis, Queen’s University, Kingston, Ontario, Canada,
40 pp.
2752 K . R U H L A N D et al.
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
Ptacnik R, Diehl S, Berger S (2003) Performance of sinking and
nonsinking phytoplankton taxa in a gradient of mixing depths.
Limnology and Oceanography, 48, 1903–1912.
Quinlan R, Hall RI, Paterson AM, Cumming BF, Smol JP (2008)
Long-term assessments of ecological effects of anthropogenic
stressors on aquatic ecosystems from paleoecological analyses:
challenges to traditional perspectives of lake management.
Canadian Journal of Fisheries and Aquatic Sciences, 65, 933–944.
Raubitschek S, Lucke A, Schleser GH (1999) Sedimentation
patterns of diatoms in Lake Holzmaar, Germany – (on the
transfer of climate signals to biogenic silica oxygen isotope
proxies). Journal of Paleolimnology, 21, 437–448.
Rautio M, Sorvari S, Korhola A (2000) Diatom and crustacean
zooplankton communities, their seasonal variability and re-
presentation in the sediments of subarctic Lake Saanajarvi.
Journal of Limnology, 59, 81–96.
Reavie ED, Smol JP, Dillon PJ (2002) Inferring long-term nutrient
changes in southeastern Ontario lakes: comparing paleolim-
nological and mass-balance models. Hydrobiologia, 48, 61–74.
Ruggiu D, Luglie A, Cattaneo A, Panzani P (1998) Paleoecologi-
cal evidence for diatom response to metal pollution in Lake
Orta (N. Italy). Journal of Paleolimnology, 20, 333–345.
Ruhland K, Priesnitz A, Smol JP (2003) Paleolimnological evi-
dence from diatoms for recent environmental changes in 50
lakes across the Canadian Arctic treeline. Arctic, Antarctic, and
Alpine Research, 35, 110–123.
Ruhland K, Smol JP (2005) Diatom shifts as evidence for recent
Subarctic warming in a remote tundra lake, NWT, Canada.
Palaeogeography Palaeoclimatology Palaeoecology, 226, 1–16.
Ruhland KM (2001) Diatom assemblage shifts relative to changes in
environmental and climatic conditions in the circumpolar treeline
regions of the Canadian and Siberian Arctic. PhD thesis, Queen’s
University, Kingston, Ontario, Canada, 265 pp.
Rusak JA, Yan ND, Somers KM (2008) Regional climatic drivers
of synchronous zooplankton dynamics in Dorset lakes. Cana-
dian Journal of Fisheries and Aquatic Sciences, 65, 878–889.
Schindler DE, Dillon PJ, Schreier H (2006) A review of anthro-
pogenic sources of nitrogen and their effects on Canadian
aquatic ecosystems. Biogeochemistry, 79, 25–44.
Schindler DW (1997) Widespread effects of climatic warming on
freshwater ecosystems in North America. Hydrological Pro-
cesses, 11, 1043–1067.
Schindler DW, Bayley SE, Parker BR et al. (1996) The effects of
climatic warming on the properties of boreal lakes and streams
at the Experimental Lakes Area, northwestern Ontario. Lim-
nology and Oceanography, 41, 1004–1017.
Smith IR (2002) Diatom-based Holocene paleoenvironmental
records from continental sites on northeastern Ellesmere Is-
land, high Arctic, Canada. Journal of Paleolimnology, 27, 9–28.
Smol JP (1988) Paleoclimate proxy data from freshwater arctic
diatoms. Verhandlungen der Internationalen Vereinigung von
Limnologen, 23, 837–844.
Smol JP (2008) Pollution of Lakes and Rivers: A Paleoenvironmental
Perspective, 2nd edn. Blackwell Publishing, Oxford.
Smol JP, Douglas MSV (2007) From controversy to consensus:
making the case for recent climate change in the Arctic
using lake sediments. Frontiers in Ecology and the Environment,
5, 466–474.
Smol JP, Wolfe AP, Birks HJB et al. (2005) Climate-driven regime
shifts in the biological communities of arctic lakes. Proceedings
of the National Academy of Sciences, 102, 4397–4402.
Solovieva N, Jones V, Birks HJB, Appleby P, Nazarova L (2008)
Diatom responses to 20th century climate warming in lakes from
the northern Urals, Russia. Palaeogeography, Palaeoclimatology,
Palaeoecology, 259, 96–106.
Solovieva N, Jones VJ, Nazarova L et al. (2005) Palaeolimnological
evidence for recent climatic change in lakes from the northern
Urals, arctic Russia. Journal of Paleolimnology, 33, 463–482.
Sorvari S, Korhola A (1998) Recent diatom assemblage change in
subarctic Lake Saanajarvi, NW Finnish Lapland. Journal of
Paleolimnology, 20, 205–215.
Sorvari S, Korhola A, Thompson R (2002) Lake diatom response
to recent Arctic warming in Finnish Lapland. Global Change
Biology, 8, 171–181.
Stoermer EF, Kociolek JP, Schelske CL, Conley DJ (1985) Siliceous
microfossil succession in the recent history of Lake Superior.
Proceedings of the Academy of National Sciences Philadelphia, 137,
106–118.
Stoermer EF, Smol JP (1999) The Diatoms: Applications for the
Environmental and Earth Sciences. Cambridge University Press,
Cambridge, UK.
Straile D, Adrian R (2000) The North Atlantic Oscillation and
plankton dynamics in two European lakes – two variations on
a general theme. Global Change Biology, 6, 663–670.
Warren FJ, Egginton PA (2008) Background Information. In: From
Impacts to Adaptation: Canada in a Changing Climate 2007 (eds
Lemmen DS, Warren FJ, Lacroix J, Bush E), pp. 27–56. Govern-
ment of Canada, Ottawa, ON.
Werner P (2003) Reconstructing trophic state changes for shallow and
deep lakes in a limestone region of Ontario, Canada, using paleo-
limnological indicators. Unpublished PhD dissertation, Queen’s
University, Ontario, Canada, 211 pp.
Werner P, Chaisson M, Smol JP (2005) Long-term limnological
changes in six lakes with differing human impacts from a
limestone region in southwestern Ontario, Canada. Lake and
Reservoir Management, 21, 436–452.
Weyhenmeyer GA, Adrian R, Gaedke U, Livingstone DM, Ma-
berly SC (2002) Response of phytoplankton in European lakes
to a change in the North Atlantic Oscillation. Verhandlungen der
Internationalen Vereinigung von Limnologen, 28, 1436–1439.
Winder M, Schindler DE (2004a) Climate change uncouples trophic
interactions in an aquatic ecosystem. Ecology, 85, 2100–2106.
Winder M, Schindler DE (2004b) Climatic effects on the phenol-
ogy of lake processes. Global Change Biology, 10, 1844–1856.
Wolfe AP (1997) On diatom concentrations in lake sediments:
results from an inter-laboratory comparison and other tests
performed on a uniform sample. Journal of Paleolimnology, 18,
261–268.
Wolfe AP, Cooke CA, Hobbs WO (2006) Are current rates of
atmospheric nitrogen deposition influencing Lakes in the
Eastern Canadian Arctic? Arctic, Antarctic, and Alpine Research,
38, 465–475.
Wolfe AP, van Gorp AC, Baron JC (2003) Recent ecological and
biogeochemical changes in alpine lakes of Rocky Mountain
National Park (Colorado, USA): a response to anthropogenic
nitrogen deposition. Geobiology, 1, 153–168.
P L A N K T O N I C D I AT O M R E S P O N S E S T O R E C E N T WA R M I N G 2753
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754
Supplementary material
The following supplementary material for this article is avail-
able online:
Appendix S1. List of diatom taxa including taxonomic
authorities, taxonomic synonyms, and the species complexes
referred to in manuscript.
Appendix S2. Results of literature search for paleolimnological
studies containing Cyclotella species. Lake names in bold
indicate studies with suitable chronology and that reported
increases in Cyclotella species 45%. Rows shaded in grey
indicate all studies that did not have suitable chronology.
Studies without a suitable chronology but that had 45%
relative abundance of Cyclotella taxa (i.e. 12 studies)
were tallied in the chronology column. HL 5 high latitude;
HA 5 high altitude; T 5 temperate lakes are sorted by
author(s) in alphabetical order. Lakes in bold that had
suitable chronology and that showed 45% increase in the
relative abundance of Cyclotella taxa were plotted in Fig. 1.
In general, lakes considered to be unpolluted had total
phosphorus (TP) concentrations o20mg L�1 and pH46.0.
w/ 5 with; w/o 5 without; ON 5 Ontario; BC 5 British
Columbia; NW 5 Northwest; NS 5 Nova Scotia; NB 5 New
Brunswich; QC 5 Quebec; USA 5 United States of America.
This material is available as part of the online article from
http://www.blackwell-synergy.com/doi/abs/10.1111/
j.1365-2486.2008.01670.x
Please note: Blackwell Publishing is not responsible for the
content or functionality of any supplementary materials
supplied by the authors. Any queries (other than missing
material) should be directed to the corresponding author for
the article.
2754 K . R U H L A N D et al.
r 2008 The AuthorsJournal compilation r 2008 Blackwell Publishing Ltd, Global Change Biology, 14, 2740–2754